SUSTAINING

ECONOMIC
GROWTH
IN ASIA
JÉRÉMIE COHEN-SETTON
THOMAS HELBLING
ADAM S. POSEN
CHANGYONG RHEE
editors

Peterson Institute for International Economics

Korean Ministry of Strategy and Finance
Bank of Korea
International Monetary Fund
SUSTAINING
ECONOMIC
GROWTH
IN ASIA
JÉRÉMIE COHEN-SETTON
THOMAS HELBLING
ADAM S. POSEN
CHANGYONG RHEE
editors

Peterson Institute for International Economics

Korean Ministry of Strategy and Finance
Bank of Korea
International Monetary Fund

Washington, DC
December 2018
SUSTAINING
ECONOMIC
GROWTH
IN ASIA
JÉRÉMIE COHEN-SETTON
THOMAS HELBLING
ADAM S. POSEN
CHANGYONG RHEE
editors

Peterson Institute for International Economics

Korean Ministry of Strategy and Finance
Bank of Korea
International Monetary Fund

Washington, DC
December 2018
Jérémie Cohen-Setton is research fellow at the Printed in the United States of America
Peterson Institute for International Economics. 20 19 18 5 4 3 2 1

Thomas Helbling is a division chief in the Library of Congress

International Monetary Fund’s Asia and Pacific Cataloging-in-Publication Data
Department, covering Australia and New Zealand. Names: Cohen-Setton, Jérémie, editor. | Helbling,
Thomas, editor. | Posen, Adam S., editor. |
Adam S. Posen is the president of the Peterson Changyong Rhee, editor. Title: Sustaining economic
Institute for International Economics. growth in Asia / Jérémie Cohen-Setton, Thomas
Changyong Rhee is the director of the Asia and Helbling, Adam S. Posen, editors. Description:
Pacific Department at the International Monetary Washington, DC : Peterson Institute for
Fund. International Economics, 2018. | Includes biblio-
graphical references and index.
Identifiers: LCCN 2018006494 (print) | LCCN
PETERSON INSTITUTE FOR
2018007715 (ebook) | ISBN 9780881327342
INTERNATIONAL ECONOMICS
(ebook) | ISBN 9780881327335
1750 Massachusetts Avenue, NW
Subjects: LCSH: Economic development—Asia.
Washington, DC 20036-1903
| Asia—Economic conditions—21st century. |
+1.202.328.9000 www.piie.com
Business cycles—Asia. | Stagnation (Economics)
Adam S. Posen, President Classification: LCC HC412 (ebook) | LCC HC412
Steven R. Weisman, Vice President for .S883 2018 (print) | DDC 338.95—dc23
Publications and Communications LC record available at https://na01.safelinks.protec-
tion.outlook.com/?url=https%3A%2F%2Flccn.loc.
Cover Photo by © CJ Nattanai—Fotolia gov%2F2018006494&data=02%7C01%7Csluetje
Printing by Ironmark n%40piie.com%7Cc11425f8858e4c30ab4108d57
Copyright © 2018 by the Peterson Institute 5356fb7%7C55339d36654f44c1a1df1ed0983fad
for International Economics. All rights reserved. No cd%7C1%7C0%7C636543794751849480&sdata=
part of this book may be reproduced or utilized in oUSLWYOGvpfumicnf8BDAFVnhRr8bo5CskA6-
any form or by any means, electronic or mechanical, A%2BquahQ%3D&reserved=0
including photocopying, recording, or by informa-
tion storage or retrieval system, without permission
from the Institute.

For reprints/permission to photocopy please

contact the APS customer service department at
Copyright Clearance Center, Inc.,
222 Rosewood Drive, Danvers, MA 01923;
or email requests to: info@copyright.com

This publication has been subjected to a prepublication peer review intended to ensure analytical quality.
The views expressed are those of the authors. This publication is part of the overall program of the Peterson Institute for
International Economics, as endorsed by its Board of Directors, but it does not necessarily reflect the
views of individual members of the Board or of the Institute’s staff or management.
The Peterson Institute for International Economics is a private nonpartisan, nonprofit institution for rigorous, intellectually open,
and indepth study and discussion of international economic policy. Its purpose is to identify and analyze important issues to make
globalization beneficial and sustainable for the people of the United States and the world, and then to develop and communicate
practical new approaches for dealing with them. Its work is funded by a highly diverse group of philanthropic foundations, private
corporations, and interested individuals, as well as income on its capital fund. About 35 percent of the Institute’s
resources were provided by contributors from outside the United States. A list of all financial supporters is
posted at https://piie.com/sites/default/files/supporters.pdf.
Contents

Preface vii
1 Introduction: Toward a New Long-Term Growth Model 1
for Asia
Jérémie Cohen-Setton, Thomas Helbling, Adam S. Posen, and
Changyong Rhee
2 Secular Stagnation: History and Reality 9
Joel Mokyr
3 Is Asia at Risk of Growing Old before Becoming Rich? 33
Serkan Arslanalp, Jaewoo Lee, Minsuk Kim, Umang Rawat,
Jacqueline Pia Rothfels, Jochen Markus Schmittmann, and
Qianqian Zhang
4 Invention, Productivity, and the Evolution of East Asia’s 73
Innovation Systems
Lee Branstetter and Namho Kwon
Comment: Jong-Wha Lee
5 Secular Stagnation and Asia: International Transmission 103
and Policy Spillovers
Olivier Jeanne
6 Getting Out of Secular Stagnation: Turning Japanese May 129
Be a Good Thing Now
Adam S. Posen
7 Secular Stagnation and the Labor Market in Japan 139
Kyoji Fukao
Comment: Jonathan Woetzel
8 Declining Potential Growth in Korea 165
Dongchul Cho and Kyooho Kwon
9 A New Macroeconomic Policy Framework for Prudence 181
and Higher-Quality Growth in China
Ma Jun
10 Do India’s Exports Reflect the New Normal? 193
Prachi Mishra and Siddhartha Nath
Comment: Kenneth Kang
11 How Has Indonesia Fared in the Age of Secular Stagnation? 235
Mitali Das
Comment: Muhamad Chatib Basri
12 Twenty-Five Years of Global Imbalances 269
Maurice Obstfeld
13 Monetary and Exchange Rate Policies for Sustained 285
Growth in Asia
Joseph E. Gagnon and Philip Turner
14 Monetary Policy in the New Mediocre 315
Rania Al-Mashat, Kevin Clinton, Benjamin Hunt,
Zoltan Jakab, Douglas Laxton, Hou Wang, and
Jiaxiong Yao
15 Avoiding a New Mediocre in Asia: What Can Fiscal 343
Policy Do?
Ana Corbacho, Dirk Muir, Masahiro Nozaki, and Edda Zoli
16 Global Imbalances and the Trade Slowdown: Implications 387
for Asia
Caroline Freund
Comment: Davin Chor
17 Toward a “New” Asian Model 407
Subir Gokarn
Index 413
Preface

Asia largely avoided the worst of the global financial crisis of 2008–10 and
remains one of the most dynamic regions of the world economy. Yet, signs
of secular stagnation in the form of depressed levels of economic growth
and low interest rates are now appearing in emerging Asia, not just Japan.
Demographic trends and declining openness to trade in the advanced econ-
omies mean that Asian policymakers must prepare now to prevent such
stagnation taking hold.
In this volume, scholars affiliated with the International Monetary
Fund (IMF), the Peterson Institute for International Economics (PIIE), and
other institutions around the world explore the relevance and implications
of the secular stagnation hypothesis in Asia. The volume contains research
focused by subject area (demographics, innovation, imbalances, spillovers,
trade, and fiscal and monetary policies) and by country (China, South Korea,
India, Indonesia, and Japan). This two-pronged comparative approach yields
a compelling and multifaceted case for getting ahead of developments.
A major finding of this book is that in a secular stagnation environment,
the highly successful Asian growth model of the past few decades is unlikely
to remain the blueprint for future growth or for further convergence among
emerging and developing economies in the region. Instead, economic inte-
gration within the region is likely to become the critical source for steady
demand and productivity improvements. This regional opening should be
accompanied by structural reforms aimed at offsetting some of the predict-
able forces of demographic change to keep the growth engine powerful. Ad-

vii
aptation of macroeconomic frameworks to be ready to undertake aggressive
stimulus when necessary will be needed in Asia as well.
The Bank of Korea and the Korean Ministry of Strategy and Finance
sponsored and hosted a conference in Seoul in September 2017, where the
earlier versions of the research included in this volume were initially present-
ed. We would like to express our sincere gratitude to economists from the
Bank of Korea, IMF, PIIE, and the other organizations who have contribut-
ed their research to this ambitious project. We would also like to especially
thank the IMF Asia and Pacific Department’s Thomas Helbling and PIIE’s
Jérémie Cohen-Setton, who jointly managed and coordinated the project.
The IMF’s Nadine Dubost and Medha Madhu Nair and PIIE’s Madona
Devasahayam, Egor Gornostay, Susann Luetjen, and Steven R. Weisman
together made possible the publication of this volume, with full documen-
tation disclosure. Finally, we would like to acknowledge the generous finan-
cial support for the whole project by the Korean Ministry of Strategy and
Finance. We believe that this highly forward-looking book will be a valuable
resource for analysts, investors, and policymakers in Asia and beyond.

***
The Peterson Institute for International Economics is a private nonpartisan,
nonprofit institution for rigorous, intellectually open, and indepth study
and discussion of international economic policy. Its purpose is to iden-
tify and analyze important issues to making globalization beneficial and
sustainable for the people of the United States and the world and then to
develop and communicate practical new approaches for dealing with them.
The Institute’s work is funded by a highly diverse group of philanthropic
foundations, private corporations, and interested individuals, as well as
income on its capital fund. About 35 percent of the Institute resources in
our latest fiscal year were provided by contributors from outside the United
States. A list of all our financial supporters for the preceding year is posted
at www.piie.com/supporters.cfm.
The Executive Committee of the Institute’s Board of Directors bears
overall responsibility for the Institute’s direction, gives general guidance and
approval to its research program, and evaluates its performance in pursuit
of its mission. The Institute’s President is responsible for the identification
of topics that are likely to become important over the medium term (one
to three years) that should be addressed by Institute scholars. This rolling
agenda is set in close consultation with the Institute’s research staff, Board
of Directors, and other stakeholders.
The President makes the final decision to publish any individual In-
stitute study, following independent internal and external review of the

viii
work. Interested readers may access the data and computations underly-
ing Institute publications for research and replication by searching titles at
www.piie.com. The Institute hopes that its research and other activities will
contribute to building a stronger foundation for international economic
policy around the world. We invite readers of these publications to let us
know how they think we can best accomplish this objective.

***
The International Monetary Fund’s Asia and Pacific Department is one of
the IMF’s five area, or regional, departments. It is responsible for advising
member countries in the Asia Pacific Region on macroeconomic policies
and the financial sector, putting together, when needed, financial arrange-
ments to support economic reform programs, and capacity building, mainly
through its regional technical assistance centers. To accomplish its mission,
the Department seeks to engage in policy debates in the Asia-Pacific and to
distill lessons from the experience in the region. In this mission, it collabo-
rates with experts from and outside region.

ADAM S. POSEN CHANGYONG RHEE

President Asia and Pacific Department Director
Peterson Institute for International Monetary Fund
International Economics

ix
PETERSON INSTITUTE FOR INTERNATIONAL ECONOMICS
1750 Massachusetts Avenue, NW, Washington, DC 20036-1903 USA
+1.202.328.9000 www.piie.com

Adam S. Posen, President

BOARD OF DIRECTORS
*Michael A. Peterson, Chair of the Board
* Lawrence H. Summers, Vice Chair of the Board
* Stephen Freidheim, Chair of the Executive Committee

Caroline Atkinson * Richard E. Salomon

Ajay Banga Sheikh Hamad Saud Al-Sayari
* C. Fred Bergsten Mostafa Terrab
Mark T. Bertolini Mark Tucker
Ben van Beurden D. James Umpleby III
Neeti Bhalla Johnson Ronald A. Williams
Frank Brosens * Robert B. Zoellick
Ronnie C. Chan Min Zhu
* Susan M. Collins
HONORARY DIRECTORS
Richard N. Cooper
Jason Cummins George David
Barry Eichengreen Alan Greenspan
* Stanley Fischer Carla A. Hills
Peter Fisher Frank E. Loy
Jacob A. Frenkel George P. Shultz
Evan G. Greenberg Jean-Claude Trichet
Maurice R. Greenberg Paul A. Volcker
Herbjorn Hansson Ernesto Zedillo
Stephen Howe, Jr. FOUNDING CHAIR
Jay Jacobs Peter G. Peterson (1926–2018;
Hugh F. Johnston Institute Chair 1981–2018)
* Michael Klein
Nobuyori Kodaira
Charles D. Lake II
Andrew N. Liveris
Pip McCrostie
* Hutham S. Olayan
Peter R. Orszag
James W. Owens
Jonathan Pruzan
John G. Rice
Ginni M. Rometty
* Lynn Forester de Rothschild

* indicates Executive Committee member

 1
Introduction: Toward a
New Long-Term Growth
Model for Asia
JÉRÉMIE COHEN-SETTON, THOMAS HELBLING,
ADAM S. POSEN, AND CHANGYONG RHEE

There is a false sense of security among many Asian economic policymakers

regarding their own and the region’s growth prospects. This is to some
degree understandable, because Asia overall has continued to enjoy high
rates of economic growth since the global financial crisis in 2008–10. In
fact, the region has become the main driver of global economic activity,
accounting for some two-thirds of global growth. In contrast, economic
recovery from the crisis in advanced economies outside Asia has entailed a
decade of mostly subpar growth and persistent labor market slack, despite
sustained loose monetary policy. This experience has motivated the revival
of Alvin Hansen’s secular stagnation hypothesis by Lawrence Summers
(2013) and others, as well as more structural explanations (notably Gordon
2016), for the lasting downshift in trend productivity growth at the tech-
nological frontier. Given the continued contrast in growth rates and ample
room for most countries to continue catch-up growth, the secular stagna-
tion hypothesis usually is not associated with Asia, except for Japan with its
aging population and high per capita income.
Nevertheless, the danger from secular stagnation to Asian economies
of an extended and difficult-to-exit growth slowdown, as has arguably beset

Jérémie Cohen-Setton is research fellow at the Peterson Institute for International Economics. Thomas Helbling
is a division chief in the International Monetary Fund’s Asia and Pacific Department, covering Australia and
New Zealand. Adam S. Posen is the president of the Peterson Institute for International Economics. Changyong
Rhee is the director of the Asia and Pacific Department at the International Monetary Fund. The views ex-
pressed in this chapter are those of the authors and do not necessarily represent the views of the IMF, its Executive
Board, or its management.

1
the G-7 economies, is real and approaching. This threat is not just the result
of the difficulties experienced through trade and financial linkages with the
advanced economies, their low growth and inflation, and the resultant very
low interest rates. When considering the broader macroeconomic picture,
several of the symptoms associated with secular stagnation are already
prominent in several Asian economies. Growth, for example, while strong in
cross-national comparison, has remained below the rates recorded before the
global financial crisis in Asia as well as the West. More troublingly, average
productivity growth has declined relative to precrisis rates in many countries
in the region, as it has elsewhere. Rapid population aging and lower popu-
lation growth are already shaping the demographics in many countries of
the region; turning points toward rising dependency ratios and shrinking
workforces will soon be a reality well beyond Japan. Inflation has been below
target for an unusually long period in several countries in the region, while
interest rates, nominal and real, have declined to very low levels in many
economies, even in countries with still high real growth. Therefore, risks of
nominal policy interest rates reaching the effective lower bound after a large
shock and related problems of macroeconomic adjustment might be likely
rather than tail risks.
For this reason, the Asia and Pacific Department of the International
Monetary Fund (IMF) and the Peterson Institute for International Econom-
ics launched a joint research project to explore the relevance of the secular
stagnation hypothesis for Asia and the policy implications. The papers
published in this volume were presented at a conference organized jointly
with the Korean Ministry of Strategy and Finance and the Bank of Korea
in September 2017. The project combined research organized by econom-
ic themes and by countries. For the country case studies, the project dis-
tinguishes three groups of countries in Asia. The first category consists of
Japan, a country in which symptoms of secular stagnation have already been
present for some years. The second group includes Korea and other mid-
dle-income economies in which the reversal of the demographic dividend is
imminent and which are already exhibiting some of the features of secular
stagnation but not yet to an extent that would support wide recognition
of the problem. In these transitionally threatened economies, fundamentals
point to increasing risks of stagnation over the next few years but also some
advance warning, meaning time to prepare and change policy. Finally, in
many of the emerging and lower-income economies in Asia, the exposure
to secular stagnation is primarily through international spillovers to date,
but these spillovers are likely to last and to constrain growth and policy in
an ongoing way.

2 SUSTAINING ECONOMIC GROWTH IN ASIA

Summary of Contributions to This Volume
In chapter 2 Joel Mokyr reminds us that secular stagnation was the rule
rather than the exception until 1800. Negative demographic feedbacks,
predatory rent-seeking institutions, and limitations on the accumulation of
human knowledge made growth imperceptibly slow. Today, the economy is
facing headwinds, but progress made in a year no longer melts away subse-
quently. A first headwind is the risk of growing old before becoming rich in
parts of Asia. In chapter 3, Serkan Arslanalp et al. find that rapid aging is
now set to create a demographic tax on growth that could subtract ½ to 1
percentage point from annual GDP growth over the next three decades in
countries such as China and Japan. A second headwind is the slowdown in
R&D productivity growth. Based on the experience of Japan, South Korea,
and Taiwan, Lee Branstetter and Namho Kwon argue in chapter 4 that these
countries’ innovation systems have become a drag on productivity growth
because of their pro-incremental and pro-incumbent biases.
Headwinds for the region are not only domestic. In fact, the concomi-
tance of low growth and inflation in a large set of countries at different
stages of development suggests a role for contagion. In chapter 5 Olivier
Jeanne explores several possible channels of contagion but finds that post-
crisis total factor productivity growth is surprisingly orthogonal to most
measures of international economic integration. There is, however, evidence
that the transmission of secular stagnation to emerging economies has
been delayed, with rebalancing in current accounts happening against the
backdrop of domestic credit booms.
When economists want to draw lessons from recent historical expe-
rience, they study Japan’s two lost decades. The problem, writes Adam S.
Posen in chapter 6, is that Japan had only one lost decade, not two. Since
2002, its per capita GDP growth and productivity growth have been decent.
And its difficulty in generating inflation is not Japan specific. Explaining
the empirical puzzles of the 21st century requires a more general rethinking
of our understanding of macroeconomics, where factors other than just
expectations can influence inflation. In chapter 7 Kyoji Fukao illustrates
the importance of changes in workforce composition such as the increase
in nonregular employment in the Japanese labor market to explain both
inflation and productivity developments.
In chapter 8 Dongchul Cho and Kyooho Kwon observe that develop-
ments in South Korea seem to mimic those of Japan with a 20-year lag. They
project that potential output growth will fall further, given demographic
trends and capital deepening, and call for bold structural reforms to raise TFP
and to lower the risks of Korea experiencing the deflation pressures Japan
has faced. In China, writes Ma Jun in chapter 9, the slow down in economic

INTRODUCTION 3
growth will be driven by demographics, environmental costs, and changing
consumer preferences. Altogether, these factors are expected to slow annual
growth by 2 percentage points in the coming decade. While ongoing tech-
nological innovations and reforms can partially cushion this deceleration, a
further slowdown in growth appears inevitable over the medium term. This
trend, together with the growing financial risks due to a high macro leverage
ratio, calls urgently for a new macroeconomic management framework that
is less growth-centric but focuses more on macro and financial stability.
In chapter 10 Prachi Mishra and Siddhartha Nath write that India’s
merchandise exports have undergone a quiet revolution with new-age engi-
neering, electronic, and pharmaceutical exports gradually replacing India’s
traditional exports of leather, textiles, gems, and jewelry. But structural ri-
gidities are still holding back diversification and thus make India vulnerable
to global risks. The contrast with Indonesia is striking as it has progressively
turned inward. This retrenchment, writes Mitali Das in chapter 11, has in-
sulated Indonesia against the vicissitudes of global developments but at the
price of a reduction in productivity-enhancing spillovers. In fact, potential
output growth has declined in recent years despite strong demographic tail-
winds and steady capital accumulation.
In reviewing the past 25 years of global imbalances, Maurice Obstfeld
emphasizes in chapter 12 that excessive imbalances declined after the global
financial crisis but are still present. Monocausal explanations for excessive
global imbalances rarely apply. Instead, such imbalances usually reflect
global forces and multiple distortions in many countries, notably diverse
financial sector distortions. Notwithstanding the need for collective action,
excess surplus countries still face little that would force them to adjust—
outside of the threat of protectionist responses—whereas most deficit coun-
tries face the risk that lenders will withdraw. Reducing global imbalances
should be a collective effort based on a shared appreciation of the roles indi-
vidual countries need to play.
In chapter 13 Joseph Gagnon and Philip Turner study the implications
of a slowdown in growth for monetary policy. A first implication is the
importance of keeping core inflation at or above 2 percent. A second impli-
cation is that central banks should not shy away from using balance sheet
policies. The view that central banks should stick to setting the overnight
rate is unhistorical. If anything, balance sheet policies have more poten-
tial today than before given the increased size and importance of domestic
Asian bond markets in influencing financial conditions. Rania Al-Mashat
et al. make the case in chapter 14 for a risk avoidance approach to monetary
policy that heavily penalizes large deviations of the target variables (output
and inflation) from their preferred values. In a world of secular stagnation
where the probability of entering low-inflation traps is higher, a prudent

4 SUSTAINING ECONOMIC GROWTH IN ASIA

approach to a recessionary shock from a weak starting point is to take a
more aggressive stance than usual. The authors of both chapters 13 and 14
argue that financial stability worries do not in general justify keeping the
policy rate higher than warranted by macroeconomic conditions.
In chapter 15 Ana Corbacho et al. analyze the implications of subdued
growth and rapid aging for fiscal sustainability and discuss options for fiscal
policy to support growth potential. Left unchecked, age-related spending
would be an overwhelming force driving up public debt in many countries in
the region. A reversal of the favorable interest rate–growth differential that
helped keep debt ratios in check in recent times would put debt dynamics
at even further risk. This being said, model-based simulations show that
many countries in Asia have scope to anchor drivers of growth potential by
redirecting budgets to infrastructure, human capital, and R&D.
In chapter 16 Caroline Freund studies the relationship between global
imbalances and trade. In the 1990s and early 2000s, increased borrowing
from abroad allowed high-demand countries to import more than if they
were constrained by their exports, stimulating trade growth. The relation-
ship between trade imbalances and trade flows is especially strong for the
United States and East Asia. In recent years, moderating imbalances have
been associated with slowing trade growth, especially in the United States
and East Asia. Going forward, even as global growth picks up, the new
weaker relationship between trade and income may remain in place if global
imbalances remain constrained.
In chapter 17 Subir Gokarn writes that the “Asian model” as we have
come to recognize it is obsolete. Countries in the region have to recognize
this and reorient their strategies for sustaining growth. Changing demand
patterns in the advanced economies clearly indicate that these markets will
become less and less important sources of demand for Asian products. But
the key element of the “old” Asian model—the promotion of efficiency and
competitiveness up the supply chain—provides the basis for a “new” model
where the focus would be on meeting the needs of the growing markets in
the region itself. Whether the “new” Asian model is a move toward medi-
ocrity or renewed excellence depends on how effectively the countries in
the region recognize and act on the opportunities and respond to the chal-
lenges.

Agreeing on the Need for a Forward-Looking Policy

There was broad agreement among conference participants that forward-
looking adaptation of policy frameworks is needed to sustain growth in Asia
for the medium term. So, even though the conference took place against the
backdrop of an improved global outlook compared with the time when the

INTRODUCTION 5
research program was conceived, and the global expansion has continued
in the months since, the best tradition of Asian economic policymaking to
plan with a long time horizon was affirmed. In particular, the profound chal-
lenges from the impending demographic transitions in many Asian econo-
mies to rapidly aging populations and lower population growth resonated
with participants. The demographic dividend has already started to reverse
in many countries in the region, including in China, the engine of growth
for the region and the world in the past two decades.
Recent experience in Japan and in Western Europe suggests that aging
and lower population growth could reduce growth in per capita incomes
and productivity, not just aggregate growth in output and investment.
There are strong hypotheses suggested by Japan and Western Europe as
well that there might even be negative feedback loops from demographics
to technical progress: Lower rates of innovation in the frontier economies
and eroding intellectual property rights could interact with aging to slow
diffusion of progress within and across countries (Posen 2012); aging could
change the structure of household demand toward less capital-intensive
services, providing yet another reason for lower investment, thus limiting
positive productivity spillovers; structural reforms are needed to strengthen
domestic innovation capacity and productivity in service sectors.
The reversal of the demographic dividend would therefore likely result
in more than just persistent even if small declines in aggregate demand.
So, even absent macroeconomic imbalances, which arguably made the
problems more acute in the G-7 economies, Asian policymakers are right
to be concerned about the negative impact of the demographic transition
on productivity and medium-term growth. The conference participants
emphasized that the transition reinforced the need to change the Asian
growth paradigm.
Similarly, the international dimension of the challenges facing Asia for
sustaining growth go well beyond the cyclical and direct immediate spill-
overs from insufficient demand and low growth in advanced economies
since the global financial crisis. After a decade of very low growth since the
global financial crisis, external demand growth from advanced economies
outside Asia is unlikely to return to providing the momentum it had before
the global financial crisis. An implication is that manufacturing exports are
unlikely to be the motor for growth and development that they were over
the past few decades. This development intensifies the phenomenon of “pre-
mature deindustrialization” identified by Dani Rodrik (2016). Moreover,
the trade tensions taking shape between the United States and many of its
largest trading partners in Asia and elsewhere are further eroding the room
for export-led growth. Even if these conflicts are resolved or at least stepped
down, legitimate doubts about the wisdom of relying on access to advanced

6 SUSTAINING ECONOMIC GROWTH IN ASIA

economies’ markets for large-scale exports will remain. If so, the highly suc-
cessful Asian growth model of the past few decades is unlikely to remain the
blueprint for future growth and convergence in emerging and developing
economies in the region.
A third major theme is that macroeconomic policy frameworks in Asia
had to be prepared for the possibility of policy rates hitting the lower bound
on nominal interest rates after contractionary or deflationary shocks, even if
the source of shock or of the low level of interest rates going in was external,
and the domestic macroeconomy appeared well-balanced. This means that
in the case of a downturn, there is a need for determined aggressive macro-
economic policy responses, given higher risks of protracted adjustment to
shocks. Many conference participants were concerned about the risks to
financial stability from such policy responses, noting the possible limited
effectiveness of prudential policies to offset those side effects.
This scenario is relevant for Asian policymakers, even if not induced
by their own economies’ actions. Admittedly, there is no easy answer to
the tradeoff between demand stimulus and financial stability, though to
some degree the risks from financial side effects can be exaggerated during
recovery phases. As argued in Summers (2013), such a situation presents the
risk not only of more severe downturns but also of getting stuck in a trap of
extended need for still lower rates and substantial fiscal stimulus. Just as for
their counterparts in the advanced economies, the risk of getting trapped
in secular stagnation with only monetary policy to get the economy out is
that it will be ineffective as well as distortionary. Thus, it is important that
Asian policymakers get ahead of the curve to build fiscal space ahead of
such situations and be prepared to use that fiscal space in productive as well
as stimulative ways (such as ready infrastructure projects). The failure of US
and European policymakers to do so, and apparent benefits with hindsight
of such aggressive fiscal action in Japan, is an object lesson.
Asia remains the most vibrant, the most diverse, and the largest economic
region of the world economy. Its citizens’ economic well-being remains
largely in the hands of their own markets and governments. The economic
policymakers of the region could take solace from their economies’ rela-
tively good performance compared with their counterparts in other parts
of the world, even as their growth rates of income and productivity have
slowed. They could even blame those slowdowns on the spillovers from the
advanced economies’ own crises and policy mistakes. That, however, would
be a mistaken approach. Significant unfavorable demographic trends, a less
open and slower-growing rest of the world to trade with, and ongoing risks
of falling into stagnation from persistently low levels of nominal interest
rates and growth in the major economies combine to demand a new growth
model for Asia.

INTRODUCTION 7
 The broad outlines of such a model for sustaining growth are applicable
for all Asian economies and should be put in place without undue delay.
n Undertake a structural reform agenda—including making better use of
the female labor force, promoting innovation and technological diffu-
sion internally, and encouraging productivity gains in service sectors—
that can robustly offset some of the predictable forces of demographic
change.
n Pursue international economic integration within the Asian region,
and openness to South-South trade and investment more broadly, to
maintain the benefits for productivity and steady demand that go with
export competition.
n Prepare macroeconomic frameworks to be ready to undertake aggres-
sive stabilizing stimulus when needed in the likely-to-persist global
low interest rate environment. This preparation requires a recognition
both of the risks of excessively anti-inflationary monetary policy and
of the need to build up fiscal space in advance of problems.

References
Gordon, Robert J. 2016. The Rise and Fall of American Growth: The U.S. Standard of Living since the
Civil War. Princeton, NJ: Princeton University Press.
Posen, Adam S. 2012. What the Return of 19th Century Economics Means for 21st Century
Geopolitics. Lecture, Chatham House, London, January 17. Available at www.chatham-
house.org/sites/default/files/public/Meetings/Meeting%20Transcripts/170112aposen.
pdf (accessed on September 7, 2018).
Rodrik, Dani. 2016. Premature Deindustrialization. Journal of Economic Growth 21, no. 1: 1–33.
Summers, Lawrence H. 2013. Remarks at the IMF Fourteenth Annual Research Conference
in Honor of Stanley Fischer, Washington, DC, November 8. Available at http://larry-
summers.com/imf-fourteenth-annual-research-conference-in-honor-of-stanley-fischer/
(accessed on September 7, 2018).

8 SUSTAINING ECONOMIC GROWTH IN ASIA

 2
Secular Stagnation: History
and Reality
JOEL MOKYR

Should we be concerned with “secular stagnation,” the decline of long-term

economic growth (Summers 2016)? Modern economic growth, or as Deirdre
McCloskey has termed it “the Great Enrichment,” has been a relatively recent
phenomenon. If recorded human history is approximately 6,000 years long,
sustained growth that doubled income every generation or so has been ex-
perienced only for at most 3 percent of human history, with the remaining
97 percent characterized by stagnant economies. No one before the Indus-
trial Revolution complained about secular stagnation—practically nobody
imagined that economic growth and continuous progress in the material
conditions of life were at all possible. Whatever progress made in one year
would have melted away subsequently, and a random individual born on
this planet at the time of Napoleon would not have been significantly richer
than a random individual born at the time of Nebuchadnezzar.1 In that
sense, secular stagnation would have been a return to normal.

Joel Mokyr is Robert H. Strotz Professor of Arts and Sciences and Professor of Economics and History at North-
western University and Sackler Professor (by special appointment) at the Eitan Berglas School of Economics,
Tel Aviv University. He is indebted to Ashish Aggarwal for research assistance, Jérémie Cohen-Setton and Deir-
dre McCloskey for helpful comments, and Madona Devasahayam for outstanding editing. The Balzan Founda-
tion and Northwestern’s Center for Economic History provided financial support.
1. For powerful, if somewhat overstated, descriptions of this absence of growth, see Clark
(2007) and Galor (2011). There were some exceptions to the rule of stagnation in northwest
Europe and Italy, where growth had already raised living standards significantly by 1800, but
even that growth had been slow and uneven, and it affected a tiny part of humanity.

9
 But what was normal for most of human history is not normal in the
20th and 21st centuries. Growth has been the norm in the more recent past,
despite two world wars, a wave of totalitarian regimes, repeated genocide,
a nuclear arms race, and other assorted disasters. In what follows, I explain
how the modern era was different from everything that came before, which
will give us a unique perspective to assess the chances of past conditions
returning. Needless to say, all such assessments need to be phrased in the
conditional and subjunctive modes, since the entire point of this chapter
is to argue that we are still in the midst of a phase transition in economic
history, and hence all bets are off. But some scenarios can still be assigned
a low probability.

The Malthusian Curse

The most widely cited explanation for why growth could not take place
in the past is population dynamics. The idea was famously enunciated by
Thomas Malthus and has since carried his name, although it was perhaps
most eloquently expressed by H. G. Wells (1923, 68): Humanity “spent the
great gifts of science as rapidly as it got them in a mere insensate multipli-
cation of the common life.” Every economist knows the argument. Under
some fairly reasonable assumptions, any rise in income per capita or produc-
tivity, whether derived from the “great gifts of science” or any other source,
will lead to a decline in deaths and a rise in births, and as population growth
sets in, diminishing returns to a fixed factor such as land or more gener-
ally the environment will set in, and income will decline (Clark 2007, Galor
2011, Ashraf and Galor 2011). This kind of demographic negative feedback
explains why in the very long run any improvement in the human condition
was quite ineffectual, and the only indicator of technological progress in
the very long run is population size (Kremer 1993).
Whether the classical “iron law” is a good description of historical reality
remains to be seen; some parts of it have been confirmed using historical
data (Kelly and Ó Gráda 2012, 2014), but it is also clear that the stark impli-
cations of the fundamentalist Malthusian position are a simplification of
history (Voigtländer and Voth 2012, 2013). Either way, there is a general
consensus that the Malthusian model had ceased to be a good description
of the world at just about the time of the writing of Malthus’s famous book,
and that the stagnation it implied no longer held in the second half of the
18th century, when demographic growth took off in most parts of Europe.
The three basic assumptions that the Malthusian theory rested on have
been undone in modern times. The first is that birth rates rise with income:
On both global and intranational levels, the correlation is reversed. The
richer a nation, or a person, the lower their fertility tends to be, although it

10 SUSTAINING ECONOMIC GROWTH IN ASIA

is far from obvious that there is a causal connection (Black et al. 2013). Four
major explanations have been proposed for this correlation, all of them the
direct consequence of some form of economic modernization: the wide
availability of inexpensive means of effective fertility control; the desire to
have higher-quality children at the expense of quantity; the many substi-
tutes that have emerged for children in their various economic and social
functions; and the decline of religion in much of the secular and humanistic
West. In the richest part of the world, Europe, fertility rates are now below
replacement levels in every nation (except in France), and in North America,
both the United States and Canada are now in the same situation.
The second assumption of the Malthusian story, that mortality rates
will decline with income, is less clearly refuted. Rich countries or people
still have lower mortality rates. But that curve is flattening out, and for a
large range of incomes a negative correlation between income and mortality
rates is not observable.2 Malthusian “positive checks” (famines, epidemics,
and wars that would have reduced the population if income was too low)
have by and large disappeared from most of the world, and when famines
occurred in the 20th century, they tended to be entirely man-made and not
the result of overpopulation (Ó Gráda 2015, Alfani and Ó Gráda 2017).3
Spikes in mortality that in the preindustrial past were correlated with
poor harvests and/or the outbreaks of epidemics and have been seen as
Malthusian responses to population pressure have more or less disappeared
in most of the world.
The third leg of the Malthusian model is that overpopulation presses
income down through diminishing returns from land and natural resources.
The responses to population growth have always been a subject of debate;
scholars have argued that population growth basically triggered technolog-
ical responses that adapted to the labor-land ratio through the intensifica-
tion of agriculture (Boserup 1965, 1981) or even that population growth
led to a high rate of technological progress simply because the likelihood
of a technological genius appearing in a larger population is higher (Simon
2001, Kremer 1993). Yet the main argument for diminishing returns that

2. Simple regressions of mortality rates on log GDP per capita for 168 nations still show a
significant negative coefficient. However, once we limit the sample to the top 60 countries,
the significance vanishes and the adjusted R-square is essentially zero (author’s calculations;
all data from CIA World Factbook, www.cia.gov/library/publications/resources/the-world-
factbook/index.html).
3. The evidence for the existence of Malthusian checks in which disasters occur as a result
of overpopulation is scant. Nineteenth century writers believed that the Irish Famine was a
clear-cut example, but the evidence for overpopulation as a clear cause of the famine is far
from persuasive (Mokyr 1985).

SECULAR STAGNATION: HISTORY AND REALITY 11

Malthusians or neo-Malthusians have relied upon is the finiteness of
nonreproducible resources. As the shares of land and natural resources in
national income have declined, the assumption that a permanently “fixed”
factor in the economy limits the growth of per capita income is weakened
by the enormous land- and resource-augmenting technological progress in
the past century.4

Rent-Seeking and Predation

Perhaps the least noticed reason why incomes before the Industrial Revolu-
tion were more or less stagnant is that most economies were “extractive”
in the nomenclature popularized by Acemoglu and Robinson (2012) and
found themselves in a “natural state” (North, Wallis, and Weingast 2009).
In these states, dominant elites extracted rents from the rest of the popu-
lation. The institutional weakness of these societies was not so much the
absence of a “rule of law” as the rule of bad law, a legal system that was
biased toward the politically powerful and favored a rapacious rent-seeking
minority. The absence of any concept of “equality before the law” in most
societies meant that greedy strongmen, above all of course the rulers, could
easily expropriate the wealth of successful individuals. This risk remained
a constraint on the incentives of would-be entrepreneurs.5 Only in areas
where the political influence of successful merchants, industrialists, and
financiers could protect them from such expropriations did entrepreneurs
feel comparatively safe from the threat of expropriation or confiscatory
fiscal extraction. There were many such places in Europe, to be sure, but
rent-seeking remained pervasive in the age of mercantilism and was clearly
an obstacle to economic growth (Ekelund and Tollison 1981) even when
property rights became more secure.

4. The most striking land-augmenting inventions were fertilizers and pesticides, the most
dramatic of which by all accounts was the Haber-Bosch nitrogen-fixing process. According
to Vaclav Smil (2001, 204), by the end of the 20th century at least 2.4 billion people were alive
because the proteins in their bodies could be synthesized from amino acids whose nitrogen
originated in this invention.
5. A notorious example was the French merchant and entrepreneur Jacques Cœur (1395–
1456), an immensely rich and powerful merchant who basically monopolized France’s
Mediterranean trade and was wealthy enough to bankroll many of Charles VII’s triumphs in
the final stages of the Hundred Years’ War against England. His wealth and power attracted
the greed and envy of many, including eventually the king. Cœur was arrested and tried on
what historians have deemed were trumped-up charges; his possessions were confiscated and
distributed among the king and his cronies. Two centuries later, a similar fate befell Nicolas
Fouquet (1615–1680), the fabulously wealthy tax collector who had enriched himself during
the ministry of Mazarin during Louis XIV’s minority. He was arrested in 1661 by Louis XIV
and imprisoned for life, and the king’s servants stripped his sumptuous chateau.

12 SUSTAINING ECONOMIC GROWTH IN ASIA

 Such legal protections were insufficient, however. Many of Europe’s
most successful pre–Industrial Revolution economies were located in rela-
tively small political entities, some of which were classic city-states. Such
entities became rich because they were able to set up city governments that
protected merchants’ properties (Gelderblom 2013). Many urban areas
were able to raise living standards above the subsistence levels that Malthu-
sian fundamentalists believe constrained living standards. In the highly
fragmented and violent Europe before the Enlightenment, however, such
wealth attracted strong but poor foreign predators as honey attracts flies.
The wealthy towns of northern Italy were ravaged in the early 16th century
by Spanish and French soldiers (not to mention condottieri from a variety
of origins), and most of them did not recover. German trading towns such
as Augsburg, Magdeburg, Frankfurt an der Oder, Stralsund, and Prague,
among many others, were mercilessly looted and sacked by the mercenary
armies that roamed Germany during the Thirty Years’ War. Since none of
those armies could provide their supplies, they imposed enormous extor-
tions on the territories they moved through.6 As a recent essay (Béaur and
Chevet 2017) points out, not only the war itself but also the actions of
the military had devastating effects on the countryside. Soldiers looted all
around them, which discouraged farmers from cropping land and raising
animals too exposed to danger. The authors note that this was especially
disastrous in France during the wars of religion (1572–98) and the wars of
the Fronde (1648–52) (Béaur and Chevet 2017, 85). The deadweight losses
of such actions were thus enormous. Cities that were spared had to invest
heavily on fortifications, such as Hamburg, or they paid off the soldiers to
prevent looting.
The most striking example of what could best be called negative insti-
tutional feedback is provided by the Low Countries, a region of Europe
that successfully broke through the Malthusian barrier to provide its citi-
zens with a standard of living considerably higher than the subsistence
level implied by the iron law. Yet powerful neighbors threatened their pros-
perity: France and above all, but after 1648, England just as much. The
wealthy medieval towns of Ghent, Bruges, and Courtray had to repeatedly
fend off French invaders in the Middle Ages; the Burgundian rulers of the
15th century allowed these towns to return to prosperity, which was over-
shadowed by the meteoric rise of Antwerp in the 16th century. But after
1555 these provinces were part of Spain, and Spain was a violent and pred-

6. These extortions, known as Kontributions or Brandschatzung, were confiscatory taxation and

extortion imposed on local towns and countrysides. Armies often chose itineraries that led
through the most prosperous areas. See, for instance, Guthrie (2003, 30–31).

SECULAR STAGNATION: HISTORY AND REALITY 13

atory master. By 1585 (the notorious sack of Antwerp), much of Flanders
and Brabant’s prosperity had declined and the southern Netherlands
entered two centuries of sustained stagnation. In their place, of course,
came the northern Netherlands, which was more favorably located (as they
could fend off invading armies by flooding the lands). Yet it, too, had to
spend inordinate amounts on defense, which by the 18th century contrib-
uted to the economic stagnation of the United Provinces. Britain’s success
in the Industrial Revolution was at least in part due to the fact that no
hostile ruler after William the Conqueror had successfully invaded it or
could even threaten to do so credibly.
Part of the reason why growth was so vulnerable in the pre–Industrial
Revolution era is that most growth was “Smithian” in nature, that is, it was
fueled by gains from trade and specialization. Such growth tended to be
vulnerable to political shocks, including the use of state-sponsored pirates
(known as privateers) and economic embargoes of various kinds. Mercantil-
ist protectionist policies, whose primary purpose was political rather than
economic, diverted trade flows and even distorted consumption patterns in
the long run (Nye 2007).
In short, through much of history, both internal rent-seekers and pre­
datory foreigners were attracted to the riches that economic success had
allowed local populations to accumulate. In so doing, they were of course
killing the goose that laid the golden egg, but especially during war, when
the future was discounted heavily, such expropriation was often resorted to.
A high rate of discount—as would occur during war—could turn a stationary
bandit into a roving one. As a result, even when economic growth occurred,
it was often undone by its own success in a dialectic way.
This negative feedback has sharply declined in modern democratic
economies. Between 1750 and 1850 much of mercantilist rent-seeking dis-
appeared in Britain (Mokyr 2009, 423–27) and declined elsewhere. In most
developed Western democracies, rent-seeking today is at historically low
levels, but it has far from disappeared worldwide. For example, in countries
like Venezuela, Nigeria, and Russia, governments have degenerated into
modern kleptocracies, where massive theft and corruption by ruling elites
have condemned large economies to economic stagnation or worse. Yet on
the whole in many of the advanced and successful countries rent-seeking
has been kept within limits, and corruption does not reach levels at which it
seriously threatens entrepreneurship and economic progress. It is not sur-
prising that many of these countries rank well on the corruption indices
compiled by Transparency International and the World Bank. Not all rent-
seeking takes the form of corruption, of course, and no simple index of rent-

14 SUSTAINING ECONOMIC GROWTH IN ASIA

seeking exists.7 These indices are quite strongly correlated, and one could
conceptualize some principal component that would combine the six into
some “quality of institutions” variable.8 It is fair to state that the quality of
these institutions in some of the most economically successful economies
such as Denmark, Canada, and Singapore is higher than it has ever been in
history. At the same time, the threat of economic nationalism and notions
of “fair” trade (a euphemism for protectionism) seem indestructible weeds
in the garden of trade.

Technology and Knowledge

Many past economies experienced some technological progress, including
many in which growth was so slow that one would surely refer to them
as suffering from secular stagnation. Inventions did occur, and in some
cases they were game-changing (think of the mechanical clock and the
moveable-font printing press invented in the late Middle Ages). The medi-
eval economy, far from being technologically backward or even stagnant,
was actually quite dynamic in many areas, including agriculture, manu-
facturing, civil engineering, shipbuilding, and metallurgy. Yet these break-
throughs usually had only a small and temporary impact on economic
well-being. One may wonder why none of these advances translated into
any form of sustained technological progress and productivity growth.
The effectiveness of sustained technological progress depends largely
on the epistemic base of the invention, namely the extent to which it is un-
derstood why and how the new technique works (Mokyr 2002, 31–32).
Technological progress occurred through much of history, but there
was something deeply different in the innovations during and after the
Industrial Revolution. Before 1750, the vast bulk of inventions had oc-
curred through serendipity, experience, continuous trial and error, and the
gradual accumulation of small incremental improvements. The techniques
worked, but those who employed them did not understand how and why
they did so. Blacksmiths made steel, brewers brewed beer, and farmers fer-
tilized their fields, but nobody had much of an idea of the chemical and
biological processes at work. The nature of combustion, one of the most
elementary techniques used throughout manufacturing, was not under-

7. The World Bank indices are Voice and Accountability, Political Stability and Absence
of Violence, Government Effectiveness, Regulatory Quality, Rule of Law, and Control of
Corruption. See http://info.worldbank.org/governance/wgi/index.aspx#home.
8. The pairwise correlation coefficients of three of the measures that are closest to rent-
seeking are especially high: between “control of corruption” and “government effectiveness”:
R = .909; between “control of corruption” and “rule of law”: R = .962; and between “govern-
ment effectiveness” and “rule of law”: R = .940 (author’s calculations).

SECULAR STAGNATION: HISTORY AND REALITY 15

stood until the late 18th century. Even during the Industrial Revolution
itself, some of the great breakthroughs—especially those in the mechanical
parts of the cotton industry—were purely the result of mechanical intuition
and tinkering rather than any profound understanding of the underlying
physics or biology. It was possible to discover new products and techniques
that worked better when their underlying natural rules were not known, but
without such knowledge it was much harder to debug, improve, and adapt
a technique and to keep raising productivity. Hence most inventions were
one-off events followed by slow, artisan-driven improvements that eventu-
ally petered out. At times such incremental improvements substantially in-
creased productivity, as was the case in the British watch industry in the
18th century (Kelly and Ó Gráda 2016). In other cases, a minimum epis-
temic base was required before a technological advance could be realized, as
was the case with electricity and nitrogen-fixing.
The problem with premodern technology, then, was that even the very
best natural philosophers (as scientists were known then) did not know
enough to give farmers, brewers, shipbuilders, blacksmiths, physicians, and
all other fabricants much sound advice on why their techniques worked,
and hence they could not tell them how to make the techniques work better,
nor could they suggest to would-be inventors what would work and what
would not. As a result, even when a technique clearly worked, it was often
used inappropriately. The great transition of the Industrial Revolution was
not the growth of invention as such but its changing nature. What made it
“modern” is the gradual widening epistemic base of technology.9 Some of it
came from the autonomous growth of science, often inspired by practical
problems. After many centuries of experience-based steelmaking, in 1786
three French followers of Lavoisier wrote a pioneering paper proposing the
chemical composition of steel.10 But its quaint terminology may have made
it “incomprehensible except to those who already knew how to make steel”
(Harris 1998, 220). It took decades for their findings to actually be of use to
steelmakers (Wertime 1961). Water power had been known since antiquity

9. Lin (1995) points out that the difference between Chinese and modern Western technology
was that both Chinese and pre-1700 Western technological progress were experience-based
and advanced through learning by doing and trial and error. The growth of what he calls
knowledge-based technological change changed the historical balance in favor of the West.
This was already realized in the 18th century. Condorcet, in his famous 1795 essay on human
progress, argued that although “isolated nations long influenced by despotism and supersti-
tion” (he means in all likelihood Asia) had been the origin of many important new technolo-
gies in textiles, ceramics, and metal, they were being surpassed by Europe in his age because
there is nothing these countries did that “announces the presence of genius—all improve-
ments there appear as the slow and painstaking work of a long routine” (Condorcet 1796, 51).
10. The famous paper was by Vandermonde, Berthollet, and Monge (1786).

16 SUSTAINING ECONOMIC GROWTH IN ASIA

yet advanced painstakingly slowly, through trial and error until about 1700.
Its accelerated improvement was due to a combination of better experi-
mental techniques and the systematic use of formal science in its analysis
(Reynolds 1983, 204–65; Wootton 2015, 486–89). Gaslighting, one of the
paradigmatic inventions of the Industrial Revolution, was supported by
the development of the concepts, materials, experience, and apparatus of
pneumatic chemistry (Tomory 2012, 13–36). The same was true of fertilizer.
Farmers had been fertilizing fields for millennia, but nobody knew why and
how fertilizers made crops grow more abundant. Once that became known
through the emergence of organic chemistry in Germany in the 1830s and
1840s, farmers could apply chemical manures as needed for specific crops.11
By no means does this suggest that the modus operandi of a technique
was fully understood when it was first developed—but in many cases some
minimum had to be in place. Once operational, however, the technique
stimulated scientists to develop the underlying science and widen the epis-
temic base endogenously to allow for further improvements. The classic
example remains the steam engine, the invention of which in the early
18th century motivated scientists to develop the theory of thermodynam-
ics (i.e., the physics of steam power) over a century later. The old gag that
science owes more to the steam engine than the other way around, however,
remains one of those oft-repeated half-truths that confuse as much as they
enlighten. Some formal scientific understanding mattered in the invention
of the steam engine. Torricelli’s discovery of atmospheric pressure in 1643
spawned the first atmospheric steam engines in the early 18th century.
Then in 1765 James Watt revolutionized the steam engine by adding a
separate condenser. Long before the birth of thermodynamics, Watt had
learned from the Scottish chemist William Cullen that in a vacuum water
would boil at much lower, even tepid, temperatures, releasing steam that
would ruin the vacuum in a cylinder. That piece of knowledge was essential
to Watt’s realization that he needed a separate condenser (Hills 1989, 53).12

11. Another example is hydrostatics needed for ship design, with which 18th century scien-
tists struggled mightily. Isaac Newton’s theory on hydrostatics, known later as impact
theory, was embraced by many contemporaries, but it turned out to be unsatisfactory and
was corrected by Leonhard Euler and the Bernoullis, who designed a theory that could deal
with the physical state variables in the whole domain of fluid. See Nowacki (2008).
12. Another paradigmatic example of how practical invention stimulated the underlying
science rather than the reverse can be found in the evolution of aerodynamics. The formal
theory of aerodynamics was laid out in 1918 by Ludwig Prandtl, 15 years after Kitty Hawk
(Constant 1980, 105; Vincenti 1990, 120–25). Even then, the ancient method of trial and
error was still widely used in airplane design. The search for the best use of flush riveting
in holding together the body of the plane or the best way to design landing gear remained
highly experimental (Vincenti 1990, 170–99; Vincenti 2000).

SECULAR STAGNATION: HISTORY AND REALITY 17

 When did the connection between propositional knowledge (knowl-
edge of “what”) and prescriptive knowledge (knowledge of “how”) begin to
get more powerful? It clearly did not erupt suddenly in 1500 and medieval
natural philosophers and inventors, as already noted, were far from the
benighted ignoramuses that subsequent writers have tried to make them
seem (Hannam 2011). Yet the notion that useful knowledge was primarily
meant, as Francis Bacon famously wrote, to be “a rich storehouse for the
glory of the creator and the relief of man’s estate” fully took hold only in the
16th and 17th centuries. To be sure, much of natural philosophy remained
cloaked in mysticism and the occult, but by the late 17th century more
and more scientists emerged following the model of John T. Desaguliers
in England and Antoine Parent in France, who applied their knowledge
of physics and mathematics to solving practical problems in mining and
watermills.
The Great Enrichment could sustain itself because between 1700 and
1850 the epistemic base of many areas in technology widened continuously,
in part through cumulative progress in science and in part as a response to
challenges and stimuli from the practical world. In the 19th century we
see an ever-increasing input of formal knowledge in the origins and—even
more so—the subsequent development of new techniques.
Where a social and informational chasm separated those who knew
things and those who made things, it was much harder to mutually reinforce
artisanal and formal knowledge and a stationary state of little or no growth
would ensue. Scholars in the Greek ancient world were still committed to
banausia, a sense of contempt for practical knowledge, a contempt that the
Romans shed to a great extent, at least for a while, as exemplified by the
practical work of writers such as Cato the Elder, Vitruvius, and Varro. In this
(late Republican) period, Macmullen (2017, 15) notes, “Mediterranean civi-
lization could be thus raised to new levels of tangible comfort, health, and
handsomeness upon the basis of knowledge necessarily shared between the
ruling classes and labor, with master craftsmen to bridge the gap between
the two.” Unfortunately, this cultural trait did not last, and Rome did not
generate a sustainable process of technological progress. Similarly, in China
the gap was becoming more formidable during the rule of the Ming and
Qing dynasties. For example, Needham pointed to the fact that the real
work in Chinese engineering was “always done by illiterate or semi-literate
artisans and master craftsmen who could never rise across that sharp gap
which separated them from the ‘white collar literati’” (Needham 1969, 27).
In Enlightenment Europe this chasm closed considerably in many areas.
There was a widely shared conviction that artisans and scientists could
benefit from each other, and in much of Western Europe places and forums
where natural philosophers and mathematicians could interact with engi-

18 SUSTAINING ECONOMIC GROWTH IN ASIA

neers and industrialists came into being (Stewart 1992, Mokyr 2009). Zilsel
(1942) pointed to artisans and craftsmen as a major source of information
for philosophers and the explicit acknowledgment of their role by many of
the leading scientists of the age.13 In the two centuries before 1700, the social
distance between “pure” natural philosophy and its application slowly but
irresistibly shrank in Western Europe.

Secular Stagnation?
If the three factors described above accounted for a substantial part of
secular stagnation in the past, they provide us with a way to reassess the
likelihood of secular stagnation as a normal condition in the foreseeable
future. As far as the specter of a Malthusian curse is concerned, the notion
that the human race is facing some risk of overpopulation (still popular-
ized by Paul Ehrlich’s Population Bomb, first published in 1968, followed
unrepentantly in 1990 by his Population Explosion) has been thoroughly
discredited. Even without the huge increase in effective and available
natural resources due to technological progress, this doomster prediction
is doomed itself by the astonishing slowdown in fertility rates worldwide.
Fertility rates in developing economies in Asia and Latin America are mostly
hovering around replacement. Fertility is declining in the most popu-
lated economies in Asia and is approaching or has fallen below replace-
ment levels. In Indonesia and Bangladesh, for instance, fertility rates today
are down to 2.11 and 2.17 births per woman, from 5.7 and 6.7 in 1960,
respectively (CIA 2017, World Bank 2017), while fertility in both China
and Iran is now significantly below replacement levels. The region where
Malthusian pressures may remain a serious limitation to growth is Africa.
African fertility rates have remained stubbornly high, and the resource base
in most of the continent is being eroded by overuse and mismanagement.
For the rest of the world, the demographic transition, whatever its exact
microeconomic foundations, is steering toward zero population growth
now predicted for some point in the 21st century.
Indeed, it is no small irony that if demography is any concern today,
it is that zero or negative population growth, coupled with the continuing
rise in the average age in most economies, is seen as a source of secular stag-
nation (Hansen 1939). Globally, the old-age dependency ratio is expected to
rise from 13 percent in 2015 to 38 percent by the end of the 21st century.14

13. For a more recent restatement of Zilsel’s ideas, see, for instance, Roberts and Schaffer
(2007).
14. “The New Old,” Economist, July 8, 2017, 3. The median age in the United States has
increased from 28.1 in 1970 to 37.6 in 2015 (Statista 2017) and is projected to rise to 41 by

SECULAR STAGNATION: HISTORY AND REALITY 19

But the notion that somehow an aging population will lead to declining
aggregate demand seems unpersuasive. For one thing, assuming retire-
ment rates do not change much, a rising dependency rate means that the
growth of aggregate demand will exceed that of aggregate supply, as rising
numbers of nonworking persons will still consume. Second, the notion
that somehow slow or zero population growth slows down investment
seems weak. It will more likely change the composition of investment, for
example, toward medical care, tourism, and home services for the elderly.
New technologies can and will substitute capital for labor in a range of such
activities. Moreover, faced with the inevitable crushing expenses of medical
care and pensions, governments will have to increase spending, offsetting
any shortfall in aggregate demand. The negative effects of aging on GDP
growth would be felt more if most of the population over 65 retired from
the labor force. Such a decline would count as “stagnation” only because
we do not count the leisure of the golden age as output.
However, the 65 years cutoff is a relic of the 19th century, when life
expectancy was much shorter. Increased labor force participation by those
in the 65–80 age bracket may hold one of the keys to future growth just
as increased female participation did in the 1960s and 1970s. Modern
medical technology has extended the number of years the post-65 cohort
can remain productive, and future technological progress suggests that
this will continue to improve.15 The more those in the 65–80 bracket can
be deployed productively, the more concerns about the economic effects
of aging will be mitigated. More than anything else, however, the concern
is with declining GNP growth, which fails to capture the huge gain in
economic welfare accruing to people who do not wish to work beyond
retirement, because of better access to entertainment, better medical tech-
nology, greater independence, and (hopefully) some measure of economic

2050. This is relatively moderate compared with projected rises in median age in other indus-
trialized countries: In Germany it will rise from 38 in 2010 to 51 in 2050, in China from 35 to
46, and in South Korea from 38 to 53 (Pew Research Center 2014, chapter 2).
15. Research suggests that in repetitive work, productivity declines with age, but in knowl-
edge-based jobs age makes no difference in performance and in jobs that require “social
skills” productivity rises with age (“The New Old,” Economist, July 8, 2017, 6). As automa-
tion and the increased use of artificial intelligence replace repetitive work but not the kind
of work that requires knowledge or social skills, labor market bias against workers over 50
may be doomed. Elderly workers will be increasingly able to participate in the “gig economy”
by, for instance, driving Uber cars, letting rooms through Airbnb, or providing babysitting
services through such sites as sitters.com. Technology also allows the elderly to cope with
many of the handicaps of old age, including on-demand services and smart appliances. As
the generation that is comfortable with smart phones and computers enters their 60s and
70s, such technologies will become easier to implement.

20 SUSTAINING ECONOMIC GROWTH IN ASIA

security if inflation can be kept under control and old-age insurance
markets adapt.
If there is any substance to Malthusian-driven stagnation for our
near future, it might be that the finite resource is not land or even natural
resources, but the overall limited capability of the planet to carry a richer
population rather than a bigger one. Even if population growth is no longer
a factor, it may be argued that rising income per capita on a global scale
is self-limiting because it exhausts the finite resources that the world can
produce—which include maintaining a stationary temperature distribution.
While some extreme environmentalist groups may be opposed to economic
growth in any form, their resistance is not based on any economic prin-
ciple and will remain a fringe movement. After all, growth can be natural-
resource-using (as it was in the first centuries of the Industrial Revolution
when fossil fuel and iron consumption increased by a huge factor), but it
can also be resource-saving, if it takes the form of renewable energy invest-
ment, public transportation, recycling of materials, and such. In other
words, resource considerations may still steer the growth vehicle but are
unlikely to become the brakes. Whether or not such resource-augmenting
technological progress will actually happen remains to be seen; in terms of
both demography and technological change, the times we live in have no
parallel in the past, and so history may be a poor guide to the future.
Politics and growth-hostile institutional change are more of a concern.
It should be obvious that in the present world order, a war of predation
against small but wealthy nations is rather unlikely even if they have
powerful but poor neighbors. Singapore is at first glance uncomfort-
ably hemmed in between two larger and poorer neighbors, Indonesia and
Malaysia. Its income per capita exceeds that of Malaysia by a factor of 3.2
and that of Indonesia by a factor of 7.4, yet its population is tiny compared
with either. All the same, the likelihood of a predatory war against it is very
low. Indeed, the last time such a predatory war took place was Saddam
Hussein’s invasion of Kuwait in August 1990, which ended disastrously for
him. The international order, through formal multinational institutions
such as the North Atlantic Treaty Organization (NATO) and the United
Nations (UN) as well as informal institutions such as US-led coalitions,
will not tolerate such predatory attacks, and potential predators know
this full well. Rather than full-scale roving-bandit raids, however, there is a
danger that medium-sized but poor nations could engage in the future in
nuclear blackmail.
The issue of internal rent-seeking is far more complex, and, as already
noted, stagnation driven by paralyzing levels of corruption and ineffective
governance can indeed threaten future growth in many countries. One of
the hardest questions to predict is whether modern countries that have

SECULAR STAGNATION: HISTORY AND REALITY 21

been able to confine cronyism, nepotism, and rent-seeking to tolerable
levels will become more corrupt over time, or whether corrupt countries
will slowly engage in institutional reform and eventually reduce corruption
to manageable levels. When highly corrupt nations interact closely with
nations with better institutions, or join supranational organizations with
them, one would hope that some reforms would take place. Romania and
Bulgaria joined the European Union in 2007; since then Romania’s gover-
nance indicators have improved somewhat but Bulgaria’s not as much.16
Data for Greece, which joined in 1981, do not go all the way back to the
time of joining, but the trend of its scores since 1996 is quite discouraging.
In China, the campaign against corruption in the past decade seems to
have led to major improvements in the three most crucial governance indi-
cators: in “control of corruption” China moved from the 37th percentile
in 2006 to the 49th percentile in 2016, in “government effectiveness” from
the 57th to the 68th percentile, and in “rule of law” from 31st to 46th
percentile. Greater integration with the world economy may thus have had
salutary effects on institutional quality—yet this has not turned China into
a more democratic and open polity. The score on the “voice and account-
ability” indicator remains very low and declined from the 8th percentile in
2005 to the 5th in 2015.
Whether and how the rise of authoritarian regimes in the 21st century
will actually affect governance quality is far from clear. In some cases a
highly authoritarian regime can turn countries away from corruption if the
autocrat—one thinks above all of Singapore’s Lee Kuan Yew—has himself or
herself high ethical standards. There is no easy way to predict whether the
new autocratic regimes established in some countries will be more like Lee
or more like the egregiously corrupt and erratic Nicolae Ceauşescu. Clearly
location is correlated with the kind of cultural foundations that help deter-
mine the quality of governance. The three Baltic nations, all former Soviet
republics, have European-level scores on their governance indicators, while
the Central Asian republics score abysmally on almost all of them.17
Yet history suggests some room for optimism. Countries can change
their political structure fairly quickly if there is a profound cultural-ideo-
logical change such as the European Enlightenment of the 18th century.

16. World Bank, Worldwide Governance Indicators, http://info.worldbank.org/governance/

wgi/index.aspx#reports.
17. At the same time, idiosyncratic and contingent factors can override both geography and
history. Consider the sharp contrast between neighboring Poland and Belarus: In 2010,
Belarus scored in the 27.14 and 15.17 percentiles on “control of corruption” and “rule of
law,” while Poland scored 70.00 and 68.25, respectively. By 2015, the gap between the two
countries had declined somewhat.

22 SUSTAINING ECONOMIC GROWTH IN ASIA

As already noted, “old corruption” declined in Britain after 1750, under the
leadership of William Pitt the Younger and Edward Burke (Mokyr 2009,
424–27). By 1850, rent-seeking in Britain had been reduced to a minimum.
The same occurred in Prussia, the Low Countries, Scandinavia, and
Switzerland. There can be little question that the countries affected most
by the Enlightenment and its spread after the French Revolution (and their
overseas offshoots) are still the ones that tend to score the best on the World
Bank and Transparency International corruption indices in our time.
The one concern that may arise is that bad institutions will support
resistance to innovation, as incumbents will try to block innovations using
a variety of political and rhetorical devices, from concerns about employ-
ment opportunities being eroded because of automation to imaginary
fears about “frankenfoods.” Such fears are a real concern (Mokyr 1998,
2008; Juma 2016). A good case in point is transgenic crops. As Juma (2016)
notes, the use of such crops not only increases agricultural productivity
but is also environmentally responsible, as it reduces the use of pesticides
and fertilizers. Yet opposition to transgenic crops led to the formulation of
the infamous Cartagena Protocol on Biosafety, adopted in 2000. The main
thrust of this emblematic manifestation of resistance to such crops, known
as the “precautionary principle,” was to reverse the burden of proof: The
originator of the biological innovation had to show it was harmless before
it could be marketed—basically an impossible task. As Juma stresses, those
who are averse to loss or risk make the logical error of assuming that the
status quo is risk-free. Yet it remains to be seen if such resistance will be
more effective now than it has been in the past. International competition
will make resistance costly, since nations that refuse to adopt the best tech-
niques clearly will be outcompeted by others. The fear of being left behind
in the technological race may outweigh the concern about possible adverse
effects of new techniques.
All in all, in our age bad governance and rent-seeking are unlikely to
stifle long-term economic growth worldwide. This is not to say that endemic
corruption and poor institutions do not impede or even stifle economic
growth in any single country or even large segments of continents. Global
competition between nations has at times induced nations to reform their
institutions, but these reforms seem especially prevalent after violent con-
flict such as the Stein-Hardenberg reforms in Prussia after 1806, the aboli-
tion of serfdom in Russia after the Crimean War, or the reintroduction of
liberal democracy in Italy and Germany after 1945. The problem, as always,
is that the decision makers in high-corruption nations are normally people
who benefit from rent-seeking and whose power depends on it; they are un-
likely to voluntarily support reforms unless they feel they have to. Yet the
overthrow of such venal dictators as Ceauşescu, Yanukovych, and Mubarak

SECULAR STAGNATION: HISTORY AND REALITY 23

shows that corrupt regimes can be deposed even if they are not invariably
replaced by better regimes.
Finally, our knowledge of nature in the past century provides us with
far better ways of developing and improving technologies. Modern science
has penetrated to the atomic and subatomic levels as well as to remote
outer space with ever more powerful tools. That is not to say that we fully
“understand” why and how nature operates or that we ever will, only that we
have a much better grip on regularities and patterns that can be exploited
and harnessed to our needs. The basic reason is quite simple: We have far
better tools and techniques to observe and analyze natural phenomena and
patterns, and those who need to know have far better access to the rapidly
expanding body of useful knowledge.
To put it differently, against the widespread concern that innovation is
getting increasingly difficult because the “low-hanging fruits” have already
been picked in the past century, it can be argued that science provides us
with taller and taller ladders. But that is not the entire story. Technology
and science coevolve in a powerful positive feedback relationship. Science
advances when it has better tools. Seventeenth century science in Europe
blossomed in large part because it produced new instruments, most
famously the telescope, the microscope, and the vacuum pump. In the 20th
century, one of the unsung heroes of science was x-ray crystallography, first
proposed by the German theoretical physicist Max von Laue (1879–1960)
and realized by William Henry Bragg (1862–1942) with his son, William
Lawrence Bragg (1890–1971). The technique has been instrumental in
discovering the structure and function of many biological molecules,
including vitamins, drugs, and proteins. Its most famous application was
Rosalind Franklin’s work in 1953, which led to the discovery of the struc-
ture of the DNA molecule, but its use has been instrumental in 29 Nobel
Prize–winning projects (International Union of Crystallography 2017).
In the past 50 years the toolkits that scientists and researchers in any
area have at their disposal have improved and expanded at an undreamed-
of rate. Old tools like microscopes and telescopes have been improved
beyond imagination, to the point where the developers of the Betzig-Hell
super-resolved fluorescent microscope—Eric Betzig, William Moerner, and
Stefan Hell—were awarded the Nobel Prize for chemistry. The telescope,
similarly, underwent such radical improvements that the much-touted
James Webb Space Telescope, to be launched in 2021, will be able to study
the origins of galaxies and search for planetary systems in the universe in
ways unimaginable before. Every laboratory in the world today uses equip-
ment that is vastly superior to what was available in the late 20th century.
Research in a plethora of fields was enriched in the past half-century by two
“general-purpose scientific instruments,” computers and lasers. The impact

24 SUSTAINING ECONOMIC GROWTH IN ASIA

of computers on science has gone much beyond simple calculations and
statistical analysis: A new era of data science has arrived, in which models are
replaced by colossal mega-data-crunching machines, which detect patterns
that the human mind could not have dreamed up or fathomed. Such deep
learning models engage in data mining in artificial neural networks. Rather
than dealing with models, such regularities and correlations are detected
by powerful computers even if they are “so twisty that the human brain
can neither recall nor predict them” (Weinberger 2017, 12). Here the slogan
might well be: Who needs causation as long as we have correlation? In some
sense, there is nothing new here: There was always an inductive method in
science, in which scientists collected data on plants, shells, and rocks and
looked for regularities without fully understanding the underlying laws.
The difference may be just in scale, but in these matters scale is everything.
Much as the much-touted James Webb is to Galileo’s first telescope, the
huge data banks of mega crunchers are to Carl Linnaeus’s notebooks.
But computers can do more than crunch data: They also simulate, and
by doing so, they can solve fiendishly complex equations and allow scien-
tists to study physiological processes and design new materials in silico and
simulate natural processes that hitherto defied human attempts. Such
simulations have spawned entirely new “computational” fields of research,
in which simulation and large data processing are strongly complementary
in areas of high complexity.
Much like computers, lasers have been used very widely in industry
and medicine, but their power as a scientific tool may be just as significant
in the long run. Among its many applications, one of the most important
is laser-induced breakdown spectroscopy (LIBS) (Thakur and Singh 2007).
It is applied in remote material assessment in nuclear power stations;
geological analysis in space exploration; diagnostics of archaeological
objects; metal diffusion in solar cells; biomedical applications to analyze
biological samples like bones, tissues, and fluids; and detection of excess
or deficiency of minerals and toxic elements in bodies. Light radar (Lidar)
is a related laser-based surveying technique that creates highly detailed
three-dimensional images used in geology, seismology, remote sensing,
and atmospheric physics. Lasers are also a mechanical tool that can ablate
(remove) materials for analysis. Among many other uses, laser interferom-
eters have been used to detect the gravitational waves Einstein postulated,
one of the holiest grails in modern physics, and awarded the Nobel Prize
in physics in 2017.
In addition to improving the tools of actually conducting research,
the modern age has revolutionized access to the existing pool of scientific
knowledge. As the body of scientific knowledge is increasing continuously,

SECULAR STAGNATION: HISTORY AND REALITY 25

the question of access looms ever larger. The reason access is important is
well understood: The only way to define “what is known” (social knowl-
edge) is as the union of all scientific knowledge known to individuals. As
scientific knowledge has expanded since 1500, an inevitable process of
specialization has taken place. Scientists know more and more about less,
and there are fewer and fewer “polymaths.”18 The “burden of knowledge” is
increasing, and the amount that scientists working at the frontier have to
learn is growing (Jones 2009). Often the relevant knowledge is quite remote
from one’s own specialization and requires access, and access requires an
access technology. Inventors need to have access to best-practice science
on the frontier to push the envelope and make sure they take technology
as far as it can possibly be taken conditional on the largest epistemic base.
Moreover, they need to know what already exists so as to avoid reinventing
an expensive wheel. Finally, many inventions recombine existing devices,
in which disparate elements are put together in new ways, in Matt Ridley’s
famous formulation, “ideas having sex.” Hence, the people working at the
frontier need to know what others know. It is no accident that the age of
Enlightenment, during which the Industrial Revolution shifted into high
gear, was also the age of the encyclopedia, which organized existing knowl-
edge and made it accessible. Many “handbooks” and “technical compendia”
were compiled by groups of experts, and they were disseminated through
cheap print editions and local libraries with the explicit purpose of making
useful knowledge accessible.19 It is thus obvious that cheap and fast access
to existing knowledge is enormously valuable in modern research and
underlines the significance of the internet to continued progress.
The race between storage-and-search techniques and the growth of
scientific content in the last few years seems so far to have been won by
the former. Anyone engaged in research can access vast banks of knowl-
edge and data. Cloud technology is just getting started. Moreover, search

18. Indeed, there are two separate books entitled The Last Man Who Knew Everything. One of
them (Findlen 2004) places the title on the Jesuit scholar Athanasius Kircher (1601–80), a
German-born polymath of prodigious scholarly productivity who wrote important books
on topics as different as natural history, mathematics, geology, and the history of ancient
Egypt. The other book (Robinson 2007) is about Thomas Young (1773–1829), a physician
who established the wave theory of light, discovered a formula for the elasticity of materials,
and helped decipher the Rosetta Stone.
19. A good example is Temple Croker’s three-volume Complete Dictionary of Arts and Sciences
(1764–66), which explicitly promised its readers that in it “the whole circle of human learning
is explained and the difficulties in the acquisition of every Art, whether liberal or mechanical,
are removed in the most easy and familiar manner.” In close to 2,000 pages, the collection
contained detailed essays on diverse topics such as architecture, botany, and hydrostatics.

26 SUSTAINING ECONOMIC GROWTH IN ASIA

and storage costs have fallen faster in the past decades than ever before.20
As Ridley (2010) has remarked, “The cross-fertilization of ideas between,
say, Asia and Europe, that once took years, decades, or centuries, can now
happen in minutes.” Needless to say, a large part of the information stored
is not in any way scientific, and even much of the scientific information
includes more chaff than wheat and requires verification and validation.
While access may become easier, validation may become increasingly more
costly as the volume of data keeps expanding. That said, the internet and
the easy and cheap access to scientific literature and public databases are
now a part of our academic infrastructure. This means that any process of
economic growth based on further expanding the epistemic base of tech-
nology will continue apace at least at the rate that we have experienced in
the past decades.
As a consequence of the exponential growth of knowledge, the nature of
scientific research is changing: Jones (2009) observes that “burden of knowl-
edge” effects make innovation more difficult and require more collabora-
tors. Training people to the point they can make important contributions
takes more time, so that “ever increasing effort may be needed to sustain
long-run growth” (p. 310). Yet Jones also notes that improved methods of
knowledge transmission may mitigate the negative effects of the growing
burden of knowledge. The emergence of virtual reality as an educational
tool may, in fact, do exactly that. He also notes that if sufficiently new tech-
nological opportunities emerge, “the output of innovators may become
sufficient, despite a rising educational burden, to sustain growth without
increasing effort.” Ever more powerful research tools will in all likelihood
raise the efficiency of capital in the production of new knowledge.
Bloom et al. (2017), in a careful empirical paper, find that productivity
in the generation of ideas has been falling across the board in a slew of
existing industries. At first glance, this finding seems worrisome, suggesting
that if we are to avoid secular stagnation and maintain a constant rate of
productivity growth, we will need to allocate more and more researchers
to the research and development (R&D) sector. Part of the problem, as
Bloom et al. recognize, is that in individual sectors diminishing returns to
research may set in but this may not be true for the economy as a whole as
new product lines are developed all the time. They deal with this by looking
at total factor productivity (TFP) growth in the economy as a whole
and find that it has actually slowed as the research effort has increased.

20. According to one estimate (Goldman 2013), in 1981 a Macintosh storage drive cost
approximately $700,000 per gigabyte. In July 2013, a Western Digital My Book external hard
drive cost $0.06 per gigabyte. The cost of flash memory, introduced in 2003, was $8,000 per
gigabyte, which fell to $0.94 in 2013 and about $0.40 in 2016.

SECULAR STAGNATION: HISTORY AND REALITY 27

Equating TFP growth based on GDP statistics with idea growth or tech-
nological progress, however, is of course hazardous. Oddly, Bloom et al.
do not note that the number of patents, a measure of output of R&D, has
been increasing economywide at a rate much faster than TFP. Whether the
authors will turn out to be correct or not, it is of course the total “output of
ideas” that matters for future economic performance, not the output per
unit of input. As Bloom et al. stress, ideas are nonrivalrous and so as long
as we produce more of them (and their quality does not decline) secular
stagnation will not become the rule. Resources allocated to R&D in the
past may have been considerably below optimal, and the increase in R&D
may have led to a decline in the marginal product of labor in R&D that was
socially optimal. It may also be true that what they observe is a race between
the growing difficulty of coming up with new ideas and the growing capa-
bility of equipment and instruments to create them.
Finally, of course, the flow of new ideas may increase even if the produc-
tivity in generating them somehow declines. As the economy gets richer and
as robotization and artificial intelligence (AI) replace even highly trained
knowledge workers, the economy can afford to keep increasing the propor-
tion of the labor force engaged in the R&D sector and thus keep the output
of new ideas from declining. What is even more tantalizing, of course, is
whether AI itself can generate some new ideas, which at some point in the
future could reverse the trends found by Bloom et al.
So far I have discussed the movement of the technological frontier or
best practice techniques. But what about the process of catching up by tech-
nological followers? European history suggests that technological followers
do quite well. Some of the smaller European economies such as the
Netherlands, Switzerland, and Ireland have been able to take advantage of
inventions made elsewhere and enjoy high standards of living. It is no secret
what makes for successful adoption: a high level of sophisticated human
capital that can absorb, operate, and exploit the techniques designed by
others (and pay the patent royalties if need be), and institutions that make
such adoption easy and smooth, including law and order, low political risk,
low levels of corruption and red tape, and openness to new ideas. These
make all the difference between a Switzerland and an Afghanistan.

Conclusion
Secular stagnation was a defining feature of most of recorded history and
has turned into sustained growth only in recent centuries. However, if the
technological frontier does not keep moving ahead, modern economies
will indeed not be able to avoid the secular stagnation that Alvin Hansen
feared. An examination of what drove the technological frontier in the past

28 SUSTAINING ECONOMIC GROWTH IN ASIA

suggests that the likelihood of reverting to a world of stasis in the foresee-
able future is not high. All the same, one could have legitimate concerns
whether open institutions that are favorable to growth can keep pace with
technological capabilities. Maintaining a high level of innovation will
imply ever-improving tools and equipment, which could change the nature
of research in ways that we cannot imagine.

References
Acemoglu, Daron, and James Robinson. 2012. Why Nations Fail: The Origins of Power, Prosperity,
and Poverty. New York: Crown.
Alfani, Guido, and Cormac Ó Gráda. 2017. Famines in Europe: An Overview. In Famine in Eu-
ropean History, ed. Guido Alfani and Cormac Ó Gráda. Cambridge: Cambridge University
Press.
Ashraf, Quamrul, and Oded Galor. 2011. Dynamics and Stagnation in the Malthusian Epoch.
American Economic Review 101 (August): 2003–41.
Béaur, Gérard, and Jean-Michel Chevet. 2017. France. In Famine in European History, ed. Guido
Alfani and Cormac Ó Gráda. Cambridge: Cambridge University Press.
Black, Dan A., Natalia Kolesnikova, Seth G. Sanders, and Lowell J. Taylor. 2013. Are Children
‘Normal’? Review of Economics and Statistics 95, no. 1: 21–33.
Bloom, Nicholas, Charles I. Jones, John Van Reenen, and Michael Webb. 2017. Are Ideas
Harder to Find? Unpublished working paper. Stanford University.
Boserup, Ester. 1965. The Conditions of Agricultural Growth. Chicago: Aldine.
Boserup, Ester. 1981. Population and Technological Change. Chicago: University of Chicago Press.
Boussingault, J. B. 1821. On the Presence of Silicia Platina and Its Presence in Steel. Repertory
of Arts, Manufactures and Agriculture, 2nd series, volume 39: 366–73.
CIA (Central Intelligence Agency). 2017. Country Comparison: Total Fertility Rate. The
World Factbook. Available at www.cia.gov/library/publications/the-world-factbook/
rankorder/2127rank.html.
Clark, Gregory. 2007. A Farewell to Alms. Princeton, NJ: Princeton University Press.
Condorcet, Marie J. A. N. C. 1796. Outlines of an Historical View on the Progress of Human Mind.
Philadelphia: Lang and Ustick.
Constant, Edward W. 1980. The Origins of the Turbojet Revolution. Baltimore: Johns Hopkins
Press.
Croker, Temple Henry. 1764–66. The Complete Dictionary of Arts and Sciences. London: Printed
for the authors.
Ekelund, Robert B., Jr., and Robert D. Tollison. 1981. Mercantilism as a Rent-Seeking Society. Col-
lege Station: Texas A&M University Press.
Findlen, Paula, ed. 2004. Athanasius Kircher: The Last Man Who Knew Everything. New York: Rout-
ledge.
Galor, Oded. 2011. Unified Growth Theory. Princeton, NJ: Princeton University Press.
Gelderblom, Oscar. 2013. Cities of Commerce: The Institutional Foundations of International Trade in
the Low Countries, 1250–1650. Princeton, NJ: Princeton University Press.
Goldman, Cooper. 2013. What’s the True Cost of Big Data? Available at www.gooddata.com/
blog/whats-true-cost-big-data.
Guthrie William P. 2003. The Later Thirty Years War: From the Battle of Wittstock to the Treaty of
Westphalia. Westport, CT: Greenwood Press.

SECULAR STAGNATION: HISTORY AND REALITY 29

Hannam, James. 2011. The Genesis of Science: How the Christian Middle Ages Launched the Scientific
Revolution. Washington: Regnery Publishing.
Hansen, Alvin H. 1939. Economic Progress and Declining Population Growth. American Eco-
nomic Review XXIX (March): 1–15.
Harris, John R. 1998. Industrial Espionage and Technology Transfer: Britain and France in the Eigh-
teenth Century. Aldershot: Ashgate.
Hills, Richard L. 1989. Power from Steam: A History of the Stationary Steam Engine. Cambridge:
Cambridge University Press.
International Union of Crystallography. 2017. Nobel Prize winners associated with crystallography.
Available at www. iucr.org/people/nobel-prize.
Jones, Benjamin F. 2009. The Burden of Knowledge and the Death of the Renaissance Man: Is
Innovation Getting Harder? Review of Economic Studies 76, no. 1: 282–317.
Juma, Calestous. 2016. Innovation and Its Enemies: Why People Resist New Technologies. New York:
Oxford University Press.
Kelly, Morgan, and Cormac Ó Gráda. 2012. The Preventive Check in Medieval and Pre-indus-
trial England. Journal of Economic History 72, no. 4: 1015–35.
Kelly, Morgan, and Cormac Ó Gráda. 2014. Living Standards and Mortality since the Middle
Ages. Economic History Review 67, no. 2: 358–81.
Kelly, Morgan, and Cormac Ó Gráda. 2016. Adam Smith, Watch Prices, and the Industrial
Revolution. Quarterly Journal of Economics 131, no. 4: 1727–52.
Kremer, Michael. 1993. Population Growth and Technological Change: One Million BC to
1990. Quarterly Journal of Economics 108, no. 4 (August): 681–716.
Lin, Justin Yifu. 1995. The Needham Puzzle: Why the Industrial Revolution Did Not Origi-
nate in China. Economic Development and Cultural Change 43, no. 2: 269–92.
Macmullen, Ramsay. 2017. The Darkening of the West: A Note. Unpublished manuscript.
Yale University.
Mokyr, Joel. 1985. Why Ireland Starved. London: Allen and Unwin.
Mokyr, Joel. 1998. The Political Economy of Technological Change: Resistance and Innova-
tion in Economic History. In Technological Revolutions in Europe, ed. Maxine Berg and Kris-
tin Bruland. Cheltenham, UK: Edward Elgar Publishers.
Mokyr, Joel. 1999 [2008]. Innovation and Its Enemies: The Economic and Political Roots of
Technological Inertia. In A Not so Dismal Science, ed. Mancur Olson and Satu Kähkönen.
Oxford: Oxford University Press. [Reprinted in Magnus Henrekson and Robin Douhan,
eds., The Political Economy of Entrepreneurship, volume 2. Cheltenham, UK: Edward Elgar.]
Mokyr, Joel. 2002. The Gifts of Athena. Princeton, NJ: Princeton University Press.
Mokyr, Joel. 2009. The Enlightened Economy. New York and London: Yale University Press.
Needham, Joseph. 1969. The Grand Titration. Toronto: University of Toronto Press.
North, Douglass C., John J. Wallis, and Barry Weingast. 2009. Violence and Social Orders: A Con-
ceptual Framework for Interpreting Recorded Human History. Cambridge: Cambridge Univer-
sity Press.
Nowacki, Horst. 2008. Leonhard Euler and the Theory of Ships. Journal of Ship Research 52, no.
4: 274–90.
Nye, John. 2007. War, Wine, and Taxes: The Political Economy of Anglo-French Trade, 1689–1900.
Princeton, NJ: Princeton University Press.
Ó Gráda, Cormac. 2015. Eating People Is Wrong, and Other Essays on Famine, Its Past, and Its Future.
Princeton, NJ: Princeton University Press.
Pew Research Center. 2014. Attitudes about Aging: A Global Perspective. Available at www.pew-
global.org/2014/01/30/attitudes-about-aging-a-global-perspective.

30 SUSTAINING ECONOMIC GROWTH IN ASIA

Reynolds, Terry S. 1983. Stronger Than a Hundred Men: A History of the Vertical Water Wheel. Balti-
more: Johns Hopkins University Press.
Ridley, Matt. 2010. When Ideas Have Sex. Available at http://designmind.frogdesign.com/
articles/and-now-the-good-news/when-ideas-have-sex.html (accessed on December 27,
2014).
Roberts, Lissa, and Simon Schaffer. 2007. Preface. In The Mindful Hand: Inquiry and Invention
from the Late Renaissance to Early Industrialization, ed. Lissa Roberts, Simon Schaffer, and
Peter Dear. Amsterdam: Royal Netherlands Academy of Arts and Sciences.
Robinson, Andrew. 2007. The Last Man Who Knew Everything: Thomas Young, the Anonymous Ge-
nius Who Proved Newton Wrong and Deciphered the Rosetta Stone, among Other Surprising Feats.
Penguin.
Simon, Julian L. 2001. The Great Breakthrough and Its Cause (edited by Timur Kuran). Ann Ar-
bor: University of Michigan Press.
Smil, Vaclav. 2001. Enriching the Earth. Cambridge, MA: MIT Press.
Statista. 2017. Median age of the resident population of the United States from 1960 to 2015.
Available at www.statista.com/statistics/241494/median-age-of-the-us-population/.
Stewart, Larry. 1992. The Rise of Public Science. Cambridge: Cambridge University Press.
Summers, Lawrence H. 2016. The Age of Secular Stagnation: What It Is and What to Do
About It. Foreign Affairs 95: 2–9.
Thakur, S. N., and J. P. Singh. 2007. Fundamentals of Laser-Induced Breakdown Spectros-
copy. In Laser-Induced Breakdown Spectroscopy, ed. J. P. Singh and S. N. Thakur. Amsterdam:
Elsevier.
Tomory, Leslie. 2012. Progressive Enlightenment: The Origins of the Gaslight Industry, 1780–1820.
Cambridge, MA: MIT Press.
Vandermonde, Alexandre-Théophile, Claude Berthollet, and Gaspard Monge. 1786. Mémoire
sur le fer considéré dans ses différens états métalliques. Paris: Académie Royale des Sciences.
Vincenti, Walter. 1990. What Engineers Know and How They Know It. Baltimore: Johns Hopkins
University Press.
Vincenti, Walter. 2000. Real-World Variation-Selection in the Evolution of Technological
Form: Historical Examples. In Technological Innovation as an Evolutionary Process, ed. John
Ziman. Cambridge: Cambridge University Press.
Voigtländer, Nico, and Hans-Joachim Voth. 2012. The Three Horsemen of Riches: Plague,
War, and Urbanization in Early Modern Europe. Review of Economic Studies 80: 774–811.
Voigtländer, Nico, and Hans-Joachim Voth. 2013. How the West “Invented” Fertility Restric-
tion. American Economic Review 103, no. 6: 2227–64.
Weinberger, David. 2017. Alien Knowledge: When Machines Justify Knowledge. Available at
https://backchannel.com/our-machines-now-have-knowledge-well-never-understand-
857a479dcc0e.
Wells, H. G. 1923. Men Like Gods. London: Cassell.
Wertime, Theodore A. 1961. The Coming of the Age of Steel. Leyden: Brill.
Wootton, David. 2015. The Invention of Science: A New History of the Scientific Revolution. London:
Allen Lane.
World Bank. 2017. Fertility Rates, Total. Available at http://data.worldbank.org/indicator/
SP.DYN.TFRT.IN.
Zilsel, Edgar. 1942. The Sociological Roots of Science. American Journal of Sociology 47, no. 4:
544–60.

SECULAR STAGNATION: HISTORY AND REALITY 31

 3
Is Asia at Risk of Growing
Old before Becoming Rich?
SERKAN ARSLANALP, JAEWOO LEE, MINSUK KIM,
UMANG RAWAT, JACQUELINE PIA ROTHFELS,
JOCHEN MARKUS SCHMITTMANN, AND
QIANQIAN ZHANG

Asia benefited significantly from demographic trends in recent decades.

Many parts of the region, particularly East Asia, reaped a “demographic
dividend,” as the number of workers grew faster than the number of depen-
dents, providing a strong tailwind for growth.
This dividend is about to end for many Asian economies. The change
will have important implications for labor markets, investment and saving
decisions, and public budgets.
This chapter examines the implications of projected demographic
changes in major Asian economies over the coming decades by looking at
the implications for growth, external balances, and financial markets in the
region.1 It presents policy options to address some of the unique challenges
arising from Asia’s demographic transition.

Serkan Arslanalp is deputy division chief, Jaewoo Lee is an advisor, Minsuk Kim, Umang Rawat, and Jochen
Markus Schmittmann are economists, and Qianqian Zhang is a research assistant at the International Mon-
etary Fund (IMF). Jacqueline Pia Rothfels is an economist at the German Ministry of Finance. All authors were
in the IMF’s Asia and Pacific Department when the chapter was written. The chapter draws on chapter 2 of the
2017 IMF Asia and Pacific Regional Economic Outlook (IMF 2017), which includes additional analysis
on Japan’s experience with adverse demographic trends in recent decades and fiscal implications of aging for
Asia, along with details on the data, methodology, and results. The views expressed in this chapter are those of the
authors and do not necessarily represent the views of the IMF, its Executive Board, or its management.
1. The chapter analyzes developments in the 13 largest Asian economies: Australia, China,
Hong Kong, India, Indonesia, Japan, Korea, Malaysia, New Zealand, the Philippines, Singa-
pore, Thailand, and Vietnam.

33
 Figure 3.1 Fertility, life expectancy, and population growth in
Asia, 1950–2050

percent years
7 90

80
6
70
5
60
4 50

3 40

30
2
20
1
10

0 0
1950 1958 1966 1974 1982 1990 1998 2006 2014 2022 2030 2038 2046

Fertility
Population growth rate
Life expectancy (right axis)

Source: IMF Staff estimates based on United Nations, World Population

Prospects: 2015 Revision (medium-fertility scenario).

Demographic Trends in Asia

Asia is undergoing a demographic transition marked by slowing population
growth and aging. The change mainly reflects declining fertility rates since
the late 1960s and to a lesser extent rising life expectancy (figure 3.1). The
population growth rate, already negative in Japan, is projected to fall to zero
for Asia by 2050. The share of the working-age population is projected to
decline over coming decades. The share of the population 65 and older will
increase rapidly, reaching almost 2.5 times the current level by 2050. East
Asia is projected to be the world’s fastest-aging region in coming decades,
with its old-age dependency ratio roughly tripling by 2050.2
The demographic outlook varies across Asia. Broadly following the
findings of the World Bank (2015), three broad groups of countries can be
distinguished (table 3.1):

2. Population projections in this chapter are based on the United Nations’ World Population
Prospects: 2015 Revision (medium-fertility scenario with unchanged net migration flows).
Projections of the various fertility scenarios do not differ much until 2030, but uncertainty
increases with the projection horizon, primarily because of different assumptions about
future fertility rates (World Bank 2015).

34 SUSTAINING ECONOMIC GROWTH IN ASIA

 Table 3.1 Demographic classification of Asia
Demographic Projected demographic
Economy characteristics in 2015 characteristics in 2030
Australia Late-dividend Late-dividend
China Post-dividend Post-dividend
Hong Kong Post-dividend Post-dividend
India Early-dividend Early-dividend
Indonesia Early-dividend Late-dividend
Japan Post-dividend Post-dividend
Korea Post-dividend Post-dividend
Malaysia Late-dividend Late-dividend
New Zealand Late-dividend Late-dividend
Philippines Early-dividend Early-dividend
Singapore Late-dividend Post-dividend
Thailand Post-dividend Post-dividend
Vietnam Late-dividend Late-dividend
Note: Post-dividend is defined as a total fertility rate 30 years prior be-
low 2.1 and a shrinking working-age population share over the next 15
years or a shrinking absolute working-age population. Late-dividend is
defined as a total fertility rate 30 years prior above 2.1 and a shrinking
working-age population share over the next 15 years. Early-dividend is
defined as an increasing working-age population share over the next
15 years.
Source: IMF Staff estimates based on United Nations, World Population
Prospects: 2015 Revision (medium-fertility scenario).

n In post-dividend economies, the size of the working-age population is

shrinking both absolutely and in terms of its share of the total popu-
lation. This group includes China, Hong Kong, Japan, Korea, and
Thailand. These economies are projected to age rapidly and reach some
of the highest old-age dependency ratios globally by 2050.3 Japan has
the world’s highest old-age dependency ratio (43 percent at the end of
2015). It will rise to 71 percent by the end of 2050. Singapore is projected
to transition to post-dividend status by 2030 (table 3.2).
n In late-dividend economies, the working-age population is declining as
a share of total population but is still growing in absolute numbers. This
group includes Malaysia and Vietnam (two moderately aging emerging
markets) as well as Australia and New Zealand (advanced economies
that experienced a demographic transition earlier than other countries

3. China began relaxing its one-child policy in 2013; since 2016 it has allowed all couples to
have two children. Demographers expect a positive but limited impact of the policy change
on fertility (Basten and Jiang 2015). The 2015 UN population projections see the fertility
rate gradually rising from 1.5 children per woman in 2010 to 1.7 by 2030 (UN 2015).

IS ASIA- CHAPTER
GRAPHICS AT RISK OF GROWING OLD
3—SUSTAINING BEFORE BECOMING
ECONOMIC ASIA 351
RICH?
GROWTH IN
2
Table 3.2 Old-age dependency ratios in Asia, 2000–50 (percent)
Country 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050

36 GRAPHICS
Japan 25.2 29.9 36.0 43.3 48.3 50.6 53.1 57.0 63.8 68.1 70.9
Hong Kong 15.3 16.5 17.2 20.6 26.5 35.1 43.7 50.1 55.6 60.4 64.6
Korea 10.2 12.7 15.0 18.0 22.2 29.4 37.6 46.1 54.4 60.7 65.8

SUSTAINING- ECONOMIC
Singapore 10.3 11.3 12.2 16.1 21.4 28.6 36.5 43.7 51.1 57.2 61.6
Thailand 9.5 11.0 12.4 14.6 18.4 23.4 29.2 35.8 42.3 48.2 52.5
New Zealand 18.0 18.1 19.6 22.9 26.3 30.2 34.9 38.0 40.6 40.5 40.7
China 9.7 10.4 11.1 13.1 17.1 20.4 25.3 32.7 39.6 43.0 46.7
Australia 18.5 19.2 20.0 22.7 25.3 28.3 31.3 32.8 34.8 35.3 37.3

CHAPTER 3—SUSTAINING
Vietnam 10.4 9.9 9.4 9.6 11.7 14.8 18.3 21.8 25.6 29.5 34.1

GROWTH IN ASIA
Malaysia 6.1 6.7 7.2 8.4 10.0 12.2 14.5 16.7 18.7 21.0 25.3
Indonesia 7.3 7.4 7.5 7.7 8.6 10.4 12.4 14.7 17.0 19.2 21.3
India 7.2 7.7 8.0 8.6 9.8 11.1 12.5 14.0 15.8 17.8 20.5
Philippines 5.5 5.8 6.7 7.2 8.0 9.1 10.3 11.5 12.5 13.5 14.5
Note: The old-age dependency ratio indicates the size of the population 65 years of age and older as a share of
the working-age population (15–64 years old).
Source: IMF Staff calculations and projections based on United Nations, World Population Prospects: 2015 Revision
(medium-fertility scenario).

ECONOMIC GROWTH IN ASIA

 in the region but maintain higher fertility rates than most East Asian
economies and have substantial immigration). Immigration has also
kept Singapore in this category, despite one of the lowest fertility rates
in the region.
n In early-dividend economies, the share of the working-age population
will rise both as a share of the total population and in absolute terms
over the next 15 years. This group includes India, Indonesia, and the
Philippines. These countries have some of the youngest populations in
the region and will see their working-age populations increase substan-
tially in coming decades. Fertility rates are projected to remain above
the replacement rate of 2.1 for India and Indonesia until 2030 and
beyond 2030 for the Philippines. Indonesia is projected to transition to
late-dividend status by 2030.

Migration is an important factor in the demographic evolution of some

Asian economies. As migrants tend to be of working age, migrant flows can
slow the demographic transition in recipient countries. Immigration has
been sizable in Australia, Hong Kong, New Zealand, and Singapore. Contin-
ued immigration in these economies is projected to substantially slow the
decline in the size of the working-age population.
Emigration of working-age people is substantial in the Philippines, but
the impact on the working-age population is small relative to the popula-
tion size. In China, Japan, Korea, and the other member countries of the
Association of Southeast Asian Nations (ASEAN), net migration is rela-
tively small.
The population of Asia is not the oldest in the world—Europe holds that
distinction—but the speed of aging in Asia is remarkable. Figure 3.2 displays
the number of years it takes for the old-age dependency ratio to increase
from 15 percent to 20 percent. This transition took 26 years in Europe
and more than 50 years in the United States. In Asia, only Australia and
New Zealand aged at similar speeds. For China, Japan, Korea, Singapore,
Thailand, and Vietnam, the same transition has taken (or will take) less
than 10 years.
The rapid speed of aging has two implications. First, countries in Asia
will have less time to adapt policies to their older population than many
advanced economies had. Second, some countries in Asia are getting old
before becoming rich; they are therefore likely to face the high fiscal costs
of aging and demographic headwinds to growth at relatively low per capita
income levels. Figure 3.3 shows per capita income at purchasing power
parity relative to the United States at the historical or projected peak of the
share of the working-age population in each country. Except for Australia
and Japan, per capita income in major Asian countries is at significantly

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 37

Figure 3.2 Number of years for the old-age dependency ratio to
increase from 15 to 20 percent

Europe
United States

New Zealand
Australia
Hong Kong
Philippines
Malaysia
India
Indonesia
Korea
Japan
Vietnam
China
Thailand
Singapore

0 10 20 30 40 50 60
number of years

Note: The old-age dependency ratio indicates the population 65 years and older as a share
of the working-age population (15–64 years). Data shown in darker shading reflect historical
data, while economies in lighter shading reflect projections.
Source: IMF Staff calculations based on United Nations, World Population Prospects: 2015
Revision (medium-fertility scenario).

lower levels than reached by mature advanced economies at the same stage
of the aging cycle. This trend underscores the need to sustain high growth
rates in these economies.
Asia’s demographic evolution has important global implications,
because of the region’s contribution to global growth, current account bal-
ances, and capital flows, as well as relative wage levels and competitiveness.
Figure 3.4 presents absolute changes in the working-age population for
different demographic country groups in Asia and the rest of the world.
Between 1970 and 2010, Asia contributed more to the growth of the global
working-age population than the rest of the world combined.
This situation is changing. Over the coming decades, rapidly aging
East Asian economies are projected to see their working-age populations
drop substantially. The largest absolute decline will be in China, where
the working-age population will fall by 170 million over the next 35 years.
Substantial absolute declines are also projected for Japan, Korea, and
Thailand. In contrast, Africa will account for most of the growth in the
global working-age population (IMF 2015a).

38 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 3.3 Per capita income level at the peak of working-age population share (based on purchasing
power parity; in percent of US per capita income at each country’s peak year)

percent

2008
100
1987 1993 2009 1987 2009 1992
80 1950
2014 2009 2056
60 2020
2040
40 2031 2013
2011
20 2014

0
a a a a
es y ly d a ce m lia an ea nd s ia a ia nd na m
t an Ita o a a e s di hi
ta m na an tr ap or al in ay In es ila C tna
S a Fr gd us J K on a ie
d er C in Ze pp al d Th V
te G K A ili M In
ni d ew Ph
U te N
ni
U
a. Based on IMF Staff projection. For Malaysia, the income level relative to the United States is calculated from the April 2017
World Economic Outlook projection for 2020. For India, Indonesia, and the Philippines, the income levels are calculated by
applying the projected purchasing power parity per capita income growth rate in 2022, starting from 2023 and up to the year
in which the working-age population share is projected to peak.
Note: For the countries shown in the figure, the working-age population (15–64 years) share of the total population has peaked, or
is projected to reach the peak, in the year indicated in parentheses.
Sources: IMF, World Economic Outlook database; IMF Staff calculations based on United Nations, World Population Prospects:

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH?

2015 Revision (medium-fertility scenario).

39
 Figure 3.4 Change in working-age population in Asia and the
rest of the world

millions
800
700
600
500
400
300
200
100
0
–100
–200
–300
1970–90 1990–2010 2010–30 2030–50

Early-dividend
Late-dividend
Post-dividend
Rest of the world

Note: The working-age population is defined as those aged 15 to 64. Early-

dividend economies include India, Indonesia, and the Philippines. Late-dividend
economies include Australia, Malaysia, New Zealand, Singapore, and Vietnam.
Post-dividend economies include China, Hong Kong, Japan, Korea, and Thailand.
Source: IMF Staff calculations and projections based on United Nations, World
Population Prospects: 2015 Revision (medium-fertility scenario).

Implications of Demographic Trends

Implications for Growth
Asia has enjoyed a substantial demographic dividend in past decades.
Rapid aging is now set to create a demographic “tax” on growth in several
economies. To quantify this effect, this section uses a template devised by
Amaglobeli and Shi (2016).

Impact of the Labor Force on Economic Growth

Demographic developments affect growth through various channels, in-
cluding the size of the labor force, productivity, and capital formation. The
analysis begins by establishing the direct impact on growth of demography-
induced changes in labor force size in a growth accounting framework. This
baseline impact rests on several assumptions:
n Total factor productivity (TFP) growth remains unchanged (based on
the historical average).

40 SUSTAINING ECONOMIC GROWTH IN ASIA

n Age- and gender-specific labor force participation rates (and employ-
ment rates) remain unchanged.
n The capital-to-effective-labor ratio remains constant.4

Long-term output is estimated using a production function approach

with capital and labor as inputs. The function is defined in logarithmic
terms as follows:

TFPt β Kt β ∑ (3.1)

where t is time; Y is real output; TFP is total factor productivity; K is the

stock of capital; β is the share of labor in output; j is the age-gender cohort;
N is the number of individuals in each cohort; LFP and E denote cohort-
specific labor force participation and employment rates, respectively; and
w is the weight factor used to adjust for the difference between the number
of employees and the effective units of labor supplied.
Population projections affect output in this framework through ag-
gregate labor and capital. To establish the baseline impact of demographic
change, we compare estimated output based on the United Nations’ medi-
um-fertility scenario (which includes migration) with a hypothetical status
quo scenario that assumes constant population size and age structure.
Separately, we also consider the UN zero-migration scenario, to assess the
impact of migration.
Figure 3.5, panel a, shows the average annual growth impact from
2020 to 2050 relative to the status quo. It reveals the following:
n Demographic trends will create strong headwinds for post-dividend
countries. In Japan the impact of aging could reduce the average annual
growth rate by almost 1 percentage point. The growth impact for China,
Hong Kong, Korea, and Thailand could be 0.5–0.75 percentage point.5
For Singapore, which transitions from late- to post-dividend status by
2030, the estimated overall impact is almost zero.
n Early- and late-dividend countries could still enjoy a substantial annual
demographic dividend, ranging from 0.5 percentage points for New
Zealand to almost 1.5 percentage points for the Philippines. Reaping

4. This assumption means that investment adjusts over time to the labor force, where labor
is expressed in efficiency units (i.e., incorporating TFP). If, for example, the capital stock
is 300 percent of GDP and the effective labor force growth declines by 1 percentage point,
the investment ratio will fall by 3 percentage points of GDP. As some substitution between
capital and labor is likely, this assumption creates an upper bound on the growth impact.
5. The drag on growth is broadly stable for Japan over the next three decades, near 1.0
percentage point in each decade. In contrast, the drag for China rises over time, from 0.4
percentage point in the first decade to about 1.1 percentage points in the last decade.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 41

 Figure 3.5 Baseline growth impact of demographic
trends in Asia
a. Impact on real GDP growth
percentage points; average, 2020–50
2.0

1.5

1.0

0.5

–0.5

–1.0

–1.5

–2.0
al s
a
do ia

us ia
N Vie alia
Ze am

H gap d
g re

ai g
K d
C a
na
l i an

ct
M ine

si

e
n

Th on

n
In Ind

A es

pa
on o

or

hi
ba Jap
Si ala

la
ay

n
tr

K
n
pp

ew t

m
ili

n
Ph

lo
G

b. Impact on real GDP per capita growth

percentage points; average, 2020–50
0.8

0.4

–0.4

–0.8

–1.2 With migration

Without migration
–1.6
es
do ia

al ia
N Vie sia
Ze am

us d

ai ia
nd

ap n
C e
H Ko a

l i ng

ct
a
n
G ong re
or
ng a
A an
In Ind

M es

Th ral

pa
in

Si ap

hi
la

ba Ko
ay

n
al
n
pp

t
ew t

m
J
ili
Ph

lo

Note: The baseline estimates are based on the assumptions of unchanged labor
force participation by age-gender cohort, constant capital-to-labor ratio, and total
factor productivity growth unchanged from historical average. Migration projections
follow historical trends. Global impact indicates the purchasing power parity–
weighted average as a percent of global GDP.
Sources: IMF Staff projections based on Amaglobeli and Shi (2016); United Nations,
World Population Prospects: 2015 Revision (medium-fertility scenario); and Penn
World Tables 9.0.

42 SUSTAINING ECONOMIC GROWTH IN ASIA

 the demographic dividend is not automatic, however. It depends on
good policies to raise productivity and create a sufficient number of
good quality jobs for the growing working-age population, as Bloom,
Canning, and Fink (2010) show.6
n Inward migration can prolong the demographic dividend or soften the
impact of rapid aging. In Australia, Hong Kong, New Zealand, and Sin-
gapore, the impact of continued immigration on the workforce could
add 0.5–1 percentage point to average annual growth.7 The impact of
net emigration from Indonesia, the Philippines, and Vietnam is small,
because of the small size of emigration as a share of population in these
countries.
n Migration can reduce but cannot reverse the negative impact of aging
on growth. Australia, for example, would need to receive immigration
equal to approximately 23 percent of the actual workforce to maintain
the same dependency ratio by 2030. For Singapore the figure would be
51 percent.

Figure 3.5, panel b, shows the growth impact on a per capita basis. The
country ordering changes slightly. The drag from demographics is smaller
for Japan but larger for Hong Kong and Singapore, because the positive
impact of immigration is partially eliminated.
We next relax several assumptions in this stylized exercise—in partic-
ular the assumptions of unchanged TFP growth and labor force partici-
pation rates—and discuss why we keep the capital-to-effective-labor ratio
assumption.

Impact of Aging on Total Factor Productivity

The first assumption in the baseline estimates is unchanged TFP growth.
But age groups differ in their productivity, because of factors such as accu-
mulation of experience, depreciation of knowledge, and age-related differ-
ences in physical and mental capabilities.

6. Long-term demographic projections are uncertain. Compared with the United Nations’
medium-fertility scenario, average annual growth is about 0.2 percentage point higher in
the high-fertility scenario and about 0.2 percentage point lower in the low-fertility scenario.
7. This effect is driven only by an increase in the size of the workforce. In addition, Jaumotte,
Koloskova, and Saxena (2016) estimate that a 1 percentage point increase in the share of
migrants in the working-age population can raise GDP per capita over the long term by up to
2 percent, by increasing labor productivity and, to a lesser extent, boosting investment. This
second-round effect is not shown in panel a of figure 3.5. The long-term UN assumptions on
net migration rates for these countries range from 2.5 percent in New Zealand to 6 percent
in Australia.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 43

 Figure 3.6 Share of older workers in working-age
population in Asia

population aged 55–64 in percent of working-age population

30
2020
25
2050
20

15

10

0
s

do ia

a
am

us a

Th lia

Ze d

ng ea

on ore

g
e

di

si

pa
n

an

on
s
in

or
hi

la
ay

ne
In

tn

ap
tr

al

Ja

K
pp

ai

K
al

ie

g
ili

A
In
Ph

Si
ew

H
N

Note: The working-age population is defined as those aged 15 to 64.

Sources: IMF Staff estimates based on United Nations, World Population
Prospects: 2015 Revision (medium-fertility scenario); and Penn World
Tables 9.0.

Several studies find evidence of a decline in worker productivity and

innovation starting between ages 50 and 60 (Feyrer 2007; Aiyar, Ebeke,
and Shao 2016; and Börsch-Supan and Weiss 2016). In contrast, Acemoglu
and Restrepo (2017) find no robust negative impact of aging on produc-
tivity, based on cross-country regressions linking aging to GDP per capita
growth. They argue that their result may reflect the more rapid adoption of
automation technologies in countries that are aging more rapidly.
Figure 3.6 shows that in most Asian countries, the share of older
workers (55–65) in the workforce is projected to increase substantially by
2050. The largest increases are projected for China, Malaysia, and Vietnam.
The impact of aging may differ across professions. Veen (2008) argues
that productivity of workers in physically demanding professions (construc-
tion, factory work) declines at older ages, whereas productivity in other
professions (law, management, medicine) may increase with age.
Figure 3.7 applies Veen’s taxonomy to selected Asian economies. Coun-
tries with lower per capita income, such as Thailand and Vietnam, tend to
have larger shares of their workforce in professions in which productivity
tends to decline with age. This finding underscores the importance of struc-
tural transformation to prepare for an aging workforce.
We estimate the effect of workforce aging (measured by the share of
workers 55–65 in the workforce) on productivity following the approach in

44 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 3.7 Share of workforce whose productivity rises or
falls with aging

EU-28
Australia
Singapore
Japan
Korea
Hong Kong
Malaysia
Thailand
Vietnam

0 20 40 60 80 100
percent of total workforce, latest available
Increases
Neutral
Decreases
Not classified

Note: Category productivity “increases” with age includes managers and

professionals; Category “neutral” includes clerical support workers and services
and sales workers; Category “decreases” includes technicians, skilled agricultural,
forestry and fishery workers, craft and related trades workers, plant and machine
operators and assemblers, elementary occupations and armed forces occupations.
Sources: Veen (2008); International Labor Organization;
J IMF Staff calculations.
log
 Yt  log TFPt   1     log  K t     log  N t j  LFPt j  Et j  wtj (3.1)
j 1
Aiyar, Ebeke, and Shao (2016) and Adler et al. (2017).8 The baseline model
fits the growth in real output per worker on the share of workers 55 and
older and the combined youth and old-age dependency ratios, with decade
(10 years) and country fixed effects. The model takes the following form:

∆𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙�� � �� 𝑤𝑤𝑤𝑤�� � �� 𝐷𝐷𝐷𝐷�� � �� � η� � ��� (3.2) (3.2)

where i indicates the country, t the decade, YW real output per worker,
w55 the share of the workforce aged 55–64, DR the dependency ratio, ui
the country fixed effect, ht the decade fixed effect, and eit the error term.
Correcting for various econometric pitfalls—such as reverse causality—the
𝑟𝑟�� � 𝛽𝛽� � 𝛽𝛽� 𝑌𝑌𝑌𝑌�� � 𝛽𝛽� �𝑌𝑌𝑌𝑌�� ∗ 𝐶𝐶𝐶𝐶�� � � 𝛽𝛽� 𝑂𝑂𝑂𝑂�� � 𝛽𝛽� �𝑂𝑂𝑂𝑂�� ∗ 𝐶𝐶𝐶𝐶�� � � 𝛽𝛽� 𝐴𝐴𝐴𝐴�� �
approach measures the impact of workforce aging on output per worker.
𝛽𝛽� �𝐴𝐴𝐴𝐴�� ∗ 𝐶𝐶𝐶𝐶�� � � 𝛽𝛽� 𝑅𝑅𝑅𝑅� � 𝛾𝛾𝐶𝐶𝛾𝛾𝛾𝛾𝛾𝛾𝑟𝑟𝛾𝛾𝑙𝑙𝛾𝛾�� � �� ����
(3.3) To address the endogeneity issue, the model also instruments the workforce
share variable and the dependency ratio with lagged birth rates (of 10, 20,
30, and 40 years), as in Jaimovich and Siu (2009).

𝑟𝑟�� � �� �8.��The �� �� studies

𝑌𝑌𝑌𝑌two � �the
� 𝐶𝐶𝐶𝐶��use �� 𝑂𝑂𝑂𝑂
same�� �� � 𝐶𝐶𝐶𝐶�� �regressing
approach, �� � TFP
� �� 𝐴𝐴𝐴𝐴��either ��
𝐶𝐶𝐶𝐶��or �� growth
TFP 𝑅𝑅𝑅𝑅� � on demo-
�� � �� � ��� .
𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾graphic (3.3a)
variables and testing whether workforce aging is associated with a permanent loss in
productivity or a slowdown in productivity growth as a result of less innovation.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 45

 For a sample of Asian and European countries, the results show that
an increase in the share of older workers is associated with a significant
reduction in labor productivity growth (table 3.3). Most of the slowdown
is caused by weaker TFP growth: Workforce aging over 2020–50 is associ-
ated with lower annual TFP growth by 0.1–0.3 percentage point a year on
average for Asia. These results are quantitatively and qualitatively in line
with the findings in Aiyar, Ebeke, and Shao (2016) for Europe and Adler et
al. (2017) for the global sample.
Figure 3.8 shows the estimated impact of projected workforce aging on
growth in selected Asian economies. On average, an older workforce is esti-
mated to reduce annual growth by 0.1 percentage point, with the biggest
impact in China (0.3 percentage point). The impact is higher for countries
that are projected to experience workforce aging faster (see figure 3.6). This
impact may be a substantial drag on future TFP growth for some coun-
tries, but it is likely to be of second-order magnitude compared with the
baseline growth impact of changes in the size of the labor force.

Impact of Aging on Labor Force Participation Rates

The second assumption in the baseline estimates is constant age- and gen-
der-specific labor force participation rates (LFPRs). LFPRs change over time
and can be affected by policies. For example, increases in life expectancy
could encourage older people to stay in the workforce. In fact, the effective
retirement age has increased in most Organization for Economic Coopera-
tion and Development (OECD) countries, although the increases have been
modest compared with increases in life expectancy (Bloom, Canning, and
Fink 2010). Alternatively, the decline in fertility rates could encourage more
women to participate in the labor force (Bloom et al. 2007). In most Asian
economies, there is scope for greater female labor force participation, al-
though unleashing the full potential of female employment requires a com-
prehensive set of policies (Steinberg and Nakane 2012, Elborgh-Woytek
et al. 2013, Kinoshita and Guo 2015). In contrast, LFPRs tend to decline
for younger workers as countries develop and average years of schooling
increase.
Figure 3.9 displays the changes in LFPRs for working-age populations
in Asian economies between 1990 and 2015. It shows that LFPRs remained
remarkably stable over this period, despite shifts for age-gender subgroups.9

9. Female LFPRs increased in the region’s advanced economies but declined in China, India,
Thailand, and Vietnam; male LFPRs declined in most countries. LFPRs for workers 15–24
declined in all countries by up to a third, reflecting longer schooling. LFPRs for workers
55–64 increased in most countries, most notably Australia, New Zealand, and Singapore.

46 SUSTAINING ECONOMIC GROWTH IN ASIA

 GRAPHICS
IS ASIA
Table 3.3 Panel regression: Demographics and labor productivitya
Change in real Change in real Change in Change in TFP Change in TFP
output per capital stock human capital (model based) (from PWT 9.0)
Dependent variables worker per worker per worker per workerb per worker

- CHAPTER
Workforce share –0.612*** 0.187*** –0.0589*** –0.740*** –0.502***

AT RISK OF
aged 55–64 (–5.309) –4.144 (–2.896) (–4.714) (–5.643)
–0.122 –0.105 0.0382 –0.0549 0.279**
Dependency ratio
(–0.695) (–1.532) –1.232 (–0.229) –2.057
Observations 571 571 571 571 571

3—SUSTAINING
Number of countries 33 33 33 33 33

GROWING OLD
PWT 9.0 = Penn World Tables 9.0; TFP = total factor productivity
a. All regressions include both country
log
log 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
log 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟 and time effects.
ɑɑ 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟

ECONOMIC
log
log �
�log
log �𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
�𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟

Note: t-statistics in parentheses; *** p<0.01, ** p<0.05, * p<0.1.

𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 b. Following a Cobb-Douglas function, the model takes the form of log (real output) = ��� �ɑɑ 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
� log �𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑇𝑇
𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟

Source: IMF Staff estimates.

BEFORE BECOMING
GROWTH INRICH? 3
ASIA 47
 Figure 3.8 Impact of aging on total factor productivity

percentage point impact on real GDP growth; average over 2020–50

0.05
0
–0.05
–0.10
–0.15
–0.20
–0.25
–0.30

on alia

Ze g

d
na

a
am

do d

es

ng ea

us n

an
si

di

si

or

pa
n

ew Ko
in

or
hi

la

tr
ay

ne

al
In
tn

ap

Ja
pp
C

ai

K
al

ie

g
Th

A
ili
M

In

Ph

Si

H
N
Note: Estimated impact of workforce aging on total factor productivity
growth follows Aiyar, Ebeke, and Shao (2016) based on a sample of Asian
and European economies.
Sources: IMF Staff estimates based on United Nations, World Population
Prospects: 2015 Revision (medium-fertility scenario) and Penn World
Tables 9.0.

Figure 3.9 Labor force participation rates ages 15–64 in

1990 and 2015

percent of population between ages 15–64 years

100
90
80
70
60
50
40
30
20
10
0
a

us n
ia

Th na

Ze d

d
am
ne
di

si

si

or

pa
on

an
al
or

hi

ew ila
ay

ne
In

tn
ap

tr
pi

Ja

al
K

C
K

a
al

do

ie
lip

ng
g
M

V
on

A
i

In
Ph

Si
H

2015
1990

Source: IMF Staff calculations based on International Labor Organization

figures.

48 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 3.10 Baseline growth impact of demographic trends
and higher labor force participation

percentage point impact on real GDP growth; average over 2020–50

2.0
1.5
1.0
0.5
0
–0.5
–1.0
–1.5
es

do ia

us a
ia

Ze m

re

Th ng

nd

ea

na

n
si

si

pa
Si lan
d

al

ew tna

o
in

or

hi
o

la
ay

ne
In

ap
tr

Ja
K
pp

C
a

ai

K
al

ie

ng

g
ili

V
A

on
In
Ph

H
N

Impact of higher labor force participation

Baseline impact

Note: The rising labor force participation rates scenario is based on the
experience of Japan from 1990 to 2015.
Source: IMF (2017).

Within the region, LFPRs increased the most in Japan, the world’s oldest
country, rising by almost 6 percentage points since 1990.
What if the LFPR of other Asian economies were to rise by as much as
it did in Japan? Figure 3.10 shows the impact of such a scenario on growth.
Annual GDP growth rises 0.2–0.3 percentage point, offsetting the lower
TFP growth that results from workforce aging. Such changes in the LFPR
are unlikely to counter the baseline growth effects induced by changes in
the overall labor force, however.10

Impact of Aging on Investment

The third assumption in the baseline impact estimates is a constant capital-
to-effective-labor ratio. Workforce aging is associated with higher capital
per worker (accounting for TFP), but economically the effect is small (table
3.3). There is no statistically significant relationship between the old-age de-

10. The impact on growth for small changes in LFPRs is close to linear. A more ambitious
scenario, in which the employment gender gap is eliminated, could add 0.25 percentage point
to annual GDP growth for Japan and up to 1 percentage point for India by 2050 (Elborgh-
Woytek et al. 2013, Cuberes and Teignier 2016, Khera 2016). Gonzales et al. (2015) show that
reducing a broader measure of gender inequality in education, political empowerment, the
LFPR, and health could lower income inequality and boost growth: A 10 percentage point
reduction in the gender inequality index is associated with almost 1 percentage point higher
per capita GDP growth.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 49

pendency ratio and capital per worker. Taken together, these results suggest
that a constant capital-to-labor ratio assumption is reasonable, especially
for thinking about the next three decades, when countries would presum-
ably be on a balanced growth path.
Overall, demographic trends could reduce growth by 0.5–1 percentage
point a year in absolute and per capita terms over the next three decades in
post-dividend countries. Over the long term, these sustained reductions in
growth rates have important welfare implications: A 0.5 percentage point
reduction in annual growth would reduce GDP by about 15 percent by
2050.

Implications for External Balances

The impact of demographics on savings, investment, and hence the current
account is examined using the external balance assessment (EBA) model
(Phillips et al. 2013, IMF 2016). The impact is captured using three vari-
ables: population growth, the old-age dependency ratio,11 and aging speed
(defined as the expected change in old-age dependency in 20 years) (see
IMF 2017 for details).12
Demographic trends affect savings, investment, and the current account
balance in the following ways:
n Savings. Countries with larger shares of dependent populations gener-
ally have lower savings.13 Therefore, both higher population growth (a
proxy for higher youth dependency) and higher old-age dependency are
linked with lower savings (table 3.4). Faster aging implies a higher prob-
ability of survival and therefore a greater need for life-cycle savings.14

11. Following the EBA model, the working-age population is defined as people 30–64, which
captures the prime-age population. As 15-year-olds are not routinely in the employed popu-
lation, the prime-age population is considered a better choice for examining savings-invest-
ment relationships. Accordingly, the old-age dependency ratio indicates people 65 and over
as a share of the prime-age population.
12. In the EBA model, population growth is a proxy for the fertility rate or youth depen-
dency ratio. Aging speed is a measure of the “probability of survival” or longevity, reflecting
the prospects of population aging; old-age dependency captures the outcome of population
aging so far (IMF 2017).
13. Grigoli, Herman, and Schmidt-Hebbel (2014) provide a comprehensive survey of saving
determinants. They find that both youth and old-age dependency lower savings in both the
theoretical and the empirical literature.
14. As people expect to live longer, they are induced to save more, counterbalancing the
effects of higher old-age dependency (Li, Zhang, and Zhang 2007). The impact of aging on
saving behavior is subject to model uncertainty, depending on whether this forward-looking
element is accounted for.

50 SUSTAINING ECONOMIC GROWTH IN ASIA

 Table 3.4 Expected impact of demographic variables on
current account
Current
Variable Savings Investment account balance
Youth dependency i h i
Old-age dependency i i Ambiguous

Aging speed h i h
Source: Authors’ illustration.

n Investment. As population growth increases, the capital-to-labor ratio

falls, raising the return on capital and boosting investment.15 Rising
old-age dependency works in the opposite direction, as it leads to a
higher capital-to-labor ratio. To the extent that it reflects expectations
of a larger future population of older people and lower future aggre-
gate demand, aging speed also results in lower investment.
n Current account balance. Population growth, aging speed, and rising
old-age dependency are expected to have negative, positive, and ambig-
uous impacts, respectively, on the current account.

A large body of literature investigates the macroeconomic impact of

population aging on savings and investment. Based on a sample of Asian
countries from the early 1950s to the 1990s, Higgins and Williamson (1997)
find that countries with relatively young populations are capital import-
ers, whereas countries with relatively old populations are capital exporters.
Higgins (1998) finds similar results based on a larger sample of 100 coun-
tries. Chinn and Ito (2007); Gruber and Kamin (2007); and Legg, Prasad,
and Robinson (2007) find large and empirically significant impacts of de-
mographics on the current account, with higher old-age dependency reduc-
ing the current account balance. Eichengreen and Fifer (2002) find the de-
mographic impact on savings and investment to be roughly similar, based
on a larger sample of 90 countries spanning 1980–94.
The significance of the negative impact of youth dependency on the
current account is less robust in the empirical literature. Jaumotte and
Sodsriwiboon (2010) find that the population growth rate has a negative
effect on the current account balance. Much of this literature misses the
impact of longevity on current account balances. Research that explicitly
captures this channel includes Li, Zhang, and Zhang (2007) and Barnes,

15. The impact of population growth (or youth dependency) on investment is less certain than
on savings. Some studies (e.g., Higgins 1998) find a positive effect; others (e.g., Williamson
2001, Bosworth and Chodorow-Reich 2007) find a negative effect.

4 GRAPHICS - CHAPTER
IS ASIA AT RISK3—SUSTAINING
OF GROWING OLD BEFORE
ECONOMIC BECOMING
GROWTH RICH? 51
IN ASIA
Lawson, and Radziwill (2010), who find that increased longevity is associ-
ated with higher savings.
Empirical results based on the EBA model support our priors on the
effect of population growth and aging speed. Higher population growth
leads to lower current account balance (as in Jaumotte and Sodsriwiboon
2010), but the effect is not empirically significant. Aging speed is strongly
associated with higher current account balances, reflecting aging-related
precautionary savings (as in Li, Zhang, and Zhang 2007 and Barnes, Lawson,
and Radziwill 2010). Higher old-age dependency is positively associated
with the current account balance when aging speed is higher than the world
average.
What changes in regional current accounts does the EBA model predict
as a result of demographic transition in the coming decade?16 By 2020
Australia, Japan, and New Zealand will have higher old-age dependency
ratios than the (GDP-weighted) world average. By 2030 Hong Kong, Korea,
and Singapore will also have higher old-age dependency ratios than the
world average.17 Several economies in the region—most notably Hong Kong,
Japan, Korea, and Singapore (advanced economies) and China, Thailand,
and Vietnam (emerging markets)—will experience very rapid aging until
2030 (IMF 2017). In contrast, some advanced economies (Australia and
New Zealand) will have lower speeds of aging than the world average.
Over 2020–30, the EBA model suggests that, all else equal, demo-
graphic trends are likely to exert upward pressure on current account bal-
ances in surplus countries, such as Thailand, Korea, and Japan given the
increase in their aging speeds between 2020 and 2030 (figure 3.11). Among
deficit countries, demographic trends are likely to exert downward pressure,
particularly in New Zealand, given its falling aging speed. Overall, demo-
graphics are projected to materially increase current account balances in
only a few economies in Asia. The total impact of demographics on global
imbalances is limited.
What will be the effect on capital flows? All else equal, demographic
factors are likely to strengthen the current dynamics of capital flows. Over
2020–30, changes in the current account as a result of demographic trends
are likely to be positively correlated with 2015 current account balances

16. The rest of this section focuses on old-age dependency and aging speed as the main
drivers of the current account because changes in population growth between 2020 and
2030 are expected to be relatively small. The contribution of population growth to changes
in current account is less than 0.1 percent of GDP for the sample period.
17. Demographic variables are expressed relative to the (GDP-weighted) world average,
reflecting the fact that countries need to be at different stages of the demographic transition
in order for it to have an impact on their external positions.

52 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 3.11 Demographic impact on current account

percent of GDP, change from 2020 to 2030

3
2
1
0
–1
–2
–3
d

ia

do a

es

nd

ea

a
si

si

di

si
pa
an

al

in

or
hi

la
ay

ne

A
In
tr
al

Ja
pp
C

ai

K
us

al
Ze

Th
ili
M
A

In

Ph
ew
N

Note: Current account norms (based on demographic variables only)

are adjusted for multilateral consistency. Asia impact indicates the sum
of projected changes in Asia’s current accounts as a percent of world
GDP. Hong Kong, Singapore, and Vietnam are not shown as they are
not included in the sample for the external balance assessment model.
Sources: United Nations, World Population Prospects: 2015 Revision
(medium-fertility scenario); IMF, World Economic Outlook; and IMF
Staff projections.

(figure 3.12): Countries with current account surpluses are expected to

remain capital exporters, and countries in deficit are expected to remain
capital importers.
These results are based on a partial equilibrium analysis, which takes
the values of other macroeconomic variables in the EBA model (the future
expected growth rate, the level of relative productivity, relative output gap,
relative fiscal balance) as fixed. The broader impact of demographics may
be smaller or larger than the estimated partial effect, depending on how
aging interacts with these variables.18

Implications for Financial Markets

The changes in savings and investment associated with aging can also have
implications for domestic financial markets. A panel regression was con-
ducted to examine the potential impact of demographic trends on domestic
interest rates, equity returns, and real estate prices in the region (IMF 2017).
The results suggest that both rising old-age dependency in post-dividend

18. For example, aging may affect fiscal balance through higher pension and healthcare
spending. As public health spending is included as a control variable in the EBA, the estimates
(figure 3.11) account for this channel based on health spending projections (Amaglobeli and
Shi 2016). The estimates do not account for the role of generosity of pension systems, which
could be an important factor behind private saving behavior.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 53

 Figure 3.12 Changes in current account induced by demographics
and current account balances in 2015

change in current account norm (percent of GDP, 2020–30)

3
Japan
2 Korea

1
Philippines Thailand
India
0 Indonesia China

Malaysia
–1 Australia

–2

New Zealand
–3
–6 –2 2 6 10
current account balance (percent of GDP, 2015)

Note: Current account norms (based on demographic variables only) are

adjusted for multilateral consistency.
Sources: United Nations, World Population Prospects: 2015 Revision (medium-
fertility scenario); IMF, World Economic Outlook; and IMF Staff estimates.

countries and falling youth dependency in early-dividend countries are

likely to put downward pressure on domestic real interest rates. The impact
of these factors diminishes with financial openness.

Impact on Interest Rates

The decline in long-term interest rates is a global phenomenon. Long-term
bond yields have declined significantly in Europe and the United States. A
similar trend is observed in Asia, particularly Korea and Australia, where
the decline has been nearly as large (figure 3.13). Besides reflecting better-
anchored inflation expectations, these declines reflect a significant decline
in world real interest rates, which drifted down from about 4 percent in the
late 1990s to about zero in 2013 (figure 3.14).
The decline in real interest rates reflected the decline in the natural
rate of interest (the interest rate that is consistent with full employment
and inflation at the central bank’s target). The estimated natural rates of
interest in Europe, the United Kingdom, and the United States declined
dramatically after the start of the global financial crisis (Lubik and Matthes
2015; Rachel and Smith 2015; Holston, Laubach, and Williams 2016).
Natural rates have fallen in Australia, Japan, and Korea while remaining
broadly stable and relatively high in emerging economies that have not yet
come under aging pressures. Natural rates in China have fallen but remain
high relative to advanced Asian economies (figure 3.15).
54 SUSTAINING ECONOMIC GROWTH IN ASIA
Figure 3.13 Change in 10-year government yield in selected countries

percentage points; end-2016 compared with 2000–07 average

0
–0.5
–1.0
–1.5
–2.0
–2.5
–3.0
–3.5
–4.0
–4.5
y

om

es

ea

lia

Th m

nd

na

a
si

si
an

or
pa

a
at

ra
or

hi
la
ne

ay
tn
gd

ap
m

Ja
St

C
t

ai
K

us

al
do

ie
er

ng
in

M
V
d

A
G

In
te

Si
d

ni
te

U
ni
U

Sources: Haver Analytics; Bloomberg LP; CEIC Asia database; IMF Staff calculations.

Figure 3.14 World real interest rates, 1985–2013

percent
6

–1
1985 1989 1993 1997 2001 2005 2009 2013

Source: King and Low (2014).

Demographic changes have been hailed as important drivers of the

secular decline in interest rates.19 In closed economies, they can affect savings

19. Other drivers of the secular decline in natural interest rates can be a slowdown in trend
productivity growth, shifts in saving and investment preferences (rising inequality), precau-
tionary saving in emerging markets, a fall in the relative price of capital goods, and a prefer-
ence away from public investment (Rachel and Smith 2015). This section focuses on real

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 55

 Figure 3.15 Real neutral interest rates in selected countries

percentage points
4.0
1990–99
3.5
2000–07
3.0 2014–16

2.5

2.0

1.5

1.0

0.5

0
United Japan Korea Australia Indonesia China Malaysia
States

Note: For Indonesia, China, and Malaysia, 1990–99 data are not available.
Sources: IMF, International Financial Statistics; Haver Analytics; IMF Staff estimates.

Table 3.5 Expected impact of demographic variables on

interest rate
Variable Savings Investment Interest rates
Youth dependency i h h
Old-age dependency i i Ambiguous

Aging Speed h i i
Source: Authors’ illustration.

and thereby interest rates, primarily through youth dependency, old-age de-
pendency, and aging speed (table 3.5):20
n Youth dependency. In principle, as the youth dependency ratio rises, the
working-age cohort saves less (to cover the rising expenditure on chil-
dren), the capital-to-labor ratio falls, and interest rates rise. Youth de-
pendency is expected to fall drastically in early-dividend countries, such
as the Philippines and India (figure 3.16, panel a).

interest rates. Low interest rates may also reflect low steady-state inflation as a result of
similar demographic pressures that weaken growth and drive up savings.
20. The previous section was based on the EBA model, which uses population growth as a
proxy for youth dependency. As youth dependency is a more direct measure of population
dynamics (and a complement to the old-age dependency ratio), we use it in this section.

56 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 3.16a Selected Asia: Youth dependency ratios

percentage points, change between 2020 and 2030

5
0
–5
–10
–15
–20
–25
us g

ew Ja ia
Ze an
ng nd

K e
ai a

C d
do na

ie a
M nam

pp a
es

ro es
ea
Th e

V esi

si

ili di
or
A on

n
al

in

Eu tat
or

ar
p

In hi
Si ala

la

ay
Ph In
ap
tr
K

n
t

S
al
g

d
on

te
H

ni
N

U
Note: Youth dependency ratio indicates the size of the population
14 years of age and younger as a share of the prime working-age
population (30–64 years old).
Sources: United Nations, World Population Prospects: 2015 Revision
(medium-fertility scenario); IMF, World Economic Outlook; IMF Staff
projections.

Figure 3.16b Selected Asia: Old-age dependency ratios

percentage points, change between 2020 and 2030

25

20

15

10

0
Ze nd

ie a
A am

ia

do an
M sia

pp a
es

ro s
ea
g

e
Th ea

Eu ate
n

si

i
an

al

Ph Ind
or

in
on

ar
hi

p
ew aila

ne

ay
tn
or

tr
al

Ja
C

St
ap
K

us

al
K

ili
ng

V
g

d
In
on

te
Si

ni
H

Note: Old-age dependency ratio indicates the size of the population 65 years of
age and over as a share of the prime working-age population (30–64 years old).
Sources: United Nations, World Population Prospects: 2015 Revision (medium-
fertility scenario); IMF, World Economic Outlook; IMF Staff projections.

n Old-age dependency. As old-age dependency rises, savings fall. As the labor

force shrinks, the capital-to-labor ratio rises and investment falls. The
impact of old-age dependency on interest rates is therefore theoretically
uncertain. Old-age dependency is expected to rise relatively quickly in
Hong Kong, Singapore, and Korea (figure 3.16, panel b).

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 57

 Figure 3.16c Selected Asia: Aging speed

percentage points, change between 2020 and 2030

8
4
0
–4
–8
–12
–16

ea
al n
a

ie a
do m
ili sia

K s

ai a
nd

ng na

us e

N ng lia
Ze ng

ro es
e
si

Th e

A or
pa

an
d
In tna

in

Eu tat
o tra

ar
or

i
la

o
ay

Ph ne

Si Ch
In

ap
Ja

al
ew K
pp

S
M

d
te
H

ni
U
Note: Aging speed is the projected change in the old-age dependency
ratio over the next 20 years. Old-age dependency ratio indicates the
size of the population 65 and older as a share of the prime working-
age population (30–64 years old).
Sources: United Nations, World Population Prospects: 2015 Revision
(medium-fertility scenario); IMF, World Economic Outlook; IMF Staff
projections.
J
log
 Yt  log TFPt   1     log  K t     log  N t j  LFPt j  Et j  wtj (3.1)
n Aging speed. Higher aging speed implies a higher probability
j 1 of survival,
which, if not matched by later retirement, is likely to have a positive
impact on life-cycle savings. Higher aging speed also lowers current
investment, thereby reducing interest rates. Aging speed is high and
expected to fall in countries in late stages of the demographic transi-
tion (figure
∆𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙3.16, panel c). Japan, where aging speed (3.2)
�� � �� 𝑤𝑤𝑤𝑤�� � �� 𝐷𝐷𝐷𝐷�� � �� � η� � ���
will continue to
increase in the next decade, is an exception.

To study the impact of demographic variables on the 10-year real inter-

est rate, we estimate the following specification:
𝑟𝑟�� � 𝛽𝛽� � 𝛽𝛽� 𝑌𝑌𝑌𝑌�� � 𝛽𝛽� �𝑌𝑌𝑌𝑌�� ∗ 𝐶𝐶𝐶𝐶�� � � 𝛽𝛽� 𝑂𝑂𝑂𝑂�� � 𝛽𝛽� �𝑂𝑂𝑂𝑂�� ∗ 𝐶𝐶𝐶𝐶�� � � 𝛽𝛽� 𝐴𝐴𝐴𝐴�� �
𝛽𝛽� �𝐴𝐴𝐴𝐴�� ∗ 𝐶𝐶𝐶𝐶�� � � 𝛽𝛽� 𝑅𝑅𝑅𝑅� � 𝛾𝛾𝐶𝐶𝛾𝛾𝛾𝛾𝛾𝛾𝑟𝑟𝛾𝛾𝑙𝑙𝛾𝛾�� � �� ���� (3.3)
(3.3)
where i indicates the country and t the year. The dependent variable, r, is
the 10-year real interest rate. Among the explanatory variables, YD and OD
denote the youth dependency ratio (the ratio of the population under 30 to
the
𝑟𝑟�� �population
�� � �� 𝑌𝑌𝑌𝑌��30–64)
�� � 𝐶𝐶𝐶𝐶and the old-age dependency ratio (the ratio of the
�� � � �� 𝑂𝑂𝑂𝑂�� �� � 𝐶𝐶𝐶𝐶�� � � �� 𝐴𝐴𝐴𝐴�� �� � 𝐶𝐶𝐶𝐶�� � � �� 𝑅𝑅𝑅𝑅� �
population
𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾�� �over
�� �64
��� .to the population 30–64), respectively, to account for
(3.3a)
the effects of changes in fertility and aging population on interest rates. AS
denotes the aging speed. These variables are separately interacted with the
capital openness index (CO) to analyze how the openness of the economy
affects the impact of demographic variables on interest rates. RW denotes
the world interest rate. The 10-year real interest rates and world interest
rate data come from IMF (2014), King and Low (2014), and the IMF World
58 SUSTAINING ECONOMIC GROWTH IN ASIA
Economic Outlook. The control variables include the ratio of a country’s per
capita GDP to that of the United States, growth in labor productivity, and
the cyclically adjusted primary balance.
The empirical estimates support our priors for the effect of youth
dependency and aging speed on interest rates. Higher old-age dependency is
found to reduce interest rates. These results are consistent with the results
presented in the previous section on the saving and investment balance.
The effects of demographic factors are not wholly channeled domesti-
cally in open economies. As one moves to an economy with an open capital
account, the savings-investment balance—and hence interest rates—are at
least partly determined by global saving and investment. In the extreme case
of perfect capital mobility, arbitrage in financial markets should equalize
interest rates across borders, and demographic factors of each country
should not have an impact on domestic interest rates (unless they are large
enough to contribute to global demographic trends).
We test this hypothesis by imposing restrictions on equation (3.3),
such that b1 = –b2, b3 = –b4, and b5 = –b6. The results hold. Hence we run a
restricted model:

. (3.3a)

The impact of domestic demographic factors on interest rates tends

to diminish as a country becomes more open. In our analysis based on the
restricted model (table 3.6), the impact of demographic variables (youth
dependency, old-age dependency, and aging speed) all become zero as an
economy becomes perfectly open (based on the Chinn-Ito index). Youth
dependency, old-age dependency, and aging speed are expressed as ratios. A
1 percentage point increase in youth dependency increases the interest rate
by 8.26 basis points when the economy is fully closed; there is no impact
in the case of a fully open economy. Hence in estimating the demographi-
cally induced changes in real interest rates over 2020–30, interest rates in
Japan, Hong Kong, New Zealand, and Singapore with full capital mobility
are decoupled from their domestic demographic trends.
In countries that are not perfectly open, the old-age dependency effect
is important for mature economies, and the youth dependency effect domi-
nates for economies that are relatively young. Increasing old-age depen-
dency is expected to decrease interest rates, especially in post-dividend
countries, such as China, Korea, and Thailand. Declining youth depen-
dency is expected to decrease interest rates, especially in early-dividend
countries, such as India, Indonesia, and the Philippines, where fertility rates
are projected to decline.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 59

 Table 3.6 Panel regression: Demographics and long-term
interest rates
10-year real interest rate
Dependent variables (15–64)
8.26***
Youth dependency ratio * (1 – capital openness)
(1.29)
–16.16***
Old-age dependency ratio * (1 – capital openness)
(5.51)
–29.26***
Aging speed * (1 – capital openness)
(9.87)
0.84***
World interest rate
(0.11)
2.43*
Ratio of GDP per capita to that of the US
(1.39)
0.00
Cyclically adjusted primary balance
(0.04)
0.07
Growth in labor productivity
(0.06)
Observations 740
Number of groups 42
*** p<0.01, ** p<0.05, * p<0.1.
Note: Standard errors in parentheses. P denotes the p-value as the probability
of obtaining a result equal to or more extreme than that observed. The regres-
sion controls for the country fixed effect.
Source: IMF Staff estimates.

A slower pace of aging can be expected to push up interest rates. They

are expected to increase most in China and Australia, as the aging speeds
in these countries fall, reducing savings and, consequently, raising interest
rates. Aging speed is projected to increase in currently young countries,
such as India and Malaysia, driving down their interest rates.
The results suggest that demographic trends could push interest rates
down by about 1–2 percentage points in the next decade, all else equal
(figures 3.17 and 3.18).21,22 The impact depends on three factors: the degree
of openness, the state of aging, and the speed of aging. For countries that
are fully open, demographic factors do not have a direct effect on domestic
long-term interest rates. For post-dividend countries (Korea and Thailand),

21. Given the low-frequency variation in demographic variables, annual real interest rates may
introduce substantial noise to any relationship with a demographic structure. To account
for this problem, we also considered three- and five-year rates for nonoverlapping periods,
as explained in IMF (2017). Such multiperiod rates emphasize the low-frequency variation
in real interest rates. The results are broadly similar to the results for the baseline scenario.
22. Rachel and Smith (2015) find that demographic factors, public investment, and a global
savings glut explain about 2 (out of 4.5) percentage points of the decline in global neutral
rates between 1980 and 2015.

6
60 GRAPHICS - CHAPTER
SUSTAINING ECONOMIC3—SUSTAINING ECONOMIC GROWTH IN ASIA
GROWTH IN ASIA
 Figure 3.17 Selected Asia: Impact of demographics on 10-year
real interest rates

percentage points, cumulative change between 2020 and 2030

1.0
0.5
0
–0.5
–1.0
–1.5
–2.0
–2.5
ia

nd

na

es

ea

lia

a
di

si

si
s

in

ra
or
hi
la
ay

ne

A
In

pp
C

t
ai

us
al

do
Th

ili
M

A
In
Ph

Aging speed effect

Youth dependency effect
Old-age dependency effect
Total demographic effect

Note: The figure for Asia reflects the nominal GDP-weighted average.
Source: IMF Staff projections.

Figure 3.18 Selected Asia: Impact of demographics on

10-year real interest rates, by country

percentage points, cumulative change between 2020 and 2030

0

–0.5

–1.0

Korea
–1.5 Philippines
China
India
–2.0 Malaysia
Thailand
–2.5
2020 2022 2024 2026 2028 2030

Source: IMF Staff projections.

rising old-age dependency is expected to depress interest rates. For early-

dividend countries (India and the Philippines), falling youth dependency
is projected to reduce interest rates. As the speed of aging slows in some
cases, the aging speed effect will attenuate the decline in interest rates.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 61

Table 3.7 Expected impact of demographic variables on asset returns
Variable Interest rate Risk appetite Equity premium Stock returns
Youth dependency h h i Ambiguous

Old-age dependency i i h Ambiguous

Aging speed i h i i
Source: Authors’ illustration.

Impact on Other Assets

A popular argument in the literature is the “asset market meltdown”
hypothesis, which postulates that as baby boomers retire and draw down
their savings, the resulting sell-off pressure in asset markets will sharply
reduce valuations. Historical experience in the United States and Japan
provides some indication of the correlation between demographics and
postwar stock market trends. But recent empirical studies find mixed
evidence of the link between demographics and asset returns (IMF 2017).
In principle, demographic trends can affect expected stock returns
through their impact on interest rates and risk premiums (table 3.7).23 In
particular, a decline in youth dependency, or an increase in old-age depen-
dency, is expected to lead to a fall in interest rates, as discussed in the
previous section (table 3.6). At the same time, these trends also reduce the
lifetime investment horizon, lower the preference for risky assets, and raise
the equity risk premium.24 The expected impact of dependency ratios on
stock returns is thus conceptually ambiguous.
The literature yields mixed findings on the empirical link between
demographics and asset returns, depending on the sample and demographic
variables used. Poterba (2001, 2004) finds evidence that US households’
asset holdings held outside defined-benefit pensions decline only gradu-
ally during retirement and that there is no significant relationship between
aging and stock returns in the postwar US data. In contrast, Geanakoplos,
Magill, and Quinzii (2004) and Davis and Li (2003) find that the middle
age to young ratio and the population share of prime savers have signifi-
cant positive relationships with real stock prices, respectively, in a group
of advanced economies. Engelhardt and Poterba (1991) show that the
empirical relationship between real housing prices and demographic vari-

23. By definition, expected stock returns are equal to the sum of the risk-free rate and the
equity risk premium.
24. If the correlation between labor income and stock returns is sufficiently low, a labor
income stream acts as a substitute for risk-free bond holdings, implying that investors
should hold a declining share of stocks in their portfolio as they get older (Jagannathan and
Kocherlakota 1996).

62 GRAPHICSECONOMIC
SUSTAINING - CHAPTER GROWTH
3—SUSTAINING ECONOMIC GROWTH IN ASIA
IN ASIA 7
Table 3.8 Panel regression: Demographics and asset returns
Percent growth
Percent growth in real property
Dependent variables in stock return price
67.28*** –11.32***
Youth dependency ratio
(23.66) (3.96)
–31.01** 0.67
Youth dependency ratio * capital openness
(14.22) (4.00)
–32.69 –53.18*
Old-age dependency ratio
(82.26) (31.77)
138.26* –5.37
Old-age dependency ratio * capital openness
(75.02) (26.99)
315.31* –100.95***
Aging speed
(163.18) (34.05)
–317.04** 60.79*
Aging speed * capital openness
(142.92) (31.86)
–1.18 –1.48***
World interest rate
(2.17) (0.53)
4.75*** 0.3
Growth in labor productivity
(0.98) (0.20)
Observations 406 716
Number of groups 14 56
*** p<0.01, ** p<0.05, * p<0.1.
Note: Standard errors in parentheses. P denotes the p-value as the probability of ob-
taining a result equal to or more extreme than that observed. The regression controls
for the country fixed effect.
Source: IMF Staff estimates.

ables that Mankiw and Weil (1989) find for the United States does not hold
for Canada. Takáts (2010) finds that an increase in the change of old-age
dependency lowers real housing price growth by about 66 basis points in a
group of 22 advanced economies.
We find that lower youth (or higher old-age) dependency is associ-
ated with lower stock returns (i.e., the interest rate channel dominates),
but the relationships are not statistically strong (table 3.8). Moreover, as
with interest rates, the impact of domestic demographic factors is partially
offset in more financially open countries.
For real estate, the relationship with demographic variables is even
more difficult to identify, because of the asset’s dual role as a durable good.
Conceptually, a fall in interest rates, triggered by a fall in youth depen-
dency (or a rise in old-age dependency), would raise housing prices. At the
same time, these trends are expected to reduce demand for housing, as
they are associated with declines in household formation.25 Empirically, we

25. Furthermore, because the housing stock responds with lags, subsequent downward price
adjustments may mask an initial price increase caused by a positive demand shock (Poterba

8 GRAPHICS - CHAPTER
IS ASIA AT RISK3—SUSTAINING ECONOMIC
OF GROWING OLD BEFOREGROWTH IN ASIA
BECOMING RICH? 63
find weak links between these variables and real estate prices. The degree
of openness plays an insignificant role, likely reflecting the local nature of
housing markets.

Policy Implications of Demographic Trends

For early-dividend countries in Asia (India, Indonesia, and the Philippines),
the main policy challenge is to harness the demographic dividend where
possible, as discussed in Bloom, Canning, and Sevilla (2003), and mitigate
any adverse spillovers from aging in the rest of Asia.
For Asian countries in the late- or post-dividend stages, adapting to
aging could be especially challenging, because of relatively low per capita
income levels. In light of this phenomenon, policies aimed at protecting
vulnerable older people and prolonging strong growth take on particular
urgency in Asia. These challenges call for adapting macroeconomic poli-
cies early, before aging sets in. Specific structural reforms can also help,
especially in the areas of labor markets, pension systems, and retirement
systems. These policies could be supplemented by productivity-enhancing
reforms (e.g., research and development and education), as discussed in
chapter 3 of the IMF’s Regional Economic Outlook: Asia and Pacific (IMF 2017).

Macroeconomic Policies
Japan’s experience shows that it is important to adapt macroeconomic
policies before aging sets in. Fiscal policy may include introducing a cred-
ible medium-term fiscal framework to secure debt sustainability, shifting
the burden of taxes from labor to consumption, and revamping the social
safety net.26 Monetary policy may involve studying how monetary trans-
mission may change with aging. If, for example, monetary transmission
works more through asset prices and household wealth than through
corporate borrowing costs, the interest rate sensitivity of output and infla-
tion may decline (Miles 2002, Bean 2004). To the extent that aging leads
to declines in the natural interest rate, regular assessment of the neutral
monetary stance by central banks is needed to avoid a potential tightening
bias. Prolonged low interest rates may also call for a strong macropruden-
tial framework to mitigate related financial stability risks.

1984, Lindh and Malmberg 2008). Moreover, higher demand for housing could manifest
itself more prominently through the rental rate, which may not always move together with
housing prices (Hamilton 1991).
26. At the same time, with a credible medium-term fiscal framework, fiscal policy can be used
more actively in the short run, given its higher potency in a low interest rate environment,
including to support aging-related structural reforms.

64 SUSTAINING ECONOMIC GROWTH IN ASIA

Structural Reforms
Labor Market Reforms
Labor market reforms aimed at tackling labor shortages and workforce
aging can help offset some of the adverse growth effects of aging discussed
in the chapter. Measures include the following:
n Increase labor force participation, especially for women and older people.
Expanding the availability of childcare facilities, removing fiscal disin-
centives against older peoples’ labor participation, and promoting
flexible employment can be especially effective at raising labor partici-
pation by women and older people (Elborgh-Woytek et al. 2013,
Kinoshita and Guo 2015, Olivetti and Petrongolo 2017). Moving
from a seniority-based to a performance-based wage system can incen-
tivize firms to relax retirement age requirements while reducing labor
market duality, which can segment the labor market into permanent
and transitory jobs (Dao et al. 2014).
n Attract foreign workers, including through guest worker programs that target
specific skills. Foreign workers could address labor shortages and have
a generally positive impact on receiving countries (Ganelli and Miake
2015; Jaumotte, Koloskova, and Saxena 2016). The experiences of
Australia, Hong Kong, New Zealand, and Singapore show that immi-
gration can prolong the demographic dividend or soften the negative
impact of rapid aging.27
n Promote active labor market policies. Making affordable health care avail-
able for older workers (who are disproportionately affected by health
risks) and facilitating the upgrading of human capital of and retraining
workers can moderate the negative effect that workforce aging can
have on productivity growth (Aiyar, Ebeke, and Shao 2016).

Pension System Reforms

Given the rapid aging and related fiscal costs in Asia, as well as the region’s
relatively low pension coverage (World Bank 2016), strengthening pension
systems should be a high priority. Policy measures could include the fol-
lowing:
n Create automatic adjustment mechanisms that link changes in the retirement
age (or benefits) to life expectancy, in order to help depoliticize pension

27. IMF research finds that key to harnessing the long-term gains of foreign workers are
actions that facilitate their integration into the labor market, including language training
and job search assistance, better recognition of migrants’ skills through the recognition of
credentials, and lower barriers to entrepreneurship (IMF 2017).

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 65

 reform and contain pension costs (Arbatli et al. 2016). Many European
countries have introduced such rules; their use has been limited in Asia
outside of Japan.
n Encourage voluntary saving (e.g., tax deductions for long-term retirement
saving), to help relieve long-term fiscal burdens.
n Increase pension coverage, by providing minimum pension guarantees. This
reform could provide a safety net for the vulnerable, mitigate the
impact of entitlement reforms, and reduce incentives for precautionary
savings (Zaidi, Grech, and Fuchs 2006).
n Reform the management of public pension funds. Asia is home to some of
the largest public pension funds in the world (OECD 2016). Reducing
home bias—along the lines of the recent Government Pension Invest-
ment Fund reforms in Japan that increased allocation to foreign
assets—could help raise the returns on these funds and secure more
sustainable resources for aging societies.

Retirement System Reforms

New financial products that help older people dissave their postretirement
savings (e.g., reverse mortgages) or insure against longevity risks (e.g., annu-
ities) could reduce the need for precautionary savings. The diverse demo-
graphic trends in Asia could also offer rich opportunities for cross-border
risk-sharing and financial integration.28 For example, savings in late- or
post-dividend countries seeking higher returns could be used to finance the
large infrastructure gaps in early-dividend Asian countries (Ding, Lam, and
Peiris 2014). Increasing the availability of “safe assets”—such as long-term
government bonds or inflation-linked securities—can be especially attrac-
tive for pension funds and insurance companies (Groome, Blancher, and
Ramlogan 2006).

Summary of Findings and Policy Recommendations

This chapter reports five main findings:
n Trends. Asia is aging rapidly. The speed of aging is especially remark-
able compared with the historical experience in Europe and the United
States. Parts of Asia risk becoming old before becoming rich. Until
recently, Asia had the largest share of the world’s working-age people.

28. Financial integration in Asia remains low, especially given its high degree of trade inte-
gration. About 60 percent of Asia’s exports and imports go to, or originate from, elsewhere
within the region; only 20–30 percent of cross-border portfolio investment and bank claims
are intraregional (IMF 2015b).

66 SUSTAINING ECONOMIC GROWTH IN ASIA

 In the coming decades, it will remove hundreds of millions of working-
age people from that population.
n Growth. Asia has enjoyed a substantial demographic dividend in past
decades. But rapid aging is now set to create a demographic tax on
growth. Demographic trends could shave 0.5–1 percentage point from
annual GDP growth over the next three decades in post-dividend coun-
tries such as China and Japan. In contrast, they could add 1 percentage
point to annual GDP growth in early-dividend countries, such as India
and Indonesia, if the transition is well managed. Overall, demographics
are likely to be slightly negative for Asian growth. They could reduce
annual global growth over the next three decades by 0.1 percentage
point (0.2 percentage point if early-dividend countries are unable to
reap the demographic dividend). If past trends continue, immigra-
tion could play an important role in softening the impact of aging or
prolonging the demographic dividend in Australia, Hong Kong, New
Zealand, and Singapore.
n Inflation. Where structural excess savings and low investment because
of demographic changes lead to such a low real neutral interest
rate that monetary policy is no longer able to act as a stimulus, the
economy may operate below potential, keeping inflation under the
central bank’s target. Asia could fall into a period of “secular stagna-
tion” at a lower income level than in advanced economies and with less
effective policy buffers.
n External flow balance. The diversity of demographic trends in the region
creates opportunities for capital flows and cross-border risk sharing,
as savings from surplus countries can be used to meet capital needs
in economies with younger populations. Projections based on the
IMF’s EBA model suggest that surpluses of some Asian economies
are projected to increase over the next decade as a result of demo-
graphic change. The impact is material only for a small set of coun-
tries, however, and the overall effect on global imbalances is likely to
be limited (about 0.1 percent of global GDP over the next decade).
n Financial markets. Demographic trends are likely to put downward pres-
sure on real interest rates and asset returns for most major countries in
Asia. These effects are likely to be less important for economies that are
financially open. For them, changes in the world interest rate—which
may also be driven by global aging trends—will likely matter more.

Three main policy implications emerge from these findings:

n Adapting to aging may be especially challenging for Asia, where aging is
taking place at relatively low per capita income levels. Policies aimed at

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 67

 protecting vulnerable older people and prolonging strong growth take
on particular urgency there.
n It is important to adapt macroeconomic policies before aging sets in. Policies
may include securing debt sustainability and monitoring potential
changes in monetary transmission related to aging.
n Structural reforms can help address these challenges. They may include labor
market reforms (promotion of labor force participation of women and
older people, guest worker programs, and active labor market policies);
pension reforms (automatic adjustment mechanisms and minimum
pension guarantees); and retirement system reforms (new financial
products to reduce precautionary saving and increase the availability of
“safe assets”). These policies could be supplemented by specific produc-
tivity-enhancing reforms (e.g., research and development and educa-
tion), as discussed in the IMF’s Regional Economic Outlook (IMF 2017).

References
Acemoglu, D., and P. Restrepo. 2017. Secular Stagnation? The Effect of Aging on Economic Growth
in the Age of Automation. NBER Working Paper 23077. Cambridge, MA: National Bureau
of Economic Research.
Adler, G., R. Duval, D. Furceri, S.K. Çelik, K. Koloskova, and M. Poplawski-Ribeiro. 2017. Gone
with the Headwinds: Global Productivity. IMF Staff Discussion Note SDN/17/04. Washing-
ton: International Monetary Fund.
Aiyar, S., C. Ebeke, and X. Shao. 2016. The Impact of Workforce Aging on European Productivity.
IMF Working Paper 16/238. Washington: International Monetary Fund.
Amaglobeli, D., and W. Shi. 2016. How to Assess Fiscal Implications of Demographic Shifts: A Granu-
lar Approach. Fiscal Affairs Department How to Note 2/2016. Washington: International
Monetary Fund.
Arbatli, E., C. Feher, J. Ree, I. Saito, and M. Soto. 2016. Automatic Adjustment Mechanisms in Asian
Pension Systems? IMF Working Paper 16/242. Washington: International Monetary Fund.
Barnes, S., J. Lawson, and A. Radziwill. 2010. Current Account Imbalances in the Euro Area: A Com-
parative Perspective. Economics Department Working Paper ECO/WKP(2010)82. Paris: Or-
ganization for Economic Cooperation and Development.
Basten, S., and Q. Jiang. 2015. Fertility in China: An Uncertain Future. Population Studies 69
(Supplement1): 97–105.
Bean, C. 2004. Global Demographic Change: Some Implications for Central Banks. Paper pre-
sented at the Federal Reserve Bank of Kansas City Annual Symposium, Jackson Hole, WY.
Bloom, D.E., D. Canning, and G. Fink. 2010. Implications of Population Ageing for Economic
Growth. Oxford Review of Economic Policy 26, no. 4: 583–612.
Bloom, D.E., D. Canning, G. Fink, and J.E. Finlay. 2007. Fertility, Female Labor Participation, and
the Demographic Dividend. NBER Working Paper 13583. Cambridge, MA: National Bureau
of Economic Research.
Bloom, D.E., D. Canning, and J. Sevilla. 2003. The Demographic Dividend: A New Perspec-
tive on the Economic Consequences of Population Change. Population Matter Series. Santa
Monica, CA: RAND.

68 SUSTAINING ECONOMIC GROWTH IN ASIA

Börsch-Supan, A., and M. Weiss. 2016. Productivity and Age: Evidence from Work Teams at
the Assembly Line. Journal of the Economics of Ageing 7: 30–42.
Bosworth, B., and G. Chodorow-Reich. 2007. Saving and Demographic Change: The Global Dimen-
sion. Working Paper. Chestnut Hill, MA: Center for Retirement Research, Boston College.
Chinn, M.D., and H. Ito. 2007. Trade Account Balances, Financial Development and Institu-
tions: Assaying the World Saving Glut. Journal of International Money and Finance 26, no. 4:
546–69.
Cuberes, D., and M. Teignier. 2016. Aggregate Costs of Gender Gaps in the Labor Market: A
Quantitative Estimate. Journal of Human Capital 10, no. 1: 1–32.
Dao, M., D. Furceri, J. Hwang, M. Kim, and T.-J. Kim. 2014. Strategies for Reforming Korea’s Labor
Market to Foster Growth. IMF Working Paper 14/137. Washington: International Monetary
Fund.
Davis, E. P., and C. Li. 2003. Demographics and Financial Asset Prices in the Major Industrial Econo-
mies. Brunel University Public Discussion Paper 03-07. London: Brunel University.
Ding, D., W. Lam, and S. Peiris. 2014. Future of Asia’s Finance: How Can It Meet Challenges of
Demographic Change and Infrastructure Needs? IMF Working Paper 14/126. Washington: In-
ternational Monetary Fund.
Eichengreen, B., and M. Fifer. 2002. The Implications of Ageing for the Balance of Payments
between North and South. In The Economics of Ageing Societies, ed. Horst Siebert, 81–105.
Tubingen, the Netherlands: Mohr.
Elborgh-Woytek, K., M. Newiak, K. Kochhar, S. Fabrizio, K. Kpodar, P. Wingender, B. Cle-
ments, and G. Schwartz. 2013. Women, Work, and the Economy: Macroeconomic Gains from
Gender Equity. IMF Staff Discussion Note SDN/13/10. Washington: International Mon-
etary Fund.
Engelhardt, G. V., and J. Poterba. 1991. House Prices and Demographic Change: Canadian
Evidence. Regional Science and Urban Economics 21, no. 4: 539–46.
Feyrer, J. 2007. Demographics and Productivity. Review of Economics and Statistics 89, no. 1:
100–109.
Ganelli, G., and N. Miake. 2015. Foreign Help Wanted: Easing Japan’s Labor Shortages. IMF Work-
ing Paper 15/181. Washington: International Monetary Fund.
Geanakoplos, J., M. Magill, and M. Quinzii. 2004. Demography and the Long-run Predictabil-
ity of the Stock Market. Cowles Foundation Discussion Papers, No. 1380R. New Haven,
CT: Cowles Foundation for Research in Economics, Yale University.
Gonzales C., S. Jain-Chandra, K, Kochhar, M. Newiak, and T. Zeinullayev. 2015. Catalyst for
Change: Empowering Women and Tackling Income Inequality. IMF Staff Discussion Note
SDN/15/20. Washington: International Monetary Fund.
Grigoli, F., A. Herman, and K. Schmidt-Hebbel. 2014. World Saving. IMF Working Paper
14/204. Washington: International Monetary Fund.
Groome, W. T., N. Blancher, and P. Ramlogan. 2006. Aging and Financial Markets. Finance
and Development 43 (September).
Gruber, J.W., and S. B. Kamin. 2007. Explaining the Global Pattern of Current Account Imbal-
ances. Journal of International Money and Finance 26, no. 4: 500–22.
Hamilton, B.W. 1991. The Baby Boom, the Baby Bust, and the Housing Market: A Second
Look. Regional Science and Urban Economics 21, no. 4: 547–52.
Higgins, M. 1998. Demography, National Savings, and International Capital Flows. Interna-
tional Economic Review 39, no. 2: 343–69.
Higgins, M., and J.G. Williamson. 1997. Age Structure Dynamics in Asia and Dependency on
Foreign Capital. Population and Development Review 23, no. 2: 261–93.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 69

Holston, K., T. Laubach, and J. Williams. 2016. Measuring the Natural Rate of Interest: Internation-
al Trends and Determinants. Federal Reserve Bank of San Francisco Working Paper 2016-11.
IMF (International Monetary Fund). 2014. Perspectives on Global Real Interest Rates. In
World Economic Outlook (April). Washington.
IMF (International Monetary Fund). 2015a. Regional Economic Outlook: Sub-Saharan Africa:
Navigating Headwinds. Washington.
IMF (International Monetary Fund). 2015b. Regional Economic Outlook: Asia and Pacific: Stabiliz-
ing and Outperforming Other Regions. Washington.
IMF (International Monetary Fund). 2016. Methodological Note on EBA-Lite. IMF Policy Paper.
Washington.
IMF (International Monetary Fund). 2017. Regional Economic Outlook: Asia and Pacific: Preparing
for Choppy Seas. Washington.
Jagannathan, R., and N.R. Kocherlakota. 1996. Why Should Older People Invest Less in
Stocks than Younger People? Federal Reserve Bank of Minneapolis Quarterly Review 20: 11–23.
Jaimovich, N., and Henry Siu. 2009. The Young, the Old, and the Restless: Demographics and
Business Cycle Volatility. American Economic Review 99, no. 3: 804–26.
Jaumotte, F., K. Koloskova, and S. Saxena. 2016. Impact of Migration on Income Levels in Advanced
Economies. Spillover Note. Washington: International Monetary Fund.
Jaumotte, F., and P. Sodsriwiboon. 2010. Current Account Imbalances in the Southern Euro Area.
IMF Working Paper WP/10/139. Washington: International Monetary Fund.
Khera, P. 2016. Macroeconomic Impacts of Gender Inequality and Informality in India. IMF Working
Paper 16/16. Washington: International Monetary Fund.
King, M., and D. Low. 2014. Measuring the “World” Real Interest Rate. NBER Working Paper
19887. Cambridge, MA: National Bureau of Economic Research.
Kinoshita, Y., and F. Guo. 2015. What Can Boost Female Labor Force Participation in Asia? IMF
Working Paper 15/56. Washington: International Monetary Fund.
Legg, A., N. Prasad, and T. Robinson. 2007. Global Imbalances and the Global Saving Glut: A Panel
Data Assessment. Research Discussion Paper 2007-11. Sydney: Reserve Bank of Australia.
Li, H., J. Zhang, and J. Zhang. 2007. Effects of Longevity and Dependency Rates on Saving
and Growth: Evidence from a Panel of Cross Countries. Journal of Development Economics
84: 138–54.
Lindh, T., and B. Malmberg. 2008. Demography and Housing Demand: What Can We Learn
from Residential Construction Data? Journal of Population Economics 21, no. 3: 521–39.
Lubik, T., and C. Matthes. 2015. Calculating the Natural Rate of Interest: A Comparison of Two Alter-
native Approaches. Economic Brief EB15-10. Federal Reserve Bank of Richmond.
Mankiw, N. G., and D. N. Weil. 1989. The Baby Boom, the Baby Bust, and the Housing Mar-
ket. Regional Science and Urban Economics 19: 235–58.
Miles, D. 2002. Should Monetary Policy Be Different in a Greyer World? In Ageing, Financial
Markets and Monetary Policy, ed. Alan Auerbach and Hermann Heinz. Heidelberg: Springer.
OECD (Organization for Economic Cooperation and Development). 2016. Annual Survey of
Large Pension Funds and Public Pension Reserve Funds. Paris.
Olivetti, C., and B. Petrongolo. 2017. The Economic Consequences of Family Policies: Lessons
from a Century of Legislation in High-Income Countries. Journal of Economic Perspectives
31, no. 1.
Phillips, S., L. Catao, L. Ricci, R. Bems, M. Das, J. Di Giovanni, D.F. Unsal, M. Castillo, J. Lee,
J. Rodriguez, and M. Vargas. 2013. The External Balance Assessment (EBA) Methodology. IMF
Working Paper 13/272. Washington: International Monetary Fund.

70 SUSTAINING ECONOMIC GROWTH IN ASIA

Poterba, J.M. 1984. Tax Subsidies to Owner-Occupied Housing: An Asset-Market Approach.
Quarterly Journal of Economics 99, no. 4: 729–52.
Poterba, J.M. 2001. Demographic Structure and Asset Returns. Review of Economics and Statistics
83 (November): 565–84.
Poterba, J.M. 2004. Portfolio Risk and Self-Directed Retirement Saving Programs. Economic
Journal 114 (March): C26–C51.
Rachel, L., and T. Smith. 2015. Secular Drivers of the Global Real Interest Rate. Staff Working Pa-
per 571. London: Bank of England.
Steinberg, C., and M. Nakane. 2012. Can Women Save Japan? IMF Working Paper 12/248.
Washington: International Monetary Fund.
Takáts, E. 2010. Ageing and Asset Prices. BIS Working Paper No. 318 (August). Basel: Bank for
International Settlements.
UN (United Nations). 2015. World Population Prospects: 2015 Revision. Department of Economic
and Social Affairs Population Division. New York. Available at https://esa.un.org/unpd/
wpp/publications/files/key_findings_wpp_2015.pdf.
Veen, S. 2008. Demographischer Wandel, alternde Belegschaften und Betriebsproduktivität. Munich:
Rainer Hampp Verlag.
Williamson, J. G. 2001. Demographic Shocks and Global Factor Flows. In Seismic Shifts: The
Economic Impact of Demographic Change, ed. J.N. Little and R.K. Triest. Conference Series 46.
Boston: Federal Reserve Bank.
World Bank. 2015. Global Monitoring Report 2015/16: Development Goals in an Era of Demographic
Change. Washington.
World Bank. 2016. East Asia and Pacific Regional Report: Live Long and Prosper: Aging in East Asia
and Pacific. Washington.
World Bank. 2017. Doing Business 2017: Equal Opportunity for All. Washington.
Yoon, J-W., J. Kim, and J. Lee. 2014. Impact of Demographic Changes on Inflation and the Macro-
economy. IMF Working Paper 14/210. Washington: International Monetary Fund.
Zaidi, A., A. Grech, and M. Fuchs. 2006. Pension Policy in EU25 and Its Possible Impact on Elderly
Poverty. CASE Paper 116. London: Centre for Analysis of Social Exclusion, London School
of Economics and Political Science.

IS ASIA AT RISK OF GROWING OLD BEFORE BECOMING RICH? 71

 4
Invention, Productivity, and
the Evolution of East Asia’s
Innovation Systems
LEE G. BRANSTETTER AND NAMHO KWON

Japan’s rapid growth from the 1950s through the early 1970s had many
drivers, but most impressive was its rapid productivity growth (Denison
and Chung 1976). Economic growth slowed in all advanced economies in
the 1970s, but Japan grew more rapidly than the other rich Western econo-
mies, and its measured productivity growth remained higher than theirs.
In the 1970s and early 1980s, Japan’s industrial structure quickly shifted
from resource- and labor-intensive manufacturing toward knowledge- and
technology-intensive industries. Japanese manufacturers of steel, autos,
industrial machines, and, increasingly, electronics began displacing long-
established Western manufacturers worldwide. Japan’s research and devel-
opment (R&D) intensity and the number of patents held by Japanese firms
at home and abroad surged.
By the early 1980s, Japanese firms, long derided as “copycat” imitators
of Western technology, managed to outengineer their Western rivals in a
range of iconic industries.1 By the late 1980s, prestigious bodies like the
US National Research Council were glumly concluding that Japanese firms
were ahead of their American rivals in 25 out of 34 “critical” areas of tech-

Lee G. Branstetter is nonresident senior fellow at the Peterson Institute for International Economics and profes-
sor of economics and public policy at Carnegie Mellon University. Namho Kwon is an associate fellow at the Ko-
rea Institute of Public Finance. The authors thank their discussant, Jong-Wha Lee, for insightful comments and
suggestions. We also acknowledge the helpful input of Jérémie Cohen-Setton, Kyoji Fukao, Ken Kang, and Adam
Posen. This paper builds on research funded by the National Science Foundation (SciSIP grant 136170) and the
Carnegie Mellon Portugal Program. However, the views expressed in this chapter are those of the authors, and
we take responsibility for all errors and omissions.

1. Mansfield (1988a, 1988b, 1988c) documented the high productivity of Japan’s corporate
R&D spending.

73
nology (National Research Council 1992), and in 1987, the president of the
US Semiconductor Research Corporation conceded, “We may never match
Japan’s R&D efficiency.”
By the end of the 1980s, South Korea and Taiwan appeared to be fol-
lowing in Japan’s footsteps. At the beginning of that decade, both nations
were regarded as low-wage manufacturers of labor-intensive goods. By
its end, both were rapidly increasing their patenting in the United States
(Branstetter and Kwon 2018), and their export mix was quickly shifting
toward labor- and knowledge-intensive goods. Analysts began predicting
that the innovation leadership in electronics and information technology
was in the early stages of a permanent shift to East Asia.
The degree to which those prophecies of innovation leadership remain
unfulfilled suggests that the East Asian approach to R&D, pioneered by
Japan and adopted by South Korea and Taiwan, had both strengths and
weaknesses (Okimoto and Rohlen 1988). This system was arguably better
than the American system at applied research, hardware, miniaturization,
process equipment and technology, product variety, and process engi-
neering, whereas the American system was stronger in basic research, soft-
ware, creating new products, high-technology systems, and creation of new
industries.2 It was perhaps natural that engineers and firms in late-devel-
oping states would focus their efforts on a technological frontier largely
defined by prior foreign invention, rather than on fundamental innovation
(Gerschenkron 1962).
Once Japan and its former colonies reached the technological frontier,
and lost their longstanding cost advantages, however, the firms based in
these economies had to shift from an innovation system appropriate for
their “catch-up” periods to one that could generate its independent frontier
innovations. This chapter argues that the policies pursued during the high
growth eras of these economies may well have contributed to the difficulties
their firms have experienced in this transition. In Japan, South Korea, and
Taiwan, the evidence suggests that public policy choices during their years
of rapid growth concentrated R&D in particular activities, industries, and
firms. And these choices have slowed the ability of these countries to adapt
to current circumstances, imposing a drag on growth.3 Other nations in the
region seeking to follow in the footsteps of these economies should learn
from these policy missteps. (See box 4.1.)

2. Okimoto and Saxonhouse (1987) viewed these differences as reflecting intelligent adapta-
tion to the circumstances in which Japanese firms found themselves in the postwar era, and
some of the observations in their essay anticipate the arguments in this chapter.
3. Nelson (1992) influenced our thinking about the evolution of these economies’ innova-
tion systems.

74 SUSTAINING ECONOMIC GROWTH IN ASIA

Box 4.1 Radical innovation, pro-incumbent/pro-
incremental bias, and the decline in innovation
productivity
Healthy innovation systems need radical and incremental ap-
proaches, and transformative innovations require incremental in-
ventions to reach their full potential (Helpman and Trajtenberg
1998a, 1998b). Most innovative activity in any economy at any time
is incremental (Harhoff, Scherer, and Vopel 2003). But incremental
invention generally runs into diminishing returns. A fundamentally
new idea brought into the marketplace can create a new techno-
logical paradigm, raising the returns on the applied R&D invest-
ment, which refines this fundamental new idea and implements it
in multiple contexts (Evenson and Kislev 1976).
Radical technological breakthroughs are fueled by rapid entry
of a large number of new firms offering competing business and
product models (Klepper 1996). Eventually, a few of these entrants
hit upon business and product models that succeed commercially.
A small number of especially successful entrants then invest in
cost-reducing, process R&D to cement their success (Cohen and
Klepper 1996).
Often this success stifles further radical experimentation and
leads to a shakeout, leaving just a few players and product models
and designs. At this point, technological change in the industry
slows substantially, and the focus of innovation shifts from radical
experimentation with new designs to incremental improvements of
existing designs. The historical development of a number of major
industries—autos, tires, televisions, lasers, and semiconductors—
has been consistent with this theory (Klepper 2010, Buenstorf and
Klepper 2009).
In all cases, rapid progress to practical and usable designs and
business models requires rapid entry of many new firms. Economic
research highlights the importance of labor mobility into and
among these new firms. One common pattern in nearly all new
industries is the early emergence of an important set of leading
firms, which “seed” the industry with a large number of “spinoffs.”
At an early stage in a new industry, a small number of firms estab-
lish themselves as market leaders and attract talented pioneers
who want to make their fortunes in this new domain. Eventually,
managers disagree over which new business models and product
designs to pursue. Disaffected managers then leave to start
their own firms, which can take a new industry in new directions
(Buenstorf and Klepper 2009).
(box continues)

EVOLUTION
GRAPHICS - CHAPTER OF EAST ASIA’S
4—SUSTAINING INNOVATION
ECONOMIC GROWTH IN ASIA 751
SYSTEMS
 Box 4.1 Radical innovation, pro-incumbent/pro-
incremental bias, and the decline in innovation
productivity (continued)
Empirical evidence and theory also suggest that established
incumbents often resist investment in experiments that render
existing business models obsolete, highlighting the importance
of making capital available to newcomers and fostering competi-
tion. Christensen (1997), for example, points out that successful
incumbents face a double bind when seeking to engage in radical
innovation. They typically arrive at their position after years, even
decades, of successful refinement of existing products and busi-
ness practices but are then ill equipped to succeed in a radically
different approach to the same product class or service.
New industry creation in the 21st century is more science-based
and more closely connected to frontier university research than
in the past, thereby highlighting the importance of labor mobility
and connections between the academic and industry sectors. The
development of successful products based on these discoveries
often requires “human bridges”—entrepreneurial faculty and grad-
uate students conversant in the new technology who can partner
with experienced managers and engineers to embed the new
ideas in successful firms and products (Agrawal and Henderson
2003; Zucker, Darby, and Brewer 1998).
In the United States, strong protection of intellectual property
rights and the willingness of American courts to levy significant
damages or even issue injunctions when inventions are infringed
by financially stronger manufacturers have strengthened the
potential for innovation (Hochberg. Serrano, and Ziedonis 2018;
Jaffe and Lerner 2004). Strong intellectual property rights also
enable holders of patents to bargain with contract manufacturers
and other business partners without fear of expropriation (Hall and
Ziedonis 2001). These protections raise the chances of success
by loosening the historical requirement that firms excel in both
product and process innovation. America’s open capital markets
allow new startups to offer outsized returns to investors willing
to take high risks, further upping the odds of success. The histori-
cally open US market for consumer and industrial goods enables
new innovators to compete with established sellers. In America’s
fluid labor markets for top managers and engineers, past asso-
ciation with an ambitious startup—even if it fails—is viewed as a
useful career experience. By contrast, the repeated downsizing
and chronic competitive woes of traditional American manufac-
turers in more technologically quiescent industries have lessened
the appeal of spending an entire career as an engineer or manager
in an established firm. For talented young American managers and
engineers, the balance of risks and rewards tilts toward smaller
firms in technologically dynamic sectors.

76 GRAPHICS
2 SUSTAINING- ECONOMIC GROWTH IN ASIA
CHAPTER 4—SUSTAINING ECONOMIC GROWTH IN ASIA
Japan
Human Capital
In a world of globalized supply chains, where it is commonplace for compo-
nents of even the most sophisticated manufactured products to be assem-
bled in low-wage developing countries, it is easy to forget what a novelty
Japan was in the early postwar era. It had been the first and only nonwestern
nation to acquire modern manufacturing capabilities on a significant
scale in the years prior to World War II. In the postwar era, it continued to
grow with surprising vigor, rapidly becoming the first nonwestern nation
to join the Organization for Economic Cooperation and Development
(OECD). These achievements were, in part, the result of Japan’s exception-
ally successful, decades-long strategy of investing in the acquisition of both
general human capital and specific technical capabilities.
The strategy took shape after Japan’s Meiji Restoration of 1868, when
the new imperial state created a modern educational system and brought
Western expertise into its leading firms and government ministries.4 The
government subsidized study abroad for thousands of Japanese students
and imported foreign instructors, consultants, and experts. Jones (1980)
describes how, in these early years, half of the Ministry of Education’s
budget and two-thirds of the national public works budget was spent on
foreign experts. As Japan’s educational system expanded, these expensive
foreigners were quickly replaced with qualified locals. Japan exported its
educational system and philosophy to its colonies, Taiwan and Korea,
which benefited from the investment.
All three economies committed themselves to educational excellence
in the postwar era. By the mid-1960s, Japanese students were outscoring
Europeans and Americans on international standardized tests of math-
ematics and science. By the 1980s, Japan graduated more than twice as
many engineers per capita as the United States, and its ordinary workers
had dramatically more competence in science and math than their Amer-
ican counterparts (Rohlen 1983). Today, standardized tests of science and
mathematics skills suggest that the average levels of skill mastery and
the fraction of students scoring one standard deviation above the OECD
average are extremely high in Japan, South Korea, and Taiwan (Hanushek
and Woessman 2015). In chapter 7, Kyoji Fukao documents the almost
embarrassing degree to which the basic skills of the average Japanese
worker exceed those of his or her American counterpart.

4. Branstetter (2017) also highlights the importance of Japan’s early investment in educa-
tion. This section draws on that earlier study.

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 77

 This investment in human capital enabled these nations to converge
technologically with the West. In fact, Romalis (2004) shows that stan-
dard models of international trade can largely explain the rise of East
Asian nations as exporters of knowledge-intensive goods once this human
capital investment is taken into account. But this national human capital
strategy had its limits. Japanese universities did well in providing a basic
engineering education to technical graduates, but they underperformed in
graduate education, which was relatively underfunded and institutionally
neglected. In the early 1980s, the United States spent six times as much as
Japan on doctoral-level training and generated six times as many PhD grad-
uates in the sciences and engineering (Okimoto and Saxonhouse 1987).
In the United States, many graduate students found employment in
industries where the top corporate R&D labs were effectively led by PhD
recipients. In Japan, in contrast, advanced training for corporate R&D
personnel took place mostly within the larger, technology-intensive firms.
While this very different approach to training an industrial R&D work-
force clearly did not prevent leading Japanese firms from quickly reaching
the technology frontier, it reduced labor mobility across firms and made
it more difficult for high-tech startups in Japan to acquire highly skilled
researchers, strengthening the pro-incumbent bias of the Japanese innova-
tion system (see box 4.1). Japanese academic culture and bureaucratic prac-
tices further constrained basic research funding from reaching the most
promising young scientists. Academic salaries lagged behind advances in
scientific productivity.
The connection between academic research and frontier innova-
tion in the United States has already been the subject of a large literature
(Mansfield 1991; Rosenberg and Nelson 1994; Zucker, Darby, and Brewer
1998). The research demonstrates that nations lacking universities in
which faculty and graduate students are pursuing frontier science lose out
in the global competition for venture capital investment and radical new
inventions. Conversely, American universities excelling in global science
lie at the core of innovation ecosystems that attract abundant inflows of
early-stage venture capital investment. In Japan—and in South Korea and
Taiwan—human capital policy choices strengthened incremental research
capabilities while constraining more fundamental research capabilities.

Intellectual Property
Prewar Japan imported the patent system of Bismarck-era Germany in the
19th century, which remained the basis of Japanese intellectual property
policy throughout the 20th century. The Japanese system, like the German
one, provided limited intellectual property protection for even modest,

78 SUSTAINING ECONOMIC GROWTH IN ASIA

incremental inventions. However, until the late 1990s, it did not protect
fundamental innovations as well as the German system did (Ordover 1991).
Until the 1990s Japanese firms could include only one claim per
patent application, and Japanese courts had no “doctrine of equivalents,”
under which minor variations on a patented invention could be ruled as
violating the original patent. So, to protect their inventions, firms were
required to file a large number of patent applications describing all the
features and possible permutations of their patents. Even a well-financed
large firm would often fail to effectively prevent rivals from using a closely
related technology, since Japanese courts interpreted claims broadly when
determining whether an invention was sufficiently novel to merit a patent
application, but they interpreted them narrowly when determining patent
infringement (Sakakibara and Branstetter 2001). Japan also required that
patent applications be published before patents were granted, and, through
the early 1990s, it allowed interested parties to contest the granting of a
patent prior to issue. Finally, until the 1990s, Japanese courts almost never
awarded large damages to inventors whose intellectual property had been
infringed.
Small firms with brilliant ideas but limited financial resources were
especially disadvantaged under this patent system. Even for larger firms,
the weakness and narrowness of the Japanese patent system undermined its
usefulness as a mechanism for appropriating the returns to R&D. Instead,
firms in postwar Japan tended to rely on their manufacturing capabili-
ties, brand names, and quickness to market as the primary mechanisms
for appropriating the returns to R&D investment. This, in turn, reinforced
the Japanese tendency to focus on incremental innovation, process engi-
neering, and inventions that were close to commercialization.5

Exchange Rates
Japan began reintegrating into the postwar global economy with a high
level of human capital, impressive manufacturing skills, and—given the
extent of wartime devastation—much lower factor costs than Western
rivals. Japan’s economy quickly recovered, but low inflation, rapid produc-
tivity convergence, and participation in the postwar system of fixed
exchange rates and limited capital flows kept Japan’s cost of production
relatively low when measured in Western currencies, especially the US
dollar (Ito 1992). Despite episodic protectionism, Japan generally found
open Western markets for its ever-expanding range of export goods. As

5. Branstetter (2017) also stressed the role of relatively weak, narrow patents in shaping the
Japanese style of innovation focused on relatively incremental invention.

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 79

Japan’s real and nominal exchange rates increasingly failed to fully reflect
its degree of productivity convergence with the Western world, Japanese
firms found that if they could close the quality gap with their Western
rivals, they would often outcompete them in the marketplace, thanks to
their cost advantage.
A long period under an undervalued exchange rate moved Japanese firms
away from fundamental innovation and toward incremental innovation. As
Japanese firms invested in incremental improvements in manufacturing
efficiency and quality, a consistently favorable exchange rate reinforced
the competitiveness of that production base. Even after the collapse of the
Bretton Woods system in the early 1970s, concerns about the impact of yen
appreciation led the Japanese government to limit exchange rate apprecia-
tion at every opportunity, while depreciation was generally welcomed. In
real terms, the yen remained much cheaper in the 1970s and 1980s than it
was for much of the 1990s.

Labor Market Institutions

An extensive literature exists on the “lifetime employment” system that
developed in postwar Japan. While some Western analyses sought to connect
the emergence of this system to traditional Japanese cultural values and
practices, in reality the system was a postwar arrangement designed to buy
labor peace and protect firms’ investments in worker skills. Prewar labor
markets were characterized by high labor turnover and mass layoffs during
downturns and recessions. The imperial government ruthlessly suppressed
efforts to form labor unions. When Japan was occupied by the New Deal
administration of Harry Truman, strong protections for labor were hard-
wired into the postwar legal regime and union organizers—many of them
radicalized by long prison terms and harsh treatment—were set free. Japan’s
labor movement sought to make up for lost time, and Japanese postwar
industrialists viewed frequent strikes and work stoppages as a threat to
their prosperity in a growing economy. Since they could no longer rely
on the secret police to suppress union organizers, they started bargaining
with them. Industrialists also realized that Japan’s industrial structure was
changing rapidly, educational institutions were not keeping up with these
changes, and effective adaptation would necessitate extensive investment in
workers’ skills—but firms needed some mechanism to ensure that workers
trained at the expense of company A would not just “jump ship” and put
those skills to work for company B.
The bargain hammered out between industrialists and the Japanese
labor movement in the 1950s resulted in what became known in English
as the lifetime employment system (Ito 1992). This system applied only to

80 SUSTAINING ECONOMIC GROWTH IN ASIA

full-time, male employees in Japan’s largest companies and covered them
through an early retirement age, forcing many of them to seek secondary
employment after they were “retired” as early as in their 50s. Still, the
system guaranteed blue- and white-collar employees employment security,
a labor representative’s presence on the corporate management team, and
profit sharing for all rank-and-file workers, not just senior executives. In
return, though, workers were expected to complete their entire career in the
firm in which they started. Firms generally promoted from within, wages
depended heavily on tenure within the firm, and any worker who moved
from one firm to another had to start at the bottom of the wage scale.
While this system helped insure workers against economic shocks, it
also drastically limited the flow of workers across firms and industries.
Furthermore, only the top firms really honored the system, which concen-
trated Japan’s top talent in the leading enterprises and drastically limited
the appeal of startups and new entrants. Any Japanese executive at a leading
firm who left to join a (risky) startup had to start all over at the bottom of
the wage scale in the (likely) event that the startup failed. The evolution of
Japanese labor market institutions thus left the nation with a pronounced
pro-incumbent bias. As Kyoji Fukao documents in chapter 7 of this volume,
even in today’s Japan, striking productivity gaps remain between large and
small firms, and Fukao concludes that Japan’s advanced human capital
remains disproportionately concentrated in large firms.

Financial Markets
Japan’s industrial evolution required access not just to skilled workers but
also to capital. An extensive literature describes the evolution of Japan’s
highly regulated postwar capital markets (Hoshi and Patrick 2000, Hoshi
and Kashyap 2001) from wartime capital controls and industrial planning.
Postwar regulatory barriers sharply limited the issuance of stocks and
bonds into the 1980s.
For most firms, it was simply not practical to obtain significant external
financing through direct sales of equity or bonds to investors. Instead,
Japan’s postwar financial system was dominated by a highly regulated
banking cartel. This cartel operated under deposit and commercial lending
rates that were set by government fiat rather than the supply and demand
of financial capital. To access sufficient external finance in Japan’s booming
postwar economy, most large Japanese manufacturing firms forged a close
connection with one of the main commercial banks.
This system provided large amounts of financial capital to a limited
set of “insider” firms at reasonably low interest rates (Hoshi, Kashyap, and
Scharfstein 1991; Hoshi and Kashyap 2001). New firms outside this system

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 81

were at a significant disadvantage in the postwar race for external finance.
Stringent regulation of the stock and bond markets effectively precluded the
rise of robust venture capital markets or junk bond markets, which would
have incentivized investment in risky projects by offering high interest rates
to investors with sufficiently high-risk tolerance. Financial deregulation in
Japan proceeded in stages. Important moves in the early 1980s enabled blue
chip manufacturers to decouple from their historical bank relationships
and issue direct stock and equity. But other important stages of the finan-
cial deregulation process were delayed until the 1990s and 2000s, by which
time the Japanese economy had been stuck in slow growth for more than a
decade (Hoshi and Patrick 2000). While this may have been unintended, the
historical evolution of Japan’s financial market policies tended to reinforce
the pro-incumbent bias in its innovation system.

Competition
Japan’s postwar development is often mischaracterized as “export-led.” In
fact, Japan did not start running consistent and significant current account
surpluses until the 1970s, and these surpluses remained limited in size rela-
tive to GDP until the first half of the 1980s. Instead, Japanese growth was
driven predominantly by domestic sources of demand.
One of the first legal changes the Japanese government instituted when
it regained full sovereignty from its American occupiers was to substantially
weaken the Anti-Monopoly Law it inherited from the New Deal govern-
ment of Harry Truman (Eads and Yamamura 1987, Uekusa 1987, Weinstein
1995). The Japanese government regularly created legal cartels when indus-
tries experienced cyclical downturns, with the government itself playing
the role of cartel enforcer (Uekusa 1987, Weinstein 1995). The market was
effectively divided up on the basis of pre-recession market shares, a policy
that—to the extent that it worked—benefitted established incumbents.
The postwar Japanese distribution system was, by contrast, fragmented
and dominated by small-scale, “mom and pop” style stores. These small
establishments could not resist efforts by much larger manufacturers to tie
them into exclusive dealing contracts. Foreign consumer goods sellers in
the 1980s regularly cited the difficulty of breaking into the Japanese retail/
distribution system as one of the most important barriers to their expan-
sion in the Japanese market (Flath 2002). Of course, this barrier was even
greater for small, startup Japanese firms trying to enter or expand into the
market for consumer goods.
Strong alliances between traditional business partners in Japan also
impeded entry of both foreign and domestic firms. So-called keiretsu
networks dominated the market for industrial goods, at least in key sectors
like automobiles. In Japan’s complicated systems of interfirm alliances
82 SUSTAINING ECONOMIC GROWTH IN ASIA
(Gerlach 1992), would-be entrants had no recourse given the relatively
weak antitrust law and practice.
Finally, in much of Japan’s service economy, heavy-handed govern-
ment regulation strictly restricted entry and competition well into the
1980s. This was true in banking, securities trading, insurance, construc-
tion, telecommunications, intercity freight transportation, and airlines. As
elsewhere in the industrialized world, regulatory capture limited competi-
tion and ensured rents for incumbents, at the expense of consumers and
would-be entrants. All of these factors tended to reinforce the pro-incum-
bent bias in the Japanese innovation system.

Decline in Japanese Industrial Research Productivity

Figure 4.1 shows the differing industrial R&D productivity trends for firms
in the electronics sector and other manufacturing firms. The figure graphs
the estimated coefficients on year dummy variables obtained by regres-
sion of a “patent production function,” as described in equation (3) in
Branstetter and Nakamura (2003):

nit  rit    t Tt   c Dic   it

t c

where, nit is innovation, rit is the firm’s own R&D investment, the Ds are
dummy variables to control for differences in the propensity to generate
new knowledge across technological fields (indicated by the subscript c),
the Ts are year dummies, and e is an error term. Branstetter and Nakamura
proxy innovation with patents and present the patterns sketched out by
their year dummies as a statistical description of trends in the productivity
of industrial R&D. The results reproduced in figure 4.1 report the coeffi-
cients of time dummies obtained from a version of equation (3) that incor-
porates firm-level fixed effects. Results are estimated separately for firms
in the electronics sector and other manufacturing firms. Measured R&D
productivity of other Japanese firms rose sharply through the 1980s, then
gradually declined through the late 1990s. R&D productivity in the elec-
tronics sector continued to increase through the mid-1990s, but began to
reverse in the late 1990s.6 These results are qualitatively robust to the use of
data on Japanese patent applications, to the weighting of US patent grants
by forward citations, and to the use of firm fixed effects.
These developments were linked to important structural changes in the
nature of private sector R&D in Japan (Branstetter and Nakamura 2003).

6. Several other scholars have also highlighted adverse shifts in the relative performance of
the Japanese innovation system since the early 1990s (Goto 2000, Goto and Odagiri 2003,
Nagaoka 2007).

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 83

 Figure 4.1 Trends in Japanese R&D productivity

R&D productivity increase relative to base year (1981)

1.2

1.0

0.8

0.6

0.4

0.2

Electronics firms
0 Firms in other manufacturing
industries
–0.2
1982 1984 1986 1988 1990 1992 1994 1996

Source: Branstetter and Nakamura (2003).

As Japanese firms reached the natural limits of an R&D strategy that was
focused on adapting foreign technology—a focus on process and incre-
mental innovation and convergence with a technology frontier rather than
an effort to go beyond it—they were forced to change their approach to
R&D. They had to build larger central R&D laboratories with more ambi-
tious agendas and focus on more fundamental, science-based research.
Unfortunately, the Japanese innovation system was itself still insufficiently
developed to make this new emphasis successful and, at least initially, the
structural changes within Japanese firms were unable to forestall a decline,
or, at best, a stagnation, in research productivity.
In the early 1990s, Japanese stock and real estate markets plunged,
inducing a significant and persistent macroeconomic slowdown. Five years
into the decade, the internet revolution was well under way in the United
States, but Japanese firms were conspicuously absent from the group of
firms—many of them new entrants—that were introducing fundamental
new innovations in IT. Even in less technologically dynamic sectors like
autos, Japanese firms seemed to be losing ground in the 1990s.
By the mid-1990s industrialists and policymakers in Japan increasingly
recognized these shortcomings. The so-called Science and Technology Basic
Law, implemented in late 1995, clearly articulated the need for fundamental
reform of the nation’s innovation system. Since the passage of this law, Japan
has not only strengthened its intellectual property system (Sakakibara and
Branstetter 2001) but also fundamentally reformed its national university

84 SUSTAINING ECONOMIC GROWTH IN ASIA

system (Amano and Poole 2005) and developed a venture capital market.
The so-called big bang financial liberalization of the mid-1990s (key compo-
nents of which were not fully completed until the 2000s) also included
ways to make it easier for innovative firms to motivate employees with
stock options. An effort was also made to make Japan’s product markets
more open to competition. Restrictions on large-scale retailers in Japan,
which had kept the balance of power between manufacturers and distribu-
tors in favor of the former, were significantly relaxed in the 1990s, and the
importance of keiretsu linkages between suppliers and assemblers in key
industries declined. Unfortunately, these reforms have neither accelerated
Japanese productivity growth at the macro level nor revived the fortunes of
Japan’s once-vaunted electronics industry.7

South Korea
Korea spent roughly the first half of the 20th century as a colony of the
Japanese Empire. Shortly after achieving its independence, it was devastated
by civil war. Economic upheaval and political turmoil followed in the wake
of this bloody conflict. The South Korean economic miracle really dates
from the military coup that brought General Park Chung-hee to power
in the early 1960s. For better and for worse, General Park left a stamp on
Korea’s political economy that endures to this day (Yi 2006).

Legacy of Park Chung-hee

The democratic constitution that occupying authorities imposed on
postwar Japan tightly constrained the scale and direction of industrial
policy (Calder 1988). A free press, open elections, and the requirement that
government policy be based on laws passed by the Diet meant that Japanese
industrial policymakers had to stay in line with popular opinion.
In striking contrast, General Park faced few constraints on his exercise
of authority. He was not answerable to a constitution, an opposition, or
a free press. Park was not limited by bureaucratic infighting because all
bureaucracies were directly answerable to him. One of Park’s first economic
policy moves was to nationalize Korea’s banks—a move that would have
been unthinkable in postwar Japan—making its financial markets an arm
of state policy. The general thus quickly acquired a degree of authority over
the allocation of resources in the Korean economy of which Japan’s indus-
trial policy bureaucrats could have only dreamed. He did not hesitate to
use it (Clifford 1997).

7. For an in-depth analysis of the loss of competitiveness of Japan’s IT sector, see Arora,
Branstetter, and Drev (2013).

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 85

 The breathtakingly rapid growth of Korea during this period—even
more rapid than that of Japan—attests to the fact that the policy mix
pursued during this period had many positive elements. But one of the
most important legacies of this period was the rise of the chaebols—Korea’s
industrial conglomerates (Joh 2014). Park used his command over Korea’s
resources to build up globally competitive national champions. Low-cost
capital, scarce foreign exchange, and industrial licenses were awarded to
conglomerates headed by business leaders who became his close allies
(Clifford 1997). Of course, Park’s favoritism tilted the playing field against
all other players and left South Korea with a degree of industrial concen-
tration that was arguably the highest in the industrialized world.
In the late 1990s, when the financial crisis hit, Korean public opinion
turned against the chaebols; their “reckless” borrowing and investing
was partly blamed for the crisis.8 Influential commentators criticized the
concentration of economic power within the large groups and the collu-
sive ties between these groups and several successive South Korean govern-
ments. The gradual political liberalization and democratization of South
Korea encouraged politicians to reform the chaebols (Haggard, Lim, and
Kim 2003). Since the crisis of 1997, nearly every South Korean president has
promised to check the power of the chaebol and encourage entrepreneur-
ship, but the results have been disappointing. Figures 4.2 and 4.3 depict the
concentration of US patent grants and R&D spending, respectively, across
Korean firms. A shockingly high share of both is concentrated in a single
chaebol—the Samsung Group.
South Koreans refer to their nation, somewhat ruefully, as the Republic
of Samsung (Pesek 2015). In addition to dominating patenting and R&D
expenditure in South Korea, the Samsung Group accounts for a high frac-
tion of South Korean exports and represents a disproportionate fraction
of the value of the South Korean stock exchange.9 Former president Lee
Myung-bak pardoned the group’s former chairman, Lee Kun-hee, who
was convicted of financial wrongdoing and tax evasion. The justification
was that Samsung was so important to the South Korean economy that it

8. The five largest chaebols—Samsung, LG, Daewoo, Hyundai, and SK—accounted for 20
percent of all outstanding debt and 75 percent of all new borrowing in 1998 (Economist, 1999,
“Survey of South Korea,” originally appeared June 3, 1995 issue, pp. 10–17).
9. The Samsung Group accounts for more than one-fifth of the value of the South Korean
stock exchange. See Steven Borowiec and Paresh Dave, “South Koreans live in ‘the Republic of
Samsung,’ where the Galaxy Note 7 crisis feels personal,” Los Angeles Times, October 11, 2016,
www.latimes.com/business/technology/la-fi-tn-samsung-note-7-korea-20161011-snap-
story.html; Zahra Ullah, “How Samsung dominates South Korea’s economy,” CNN.com,
February 17, 2017, http://money.cnn.com/2017/02/17/technology/samsung-south-korea-
daily-life/index.html.

86 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 4.2 Concentration of US patent grants to South
Korean firms

share of patents awarded to each group of firm(s)

1.0
.9
.8
.7
.6
.5
.4
.3 Samsung Electronics
.2 Top 10
Top 30
.1
Top 50
0
1991 1993 1995 1997 1999 2001 2003 2005 2007

Note: This figure measures the fraction of total US patent grants awarded to
South Korean inventors that are accounted for by Samsung Electronics and
by firms affiliated with the top 10, top 30, and top 50 chaebols. Patents are
assigned a date based on year of application. South Korean patent data also
indicate a significant concentration of patenting in the portfolios of these
chaebols. Samsung Electronics alone accounts for more than 50 percent of
all US patent grants received by South Korean firms.
Source: US Patent and Trademark Office.

would be destabilizing to jail the head of the firm, even if he were guilty of
crimes. Lee Kun-hee has since transferred effective control of Samsung to
his son, Lee Jae-yong. The younger Lee was recently convicted of bribing
Lee Myung-bak’s successor (and General Park’s daughter), Park Gyeun-hye,
who was the first South Korean president to be impeached and removed
from office. However, that sentence was suspended in February 2018, and
many observers expect the younger Lee will escape the legal consequences of
his crime. The tangled state of South Korean politics attests to the extreme
degree of pro-incumbent bias that Park’s industrial policies generated.

Policies and Pro-Incremental Bias in South Korea

Once Park took over, the Korean educational system expanded in reach
and quality even more rapidly than did the educational system of its former
colonial master. Younger Koreans are now among the most highly educated
people in the industrial world, and Korean students regularly outperform
Japanese students on the sorts of internationally comparable standardized
tests in which Japanese students regularly outperform Americans (OECD
2016).

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 87

 Figure 4.3 Concentration of R&D spending among South
Korean firms

share of R&D by each group of firm(s)

1.0
.9
.8
.7
.6
.5
.4
.3 Samsung Electronics
.2 Top 10
Top 30
.1 Top 50
0
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009

Note: This figure uses data on total R&D expenditure, as reported by the
publicly traded firms listed on South Korean equity markets, to calculate the
fraction of annual expenditure accounted for by Samsung Electronics and
the top 10, top 30, and top 50 chaebols. In 2010, Samsung Electronics
accounted for more than 50 percent of total R&D expenditure reported
by these firms.
Source: Korean Listed Companies Association (KLCA).

But South Korea’s educational system also has some of the same weak-
nesses as Japan’s. Like their Japanese counterparts, South Korea’s elite
universities are better known for the intensive competition for admission
than for the quality of research or even instruction. The same institu-
tional neglect of graduate education, lack of research funding, unwilling-
ness of the education ministry or the universities to concentrate research
resources in the hands of innovative scholars rather than senior scholars,
and indifference to immigrants have historically limited the research capa-
bilities of South Korean universities.10 While leading chaebols can set up
R&D facilities in Silicon Valley, the relative weakness of Korean universi-
ties drastically limits the access of South Korean startups to world-class
faculty entrepreneurs and graduate students capable of translating labora-
tory breakthroughs into new products and services.
South Korea inherited the Japanese patent system and, even after inde-
pendence, the evolution of the Korean system followed that of Japan’s.
After a long period of weak and narrow protection with indifferent enforce-
ment by courts, reforms in the 1980s led to stronger patents and enforce-
ment, and patent practices began converging with Western standards in

10. Only one university from South Korea is on the list of “Top 200 Universities in the
World,” www.4icu.org/top-universities-world.

88 SUSTAINING ECONOMIC GROWTH IN ASIA

the 1990s (La Croix and Kawaura 1996). Nevertheless, the persistence of
weak protection during Korea’s high growth period may have reinforced a
pro-incremental bias in Korea, as it did in Japan.
Prior to the mid-1980s, South Korean firms received few domestic
patents and almost none from international patent agencies like the US
Patent and Trademark Office. After the mid-1980s, the number of interna-
tional patents granted to South Korean firms exploded, seemingly marking
a surprisingly sudden transition by South Korean firms from imitators
to innovators. Several scholars argue that this patent explosion can be ex-
plained by the domestic patent reform, which increased the incentives for
R&D. However, our research (Branstetter and Kwon 2018) suggests a dif-
ferent interpretation. Firm-level patenting and R&D data reveal little im-
mediate growth in chemicals and pharmaceutical patenting, even though
domestic patent law strengthened the most in these industries. In contrast,
R&D and patenting rapidly expanded in exporting firms for which foreign
markets were more important than the domestic market potentially affect-
ed by the change in Korean patent law.
Branstetter and Kwon (2018) also point out that the relative exchange
rates of the yen and the won sharply changed in the mid-1980s. The yen
appreciated sharply against the US dollar after 1985. The won, still pegged
to the dollar, effectively depreciated sharply against the yen. Korea-based
producers suddenly gained a large cost advantage over Japanese rivals. By
investing in process engineering and other efforts to close the quality gap
vis-à-vis Japanese firms, they stood to outsell their rivals in the United
States and Western Europe. It was the tantalizing prospect of outselling
Japanese firms in the world’s largest markets, rather than marginally better
patent protection in the small South Korean market, that induced a South
Korean R&D and patenting explosion.
South Korea underpriced Japanese rivals in the 1980s the same way
Japanese firms undercut American and European rivals in the 1970s and
early 1980s. With cheaper factor costs and a favorable exchange rate, South
Korean firms could do to Japan what Japan had done to the industrial
West. Throughout the 1990s, low-cost Korean producers of commodity
semiconductor products like DRAM chips were edging out their Japanese
rivals in global markets.11 By the end of the decade, Korean firms were deci-
sively in the lead, and the once vaunted Japanese semiconductor industry
was a shadow of its former self (Cho, Kim, and Rhee 1998). Japanese firms
whose R&D investment strategies had been predicated on their retaining a

11. In 1993, Samsung produced more DRAM chips than any other firm. In the same year,
Hyundai and LG (formerly Lucky GoldStar) ranked ninth and tenth on the list of top global
DRAM producers (Cho, Kim, and Rhee 1998).

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 89

cost advantage vis-à-vis foreign competitors were increasingly “squeezed”
between the rising quality and manufacturing capability of lower-cost
Asian nations, like South Korea, and the persistently superior product
innovation capabilities of their American rivals (Lee 2013).
The strength of the Korean won ebbed and flowed over the next two
decades. The financial crisis of 1997–98 led to a collapse in the value of the
won, a long period of relative won weakness, and a revival of South Korean
firms’ innovation strategies built around relatively low production costs,
scale, and strength in process engineering. However, as Chinese manu-
facturing capabilities grew, Korean production sites found it increasingly
difficult to compete on price with those based in a country at a fundamen-
tally lower level of economic development. In the 2000s, Korean firms felt
the same kind of “squeeze” as they had put on their Japanese rivals in the
prior decade (Lee 2013).
This chapter’s discussant, Jong-Wha Lee, notes that a highly competitive
exchange rate could help a late-developing country transition from “pure
imitator” to “incremental innovator.” But South Korean firms are now
struggling to move from incremental innovation to more fundamental in-
novation. They remain important exporters of consumer electronics goods
and electronic components, but are rarely the source of major product in-
novations. Firms located in other countries are leading the development
of artificial intelligence (AI) and machine learning technologies. Similarly,
South Korea remains a successful exporter of autos, but the major inno-
vations in that sector, like self-driving cars and electric vehicles, are being
pioneered elsewhere.
Samsung’s prominence in the global smartphone industry is perhaps
emblematic of the strengths and weaknesses of South Korea’s innova-
tion system. Shortly after the introduction of the iPhone, a revolutionary
innovation by Apple, Samsung introduced a very similar product based
on the Android operating system and web platform created by Google.
Samsung’s products were so similar that Apple launched a global round of
patent infringement lawsuits. In the end, these lawsuits did not result in
damage judgments large enough to displace Samsung from the market,12
and Samsung’s lower-priced products have sold in large volumes around
the world, boosting Samsung’s profits.13

12. Seth Fiegerman, “Supreme Court sides with Samsung in Apple patent case,” CNN.
com, December 6, 2016, http://money.cnn.com/2016/12/06/technology/samsung-apple-
supreme-court/index.html.
13. Lance Whitney, “Samsung selling more of its cheaper smartphones,” Cnet.com, November
2, 2015, www.cnet.com/news/samsung-selling-more-phones-at-cheaper-prices-says-new-re-
search.

90 SUSTAINING ECONOMIC GROWTH IN ASIA

 Samsung’s global market share remains high in terms of handsets
sold, but its share of global profits earned from smartphones declined over
the three years ending in 2016.14 Other Asia-based electronics firms began
manufacturing even cheaper smartphones, also based on the open Android
system. These lower-cost firms took market share away from Samsung,
which had to respond by cutting prices, since it was offering little in the way
of product innovation that lower-cost Asian rivals could not quickly and
closely match.15 The real innovation that made Samsung phones so attrac-
tive—the Android platform—was not Samsung’s innovation, but Google’s,
which maximized its revenues by offering the platform to as many manu-
facturers as possible, including low-cost Chinese producers like Huawei,
Oppo, and Xiaomi. The other major revenue driver for Samsung is manu-
facture of commodity components like DRAM chips. South Korea’s inno-
vative strengths thus appear to be concentrated on the “incremental” side
of the innovation spectrum and overwhelmingly concentrated in a small
number of chaebols whose future is uncertain (IMF 2014). South Korean
policy choices may have reinforced these pro-incremental and pro-incum-
bent biases.

Taiwan
Like Korea, Taiwan was a colony of the Japanese Empire from 1895 until
the defeat of Imperial Japan in 1945. While brutal, exploitative, and author-
itarian in many ways, Japanese colonial rule was markedly less harsh in
Taiwan than in Korea. The Japanese left Taiwan with the physical and insti-
tutional infrastructure of a modern economy. When it was ceded to Chiang
Kai-shek’s Republic of China by the Allied Powers at the end of World
War II, it was the richest province in his nation by a large margin. Initially
welcomed by the Taiwanese, Chiang’s army and provincial government
quickly demonstrated a remarkable combination of corruption, brutality,
and incompetence. A popular uprising against mainland authority in
1947 was brutally suppressed, and when Chiang Kai-shek relocated his
Nationalist government to Taiwan in 1949, he presided over an authori-
tarian regime characterized by martial law, systematic oppression of any

14. Samsung’s share of global smartphone shipments decreased after 2013. However, the
share was still over 20 percent in 2016 (data from Statista.com, www.statista.com/statis-
tics/271492/global-market-share-held-by-leading-smartphone-vendors/).
15. The market shares of Chinese phone makers such as Huawei, Xiaomi, Oppo, and Vivo
have been increasing. See Scott Cendrowski, “How China’s Smartphone ‘Big Four’ Are
Fighting for Global Customers,” Fortune, January 25, 2017, http://fortune.com/2017/01/24/
china-smartphones-oppo-vivo-huawei-xiaomi.

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 91

political dissent, and concentration of economic policymaking authority in
his own hands.
Taiwan’s experience was similar to South Korea’s in several ways.
Taiwan inherited an educational system that downplayed research and
graduate education and an intellectual property system that offered, at
best, narrow protection that was weakly enforced. The much lower factor
prices in Taiwan fostered process-oriented, cost-reducing, and incremental
R&D. The sharp appreciation of the Japanese yen in the mid-1980s gifted
Taiwanese producers with the opportunity to undercut their traditional
Japanese rivals in the world’s most important export markets. Just as
the South Koreans, Taiwanese firms responded to this opportunity with
a surge of investment in R&D, a surge in international patents, and an
export boom.
However, the direction in which the Chiang dynasty moved Taiwan
differed in important ways from that of Park’s South Korea. First, Chiang
Kai-shek’s economic planners came to believe that the devastating bouts of
inflation that wracked the Chinese mainland were one of the chief reasons
his Nationalist Party lost the Chinese civil war to the Communists. To avoid
repeating this fatal error, the technocrats in charge of monetary and finan-
cial policy adopted the so-called high interest rate policy (Wade 1990). This
policy increased saving because it rewarded Taiwanese savers for placing
funds in the financial system. Because all firms, even large ones favored by
the government-run banking system, paid relatively high interest rates, the
financial playing field was never tilted toward favored conglomerates to
the same extent as it was in South Korea.
Taiwanese authorities also tolerated the “gray market” for capital that
leaked out of the formal banking system to fund highly profitable enter-
prises that could not secure formal bank loans. In fact, the authorities used
the interest rates charged in this gray market as a benchmark for setting
official interest rates. Smaller firms outside favored corporate groups
and sectors had much easier access to capital than was possible in South
Korea. While some degree of official favoritism existed, the government of
Taiwan was not nearly as attached to the idea of building national cham-
pions as was the government of South Korea, and Taiwan’s level of indus-
trial concentration was much lower. As the Taiwanese economy matured,
it came to host a fairly robust domestic venture capital market, in which
local firms were able to find sufficient financing to enter even reasonably
complex industries. The pro-incumbent bias that was so overwhelming in
South Korea was not as significant in Taiwan.
Taiwan’s industrial structure was different also because of the entre-
preneurial character of Taiwanese managers, and the greater openness and

92 SUSTAINING ECONOMIC GROWTH IN ASIA

flexibility of Taiwanese markets for electronic products, labor, and capital.
The Taiwanese electronics industry, which eventually became the main-
stay of Taiwanese manufacturing, is characterized by a very high degree
of entry and exit (Aw, Chen, and Roberts 2001). The extensive government
involvement in Taiwan also helped differentiate the industrial structure. In
Japan, government subsidization of R&D was quite limited; firms were left
to fend for themselves. Only larger enterprises had the private resources
to make large, risky investments in adopting and modifying (or creating)
technology and to construct the mechanisms to appropriate the returns
from such investments. In Taiwan, the state bore a considerable fraction
of the fixed costs of technology adoption and refinement and provided
the results to local firms on favorable terms. These efforts had a strong
industry focus and, to some extent, a pro-incremental bias.
It is perhaps inevitable that an economy like Taiwan, which is substan-
tially smaller in geographic size and population than South Korea, would be
less diversified than its larger neighbor. South Korea is a significant global
player in capital-intensive sectors such as autos, shipbuilding, and steel,
as well as electronics. In contrast, Taiwan’s exports and industrial struc-
ture are more concentrated in electronics and information technology (IT).
The promotion of the electronics industry has been a key priority of the
Taiwanese government for decades. The Industrial Technology Research
Institute (ITRI), established in 1973 to increase national R&D capacity,
accounted for roughly 25 percent of the government’s total R&D expendi-
tures by 1987—a research budget of some US$215 million (Wade 1990, 98).
In 1974, the Electronic Research and Service Organization (ERSO)
was established under ITRI to recruit foreign partners to develop and
commercialize semiconductor fabrication technology. The government
was heavily involved in the 1986 formation of the highly successful Taiwan
Semiconductor Manufacturing Corporation (TSMC), initially a joint
venture between Philips and several domestic public and private firms.
ERSO was deeply involved in international technology transfer through
the 1980s, often identifying key foreign technologies itself, then subli-
censing them to local firms (Wade 1990, 103–107).16 Thus, the Taiwanese
government was very successful in directing and subsidizing substantial
international technology transfer in this industry.17 Given the special

16. This practice of sublicensing could mean that our plant-level data fail to capture all
“effective” foreign technology purchases.
17. In contrast, the Japanese government began the postwar period with a policy of limiting
foreign technology purchases (Wakasugi 1997). As the Japanese economy developed and
the potential benefits of technology imports became clear to government authorities, these
limits were progressively relaxed. Wakasugi (1997) has argued that the government allowed

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 93

 Figure 4.4 Concentration of US patent grants to Taiwanese firms
in electronics hardware, 1981–2014

proportion of patents
0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014

Note: This graph uses data from the US Patent and Trademark Office to calculate
the share of patents awarded to Taiwanese inventors that are in patent classes
associated with electronics and IT hardware. Software-related patents are omitted.
Since the late 1990s, patents in these classes have accounted for more than 60
percent of the total. Taiwanese patents and data on Taiwanese R&D spending also
point to a strong concentration of activity in the electronics industry.
Source: Branstetter and Chen (2006).

favors showered on electronics, it is not surprising that, as Taiwan’s indus-

trial structure shifted from labor-intensive goods to more knowledge-
intensive goods, it became increasingly concentrated in that sector. Figure
4.4 illustrates the high concentration of Taiwanese firms’ US patents in
classes associated with electronics hardware.
ITRI/ERSO was not the only source of Taiwanese government largesse
for electronics relative to other industries. Taiwanese industrial planners
sought to build an electronics-based industrial park south of the capital,
Taipei, in Hsinchu, that would become the island’s answer to Silicon Valley.
First proposed in 1976, the park was initially opened in 1980 and grew
rapidly over the next two decades. The government made massive invest-
ments in infrastructure, offered special financing to new firms moving into
the park, and even constructed American-style schools for families of expa-
triate Taiwanese it hoped to lure back to Taiwan from the United States.

only “favored” firms to bid for foreign technology to prevent competition among potential
Japanese purchasers from driving up the price of the license. This may have lowered the price
of foreign technology for these firms, but it effectively increased the price (to infinity) for the
firms that were not favored, and, in Wakasugi’s view, it is unlikely that this intervention was
welfare-enhancing.

94 SUSTAINING ECONOMIC GROWTH IN ASIA

Taiwan’s drive to build the Hsinchu Science Park kicked into high gear just
as rising living standards and a shift to democracy were making Taiwan
more attractive to expatriates seeking career advancement. As American
high-tech companies began to offshore their manufacturing to Asia,
Taiwanese expatriate engineers facing weak career prospects (or layoffs) in
the United States were increasingly inclined to move “home” to Hsinchu.
By the mid-1990s, social scientists and international investors were hailing
the success of the park.18
The limitations of Hsinchu’s success, and of the more general nexus of
policies that nurtured Taiwan’s high-tech industries, however, have become
more apparent over the years. Relatively small firms that historically relied
on ITRI/ERSO to do the “heavy lifting” of identifying key foreign tech-
nologies and underwriting assimilative R&D were naturally focused on
implementing and refining technologies developed by others, rather than
breaking new ground. The new firms taking shape in Hsinchu were more
focused on the efficient contract manufacturing of Silicon Valley designs
than creating “breakthrough” new products. Like a “helicopter parent”
who does too much to position their children for success, the institu-
tions and policies designed to nurture Taiwan’s high-tech sector may have
inadvertently undermined the incentives for these firms to develop within
themselves more fundamental capabilities for research. The industry that
took root in Hsinchu was focused on hardware, not software, and incre-
mental, rather than transformative, R&D.
Despite government fears of excessive economic dependence on main-
land China, Taiwanese electronics firms could not resist the attraction of
China’s low costs and apparently limitless skilled labor. The controversial
Chen Shui-bian administration in Taiwan liberalized cross-straits invest-
ment, and a tidal wave of investment ensued. Branstetter et al. (2017) analyze
the impact of this relocation of manufacturing on the R&D of Taiwanese
firms in the technological domains related to the products and components
that were offshored. They find a statistically robust negative effect. Taiwan’s
trade flows to and from China first rose sharply, then fell substantially as an
increasingly large fraction of the entire value chain moved to China.
Taiwan’s leading contract manufacturing firms have managed to stay
one step ahead of their “indigenous” Chinese competitors on the main-
land, but the Taiwanese economy has slowed sharply since the go-go 1990s,
incomes have stagnated, and the facts on the ground no longer support the
notion that Hsinchu Science Park is going to match Silicon Valley. As was
the case in Japan and South Korea, it seems likely that Taiwanese govern-

18. See Saxenian and Hsu (2001) for a typical positive assessment of the venture.

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 95

ment policies reinforced and deepened the pro-incremental bias in Taiwan’s
innovation system. While these policies did not leave Taiwan with the same
pro-incumbent bias as its larger neighbors, they appear to have concentrat-
ed Taiwanese activity in sectors that no longer hold the most significant
technological opportunities. Like Japan and South Korea, Taiwan seems in-
creasingly squeezed between Silicon Valley and China. American firms con-
tinue to generate fundamental innovations Taiwan cannot match. China
continues to offer a mix of lower costs and rising skills that Taiwanese
manufactures cannot resist. For Taiwan, the increasingly tight economic
embrace of the Chinese mainland raises existential political questions about
the future of the island as a de facto independent state.

Conclusion
Japan, South Korea, and Taiwan have successfully climbed the development
ladder, yet, when measured by their own past performance, one cannot
help but be struck by the degree to which they have each fallen short of
what might have been reasonably expected a decade or two ago. Compared
with the United States, Japan’s overall total factor productivity has actu-
ally declined substantially over the past 20 years, even as it has become a
much more R&D-intensive economy than the United States (IMF 2017).
Despite emerging as the most R&D-intensive nation in the OECD, South
Korea’s level of productivity relative to the United States has essentially
remained the same over the past 20 years—stuck at around 60 percent of
the US aggregate TFP level.
The three countries have suffered, to varying degrees, from declining
R&D productivity because the national innovation systems designed for
their high growth periods are no longer as effective in their maturing econ-
omies. The pro-incremental and pro-incumbent biases that government
policies created or reinforced in these countries have made it difficult to
change these innovation systems. Other nations in the region, including
China, seeking to emulate the success of these three nations should avoid
reinforcing the same kinds of biases in their own emerging innovation
systems. The costs of not doing so appear to be high. Once these biases
penetrate the fabric of an economy’s innovation system, they are difficult
to eliminate.

References
Agrawal, A., and R. Henderson. 2003. Putting Patents in Context: Exploring Knowledge
Transfer from MIT. Management Science 48, no. 1: 44–60.
Amano, I., and G. Poole. 2005. The Japanese University in Crisis. Higher Education 50: 685–711.

96 SUSTAINING ECONOMIC GROWTH IN ASIA

Arora, A., L. Branstetter, and M. Drev. 2013. Going Soft: How the Rise of Software-Based In-
novation Led to the Decline of Japan’s IT Industry and the Resurgence of Silicon Valley.
Review of Economics and Statistics 95, no. 3: 757–75.
Aw, B., X. Chen, and M. Roberts. 2001. Firm-Level Evidence on Productivity Differentials,
Turnover, and Exports in Taiwanese Manufacturing. Journal of Development Economics 66:
51–86.
Branstetter, L. 2017. Intellectual Property Rights, Innovation, and Development: Is Asia Dif-
ferent? Millennial Asia 8, no 4: 5–25.
Branstetter, L., and J.-R. Chen. 2006. The Impact of Technology Transfer and R&D on Pro-
ductivity Growth in Taiwanese Industry: Microeconometric Analysis Using Plant and
Firm-Level Data. Journal of the Japanese and International Economies 20, no. 2: 177–92.
Branstetter, L., and N. Kwon. 2018. South Korea’s Transition from Imitator to Innovator: The Role of
External Demand Shocks. Forthcoming in Journal of the Japanese and International Economies.
Branstetter, L., and Y. Nakamura. 2003. Has Japan’s Innovative Capacity Declined? In Structur-
al Impediments to Growth in Japan, ed. A. Kashyap, J. Corbett, M. Blomström, and F. Hayashi.
University of Chicago Press and National Bureau of Economic Research.
Branstetter, L., J. Chen, B. Glennon, C. Yang, and N. Zolas. 2017. Does Offshoring Manufacturing
Harm Innovation in the Home Country? Evidence from Taiwan and China. Working Paper (July).
Carnegie Mellon University.
Buenstorf, G., and S. Klepper. 2009. Heritage and Agglomeration: The Akron Tire Cluster Revisited.
Economic Journal 119, no. 537: 705–33.
Calder, K. 1988. Crisis and Compensation: Public Policy and Political Stability in Japan. Princeton,
NJ: Princeton University Press.
Cho, D. S., D. J. Kim, and D. K. Rhee. 1998. Latecomer Strategies: Evidence from the Semicon-
ductor Industry in Japan and Korea. Organization Science 9, no. 4: 489–505.
Christensen, C. 1997. The Innovator’s Dilemma: When New Technology Causes Great Companies to
Fail. Cambridge, MA: Harvard University Press.
Clifford, M. 1997. Troubled Tiger: Businessmen, Bureaucrats, and Generals in South Korea. New
York: Routledge.
Cohen, W., and S. Klepper. 1996. Firm Size and the Nature of Innovation within Industries:
The Case of Process and Product R&D. Review of Economics and Statistics 78, no. 2: 232–43.
Denison, E., and W. Chung. 1976. How Japan’s Economy Grew So Fast. Washington: Brookings
Institution.
Eads, G., and K. Yamamura. 1987. The Future of Industrial Policy. In The Political Economy of
Japan, Volume 1: The Domestic Transformation, ed. K. Yamamura and Y. Yasukichi. Stanford,
CA: Stanford University Press.
Evenson, R., and Y. Kislev. 1976. A Stochastic Model of Applied Research. Journal of Political
Economy 84: 265–82.
Flath, D. 2002. Distribution Keiretsu, FDI, and Important Penetration in Japan. Working Paper. Co-
lumbia University.
Gerlach, M. 1992. Alliance Capitalism: The Social Organization of Japanese Business. Berkeley, CA:
University of California Press.
Gerschenkron, A. 1962. Economic Backwardness in Historical Perspective: A Book of Essays. Cam-
bridge, MA: Harvard University Press.
Goto, A. 2000. Japan’s National Innovation System: Current Status and Problems. Oxford Re-
view of Economic Policy 16, no. 2: 103–13.
Goto, A. and H. Odagiri, eds. 2003. Science-Based Industries. Tokyo: NTT Publishing.

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 97

Haggard, S., W. Lim, and E. Kim, eds. 2003. Economic Crisis and Corporate Restructuring in Korea:
Reforming the Chaebol. Cambridge: Cambridge University Press.
Hall, B., and R. Ziedonis. 2001. The Patent Paradox Revisited: An Empirical Study of Patent-
ing in the U.S. Semiconductor Industry, 1979–1995. RAND Journal of Economics 32, no. 1:
101–28.
Hanushek, E., and L. Woessman. 2015. The Knowledge Capital of Nations: Education and the Eco-
nomics of Growth. Cambridge, MA: MIT Press.
Harhoff, D., F. Scherer, and K. Vopel. 2003. Citations, Family Size, Opposition, and the Value
of Patent Rights. Research Policy 32, no. 8: 1343–63.
Helpman, E., and M. Trajtenberg. 1998a. A Time to Sow and a Time to Reap. In General Pur-
pose Technologies and Economic Growth, ed. E. Helpman. Cambridge, MA: MIT Press.
Helpman, E., and M. Trajtenberg. 1998b. The Diffusion of General Purpose Technologies.
In General Purpose Technologies and Economic Growth, ed. E. Helpman. Cambridge, MA: MIT
Press.
Hochberg, Y., C. Serrano, and R. Ziedonis. 2018. Patent Collateral, Investor Lending, and the
Market for Venture Lending. Journal of Financial Economics (forthcoming).
Hoshi, T., and A. Kashyap. 2001. Corporate Finance and Governance in Japan: The Road to the Fu-
ture. Cambridge, MA: MIT Press.
Hoshi, T., A. Kashyap, and D. Scharfstein. 1991. Corporate Structure, Liquidity, and Invest-
ment: Evidence from Japanese Industrial Groups. Quarterly Journal of Economics 106: 33–36.
Hoshi, T., and Patrick, H., eds. 2000. Crisis and Change in the Japanese Financial System. Boston,
MA: Kluwer.
IMF (International Monetary Fund). 2014. Article IV Consultation: South Korea. IMF Country
Report No. 14/101. Washington.
IMF (International Monetary Fund). 2017. The “New Mediocre” and the Outlook for Produc-
tivity in Asia. In Regional Economic Outlook: Asia and Pacific, April 2017: Preparing for Choppy
Seas. Washington.
Ito, T. 1992. The Japanese Economy. Cambridge, MA: MIT Press.
Jaffe, A., and J. Lerner. 2004. Innovation and Its Discontents. Princeton, NJ: Princeton University
Press.
Joh, S. W. 2014. Chaebols as South Korean Entrepreneurship. In Handbook of East Asian Entre-
preneurship: 157–68.
Jones, H. 1980. Live Machines: Hired Foreigners and Meiji Japan. Vancouver: University of British
Columbia Press.
Klepper, S. 1996. Entry, Exit, Growth, and Innovation over the Product Life Cycle. American
Economic Review 86, no. 3: 562–83.
Klepper, S. 2010. The Origin and Growth of Industry Clusters: The Making of Silicon Valley
and Detroit. Journal of Urban Economics 67, no. 1: 15–32.
La Croix, S., and A. Kawaura. 1996. Product Patent Reform and Its Impact on Korea’s Phar-
maceutical Industry. International Economic Journal 10, no. 1: 109–24.
Lee, K. 2013. Schumpeterian Analysis of Economic Catch-Up: Path-Creation and the Middle Income
Trap. New York: Cambridge University Press.
Mansfield, E. 1988a. Industrial Innovation in Japan and the United States. Science 241: 1769–
74.
Mansfield, E. 1988b. Industrial R&D in Japan and the United States: A Comparative Study.
American Economic Review 78, no. 2: 223–28.
Mansfield, E. 1988c. The Speed and Cost of Industrial Innovation in Japan and the United
States: External vs. Internal Technology. Management Science 34, no. 10: 1157–68.

98 SUSTAINING ECONOMIC GROWTH IN ASIA

Mansfield, E. 1991. Academic Research and Industrial Innovation. Research Policy 20, no. 1:
1–12.
Nagaoka, S. 2007. Assessing the R&D Management of a Firm in Terms of Speed and Science
Linkage: Evidence from U.S. Patents. Journal of Economics, Management, and Strategy 16, no.
1: 129–56.
National Research Council. 1992. Japan’s Growing Technological Capability: Implications for the
U.S. Economy. Washington: National Academy Press.
Nelson, R. 1992. National Innovation Systems: A Retrospective on a Study. Industrial and Cor-
porate Change 1, no. 2: 347–74.
OECD (Organization for Economic Cooperation and Development). 2016. Korea. In Educa-
tion at a Glance 2016: OECD Indicators. Paris: OECD Publishing.
Okimoto, D. I., and T. P. Rohlen. 1988. Inside the Japanese System: Readings on Contemporary
Society and Political Economy. Stanford, CA: Stanford University Press.
Okimoto, D., and G. Saxonhouse. 1987. Technology and the Future of the Economy. In The
Political Economy of Japan, Volume 1: The Domestic Transformation, ed. K. Yamamura and Y.
Yasukichi. Stanford, CA: Stanford University Press.
Ordover, J. A. 1991. A Patent System for Both Diffusion and Exclusion. Journal of Economic
Perspectives 5, no. 1: 43–60.
Pesek, W. 2015. Life in the Republic of Samsung. Barron’s Online, December 3. Available at
www.barrons.com/articles/life-in-the-republic-of-samsung-1449106291.
Rohlen, T. 1983. Japan’s High Schools. Berkeley, CA: University of California Press.
Romalis, J. 2004. Factor Proportions and the Structure of Commodity Trade. American Eco-
nomic Review 94, no. 1: 67–97.
Rosenberg, N., and R. Nelson. 1994. American Universities and Technical Advance in Indus-
try. Research Policy 23, no. 3: 323–48.
Sakakibara, M., and L. Branstetter. 2001. Do Stronger Patents Induce More Innovation? Evi-
dence from the 1988 Japanese Patent Law Reforms. RAND Journal of Economics 32, no. 1:
77–100.
Saxenian, A., and J. Hsu. 2001. The Silicon Valley-Hsinchu Connection: Technical Communi-
ties and Industrial Upgrading. Industrial and Corporate Change 10, no. 4.
Uekusa, M. 1987. Industrial Organization: The 1970s to the Present. In The Political Economy of
Japan, Volume 1: The Domestic Transformation, ed. K. Yamamura and Y. Yasukichi. Stanford,
CA: Stanford University Press.
Wade, R. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian
Industrialization. Princeton, NJ: Princeton University Press.
Wakasugi, R. 1997. Technology Importation in Japan. In Innovation in Japan, ed. A. Goto and
H. Odagiri. Oxford: Clarendon Press.
Weinstein, D. 1995. Evaluating Administrative Guidance and Cartels in Japan, 1957–1988.
Journal of the Japanese and International Economies 9: 200–23.
Yi, P. C., ed. 2006. Developmental Dictatorship and the Park Chung-hee Era: The Shaping of Modernity
in the Republic of Korea. Homa & Sekey Books.
Zucker, L., M. Darby, and M. Brewer. 1998. Intellectual Capital and the Birth of U.S. Biotech-
nology Enterprises. American Economic Review 88: 290–306.

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 99

Comment
Jong-Wha Lee

Lee Branstetter and Namho Kwon introduce the concepts of “pro-incre-

mental” and “pro-incumbent” biases in the innovation systems of three
East Asian economies—Japan, South Korea, and Taiwan—using case studies.
However, their illustration is limited on how incremental and incumbent
innovation affects research productivity, research input, and output dif-
ferently than radical innovation. The determinants of these two different
types of innovation are also unclear. Further discussion would be useful
to support their main arguments on the role of pro-incremental and pro-
incumbent innovation.
The chapter also does not provide robust evidence on the role of pro-
incremental and pro-incumbent innovation in declining research produc-
tivity. In this regard, Branstetter and Kwon use firm-level data on patenting
and R&D spending in Japan in the 1990s as evidence. The decline in research
productivity in Japan, however, might have been caused by other factors
both domestic and global. Productivity growth in Japan and many other
East Asian economies has been impeded by many more important country-
specific factors. They include slower convergence in productivity, limited
innovative capability, low productivity growth in the services sector, low
efficiency in resource allocation between small and large enterprises and
between manufacturing and service sectors—mainly due to inefficiencies in
labor and financial markets—and low quality of institutions (i.e., govern-
ment regulation, intellectual property protection, corporate governance,
and policy uncertainty). Thus, it is important to identify the exact extent
to which pro-incremental and pro-incumbent biases, rather than other
factors, cause diminishing returns to R&D investment. Furthermore, the
decline in productivity growth might be attributable to global factors such
as the absence of revolutionary invention and slow diffusion of new tech-
nologies across industries and firms globally in recent decades.
Some studies point out that declining productivity growth is not
unique to East Asian economies. For example, Bloom et al. (2017) show
that despite the increase in the effective number of researchers, research
productivity has continuously declined in the United States since the
1940s. Also, given that research productivity is different from total factor
productivity (TFP), it would be helpful to discuss whether research produc-

Jong-Wha Lee is a professor of economics and director of the Asiatic Research Institute at Korea University. He
was chief economist and head of the Office of Regional Economic Integration at the Asian Development Bank
and an economist at the International Monetary Fund.

100 SUSTAINING ECONOMIC GROWTH IN ASIA

tivity has declined more rapidly in East Asian economies compared with the
United States or Western Europe. One should also note that it is natural
for economies like those in East Asia—which successfully transformed
themselves from imitators to innovators—to create pro-incremental and
pro-incumbent biases in the early stages of development and then build up
innovative capabilities with the accumulation of knowledge.
Branstetter and Kwon explain how policies, such as human capital
policy, intellectual property policy, and exchange rate policy in the three East
Asian economies reinforced pro-incumbent and pro-incremental biases in
their innovation systems, contributing to a slowdown in R&D productivity.
On exchange rate policy, lack of evidence makes it hard to believe their
argument that an undervalued exchange rate policy turned Japanese and
Korean firms away from fundamental innovation and toward incremental
innovation. The “export-oriented policy” did indeed support export firms,
which are seen to be more innovative than domestic ones. This policy was
also effective in pushing the pace of change in comparative advantage, from
labor-intensive manufacturing to more capital-intensive and then to tech-
nology-intensive industries. These East Asian countries switched to more
market-based flexible exchange rate systems a few decades ago. Therefore,
more discussion is needed on the role of exchange rate policy in creating
pro-incremental and pro-incumbent biases.
For East Asian economies, technological progress is most critical to
maintaining strong growth. Branstetter and Kwon suggest reforming
outdated policies and emulating the Israeli or Silicon Valley model as an
alternative. These economies should make the effort to build innovative
capabilities on their own because the role of government in promoting
technological innovation, especially fundamental, radical innovation, is
not always clear. To do so, policies should be prioritized to tackle coun-
try-specific factors impeding productivity growth. For instance, in South
Korea, the government can nurture research capabilities by upgrading the
quality of tertiary education and human capital in science and technology
and increasing investment in basic research. The government should also
develop the venture capital industry and startups, improve resource allo-
cation across enterprises and sectors, and reduce political instability and
policy uncertainty.

References
Lee, Jong-Wha. 2016. Korea’s economic growth and catch-up: Implications for China. China
& World Economy 24, no. 5: 71–97.
Bloom, Nicholas, Charles I. Jones, John Van Reenen, and Michael Webb. 2017. Are ideas get-
ting harder to find? NBER Working Paper 23782. Cambridge, MA: National Bureau of Eco-
nomic Research.

EVOLUTION OF EAST ASIA’S INNOVATION SYSTEMS 101

 5
Secular Stagnation and Asia:
International Transmission
and Policy Spillovers
OLIVIER JEANNE

Depressed economic growth and interest rates are the two main symptoms
of secular stagnation. Low interest rates can have problematic side effects,
such as liquidity traps or unsustainable booms in credit and asset prices.
The relationship between these symptoms and their underlying causes are
a matter of debate. If secular stagnation were a health condition, it would
more likely be called a syndrome than a disease.1
To extend the medical metaphor, this chapter explores the contagion
of secular stagnation across countries. The symptoms of secular stagna-
tion have not been limited to a particular group of countries. Many coun-
tries at different stages of economic development and in different regions
have displayed these symptoms. The concomitance of the symptoms could
reflect domestic factors that would have appeared independently in many
countries, but it may suggest that some form of international transmission
or contagion has been at work, raising the question of the mechanisms
and processes by which secular stagnation can spread from a country to
its neighbors.
In a globally integrated economy it is not difficult to imagine channels
of international transmission of secular stagnation. In a financially inte-
grated world a surplus of savings in one part should depress interest rates

Olivier Jeanne is senior fellow at the Peterson Institute for International Economics and professor of economics
at Johns Hopkins University. He thanks the conference participants and in particular the chapter’s discussant,
Heenam Choi, for their insightful comments, as well as Medha Nair for her research assistance.

1. A syndrome is a set of correlated symptoms that may or may not be associated with a well-
defined disease.

103
globally. A country attempting to resist lowering its interest rate would be
doing so at the cost of currency appreciation. In a world integrated via trade
a negative demand shock in one part of the world may reduce growth in the
rest of the world. Textbook macroeconomic theory does not say much about
spillovers in productivity growth, but channels of transmission related to
international finance and international trade can explain such spillovers.
This chapter first provides a quick review of the literature on secular
stagnation. It then tries to better understand the channels by which low pro-
ductivity growth can spill over across countries. In principle this spillover
can occur through several channels, related to finance or trade. For example,
if efficient financial intermediation is important for firms to finance pro-
ductivity-enhancing expenditures, then an internationally contagious
banking crisis could depress productivity growth in many countries at the
same time. Or if learning-by-exporting is important for firms to adopt more
efficient foreign technologies, then a slowdown in international trade could
explain a general decrease in productivity growth.
The chapter then investigates how the other symptom of secular stagna-
tion—low interest rates—was transmitted across countries. A simple open-
economy model is presented to set a benchmark for what one should look
for in the data. The model features a country that is an “innocent bystander”
affected by secular stagnation in the rest of the world. Secular stagnation
abroad takes the form of lower real interest rates and lower demand in the
rest of the world. The model shows that the domestic response must involve
a fall in the domestic real rate of interest and net capital inflows. These spill-
overs are part of the natural adjustment process of the domestic economy
to secular stagnation abroad and should not a priori be resisted. However,
they may have problematic side effects, such as drawing monetary policy
closer to the zero lower bound or increasing domestic credit. The chapter
then looks for these side effects in the data for emerging-market economies
as a whole and for the Asian region, taking the global financial crisis as a
watershed event.
Although secular stagnation is by definition a protracted phenom-
enon, the crisis was an important step in its diffusion. Furthermore it is an
event in which the direction of transmission is not ambiguous: The impulse
originated in the advanced economies at the epicenter of the banking crisis
and was transmitted to the rest of the world.2 This chapter is an attempt to
learn something about the transmission of secular stagnation from differ-

2. This is not to say that secular stagnation is in general transmitted from advanced econo-
mies to the rest of the world. For example, many have argued that China has been an impor-
tant source of excess saving in the global economy.

104 SUSTAINING ECONOMIC GROWTH IN ASIA

ences in the way the global financial crisis affected countries depending on
their circumstances and policies.

Literature
The literature on secular stagnation has been growing since Lawrence
Summers popularized the concept in 2013. A large part of this litera-
ture considers this topic in a closed-economy context—see, for example,
Andrews, Criscuolo, and Gal (2016) or IMF (2017a) for recent discussions
of internal factors of stagnation such as the information technology revo-
lution, population aging, slowing human capital accumulation, or fading
structural reform efforts. These factors may exercise their effects concomi-
tantly, which could explain why the symptoms of secular stagnation have
affected many countries at the same time without its being the result of con-
tagion. This chapter is related more to the literature that considers interna-
tional transmission of secular stagnation in an open economy. One part of
that literature has focused on the financial channels by which excess saving
in some countries can depress the real interest rate in others. That litera-
ture is mostly theoretical. Eggertsson et al. (2016) and Caballero, Farhi, and
Gourinchas (2015) study global equilibria in which negative spillovers can
spread secular stagnation worldwide.3 In the model of Caballero, Farhi, and
Gourinchas (2015) secular stagnation results from a shortage of safe assets.
Corsetti et al. (2017) build on the model of Eggertsson et al. (2016) to study
the behavior of exchange rates and capital flows in a global liquidity trap.
On the empirical side there is a large literature on the channels of con-
tagion during the global financial crisis. This literature has found it surpris-
ingly difficult to relate the strength of contagion to conventional measures
of economic openness. Rose and Spiegel (2011) considered a wide range of
indicators in the precrisis data that might predict the cross-country inci-
dence of the Great Recession and generally found no significant link between
these indicators and a variety of financial and real manifestations of the 2008
crisis. Lane and Milesi-Ferretti (2011) find similar results. Kalemli-Ozcan,
Papaioannou, and Perri (2013) find that economies with stronger financial
ties to the United States were more affected by the global financial crisis
but establish this result only for developed economies. Hausmann-Guil,
van Wincoop, and Zhang (2016) present a model of self-fulfilling crises to
explain that integration did matter for contagion beyond some threshold.
There is a large literature on the determinants of productivity growth
and some involve international transactions. I am not aware of any studies

3. Earlier papers had studied the international propagation of liquidity traps; see, for example,
Fujiwara et al. (2013), Devereux and Yetman (2014), Cook and Devereux (2013), and Acharya
and Bengui (2016).

INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 105

 Table 5.1 Sample of Asian countries
Emerging-market
Advanced and developing
AUS Australia BGD Bangladesh
HKG Hong Kong CHN China: Mainland
JPN Japan IND India
KOR Korea IDN Indonesia
NZL New Zealand LAO Laos
SGP Singapore MYS Malaysia
TWN Taiwan MNG Mongolia
MMR Myanmar
NPL Nepal
PNG Papua New Guinea
PHL Philippines
LKA Sri Lanka
THA Thailand
VNM Vietnam
Note: This table reports the Asian countries in the
sample used in this chapter. It includes the countries
under the surveillance of the Asia and Pacific
Department of the International Monetary Fund
(IMF), excluding those with a population of less than
1 million as well as Cambodia and Timor Leste, for
which data were insufficient.
Source: IMF, World Economic Outlook.

that have harnessed the theoretical insights from that literature to explore
the international transmission of secular stagnation. I discuss the relevant
literature in the next section. Adler et al. (2017) also discuss these channels
and present empirical evidence.

Productivity Spillovers
The main feature of secular stagnation is a persistent fall in productivity
growth. Total factor productivity (TFP) growth fell at about the same
time in countries with very different levels of economic development (see
table 5.1 for the sample of Asian countries). This concomitance is not
just a feature of the last ten years: De Gregorio (2017) shows that the TFP
growth of emerging-market economies has been correlated with that of
advanced economies since the 1960s, suggesting the existence of spill-
overs from advanced to less developed economies. However, there is no
well-established theory of international spillovers in productivity growth.
This section first discusses the theoretical channels through which secular

106 SUSTAINING
GRAPHICS ECONOMIC
- CHAPTER GROWTH IN ASIA
5—SUSTAINING ECONOMIC GROWTH IN ASIA 1
stagnation in the rest of the world could affect productivity growth in a
country and then looks for evidence on these spillovers in the data.

Theory
Development economists distinguish between advanced economies that
are at the world technology frontier and less advanced economies that are
catching up to this frontier. Innovation determines productivity growth at the
frontier, while away from the frontier it is determined by the diffusion of the
technologies and processes that are used at the frontier. Technopessimists
such as Gordon (2016) or Fernald (2015) attribute secular stagnation to a
slowdown in productivity growth caused by a natural exhaustion of econ-
omies’ innovative potential. This view does not explain why productivity
growth also slowed down in countries that are catching up to the frontier
by adopting existing technologies rather than by innovating. Moreover, it
does not explain why productivity growth in these countries slowed during
the global financial crisis.
One possible explanation for the general slowdown in productivity
growth is that financial frictions may have spilled over across countries.
Financial frictions can bias business investment toward more liquid, low-
risk/low-return projects, which may in turn slow technological progress
embodied in new capital goods or resulting from risky investments. For
example, Caballero, Hoshi, and Kashyap’s (2008) model shows how banking
problems in Japan may have persistently depressed Japanese productivity
growth. Anzoategui et al. (2016) propose a model in which an increase in
demand for liquidity, as observed during the crisis, decreases investment
and productivity growth. In a financially integrated world, lower produc-
tivity growth could be transmitted across countries through financial
channels. This is a plausible channel of transmission for the global finan-
cial crisis, where global banks seem to have played a significant role in the
transmission of the crisis from advanced to emerging-market economies
(Cetorelli and Goldberg 2011).
Secular stagnation could also be diffused via international trade in
goods and services. Economic researchers have often noted the simulta-
neous slowing of productivity growth and of trade integration after the
global financial crisis (IMF 2016). According to the learning-by-exporting
hypothesis, export participation improves productivity. A firm’s expected
profits from process or product innovation rise with the size of the final
market so that increased exports allow firms to bear higher fixed costs
of research and development (Rodrik 1988, Yeaple 2005). Alternatively,
trade flows might facilitate international knowledge spillovers and thus
contribute to the adoption of new technologies (Coe and Helpman 1995).
If this is true, the fact that many advanced economies reduced their demand
INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 107
for imports after the global financial crisis could result in lower produc-
tivity growth in the rest of the world.
Finally, the spillovers could result from the fact that productivity is
endogenous to demand—consistent with the hysteresis view that cyclical
changes in demand have a permanent impact on output. For example,
Benigno and Fornaro (2017) present a model in which low demand leads
to low productivity growth because firms spend less on productivity-
enhancing activities. A fall in global demand could thus lead to a gener-
alized decrease in productivity growth even in a model where there is no
learning-by-exporting.

Evidence
Several branches of empirical literature are relevant to quantify the theo-
retical mechanisms discussed above, but they have not been systematically
harnessed to analyze the international transmission of secular stagnation.
There is a large empirical literature on the link between exporting and
productivity growth based on firm-level evidence.4 Whereas Keller’s (2004)
review of the literature concluded that there is little evidence of a strong
learning-by-exporting effect, recent research was more supportive of such
effects (see Bustos 2011, De Loecker 2007, Lileeva and Trefler 2010). One
problem with applying this research to the analysis of secular stagnation is
that the results from studies based on firm-level data are often difficult to
translate to the aggregate level.5
One line of literature investigates the financial channels of contagion
during the global financial crisis and in the global financial cycle, but it
does not make the link to productivity (Hausmann-Guil, van Wincoop,
and Zhang 2016). Another line of literature has investigated how produc-
tivity is determined by the efficiency of factor allocation and in particular
the quality of financial intermediation. For example, Adler et al. (2017)
report that TFP growth fell more in companies with weaker balance sheets
prior to the global financial crisis than in their counterparts with stronger
balance sheets. But to the best of my knowledge this literature has not
quantified the extent to which secular stagnation is transmitted interna-
tionally through financial channels.
The remainder of this section investigates the channels of international

4. There is an older literature based on cross-country macroeconomic evidence; see, for

example, Frankel and Romer (1999).
5. Many studies estimate the impact on an individual firm of a binary decision to start
exporting, which is difficult to translate into an estimate of the impact of the aggregate
volume of exports on productivity. Studies based on the difference-in-difference method-
ology do not easily lend themselves to estimating aggregate effects.

108 SUSTAINING ECONOMIC GROWTH IN ASIA

 Table 5.2 Change in total factor productivity (TFP) growth
and economic integration
Variable (1) (2) (3)
–0.614 –1.395 –1.309
Exports/GDP (10–2)
(0.831) (1.184) (0.905)
–1.808**
De jure financial openness
(0.704)
0.006
De facto financial integration (10–2)
(0.083)
0.015
De facto banking integration (10–2)
(0.134)
1.499*** 0.785* 0.667
Constant
(0.506) (0.463) (0.405)
R-squared 0.108 0.033 0.039
Number of observations 81 82 77
Note: The dependent variable is the change in TFP growth between 2000–07
and 2008–14 in percentage points. Trade openness is measured as the ratio of
exports to GDP in 2007 in percentage points. De jure financial openness is the
Chinn-Ito index of financial openness for 2007. De facto financial integration is
the ratio of foreign assets plus liabilities to GDP in 2007 in percentage points.
De facto banking integration is the ratio of “other investments” assets plus
liabilities to GDP in 2007 in percentage points. See appendix 5A for sources.
Standard errors are in parentheses. *, **, and *** indicate statistical significance
at the 10, 5, and 1 percent levels, respectively.

transmission for productivity growth with simple cross-country regressions.

My main identification assumption is that in the global financial crisis the
spillovers went from the advanced economies that were at the epicenter of
the banking crisis to the rest of the world. The theories discussed above have
different implications for which countries should have been most affected
by the crisis. For example, if the main channel of transmission is financial,
one should expect the countries that are more financially integrated to be
more affected by the crisis. If the most important channel is trade, then
countries that are most integrated with world trade should be affected the
most. What do the data show?
Table 5.2 presents the results of three cross-country regressions of the
change in the TFP growth rate associated with the global financial crisis on
measures of financial and trade integration. The country sample includes
all the countries for which data are available, excluding the countries at the
center of the global financial system (the United States, the euro area, Japan,
the United Kingdom, and Switzerland). Precrisis measures of economic inte-
gration are used to reduce the endogeneity bias in the explanatory variables.
Trade integration is measured by the ratio of exports to GDP in 2007
(a standard measure in the literature). Financial openness is more difficult
to measure and three different indicators are used, corresponding to the
three columns of the table. The first measure is the Chinn-Ito (2008) index

2 GRAPHICSINTERNATIONAL
- CHAPTER 5—SUSTAINING
TRANSMISSION AND POLICY
ECONOMIC IN ASIA 109
SPILLOVERS
GROWTH
of de jure financial openness. This index is based on binary variables that
codify the tabulation of restrictions on cross-border financial transactions
reported in the International Monetary Fund’s Annual Report on Exchange
Arrangements and Exchange Restrictions (AREAER). The other indicators are
measures of de facto financial integration. The explanatory variable used in
the second column of the table is the ratio of the sum of total foreign assets
plus total foreign liabilities to GDP based on the data in Lane and Milesi-
Ferretti (2017)—a measure that has often been used in the literature. The
last measure is similar to the second one but is limited to the “other invest-
ments” category in foreign assets and liabilities in an attempt to capture
international banking integration. All the indicators of financial integra-
tion are taken in 2007, the year before the crisis erupted.
Several observations can be made from the regression results in table
5.2. First, trade integration has a negative impact on TFP growth in all
three specifications and the point estimates are economically significant.
For example, the coefficient reported in the second column implies that,
other things equal, having a 10 percent larger exports-to-GDP ratio in 2007
lowered postcrisis TFP growth by about 0.14 percent per year. However,
this relationship is not statistically significant in any of the regressions.
Second, the impact of de jure financial integration is statistically signif-
icant whereas the impact of de facto integration is not. The impact of de
jure financial openness seems to be large. Based on my estimates, increasing
de jure financial openness from the Chinese to the US level is associated
with a 1.8 percent reduction in TFP growth after the crisis. However, this
effect seems too large to be interpreted as reflecting causality from finan-
cial openness to productivity growth and it is difficult to understand given
that the main channel through which de jure openness should matter is by
fostering de facto integration, and I find that de facto integration itself has
no impact on productivity growth.
To summarize, postcrisis TFP growth seems orthogonal to most mea-
sures of international economic integration (with the exception of de jure
financial openness). The weakness of the relationship between postcrisis
TFP growth and economic openness is robust to changes in the regression
specification. I tried many variants of the regressions reported in table 5.2.
For example, I looked at the impact on productivity growth of a decrease in
the demand for each country’s exports, where demand was measured based
on postcrisis growth in the country’s regional export markets. The country
sample was changed in different ways, for example, by excluding oil export-
ers or including only emerging-market and developing economies. Extreme
observations were excluded. I used growth in GDP per capita (instead of
TFP), which yielded several more years of data after the crisis. Standard

110 SUSTAINING ECONOMIC GROWTH IN ASIA

growth regressions were tried controlling for the initial level of GDP per
capita. Details of these regressions are not reported here for the sake of
brevity, but the important point is that they did not substantially revise the
general conclusion that there seems to be little relationship between post-
crisis TFP growth and measures of economic openness.
The results presented here are perhaps not surprising from the point of
view of the empirical literature that finds no robust relationship between
economic openness and the decline in growth during the Great Recession
(Kalemli-Ozcan, Papaioannou, and Perri 2013; Kamin and DeMarco 2012;
Hausmann-Guil, van Wincoop, and Zhang 2016; IMF 2013; Lane and
Milesi-Ferretti 2011). The main difference between these studies and the
regressions presented here is that I look at TFP growth rather than GDP
growth, and I look at it over a longer period after the crisis. Nevertheless
my results are puzzling for the view that international spillovers are impor-
tant in explaining why secular stagnation affected so many countries at
about the same time.

Macroeconomic Spillovers: A Model

The rest of the chapter takes productivity as exogenous and focuses on the
macroeconomic spillovers in the transmission of secular stagnation across
countries. This section presents a model of how a small open economy
responds to secular stagnation abroad. Foreign secular stagnation is repre-
sented as the combination of two developments: a decrease in the foreign
real rate of interest and a decrease in foreign demand for the country’s
exports. The question is whether these developments can bring foreign stag-
nation home and through which channels.
The model is written in real terms and is intended to capture the long
run. It predicts the behavior of the real interest rate, the real exchange
rate, and capital flows. The real interest rate predicted by the model can
be interpreted as the “natural” rate of interest—the level of the real interest
rate that ensures full employment. Modern macroeconomic theory holds
that the main function of monetary policy is to keep the real rate of
interest close to the natural level so as to stabilize both employment and
inflation.
The model is simple and does not capture all the international spill-
overs of interest. In particular, it takes domestic growth as given and so
does not have anything to say about spillovers in productivity growth. The
model focuses on the rebalancing of trade and financial flows that must
take place in response to foreign secular stagnation. Some features missing
from the model are discussed later in the section.

INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 111

 The main findings of the model are that in response to foreign secular
stagnation:
n the domestic natural rate of interest falls by the same amount as the
foreign rate in the long run, but less so initially;
n there is a net capital inflow and an increase in domestic demand in the
short run (which is reversed in the long run); and
n restrictions to capital mobility may reduce the response of the domestic
interest rate but amplify that of the real exchange rate.

The model is in continuous time and can be summarized by the fol-

lowing three equations:
𝑎𝑎� � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦�∗ � � 𝑟𝑟�∗ 𝑎𝑎�
� 𝑎𝑎� � � 𝜎𝜎�𝑎𝑎��𝑟𝑟�∗ � � 𝑎𝑎� �
𝑟𝑟� � 𝑟𝑟�∗ � 𝑞𝑞� � /𝑞𝑞�

The main variables are the net foreign assets at, the net trade balance xt, the
real exchange rate qt, and the domestic and foreign real rates of interest rt
lim 𝑟𝑟�∗is�denoted
and rt*. The derivative of a variable 𝑟𝑟̅ ∗ and lim with𝑦𝑦a�∗ dot.
� 𝑦𝑦� ∗
���� ����
The first equation is the balance-of-payments equation: the rate of
increase in net foreign assets at is equal to the trade balance xt plus the
return on foreign assets. The trade balance is an increasing function of the
𝑥𝑥�𝑞𝑞�, 𝑦𝑦� ∗ � � 𝑟𝑟̅ ∗ 𝑎𝑎��𝑟𝑟̅ ∗ � � �
real exchange rate qt and of foreign demand yt*. The real exchange rate is
defined in such a way that an increase in q corresponds to a real deprecia-
tion. A real depreciation shifts domestic and foreign spending toward the
home good. 𝑟𝑟̅ � 𝑟𝑟̅ ∗
The second equation describes𝑎𝑎� � � the , 𝑦𝑦�∗ � � 𝑟𝑟�∗ 𝑎𝑎of
𝑥𝑥�𝑞𝑞�dynamics � foreign assets. Foreign
∗� �
assets converge toward a desired� 𝑎𝑎� �level
� 𝜎𝜎�𝑎𝑎��𝑟𝑟 , which
� � 𝑎𝑎� a function of the return
is
∗
𝑟𝑟
on these assets. The desired stock� of net � 𝑟𝑟 � 𝑞𝑞� /𝑞𝑞
foreign wealth is an increasing
𝜎𝜎𝜎𝜎��𝑟𝑟��∗ � � �𝑥𝑥�𝑞𝑞�� , 𝑦𝑦�∗ �
function of the interest rate. This is a shortcut to capture the foreign asset
dynamics resulting from open-economy models with overlapping genera-
tions such as Coeurdacier, Guibaud, and Jin (2015) or Eggertsson et al.
lim 𝑟𝑟�∗ � 𝑟𝑟̅ ∗ and lim 𝑦𝑦�∗ � 𝑦𝑦� ∗
(2016) (although the dynamics of net foreign 𝑞𝑞 � � 𝑞𝑞
�� assets are more complicated ����
����

��
in these models).
( � � �)The third equation is interest parity. The domestic real rate of interest
��
is equal to the foreign rate plus the rate of real exchange rate depreciation. 𝑥𝑥�𝑞𝑞�, 𝑦𝑦� ∗ � � 𝑟𝑟̅ ∗ 𝑎𝑎��𝑟𝑟̅ ∗ � � �
∗
𝑟𝑟 � 𝑟𝑟
As a starting point perfect financial integration is assumed, implying that
� �

interest parity holds between domestic and foreign bonds.

The model assumes full employment it = rtso + π. that rt is the natural rate of ∗
𝑟𝑟̅ � 𝑟𝑟̅
interest. Foreign secular stagnation takes the form of decreases in the
foreign real rate of interest, rt∗, and in foreign demand for the country’s

𝜎𝜎𝜎𝜎��𝑟𝑟�∗ � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦�∗ �

112 SUSTAINING ECONOMIC GROWTH IN ASIA

𝑞𝑞� � 𝑞𝑞��
 𝑎𝑎� � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦�∗ � � 𝑟𝑟�∗ 𝑎𝑎�
exports, yt . The question
∗
� 𝑎𝑎� � � 𝜎𝜎�𝑎𝑎of ∗
��𝑟𝑟interest
� � � 𝑎𝑎� � is how capital flows, the real interest
∗
� 𝑟𝑟�∗ 𝑎𝑎� 𝑟𝑟� � 𝑟𝑟� �
rate, and the real exchange rate𝑞𝑞� � /𝑞𝑞
must
� respond to ensure full employment
� 𝑎𝑎� � at home.
𝑎𝑎
/𝑞𝑞
� Let us summarize a few key properties of the model, starting with the
�
long run. 𝑎𝑎� � � I𝑥𝑥�𝑞𝑞
assume ∗ that ∗ foreign variables converge
𝑟𝑟�∗ 𝑎𝑎� 𝑟𝑟toward
𝑎𝑎� � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦�∗ � � lim ∗ long-run values
� , 𝑦𝑦� � � 𝑟𝑟� 𝑎𝑎� ∗ ∗
�∗
� � 𝑟𝑟̅ and lim 𝑦𝑦� � 𝑦𝑦
denoted
∗ �� 𝑎𝑎� ∗with
𝑎𝑎� � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦� � 𝑟𝑟�� � bars: ∗�
𝑎𝑎� 𝜎𝜎�𝑎𝑎��𝑟𝑟� � 𝑎𝑎� � ∗ ����
� 𝑎𝑎� � � 𝜎𝜎�𝑎𝑎��𝑟𝑟� � � 𝑎𝑎� � ����

� 𝑎𝑎� � � 𝜎𝜎�𝑎𝑎��𝑟𝑟
lim
∗
� � 𝑟𝑟� ∗ 𝑎𝑎𝑟𝑟�� �
∗
∗ 𝑟𝑟� � 𝑞𝑞� � /𝑞𝑞� ∗
� ∗. 𝑟𝑟� � 𝑟𝑟�∗ � 𝑞𝑞� � /𝑞𝑞�
∗ � �� 𝑟𝑟̅ and lim 𝑦𝑦� � 𝑦𝑦
𝑟𝑟� � 𝑟𝑟�����
� 𝑞𝑞� � /𝑞𝑞� ����
∗
lim 𝑟𝑟� � ∗
In𝑟𝑟̅the and 𝑦𝑦�∗ the
limrun
long � 𝑦𝑦� ∗country’s foreign assets converge
𝑥𝑥�𝑞𝑞�, 𝑦𝑦� ∗ � �to
𝑟𝑟̅ ∗ 𝑎𝑎��𝑟𝑟̅ ∗ �. �
It �then
���� ����
follows from the balance-of-payments equation that the real exchange rate
lim 𝑟𝑟�∗ � 𝑟𝑟̅ ∗ and lim 𝑦𝑦�∗ �lim𝑦𝑦� ∗ 𝑟𝑟�∗ � 𝑟𝑟̅ ∗ and lim 𝑦𝑦�∗ � 𝑦𝑦� ∗
𝑥𝑥�𝑞𝑞�, 𝑦𝑦� ∗ � and
is constant � 𝑟𝑟̅ ∗ 𝑎𝑎
��𝑟𝑟̅ ∗ � � � ∗
satisfies ∗
����
∗ ∗
���� ���� ����
lim 𝑟𝑟� � 𝑟𝑟̅ and lim 𝑦𝑦� � 𝑦𝑦�
���� ����
𝑥𝑥�𝑞𝑞�, 𝑦𝑦� ∗ � � 𝑟𝑟̅ ∗ 𝑎𝑎��𝑟𝑟̅ ∗ � � �. 𝑟𝑟̅ � 𝑟𝑟̅ ∗

Secular stagnation � 𝑦𝑦𝑟𝑟̅� ∗∗ reduces the foreign �, 𝑦𝑦� ∗ � � rate

interest
𝑥𝑥�𝑞𝑞 𝑟𝑟̅ ∗ 𝑎𝑎��𝑟𝑟̅ ∗ �and � �demand 𝑥𝑥�𝑞𝑞�, 𝑦𝑦� ∗ �in�the 𝑟𝑟̅ ∗ 𝑎𝑎��𝑟𝑟̅ ∗ � � �
lim 𝑟𝑟�∗ � 𝑟𝑟̅ ∗ and lim 𝑦𝑦�𝑟𝑟̅∗ �
���� long run. ���� It follows that
∗𝑥𝑥�𝑞𝑞 𝑦𝑦�∗∗ � �∗ 𝑟𝑟̅ ∗ 𝑎𝑎��𝑟𝑟̅that
�, increases, ∗�
�is, � there is a real depreciation in
𝑎𝑎� �∗ � 𝑎𝑎� �𝑥𝑥�𝑞𝑞
� 𝑥𝑥�𝑞𝑞� , 𝑦𝑦���, 𝑦𝑦 ��∗ �𝑟𝑟�� 𝑎𝑎𝑟𝑟� � 𝑎𝑎� ∗ ∗�
the long run. 𝑟𝑟̅ � 𝑟𝑟̅ The real exchange 𝜎𝜎𝜎𝜎 � � � 𝑥𝑥�𝑞𝑞
��𝑟𝑟offset � , 𝑦𝑦�foreign
∗ � ∗ � � rate � must depreciate to lower
� 𝑎𝑎��� � 𝑎𝑎� �𝜎𝜎�𝑎𝑎
���𝑟𝑟 𝜎𝜎�𝑎𝑎 ���𝑟𝑟 � � �𝑎𝑎 � 𝑎𝑎 �
demand as∗ well 𝑟𝑟� as�𝑟𝑟�𝑟𝑟the ∗ lower
� ∗
∗𝑟𝑟��𝑞𝑞� �
domestic demand coming from the decrease ∗
𝑎𝑎� 𝜎𝜎𝜎𝜎 ��𝑟𝑟 � � 𝑥𝑥�𝑞𝑞 ,��𝑦𝑦 � /𝑞𝑞
𝑞𝑞�� /𝑞𝑞 � 𝑟𝑟̅ � 𝑟𝑟̅ ∗ 𝑟𝑟̅ � 𝑟𝑟̅
∗ � in foreign∗ ∗ � �assets. � �
� 𝑥𝑥�𝑞𝑞�, 𝑦𝑦� � 𝑟𝑟̅ 𝑎𝑎��𝑟𝑟̅ � � ∗
𝑎𝑎� � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦𝑟𝑟̅� ���𝑟𝑟̅𝑟𝑟� 𝑎𝑎� ∗∗
lim 𝑟𝑟�∗ � The
𝑟𝑟̅ ∗ �∗and
𝜎𝜎𝜎𝜎��𝑟𝑟 �� real lim
𝑥𝑥�𝑞𝑞 exchange
� ,𝑦𝑦𝑦𝑦�∗�∗��� 𝑎𝑎�𝑦𝑦��∗ �rate is constant in the long run𝑞𝑞�so
𝜎𝜎�𝑎𝑎��𝑟𝑟�∗ � � 𝑎𝑎� � � 𝑞𝑞that
�� interest
���� ����
parity implies ∗
𝑟𝑟� � 𝑟𝑟� � 𝑞𝑞� � /𝑞𝑞�
�� lim ∗ �∗ � ∗ ∗ ∗∗
𝑦𝑦�∗ � 𝑦𝑦�∗��𝑟𝑟∗ ∗ ∗
� � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦�∗ �
𝑞𝑞� � ( �𝑞𝑞��� �) 𝜎𝜎𝜎𝜎
���� �𝑟𝑟
��𝑟𝑟lim
����
𝑟𝑟̅��and
� �𝑟𝑟�𝑥𝑥�𝑞𝑞 𝑟𝑟̅, 𝑦𝑦 � lim lim
�and
���� ����
𝜎𝜎𝜎𝜎𝑦𝑦�� �𝑦𝑦
∗.
𝑟𝑟̅ � 𝑟𝑟̅ � � ∗�
𝜎𝜎𝜎𝜎��𝑟𝑟� � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦� ∗�

lim
𝑥𝑥�𝑞𝑞 �, 𝑟𝑟
𝑦𝑦
�∗∗
�
� � 𝑞𝑞∗𝑟𝑟̅�and
𝑟𝑟̅ ∗� 𝑞𝑞
𝑎𝑎
��𝑟𝑟̅ ��∗ �lim � �𝑦𝑦 ∗
� 𝑦𝑦
� ∗
���� That is, the domestic
�
����
� interest rate decreases by the ∗same 𝑟𝑟� � 𝑟𝑟�∗
amount as the
lim 𝑟𝑟� � 𝑟𝑟̅ ∗ and lim 𝑦𝑦�∗ � 𝑦𝑦� ∗
foreign interest rate in the long run. This is 𝑥𝑥�𝑞𝑞 an implication
�𝑦𝑦� �,𝑞𝑞���𝑦𝑦�� �𝑟𝑟̅�𝑎𝑎��𝑟𝑟̅
𝑞𝑞��,𝑥𝑥�𝑞𝑞
���� ∗ ∗ ∗ ∗ of
∗ perfect
∗
� ����
𝑟𝑟̅ 𝑎𝑎��𝑟𝑟̅ ���� � finan-
𝑞𝑞� � 𝑞𝑞��
𝜎𝜎𝜎𝜎��𝑟𝑟�∗ �cial � 𝑥𝑥�𝑞𝑞 integration.
� , 𝑦𝑦 ∗�
� 𝑟𝑟 � � 𝑟𝑟 ∗
� 𝑞𝑞 � 𝑞𝑞
��
lim 𝑟𝑟�∗ � 𝑟𝑟̅ ∗ and �� � limtransition 𝑦𝑦 ∗ � 𝑦𝑦�∗ ∗ dynamics ��
� it = rt + π.
���� ( �,
𝑥𝑥�𝑞𝑞 The
�𝑦𝑦�𝑟𝑟̅∗�)
���� ���� 𝑟𝑟̅𝑟𝑟̅∗ �∗𝑟𝑟𝑎𝑎��𝑟𝑟̅
∗ � � � ( � �are �) different. Assume that foreign secular
�� � � 𝑟𝑟 � � �
( � � �) stagnation is�“announced” at time t = 0 in an equilibrium 𝑥𝑥�𝑞𝑞�, 𝑦𝑦�∗ ∗ �where
� ∗ the
��𝑟𝑟̅ ∗ � initial
�� it = rt + π. ∗ 𝑟𝑟̅ 𝑎𝑎 ��
a ∗𝑟𝑟̅ �𝑟𝑟̅𝑟𝑟̅� 𝑟𝑟̅
foreign
𝑞𝑞� � 𝑞𝑞�� assets are equal to zero ( 0
= 0 ). Then, 𝑟𝑟� � 𝑟𝑟 � 𝑟𝑟� � 𝑟𝑟�∗
it =∗ rt + π. ∗ 𝑟𝑟� � 𝑟𝑟� ∗
𝑥𝑥�𝑞𝑞�, 𝑦𝑦� ∗ �𝜎𝜎𝜎𝜎
���𝑟𝑟𝑟𝑟̅ ∗�∗𝑎𝑎��𝑟𝑟̅
� �𝑟𝑟̅�𝑥𝑥�𝑞𝑞 � 𝑟𝑟̅��∗, 𝑦𝑦� �.
�
it = rt +∗π. ∗ 𝑟𝑟̅ � ∗
∗ 𝑟𝑟̅ ∗ �
it = rt + π.
The impact of foreign secular stagnation (lower 𝜎𝜎𝜎𝜎��𝑟𝑟 𝜎𝜎𝜎𝜎
���𝑟𝑟 �� � �𝑥𝑥�𝑞𝑞
and � , 𝑦𝑦��)�, 𝑦𝑦on
� 𝑥𝑥�𝑞𝑞 � the real
i t = rt + π.
∗
𝑟𝑟exchange
� � 𝑟𝑟�∗ rate is ambiguous in the short run. On one hand, the lower
𝑟𝑟̅ � 𝑟𝑟̅ 𝑞𝑞
𝜎𝜎𝜎𝜎��𝑟𝑟��∗demand
foreign ���𝑞𝑞��𝑥𝑥�𝑞𝑞� , 𝑦𝑦�∗ �should be offset by a real depreciation (an increase in
q ). On the other hand, the lower foreign interest rate 𝜎𝜎𝜎𝜎��𝑟𝑟�∗ �leads � 𝑥𝑥�𝑞𝑞to ∗
��
� , 𝑦𝑦capital
it =0 rt + π. 𝑞𝑞� � 𝑞𝑞�𝑞𝑞��� 𝑞𝑞��
inflows that have an expansionary effect on the domestic economy. The
𝜎𝜎𝜎𝜎��𝑟𝑟�∗real exchange
� � 𝑥𝑥�𝑞𝑞 ,��𝑦𝑦� �∗�
�(𝑞𝑞 rate
��� �𝑞𝑞��� �) depreciates only if the impact of lower foreign demand
𝑟𝑟 �� �� 𝑟𝑟(∗���) 1
dominates � �that of higher domestic demand. In general the real exchange
𝑞𝑞� � 𝑞𝑞��
rate depreciates less in the short run than in the long 𝑟𝑟 �
run,
𝑟𝑟 ∗ ∗
1 � 𝑟𝑟� �� 𝑟𝑟�
�� �
𝑞𝑞� � i𝑞𝑞t��=. rt + π. (�� � �)
1 𝑟𝑟 � 𝑟𝑟 ∗
� �
it = ritt =+ rπ.t + π. ∗
1 𝑟𝑟� � 𝑟𝑟� 1
it = rt + π. 1
𝑟𝑟� � 𝑟𝑟�∗ INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 113
it = rt + π.
 𝜎𝜎𝜎𝜎��𝑟𝑟� � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦� �
lim 𝑟𝑟�∗ � 𝑟𝑟̅ ∗ and lim 𝑦𝑦�∗ � 𝑦𝑦� ∗
���� ����

𝑞𝑞� � 𝑞𝑞��
∗�
� 𝑟𝑟̅ ∗implies
𝑥𝑥�𝑞𝑞�, 𝑦𝑦� This 𝑎𝑎��𝑟𝑟̅ ∗ � �that
� the real exchange rate must depreciate over time (�� � � �)
��
so that by interest parity
𝑟𝑟� � 𝑟𝑟�∗.
𝑟𝑟̅ � 𝑟𝑟̅ ∗
In the short run the domestic rate of interest decreases by less than the
= rt + π. rate—the interest rate pass-through is less than one-for-one.
itforeign
It is important to observe that foreign secular stagnation must lead
𝜎𝜎𝜎𝜎��𝑟𝑟�∗ � � 𝑥𝑥�𝑞𝑞� , 𝑦𝑦�∗ �
initially to a boom in capital inflows and an expansion in domestic
demand. These spillovers are part and parcel of the adjustment process to
secular stagnation abroad. Low foreign interest rates are expansionary and
𝑞𝑞� � 𝑞𝑞�� demand must increase to offset the fall in foreign demand.
domestic
However, this adjustment process may have problematic side effects.
The first one comes from the zero bound on the nominal interest rate. The
analysis has been in real terms so far but it is not difficult to introduce
𝑟𝑟� � 𝑟𝑟�∗ policy. The home central bank determines the nominal interest
monetary
rate i, which is equal to the real interest rate plus inflation
1
it = rt + π..

Let us assume that inflation is equal to the inflation target, π, which is exog-
enously determined. In general nothing guarantees that the natural rate of
interest should be positive. If the inflation target is too low, the implied
nominal interest rate may hit the zero lower bound. Then the economy falls
into a liquidity trap with some unemployment.
The second side effect is related to the booms in capital inflows and
domestic demand. These developments may lead to excessive growth in
domestic credit and asset prices. The associated risks could in principle be
kept in check by using macroprudential policy, but not all countries have
the appropriate policy instruments and frameworks in place.
1 So far the analysis here has assumed perfect financial integration, but
the analysis can easily be extended to the case where the arbitrage between
foreign and domestic bonds is limited by financial frictions, risk premia,
or capital controls. For that, one simply needs to replace rt∗ by rt∗ + τt in the
model, where the wedge τt could be interpreted as a risk premium or a tax
on capital inflows. Then it is easy to see that the country can insulate its
interest rate from foreign secular stagnation by increasing τt as the foreign
interest rate rt∗ goes down. This policy can prevent the country from falling
into a liquidity trap. However, it could be viewed as mercantilist as it
requires the real exchange rate to depreciate earlier and by a larger amount
as the country keeps foreign capital out. The real interest rate can be insu-
lated at the cost of a larger response in the real exchange rate.

114 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 5.1 Interest rates before and after the global financial
crisis, by region, 2000–16

a. Monetary policy interest rate

percent

15

10

0
2000−07 2008–16 2000−07 2008−16 2000−07 2008−16
Advanced non-Asia Emerging and Emerging and
developing Asia developing non-Asia

b. Real interest rate

percent

2000−07 2008–16 2000−07 2008−16 2000−07 2008−16

Advanced non-Asia Emerging and Emerging and
developing Asia developing non-Asia

Note: The figure compares the level of the interest rate before the crisis
(2000–07) and after the crisis (2008–16). Country classification is from the IMF
World Economic Outlook and averages are GDP-weighted.
Source: IMF, International Financial Statistics.

Interest Rates
Having established a theoretical benchmark for the macroeconomic spill-
overs associated with secular stagnation, this section looks at interest rates
in the data.
Figure 5.1 compares the levels of interest rates before the global finan-
cial crisis (2000–07) and after the crisis (2008–16). The upper and lower
panels, respectively, show the nominal policy rate and the real interest rate.
The figure compares non-Asian advanced economies with emerging-market
and developing economies, both in Asia (middle bars) and outside Asia
(right bars).
The figure shows several interesting facts. First, the zero bound con-
straint on the nominal interest rate has not prevented advanced economies
from lowering the real interest rate into negative territory after the crisis as

INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 115

they have maintained positive levels of inflation. In those economies the
real interest rate has decreased by about the same amount as the nominal
interest rate (about 2 percentage points).
Second, the crisis does not seem to have affected the nominal interest
rate in Asian emerging-market economies but real interest rates did fall on
average in that region. Remember that the averages are weighted by GDP
so that Asian emerging-market economies are dominated by China and
to a lesser extent India. In both countries the nominal interest rate barely
moved but inflation increased after the crisis (more so in India than in
China). Thus the real interest rate decreased by almost as much in Asian
emerging-market economies as in advanced economies. The same is true
of non-Asian emerging-market economies on average, although for a
different reason. In those countries inflation was significantly higher than
in advanced or Asian emerging-market economies, both before and after
the crisis, and it decreased with the crisis. The real interest rate went down
on average because of a large decrease in nominal interest rates. This differ-
ence notwithstanding, there seems to have been significant transmission
in real interest rates from advanced to emerging-market and developing
economies, both in Asia and elsewhere.
These averages mask considerable cross-country variation, as illus-
trated in figure 5.2. This figure shows a scatter plot of the policy interest
rate before the global financial crisis (averaged over 2000–07, on the hori-
zontal axis) against the same variable after the crisis (averaged over 2008–
16, on the vertical axis). Most countries are located below the 45 degree line
because their policy rates were reduced after the crisis. The figure differen-
tiates between advanced economies and emerging-market and developing
economies and labels only Asian economies.
Figure 5.2 shows that advanced and emerging-market economies have
had quite different experiences with the zero lower bound. Most advanced
economies are situated in the lower left side of the figure. These coun-
tries lowered their interest rates after the crisis, and many of them were
constrained by the zero lower bound because their rates were already rela-
tively low before the crisis. In Asia this group includes Japan (which was
already at the lower bound before the crisis), Korea, Singapore, Hong Kong,
and to a lesser extent Taiwan. Australia and New Zealand stayed away from
the zero bound because they maintained higher real interest rates than their
Asian advanced counterparts.
Most emerging-market economies also reduced their policy interest
rates but they were much less likely to come close to the zero bound
because they started from higher levels. Asian exceptions include Thailand.

116 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 5.2 Interest rates before and after the global financial
crisis, by country, 2000–16

2008−16 (percent)

15
Advanced
Emerging and developing
MNG
MMR
10
VNM

IND IDN
PNG
5 BGD
PHL
CHN
MYS AUSNZL
THA
TWN
NPL
KOR
JPN SGP HKG
0
0 5 10 15
2000−07 (percent)

Note: The figure does not include countries that had a precrisis interest rate
in excess of 15 percent. Most countries are located below the 45 degree line
because their policy rate was reduced after the crisis. See table 5.1 for country
key.
Source: IMF, International Financial Statistics.

Bulgaria is the only emerging-market economy to have a policy rate below

0.5 percent in 2016.6
Emerging-market economies had higher nominal interest rates in part
because of higher inflation, but this is not the only reason. They also had
higher real interest rates before the crisis. This is a silver lining of limited
financial integration: keeping real interest rates high leaves more room for
reducing them if necessary.
The magnitude of the fall in interest rates is very different across coun-
tries. The model presented in the previous section suggests that less finan-
cially integrated countries might be able to insulate their interest rate from
foreign secular stagnation. Is this observed in the data?
One can investigate this question with cross-country regressions of
the change in interest rate associated with the global financial crisis on
countries’ levels of financial openness. The results of four regressions are
reported in table 5.3. The changes in the real interest rates that followed
the global financial crisis were regressed on the same measures of de jure

6. Bulgaria has a currency board with the euro. No other emerging-market economy had a
policy rate below 0.5 percent in any year after 2008.

INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 117

Table 5.3 Change in real interest rate and financial integration
Variable (1) (2) (3) (4)
–0.967 4.396
De jure financial openness
(2.368) (5.577)
–0.146 –0.083
De facto financial integration
(0.234) (0.278)
–2.250 –5.083
De jure bond integration
(3.200) (5.007)
–2.12 –1.915 –1.985 –2.903
Constant
(1.519) (1.099) (2.133) (2.481)
R-squared 0.001 0.004 0.008 0.018
Number of observations 98 99 64 64
Note: The dependent variable is the change in real interest rate between 2000–07
and 2008–16 in percentage points. De jure financial openness is the Chinn-Ito index
of financial openness for 2007. De facto financial integration is the ratio of foreign
assets plus liabilities to GDP in 2007 in percentage points. De jure bond integration
is from Fernandez et al. (2016). See appendix 5A for sources. Standard errors are in
parentheses.

and de facto integration as in table 5.2, and an additional measure of bond

market integration, which is perhaps the most relevant form of integra-
tion for interest rate parity. The last measure is taken from the database
constructed by Fernandez et al. (2016). The country sample is the same as
in table 5.2.
The upshot from the table is that none of the financial openness indi-
cators explains the cross-country differences in interest rate transmission.
The absence of explanatory power is robust to many changes in the regres-
sion specification. For example, it remains true if the dependent variable is
the nominal (instead of real) interest rate, if the restrictions are to money
market flows rather than bond flows, and if de facto integration is measured
for banking specifically. It also remains true if one restricts the regression to
emerging-market and developing economies. In other words, cross-country
differences in interest rate transmission seem unrelated to countries’ open-
ness to financial flows.
In particular it does not seem that countries were willing or able to insu-
late themselves from foreign secular stagnation through capital controls.
This finding may disappoint proponents of capital flow management.
However, note that capital flow management is meant to buffer economies
against relatively high-frequency variation in capital flows, not to insulate
interest rates against persistent pressure from foreign secular stagnation.
To conclude, this analysis finds that secular stagnation affected nominal
or real interest rates in a wide range of countries irrespective of their finan-
cial openness. The implications of lower interest rates were different for
countries at different levels of economic and financial development. In
more advanced economies the main adverse effect was to bring policy rates
118 SUSTAINING
GRAPHICS ECONOMIC
- CHAPTER GROWTH IN ASIA
5—SUSTAINING ECONOMIC GROWTH IN ASIA 3
close to the zero bound constraint. Affected countries include Korea, Hong
Kong, Taiwan, and Singapore in Asia. This problem generally did not affect
emerging-market and developing economies, although Thailand is border-
line in this regard. There is no evidence that emerging-market economies
have insulated themselves against foreign secular stagnation by restricting
their financial openness.

Credit
Real interest rates have gone down in emerging-market and developing
economies. What was the impact on credit conditions? Rebalancing the
economy from foreign to domestic demand is a natural adjustment to
foreign secular stagnation, as shown in the model presented earlier. This
rebalancing could be accompanied by an increase in domestic credit. At
the same time, this expansion is problematic if it is excessive and leads to
a boom-bust cycle. The buildup of excessive credit features prominently in
discussions about financial crises. Importantly, from a policy perspective,
large credit expansions have been found to be a reliable early warning indi-
cator of banking crises or severe distress. Some models, such as Eggertsson
and Krugman (2012), explain secular stagnation in advanced economies by
the deleveraging in the private nonfinancial sector that took place after the
crisis. An excessive expansion of credit in emerging-market and developing
economies might generate concerns about a delayed transmission of secular
stagnation from advanced economies.
I look into this question by using the credit gap measures produced
by the Bank for International Settlements (BIS). BIS measures the credit-
to-GDP gap as the deviation of the credit-to-GDP ratio from its long-run
trend. This measure takes account of all sources of credit to the private
nonfinancial sector, rather than just bank credit. Drehmann (2013) finds
that total credit developments predict the risk of systemic crises better
than indicators based solely on bank credit.
Figure 5.3 shows the credit gap in advanced and emerging-market econ-
omies between 2000 and 2016. Advanced economies witnessed a sharp ex-
pansion of credit, followed by a sharp restriction after the global financial
crisis. By contrast, the credit gap expanded continuously in emerging-mar-
ket and developing economies. The magnitude of this development reflects
primarily the expansion of credit in China (the credit gap of emerging-mar-
ket economies is heavily influenced by China because the average is GDP-
weighted). But the credit gap did not expand just in China. In Asia, Hong
Kong, Singapore, and to a lesser extent Thailand, Indonesia, and Malaysia
also experienced large increases in their credit gaps.
Why do some countries have credit booms while others do not? I
found no significant cross-country correlation between credit booms and
INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 119
 Figure 5.3 Credit gap, 2000–16

percent

15 Emerging-market and
developing economies—Asia

10

5 Emerging-market
and developing
economies

−5

Advanced economies
−10
2000 2005 2010 2015

Note: The figure shows the credit gaps as measured by the Bank for International
Settlements. Country classification is from the IMF’s World Economic Outlook
(see table 5.1 for Asian countries). Averages are GDP-weighted.
Sources: IMF, World Economic Outlook; Bank for International Settlements.

low real or nominal interest rates in the data (the regression results are
not reported here). In the case of China the expansion in credit was driven
by administrative measures rather than lower market interest rates. As
shown in figure 5.4 the intensity of macroprudential measures significantly
increased after the crisis in emerging-market economies and especially in
Asia. Macroprudential policy thus appropriately leaned against the wind
but without fully containing the underlying pressure. Anecdotal evidence
suggests, in the case of China, that macroprudential restrictions often
moved the credit growth pressure from one area of the economy to another
without having serious effects on total credit.
The model presented earlier suggests that domestic financial develop-
ments may be related to the rebalancing of the economy toward domestic
demand associated with a decrease in the current account balance, i.e., an
increase in net capital inflows. Indeed, one variable that is clearly corre-
lated with credit is the current account balance. Figure 5.5 plots the change
in the credit gap between 2007 and 2016, on the vertical axis, against the
change in the current account balance over the same period, on the hori-
zontal axis. All variables are in share of GDP. There is very significant nega-
tive correlation between the two variables.
120 SUSTAINING ECONOMIC GROWTH IN ASIA
 Figure 5.4 Intensity of macroprudential policies, 2000–13

intensity

5
Emerging-market and
developing economies—Asia

Emerging-market and
3 developing economies

2
Advanced economies

2000 2005 2010 2015

Note: The figure shows the intensity of macroprudential policy as measured by

Cerrutti, Claessens, and Laeven (2015). Country classification is from the IMF’s
World Economic Outlook (see table 5.1 for Asian countries). Averages are
GDP-weighted.
Source: Cerrutti, Claessens, and Laeven (2015); IMF, World Economic Outlook.

On average the countries that had credit booms also saw an increase in
net capital inflows (or a decrease in net capital outflows). The line of best fit
implies that a 1 percent increase in net capital inflows is associated with a
3.7 percent increase in the credit gap. Several Asian emerging-market econo-
mies are in the upper-left quadrant of the figure—countries that had both a
decrease in the current account balance and an increase in domestic credit.
One might interpret this correlation as a manifestation of the global
financial cycle, with credit booms being financed by capital inflow surges.7
But there are several reasons why this is probably not the main explanation
for the correlation observed in figure 5.5. First, the current account balance
measures net capital outflows, not the much larger gross capital flows that
analysts have focused on in recent discussions of the global financial cycle.
Second, the figure focuses on a relatively long period (nine years), which
is the appropriate frequency to study secular stagnation but probably too
long for the global financial cycle. Third, several countries close to the

7. Studies on the correlation between capital inflows and credit include Magud, Reinhart,
and Vesperoni (2014) and Avdjiev et al. (2017).

INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 121

 Figure 5.5 Current account balance and credit gap, 2007–16

change in credit gap (percent)

50
MYS THA
SGP
CHN

HKG JPN
IDN

0 KOR

IND AUS
NZL

−50

−100

−20 −10 0 10 20
change in current account balance (percent)

Note: See table 5.1 for country key.

Sources: IMF, World Economic Outlook; Bank for International Settlements.

regression line were not severely affected by the global financial cycle. For
example, it is clear in the case of China that the decrease in its current
account balance and the concomitant increase in domestic credit are not
the result of a surge in capital inflows. Rather, the Chinese authorities had
to compensate for the decrease in foreign demand by raising domestic
demand, which they did through a credit stimulus.
An interpretation of figure 5.5 that is more in line with the model
presented earlier is that changes in credit reflect demand rebalancing that
followed the global financial crisis. The crisis was followed by a contrac-
tion in current account surpluses and deficits. The countries in the upper-
left quadrant of the plot are countries that, like China, had to stimulate
domestic demand in order to offset the reduction in foreign demand. The
countries in the lower-right quadrant of the plot are advanced economies
that, like the United States, had to reduce their current account deficits
and their domestic demand after the crisis (in Asia this group includes
Australia and New Zealand). The main takeaway from the figure is that
changes in domestic demand were strongly correlated with changes in
domestic credit during that period.

122 SUSTAINING ECONOMIC GROWTH IN ASIA

 Whether the credit growth observed in emerging-market economies
such as China is sustainable is a source of concern for the international
community. In its latest Article IV consultations with China the IMF
observes that among the 43 cases of credit booms in which the credit-to-
GDP ratio increased by more than 30 percentage points over a five-year
period, only five cases ended without a major growth slowdown or a finan-
cial crisis (IMF 2017b). The IMF staff recommend deep reforms to transi-
tion from the current growth model that relies on credit-fed investment
toward less credit-intensive growth. This suggests that the international
transmission of secular stagnation may go beyond what has been observed
so far since the global financial crisis. The demand rebalancing that took
place after the global financial crisis may have come at the cost of financial
developments that plant the seeds of secular stagnation in the future.

Conclusion
This analysis revolved around two symptoms of secular stagnation, low
productivity growth and low interest rates. Productivity growth and interest
rates have decreased everywhere since the early 1990s, with the global finan-
cial crisis marking a watershed in these developments. During the crisis,
spillovers went from the advanced economies that were at the epicenter of
the banking crisis to the rest of the world. Thus this chapter uses the global
financial crisis as a “natural experiment” that can shed light on the mecha-
nisms of international transmission of secular stagnation.
Productivity growth has decreased in advanced economies as well as
countries that are far from the technology frontier. This is puzzling because
the determinants of growth should not be the same in the former and in
the latter. The chapter discussed several channels involving trade and finan-
cial flows by which low productivity growth could spill over from advanced
to developing economies. However, the analysis of the data in this chapter
does not support the importance of these channels, and the reason why TFP
growth fell in so many countries at about the same time remains a mystery. I
doubt that much more information can be obtained from the type of cross-
country regressions presented in this chapter and believe that more efforts
should be put into extracting information from firm-level evidence, along
the lines of Adler et al. (2017).
Nominal interest rates hit the zero lower bound in most advanced
economies at the time of the global financial crisis, whereas they did not
seem to change very much in the rest of the world. However, this analysis
found that real interest rates decreased by the same amount in emerging-
market economies as in advanced economies. There is no evidence that
countries that were more closed financially were willing or able to insulate
their interest rates from foreign secular stagnation.
INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 123
 Low interest rate environments can have a number of adverse effects.
One is to push the economy into a liquidity trap where monetary policy
loses a lot of its traction. Another potential problem is the emergence of
unsustainable booms in capital inflows, credit, and asset prices. To some
degree credit and asset price increases should be expected as they are part
of the natural process of adjustment to foreign secular stagnation. But
these developments, if excessive and left unchecked, can also lead to costly
financial crises down the road.
Asian advanced economies behaved similarly to advanced economies
elsewhere. They had relatively low levels of inflation and nominal interest
rates before the crisis and were pulled toward liquidity traps at the time
of the crisis. Leaving Japan aside, nominal interest rates fell closest to zero
in Hong Kong and Singapore, followed by Korea and Taiwan. Australia
and New Zealand were somewhat less affected because these countries had
higher real and nominal interest rates before the crisis.
Emerging-market and developing economies generally remained at a
safe distance from liquidity traps (with Thailand perhaps an exception in
Asia). These countries were protected not so much by lower levels of inter-
national financial integration as by the fact that they had higher levels of
nominal interest rates before the crisis. This is due in part to higher infla-
tion than in advanced economies but also to higher real interest rates. Thus
what could be construed as a cost of low financial integration (a higher
cost of capital) became somewhat of a blessing in an environment of global
secular stagnation.
As for the other potential cost of low interest rates—the risk of unsus-
tainable credit booms—evidence shows that the credit boom that turned
to a bust in advanced economies was replaced by a credit boom of even
greater magnitude in emerging-market economies. Excessive expansion of
credit generates concerns of a delayed transmission of secular stagnation
from advanced economies. Global secular stagnation could enter a new
phase if the emerging-market credit boom turned into a bust.

124 SUSTAINING ECONOMIC GROWTH IN ASIA

Appendix 5A Data Sources
Interest Rates. The source for the nominal interest rate is the IMF Interna-
tional Financial Statistics. The central bank interest rate is the minimum of
the discount rate and the central bank policy rate when one or the other is
available. In selected economies where neither rate was available (Argentina,
the Czech Republic, Malaysia, Mexico, Poland, and Russia), the minimum
of the money market rate and the treasury bill rate was used.
The real exchange rate was derived by subtracting the inflation rate
(measured as the annual change in the consumer price index from the IMF
World Economic Outlook).
Productivity. Total factor productivity (TFP) is from the Penn World
Tables version 9.0 (Feenstra, Inklaar, and Timmer 2015). I used TFP at con-
stant national prices (RTFPNA), which is the recommended measure to
compare productivity growth over time in a given country. This variable is
normalized to 1 in 2005 for all countries. To obtain the TFP levels reported
in figure 5.1 RTFPNA is multiplied by the value of CTFP in 2005 (CTFP
measures the productivity level across countries in each year).
Credit. The source is credit-to-GDP gap statistics from the Bank for Inter-
national Settlements. See Dembiermont, Drehmann, and Muksakunratana
(2013).
Financial Openness. The Chinn-Ito index of financial openness is
described in Chinn and Ito (2008). This index is based on the binary dummy
variables that codify the tabulation of restrictions on cross-border financial
transactions reported in the IMF’s Annual Report on Exchange Arrangements
and Exchange Restrictions. The measure of bond market integration is from
Fernandez et al. (2016). The data to construct de facto measures of inter-
national financial integration come from the External Wealth of Nations
dataset, constructed by Lane and Milesi-Ferretti (2007).

References
Acharya, Sushant, and Julien Bengui. 2016. Liquidity Traps, Capital Flows and Currency
Wars. University of Montreal. Photocopy.
Adler, Gustavo, Romain Duval, Davide Furceri, K. Sinem, Ksenia Koloskova, and Marcos
Poplawski-Ribeiro. 2017. Gone with the Headwinds: Global Productivity. IMF Staff Discussion
Note 17/04. Washington: International Monetary Fund.
Andrews, Dan, Chiara Criscuolo, and Peter N. Gal. 2016. The Best versus the Rest: The Global
Productivity Slowdown, Divergence across Firms and the Role of Public Policy. OECD Productiv-
ity Working Paper 5. Paris: Organization for Economic Cooperation and Development.
Anzoategui, Diego, Diego Comin, Mark Gertler, and Joseba Martinez. 2016. Endogenous Tech-
nology Adoption and R&D as Sources of Business Cycle Persistence. NBER Working Paper 22005.
Cambridge, MA: National Bureau of Economic Research.

INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 125

Avdjiev, Stefan, Bryan Hardy, Sebnem Kalemli-Ozcan, and Luis Servén. 2017. Gross Capital
Inflows to Banks, Corporates and Sovereigns. NBER Working Paper 23116. Cambridge, MA:
National Bureau of Economic Research.
Benigno, Gianluca, and Luca Fornaro. 2017. Stagnation Traps. ECB Working Paper 2038.
Frankfurt: European Central Bank.
Bustos, Paula. 2011. Trade Liberalization, Exports and Technology Upgrading: Evidence on
the Impact of MERCOSUR on Argentinean Firms. American Economic Review 101, no. 1:
304–40.
Caballero, Ricardo J., Emmanuel Farhi, and Pierre-Olivier Gourinchas. 2015. Global Imbalances
and Currency Wars at the ZLB. NBER Working Paper 21670. Cambridge, MA: National Bu-
reau of Economic Research.
Caballero, Ricardo J., Takeo Hoshi, and Anil K Kashyap. 2008. Zombie Lending and De-
pressed Restructuring in Japan. American Economic Review 98, no. 5: 1943–77.
Cerrutti, Eugenio, Stijn Claessens, and Luc Laeven. 2015. The Use and Effectiveness of Macropru-
dential Policies: New Evidence. IMF Working Paper 15/21. Washington: International Mon-
etary Fund.
Cetorelli, Nicola, and Linda S. Goldberg. 2011. Global Banks and International Shock Trans-
mission: Evidence from the Crisis. IMF Economic Review 59: 41–6. Washington: Interna-
tional Monetary Fund.
Chinn, Menzie D., and Hiro Ito. 2008. A New Measure of Financial Openness. Journal of Com-
parative Policy Analysis 10, no. 3: 309–22.
Coe, David T., and Elhanan Helpman. 1995. International R&D Spillovers. European Economic
Review 39, no. 5: 859–87.
Coeurdacier, Nicolas, Stéphane Guibaud, and Keyu Jin. 2015. Credit Constraints and Growth
in a Global Economy. American Economic Review 105, no. 9: 2838–81.
Cook, David, and Michael B. Devereux. 2013. Sharing the Burden: Monetary and Fiscal Re-
sponses to a World Liquidity Trap. American Economic Journal: Macroeconomics 5, no. 3:
190–228.
Corsetti, Giancarlo, Eleonora Mavroeidi, Gregory Thwaites, and Martin Wolf. 2017. Step
Away from the Zero Lower Bound: Small Open Economies in a World of Secular Stagna-
tion. Cambridge University. Photocopy.
De Gregorio, José. 2017. Productivity in Emerging-Market Economies: Slowdown or Stagna-
tion? Paper presented at conference on the Policy Implications of Sustained Low Produc-
tivity Growth, Peterson Institute for International Economics, Washington, November 9.
De Loecker, Jan. 2007. Do Exports Generate Higher Productivity? Evidence from Slovenia.
Journal of International Economics 73: 69–98.
Dembiermont, Christian, Mathias Drehmann, and Siriporn Muksakunratana. 2013. How
much does the private sector really borrow? A new database for total credit to the private
non-financial sector. BIS Quarterly Review (March): 65–81. Basel: Bank for International
Settlements.
Devereux, Michael B., and James Yetman. 2014. Capital Controls, Global Liquidity Traps, and
the International Policy Trilemma. Scandinavian Journal of Economics 116, no. 1: 158–89.
Drehmann, Mathias. 2013. Total Credit as an Early Warning Indicator for Systemic Banking
Crises. BIS Quarterly Review (June): 41–45. Basel: Bank for International Settlements.
Eggertsson, Gauti B., and Paul Krugman. 2012. Debt, Deleveraging, and the Liquidity Trap: A
Fisher-Minsky-Koo Approach. Quarterly Journal of Economics 127, no. 3: 1469–513.
Eggertsson, Gauti B., Neil R. Mehrotra, Sanjay R. Singh, and Lawrence H. Summers. 2016. A
Contagious Malady? Open Economy Dimensions of Secular Stagnation. IMF Economic
Review 64, no. 4: 581–634.

126 SUSTAINING ECONOMIC GROWTH IN ASIA

Feenstra, Robert C., Robert Inklaar, and Marcel P. Timmer. 2015. The Next Generation of the
Penn World Table. American Economic Review 105, no. 10: 3150–82.
Fernald, John G. 2015. Productivity and Potential Output Before, During, and After the Great
Recession. NBER Macroeconomics Annual 29, no. 1: 1–51. Cambridge, MA: National Bureau
of Economic Research.
Fernandez, Andrés, Michael W. Klein, Alessandro Rebucci, Martin Schindler, and Martin
Uribe. 2016. Capital Control Measures: A New Dataset. IMF Economic Review 64, no. 3:
548–74. Washington: International Monetary Fund.
Frankel, Jeffrey, and David Romer. 1999. Does Trade Cause Growth? American Economic Re-
view 89, no. 3: 380–99.
Fujiwara, Ippei, Tomoyuki Nakajima, Nao Sudo, and Yuki Teranishi. 2013. Global Liquidity
Trap. Journal of Monetary Economics 60, no. 3: 936–49.
Gordon, Robert J. 2016. The Rise and Fall of American Growth: The US Standard of Living since the
Civil War. Princeton, NJ: Princeton University Press.
Hausmann-Guil, Guillermo, Eric van Wincoop, and Gang Zhang. 2016. The Great Recession:
Divide between Integrated and Less Integrated Countries. IMF Economic Review 64, no. 1:
134–76. Washington: International Monetary Fund.
IMF (International Monetary Fund). 2013. Dancing Together? Spillovers, Common Shocks,
and the Role of Financial and Trade Linkages. Chapter 3 in World Economic Outlook (Oc-
tober). Washington.
IMF (International Monetary Fund). 2016. Global Trade: What’s Behind the Slowdown?
Chapter 2 in World Economic Outlook (October). Washington.
IMF (International Monetary Fund). 2017a. Asia at Risk of Growing Old before Becoming
Rich? In Regional Economic Outlook, Asia and Pacific (April). Washington.
IMF (International Monetary Fund). 2017b. People’s Republic of China. IMF Country Reports
17/247 and 17/248. Washington.
Kalemli-Ozcan, Sebnem, Elias Papaioannou, and Fabrizio Perri. 2013. Global Banks and Cri-
sis Transmission. Journal of International Economics 89, no. 2: 495–510.
Kamin, Steven B., and Laurie Pounder DeMarco. 2012. How did a domestic housing slump
turn into a global financial crisis? Journal of International Money and Finance 31, no. 1: 10–41.
Keller, Wolfgang. 2004. International Technology Diffusion. Journal of Economic Literature 42,
no. 3 (September): 752–82.
Lane, Philip R., and Gian Maria Milesi-Ferretti. 2007. The External Wealth of Nations Mark
II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970–2004. Journal of
International Economics 73, no. 2: 223–50.
Lane, Philip R., and Gian Maria Milesi-Ferretti. 2011. The Cross-Country Incidence of the
Global Crisis. IMF Economic Review 59, no. 1: 77–110. Washington: International Mon-
etary Fund.
Lane, Philip R., and Gian Maria Milesi-Ferretti. 2017. International Financial Integration in the
Aftermath of the Global Financial Crisis. IMF Working Paper 17/115. Washington: Interna-
tional Monetary Fund.
Lileeva, A., and D. Trefler. 2010. Improved Access to Foreign Markets Raises Plant-level Pro-
ductivity … For Some Plants. Quarterly Journal of Economics 125, no. 3: 1051–99.
Magud, Nicolas E., Carmen M. Reinhart, and Esteban R. Vesperoni. 2014. Capital Inflows,
Exchange Rate Flexibility and Credit Booms. Review of Development Economics 18, no. 3:
415–30.
Rodrik, Dani. 1988. Closing the Technology Gap: Does Trade Liberalization Really Help? NBER
Working Paper 2654. Cambridge, MA: National Bureau of Economic Research.

INTERNATIONAL TRANSMISSION AND POLICY SPILLOVERS 127

Rose, Andrew K., and Mark M. Spiegel. 2011. Cross-Country Causes and Consequences of the
Crisis: An Update. European Economic Review 55, no. 3: 309–24.
Yeaple, Stephen Ross. 2005. A Simple Model of Firm Heterogeneity, International Trade, and
Wages. Journal of international Economics 65, no. 1: 1–20.

128 SUSTAINING ECONOMIC GROWTH IN ASIA

 6
Getting Out of Secular
Stagnation: Turning Japanese
May Be a Good Thing Now
ADAM S. POSEN

Economists used to play a game asking which country was turning Japa-
nese. Back in 2003, I wrote a paper playing off this game, where I assessed
whether Germany might fall into a Japan-like trap of persistent low growth
(Posen 2003). Today, the fear of suffering Japan’s fate persists for advanced
economies, including those in Asia (see chapter 8 by Dongchul Cho and
Kyooho Kwon in this volume). But the analogy is mistaken. Rather than
focusing on the risk of turning Japanese in the late 1990s sense, it would be
better for countries in Asia to learn how Japan got itself off the growth floor
from 2003 to 2018. It is true that both productivity growth and labor force
growth have slowed in most of the advanced economies and an increasing
number of emerging markets, even in Asia.
There is a popular but incorrect characterization of Japan as having suf-
fered two lost decades. Japan lost one decade, where the effect of an asset
price collapse was prolonged and deepened by a set of idiosyncratic Japan-
specific factors, many of which were due to policy (Posen 2010, Kuttner
and Posen 2002). The lost decade ended with the Koizumi administration
and the cleanup of the country’s banks in late 2002/early 2003. Since then,
per capita GDP growth in Japan has actually been quite good, especially in
comparison to the rest of the G-7 nations, even leaving aside the global fi-
nancial crisis. Productivity growth has been decent. If anything, the macro-
economic challenges that Japan faced in failing to increase inflation and to

Adam S. Posen is the president of the Peterson Institute for International Economics.

129
reduce government debt have been faced simultaneously by other countries
and so are no longer Japan-specific. The puzzling failure of inflation to in-
crease, despite best-practice aggressive monetary policy, remains troubling
but should not obscure the progress in the real economy.
Japan’s growth and productivity performance stands out against demo-
graphic decline at home and increasing stagnation and even crisis among
many of its trading partners. The eventual cleanup of the financial system
and reversal of mistaken contractionary macroeconomic policies were nec-
essary to end the decline (Posen 2010). But they are not sufficient to explain
the long and resilient subsequent recovery. Instead, Japan took major steps
to address the sources of secular stagnation that were open to domestic ame-
lioration: using the fiscal room still available to smooth exogenous shocks
like the Fukushima natural disasters and the spillovers from the global fi-
nancial crisis; partially offsetting the demographic-driven shrinkage of the
labor force by increasing female labor force participation and employment
flexibility; and in recent years, under Prime Minister Shinzō Abe, pursu-
ing greater economic integration with the world, notably via the Compre-
hensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP)
and EU-Japan trade agreements. These steps were taken slowly and did not
fulfill all the promises of Abenomics, let alone the free market fantasies of
some Western commentators. These three initiatives, however, when accel-
erated by Abe, did materially improve Japan’s real economic performance
and resilience. As such, this three-part framework—fiscal stabilization, labor
market reform, and trade integration—provides a useful model for emerg-
ing Asia today.
Reconsidering the Japanese experience along these lines raises three
questions. First, has recent underperformance in Japan been a puzzle for
modern macroeconomics? In other words, since Abenomics has basically
done many of the things that we were all calling for—including aggressive
monetary stimulus with a forward-looking positive inflation target—why
has it not resulted in consistently positive inflation near the target? I believe
that this can in part be explained by the overemphasis on a credibility-cen-
tered view of macroeconomic policy (Posen 2012) and by a failure to prop-
erly account for the effects of nonmonetary factors on inflation.
Second, does the apparent failure to raise the inflation rate in Japan
matter for economic performance? My sense is that it does, but much less
significantly than many of us initially thought. Despite consistently failing
to reach its inflation target, Japan never fell into the much feared deflation
spirals that the United States, for example, experienced in the 1930s. For
this reason, the growth and welfare costs of low inflation are much lower
than we initially believed, although they are not negligible.

130 SUSTAINING ECONOMIC GROWTH IN ASIA

 Third, how should we think about the political economy in a context
where the punishments for bad policies and the rewards for good policies
are, if not completely muted, at least substantially delayed? How should we
think about the political economy of reform if inadequate monetary poli-
cies do not translate into deflation spirals and if long-term interest rates do
not go up against the backdrop of unsustainable fiscal policies? In Posen
(2003), I argued that the pressure provided by European regional integra-
tion saved Germany from itself. If anything, this lesson applies more to Asia
today than to Europe then.

Fiscal Room and Expectation-Centered View of Macro

Policies
It is often said that Abenomics has generated a series of weird and hard
to rationalize economic facts. Monetary policy appears unable to offset the
impact of fiscal shocks (e.g., as in 2014 with the rise in the consumption tax).
Inflation and imports appear to be insensitive to the long and sustained
depreciation of the yen. If not dead, the Phillips curve is at least dormant.
And most stunningly of all, Japan continues to rack up public debt and its
long-term interest rates do not rise. The discrepancy between many econo-
mists’ predictions and essential observations reflects two shortcomings of
our standard models. First, too much emphasis is put on forward-looking
expectations. Second, not enough emphasis is put on the short- and medi-
um-term impacts of nonmonetary factors on inflation and output.
Take fiscal policy, where our academic models assume that, at the
long-run horizon, all debts must be paid off completely somehow and that
implies some Ricardian equivalence affecting today’s behavior of savers. In
this world, the impact of an increase in the consumption tax should have
been particularly small given Japan’s large stock of public debt. After all,
these debts would have to be repaid at some point by this or near future
generations of consumers. To the extent that this generation cares about
future generations, changing the particular time profile for repaying this
liability should not have a large economic impact. Unfortunately, the tax
hikes of 1997 and 2014, among others, did have a large recessionary impact.
Yet those of us who discounted the importance of this notion of perfect
foresight were not surprised (Kuttner and Posen 2002).
In the spirit of the view that inflation is “always and everywhere a mon-
etary phenomenon” (Friedman 1963), market monetarists also argued that
the Abe 2014 consumption tax hike should have had no negative impact
as long as the fiscal shock could be fully offset by monetary policy. In this
view, the negative impact of the value-added tax hike on output and infla-
tion reflected a failure by the Bank of Japan (BoJ) to provide enough mon-

GETTING OUT OF SECULAR STAGNATION 131

etary stimulus. At this point, it is, however, worth remembering that since
the bank’s new leadership took over in April 2013, the BoJ has done pretty
much everything that the likes of Paul Krugman, Ben Bernanke, Lars Svens-
son, and I asked them to do a decade earlier.
The BoJ has announced a positive inflation target. It has bought long-
term rather than short-term bonds. It has let the exchange rate go. It has
coordinated with the government when fiscal expansion occurred. Some
would even say that it has compromised its independence, which I think is
inaccurate but should be inflationary as a market perception. In any case,
this is what the liquidity trap literature recommended (Krugman 1998).
Despite this, we have not seen huge inflation, and the contraction in growth
due to the consumption tax increase was not offset.
Remember, the fact that we are not seeing major moves in interest rates
in response to much higher public debt levels is not specific to Japan. We are
not just disproving Reinhart and Rogoff’s (2010) mythical 90 percent debt-
to-GDP ratio as a tipping point for government debt sustainability but also
learning that our simple framework for distinguishing permanent and tem-
porary tax policy impacts is inadequate. This is not to say that even Japan’s
net debt (currently at 160 percent of GDP) is in any sense on a sustainable
path, given demographics and healthcare commitments. It is to say that
some measures of financial repression and monetary financing are clearly
more effective at calming individual behavior than one would have been led
to believe by much of the economics profession in recent years.
Arguably, what governments spend or tax, and when they do it, may
matter more than debt levels. Governments in emerging Asia have long been
predisposed to being fiscally conservative for good reason. The history of
their counterparts among the economies of Latin America and Eastern and
Southern Europe serves as a warning against budgetary laxity. Japan’s expe-
rience, however, is an equally valid caution against abjuring fiscal stabiliza-
tion too blindly. There are shocks, often financial, that are too sizable and
panic inducing to be dealt with by monetary policy alone. When Japan reso-
lutely refused to use the fiscal means it had in the 1990s and early 2000s, it
made matters far worse (Posen 1998, 2010). Of course, the problems caused
by premature austerity starting in 2010 for most of the European Union,
including the United Kingdom, and to a lesser degree the United States are
self-evident. While evidence on labor market hysteresis is mixed, it is clear
that the negative effects of persistent downturns on investment and politics
are asymmetrically larger than the costs of extended booms. Part of Japan’s
success since 2003 has been a more flexible fiscal response when needed.
Determining available fiscal room is challenging both analytically and
politically. Even robust clear analysis will not always restrain political desire
for wasteful procyclical fiscal policies (as exemplified by the US fiscal stance

132 SUSTAINING ECONOMIC GROWTH IN ASIA

of 2018–19). Temporary targeted stimulus programs linked to a clear shock
are different, however, and seem to be recognized as such by markets and
households. The first important thing is to try to keep fiscal policy on a
broadly sustainable trend in terms of limits on primary deficits so that there
is room to aid. The second key measure is to have savers who maintain faith
in domestic currency–denominated assets, if not excessive home bias. On
both these counts, much of emerging Asia shares the attributes that enable
Japan to maintain fiscal stabilization capacity even today: high household
saving rates, local institutions and vehicles for safety depositing those
savings, monetary (a.k.a. price) stability, common support for public invest-
ment projects, and the ability to collect taxes when necessary. Given these
widespread attributes, it would be wise for Asian governments to advance
their fiscal conservatism to allow for stabilization policy against negative
shocks, as Japan has rightly done.

Reemphasizing the Role of Demographic and Structural

Factors in Growth
The view that the BoJ has not done enough to raise inflation expectations is
just not convincing. An alternative hypothesis is that our standard models
have simply a too naïve view of the inflation process. In particular, they fail
to recognize that nonmonetary factors have a major influence on inflation
in the short and medium runs. Once we account for these factors and rec-
ognize that they have over the past decade all been disinflationary, then
the difficulty faced by Japan and its spread to other advanced economies
become less puzzling.
Take recent labor market changes, notably the addition of many female
workers and workers working flexible hours rather than traditional full-
time Japanese employment patterns (see chapter 7 by Kyoji Fukao in this
volume). People generally fail to comprehend the scale of these rapid changes
in Japan. “Womenomics,” as the policy of incentives to bring more women
into the workforce has been dubbed, has raised female labor force participa-
tion by over 3 percentage points in about five years. This means that more
than 1.5 million women have rejoined a shrinking workforce of about 60
million. Many of these women have joined the workforce on a part-time or
flexible basis, but we know from the Nordic experience that this is one way
of retaining women in the workforce, so it should not be discounted.
Such a large structural shock is disinflationary, if not deflationary, in
the short run. First, in the absence of an increase in labor demand, increases
in the supply of labor decrease wages and thus prices. Already in Germany
in 2003, it was clear that the Hartz IV labor market reform had this side
effect. Second, and often less understood, it is deflationary because of how

GETTING OUT OF SECULAR STAGNATION 133

it changes the composition of the labor force. Bringing more women into
the labor force means bringing in people who are paid less than comparable
male workers because of both differentials in seniority and unfair treatment.
An increase in part-time workers also has a similar effect on the aggregate.
We do not yet have a specific quantification of this effect for Japan, but a
recent decomposition by Daly, Hobijn, and Pyle (2016) for the United States
suggests that these composition effects actually account for the main drag
on wage growth between 2010 and 2016. Over these years, the United States
has seen large outflows of baby boomers from the labor force, large inflows
of new entrants into the labor force, and large flows of workers from part-
time to full-time jobs. As the retiring baby boomers typically earn above the
median, these margins pull down median earnings growth. The effects are
much larger in Japan given the scale of the changes.
From a productivity point of view, the increase in female labor is hope­
ful because the most productive women are the first to come back into the
workforce. These women are presumably displacing or competing with
some men who are underqualified for the jobs they hold (even if sexism
dampens this effect). So it is a positive supply shock, with an additional
deflationary effect on aggregate wages.
There is a tendency to take demographics as destiny. This view is related
in spirit as well as substance to the idea of the middle-income trap. Both
compel policymakers to be realistic about the inability to sustain high GDP
growth rates as economies age and develop. In chapter 8 in this volume, Cho
and Kwon make a strong case of this sort for Japan, Korea, and China in
sequence. It is better as well as quite feasible to treat these factors like a golf
handicap—one cannot expect to putt or drive in one’s 50s as one did in one’s
20s, let alone score the same as a touring pro. A golfer can, however, con-
tinue to work on her or his swing, get the right clubs, and work to improve
even if their handicaps do not go all the way to zero. Not only is the game
still worth playing but also expectations can be beaten.
The transformation of the Japanese labor supply has been tremendous
through a combination of female labor force participation increases, greater
flexibility in work hours, and contracting and informal acceptance of more
migrant workers. This has all been achieved with little fiscal effect in terms
of changes to benefits and taxes and with no social dislocation of note. The
result has been a postponement of labor force shrinkage by more than five
years and improved productivity to boot. Though Japan and Korea were and
remain laggards in gender equality in employment, all of emerging Asia can
offset the effects of aging by better utilizing female labor supply. Enhance-
ments to labor market flexibility, particularly for young people—if their em-
ployment remains in the formal sector of the economy—can serve a similar
goal.

134 SUSTAINING ECONOMIC GROWTH IN ASIA

 Unlike during the 1990s, many of the observed phenomena in Japan
today—low wage growth, low exchange rate pass-through to inflation, a
barely visible short-run Phillips curve, low long-term interest rates despite
debt accumulation, low response of imports to depreciation, among others—
are simultaneously occurring worldwide. The depreciation of the yen
between 2013 and 2016 from ¥79 to ¥120 against the dollar without appre-
ciable inflation has raised issues to do with exchange rate pass-through and
the trade balance. But this issue of diminished exchange rate pass-through
is hardly specific to Japan. For economies at the upper end of the interna-
tional division of labor, this is the result of global supply chains, network
effects, and trade in services. Like Japan, the United Kingdom showed a lack
of sustained inflation passthrough following a 25 percent depreciation of
the sterling in 2008.
Integration of production is certainly not enough to fully explain the
limited net export improvement following a depreciation, especially in so-
cieties where household consumption of imported goods remains high and
where the manufacturing share of GDP is declining. But if this is a puzzle, it
is a more interesting and deeper puzzle than just one about understanding
Japan.
Even the absence of large balance sheet effects (Koo 2003) as an engine
for investment is not specific to Japan. The situation in the United States
and some Western European countries is similar. There is no question that
the resolution of the banking crisis in 2003 undertaken by Heizo Takenaka,
then minister of financial services, which included recapitalization and con-
solidation of the banking sector, was a necessary condition to get Japan out
of its lost decade. But it must be reckoned that while fixing a banking crisis
prevents bad outcomes, it does not seem to be sufficient to stimulate good
outcomes. Clean balance sheets across the Japanese economy seem to have
brought less benefit in terms of growth and investment than we expected.
Demand had to come from abroad, as did supply-enhancing competition
and efficiencies.

Can Regional Integration Prevent Domestic Stagnation

Traps?
In Posen (2003) I argued that for an advanced economy to perpetually stag-
nate, its political economy must have four syndromes: (1) incomplete finan-
cial liberalization where nonperforming banks are not allowed to fail despite
deregulation; (2) uncoordinated deflationary macroeconomic policy; (3) fi-
nancially and politically passive citizens; and (4) a lack of openness to trade
or capital flows or foreign ideas.

GETTING OUT OF SECULAR STAGNATION 135

 What made Japan different was that obvious economic underperfor-
mance, fed by regulatory neglect and deflationary policies (not to mention
overt corruption), would not and did not turn into public outcry. Govern-
ment and established interests had eluded most pressures for change. Pas-
sivity was reinforced by the closed nature of much of Japan’s economy and
society, as reflected by its discouragement of immigration and the virtual
absence of regional security or trade integration.
Germany, in contrast, was spared from this passivity by its long-stand-
ing openness and commitment to international economic integration. If
anything, it was forced to be more open, it was forced to be more responsive,
it was forced to better coordinate monetary and fiscal policies by the Euro-
pean Central Bank, albeit against its will.
That then raises the question of whether there is now enough open-
ness and integration in Asia to forestall falling into the trap of repressing
savings, in which an economy’s savers cannot force accountability either by
exiting the system (taking capital out) or by voicing complaints effectively.
It also raises the question whether aging groups will push for deflation and
protectionism. I agree with Subir Gokarn (see chapter 17 in this volume)
that much of emerging Asia isn’t there yet. But a new Asian growth impetus
focused on regional integration makes economic sense. From a political
economy point of view, integration would also be beneficial as it would help
generate external pressure against the status quo, which is still missing in
many Asian economies. The CPTPP, Regional Comprehensive Economic
Partnership, and other plurilateral trade arrangements, as well as decentral-
ized efforts to increase intraregional trade and investment, will pay off along
multiple dimensions—as Japan is belatedly but now decidedly experiencing.

References
Daly, Mary C., Bart Hobijn, and Benjamin Pyle. 2016. What’s Up with Wage Growth? FRBSF
Economic Letter 2016-07 (March 7). San Francisco, CA: Federal Reserve Bank of San Fran-
cisco.
Friedman, Milton. 1963. Inflation: Causes and Consequences. Bombay: Asian Publishing House.
Koo, Richard. 2003. Balance Sheet Recession: Japan’s Struggle with Uncharted Economics and Its Glob-
al Implications. Hoboken, NJ: Wiley and Sons.
Krugman, Paul. 1998. It’s Baaack: Japan’s Slump and the Return of the Liquidity Trap. Brookings
Papers on Economic Activity 29, no. 2: 137–206. Washington: Economic Studies Program,
Brookings Institution.
Kuttner, Kenneth N., and Adam S. Posen. 2002. Fiscal Policy Effectiveness in Japan. Journal of
the Japanese and International Economies 16, no. 4 (December): 536–58.
Posen, Adam S. 1998. Restoring Japan’s Economic Growth. Washington: Institute for Interna-
tional Economics.
Posen, Adam S. 2003. Is Germany Turning Japanese? Working Paper 03-2. Washington: Institute
for International Economics.

136 SUSTAINING ECONOMIC GROWTH IN ASIA

Posen, Adam S. 2010. The Realities and Relevance of Japan’s Great Recession: Neither Ran Nor
Rashomon. PIIE Working Paper 10-7. Washington: Peterson Institute for International
Economics.
Posen, Adam S. 2012. Comments on “Methods of Policy Accommodation at the Interest-Rate Lower
Bound” by Michael Woodford. FRBKC Economic Policy Symposium on the Changing Policy
Landscape, Jackson Hole, WY, August 31. Available at www.kansascityfed.org/publicat/
sympos/2012/posen.pdf.
Reinhart, C. M., and K. S. Rogoff. 2010. Growth in a Time of Debt. American Economic Review
100: 573–78.

GETTING OUT OF SECULAR STAGNATION 137

 7
Secular Stagnation and the
Labor Market in Japan
KYOJI FUKAO

Japan has been suffering from secular stagnation since the economic bubble
burst in 1989–90. Its labor productivity stopped catching up with that
of the United States around 1990 (figure 7.1). In fact, labor productivity
(measured in purchasing power parity [PPP]) in Japan today is much lower
than in other major OECD countries except South Korea (figure 7.2). This
chapter examines the structural causes of Japan’s secular stagnation, focus-
ing on labor market issues.
A number of studies have attempted to identify the causes of Japan’s
secular stagnation. Most scholars seem to agree that there are two major
structural causes of the stagnation, and they are related to each other: in-
sufficient final demand (excess saving problem) and slow total factor pro-
ductivity (TFP) growth.1 When TFP growth is slow, fixed capital formation
becomes sluggish, which reduces final demand. On the other hand, when
there is excess supply, firms are reluctant to invest in intangibles, which
reduces TFP growth.
Based on this diagnosis, in early 2013 the Japanese government em-
barked on so-called Abenomics, which consists of three arrows: aggressive
monetary easing, fiscal stimulus, and a growth strategy (structural reforms)
to tackle the twin problems of insufficient demand and slow TFP growth

Kyoji Fukao is professor at the Institute of Economic Research, Hitotsubashi University, and president of the
Institute of Developing Economies, Japan External Trade Organization.
1. For a more general discussion of these structural causes, see Fukao (2013) and Fukao et
al. (2016b).

139
 Figure 7.1 Labor productivity in Japan and the United States,
1947–2012

US dollars in 2012 prices per hour worked

70
60
50
40
30
20
10
0
1947 1952 1957 1962 1967 1972 1977 1982 1987 1992 1997 2002 2007 2012

United States, gross value added per hour worked, US dollars

in 2012 prices
Japan, gross value added per hour worked, converted into
US dollars in 2012 prices by market exchange rate of 2012
(79.8 yen/dollar)
Japan, gross value added per hour worked, converted into
US dollars in 2012 prices by 2012 PPP (106.0 yen/dollar)

PPP = purchasing power parity

Source: OECD Statistics, https://stats.oecd.org.

Figure 7.2 Labor productivity in major OECD countries, 2014

Korea
Japan
OECD—total
European Union
(28 countries)
Italy
United Kingdom
Canada
Euro area
(19 countries)
Australia
G-7
Germany
France
United States

0 10 20 30 40 50 60 70
US dollars in constant prices, 2010 PPP

PPP = purchasing power parity

Source: OECD Statistics, https://stats.oecd.org.

140 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 7.3 Total factor productivity of Japan’s macroeconomy,
1995–2015

index (1995 = 1)
1.02
1.01
1.00
0.99
0.98
0.97
0.96
0.95
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Source: Conference Board, Total Economy Database: Growth Accounting and

Total Factor Productivity, 1995–2015 (adjusted version).

simultaneously. The first two arrows seem to have resolved the problem of
insufficient demand to some extent. The recovery of major economies such
as the United States and China has helped gradually shrink the GDP gap,
which the Cabinet Office estimated had declined to minus 0.4 percent by
the fourth quarter of 2016. Japan’s labor market has also been tightening,
with the ratio of job openings to applicants reaching 1.45 in March 2017,
the highest since November 1990.
Major structural reforms of the third arrow are (1) stimulating private
investment in targeted sectors such as medical services, renewable energy,
and agriculture; (2) promoting female labor participation; (3) pushing for
the Trans-Pacific Partnership (TPP); and (4) deregulating targeted sectors,
such as healthcare, agriculture, and energy. Despite these reforms, the TFP
of Japan’s macroeconomy has hardly changed since 2013 (figure 7.3), and
private investment has not accelerated.
The disappointing results of the structural reforms prompted the
Japanese government in October 2015 to introduce Abenomics 2.0, labeled
“All 100 Million Playing an Active Role” [Ichi-Oku Sou-Katsuyaku], which
consists of four pillars: (1) a strong economy (continuation of active mone-
tary and fiscal stimulus); (2) support for child rearing; (3) improvement of
the social security system; and (4) introduction of an “equal work, equal
pay” rule. The government also introduced life-work balance policies in
an attempt to reduce overtime work. These reforms focus more on labor
market and social policy issues than did the first round of reforms. This
change seems to partly reflect the new strategy of the ruling parties, the
Liberal Democratic Party and Komeito, to incorporate traditional policies
of the opposition party, the Democratic Party. At the same time, it also

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 141

 Figure 7.4 Literacy proficiency among 16–65-year-olds, major
OECD countries, 2011–12

Japan
Australia
Canada
Average
Korea
England/Northern
Ireland (United Kingdom)
Germany
United States
Austria
France
Italy

220 230 240 250 260 270 280 290 300

literacy score

Source: OECD (2016a).

Figure 7.5 Literacy proficiency by country and gender, 2011–12

deviation from average divided by standard deviation

0.6

0.4

0.2

–0.2

–0.4

–0.6

–0.8

–1.0
Male Female Male Female Male Female

Japan United Kingdom United States

Note: The average and the standard deviation are calculated from all the microdata
of about 10,000 adults (both males and females) in the three countries.
Source: Kawaguchi (2017).

reflects the widespread view in Japan that labor market issues are the key
to Japan’s revitalization.
The OECD’s Survey of Adult Skills (PIAAC) ranks Japan at the top
among OECD countries in terms of adult proficiency in key information-
processing skills such as literacy, numeracy, and problem solving in tech-
nology-rich environments (figures 7.4 and 7.5 show results on literacy).

142 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 7.6 Use of literacy proficiencies at work, by country and
gender, 2011–12

deviation from average divided by standard deviation

0.15
0.10
0.05
0
–0.05
–0.10
–0.15
–0.20
–0.25
–0.30
–0.35
Male Female Male Female Male Female

Japan United Kingdom United States

Note: Based on the item response theory (OECD 2013), Kawaguchi estimated these
scores from responses to questions on use of literacy proficiency at work. The
average and the standard deviation are calculated from all the microdata of about
10,000 adults (both males and females) in the three countries.
Source: Kawaguchi (2017).

Japan’s problem is that it does not fully utilize its labor, especially female
workers, whose proficiency is barely utilized at work (figure 7.6).
Greater utilization of workers’ proficiency would raise Japan’s productiv-
ity and labor efficiency. Since one of the main causes of insufficient demand
in Japan is the stagnation of capital formation, which is partly caused by
the shrinking working-age population and low TFP growth, labor market
reforms will also contribute to relieving Japan’s two structural problems.
This chapter analyzes two key issues in Japan’s labor market—nonreg-
ular employment and productivity and wage gaps between large and small
firms (dual labor market problem)—and derives policy implications.

Japan’s Nonregular Employment Problem

The share of nonregular employees in total workers has increased substan-
tially since the end of the 1980s.2 In the nonmanufacturing sector (in both
the nonmarket and market economies),3 the share of nonregular employees

2. In Japan, a regular employee “is generally considered as an employee who is hired directly
by his/her employer without a predetermined period of employment, and works for sched-
uled hours…. Consequently, a ‘nonregular employee’ is an employee who does not meet one
of the conditions for regular employment” (Asao 2011).
3. Nonmarket economy comprises general government and nonprofit institutions of educa-
tion, medical services, and other services.

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 143

Figure 7.7 Share of nonregular employees in total workers, by sector,
1970–2012

percent
30
Nonmanufacturing (nonmarket economy)
25 Nonmanufacturing (market economy)
Manufacturing
Primary sector
20

15

10

0
1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 2010

Source: Japan Industrial Productivity Database 2015, Hitotsubashi University and Research
Institute of Economy, Trade and Industry (RIETI).

is around 30 percent (figure 7.7). The wage gap between nonregular and
regular employees is large (figure 7.8). Wages of regular workers increase as
they get older and accumulate experience in the workplace.
Labor economists have argued that two factors are behind this upward
slope of the age-wage profile. First, as workers accumulate human capital,
their marginal contribution to production increases, which raises their
wages. Second, to enhance worker loyalty and incentivize them, firms defer
compensation until workers are older. Japanese firms provide active on-the-
job training and some off-the-job training for regular employees (Fukao et
al. 2009) but not much for nonregular employees. This may be why the age-
wage profile of regular employees in Japan is steeper than that in other de-
veloped countries (Fukao et al. 2006).
If the upward slope of the age-wage profile is mainly caused by an
increase in the marginal productivity of regular employees through the
accumulation of human capital, then the wage gap between regular and
nonregular workers in figure 7.8 can be regarded as representing the differ-
ence in labor quality between the two, meaning that the recent increase in
nonregular workers has important implications. Using the Japan Industrial
Productivity (JIP) Database,4 one can estimate the impact of the increase

4. The database is available at www.rieti.go.jp/en/database/jip.html.

144 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 7.8 Wage level as a percent of the average wage of regular
employees, June 2015

percent
140
Regular
120 Nonregular

100

80

60

40

20

0
20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69
age group

Source: OECD (2017). The original data are from Ministry of Health, Labour, and
Welfare, Basic Survey on Wage Structure 2015. Overtime payments and bonuses
are not included.

in nonregular employees on Japan’s labor quality through counterfactual

simulation. If the share of nonregular employees in total workers had not
increased since 1988, the average quality of Japanese labor would have been
8 percent higher than it is now.5
To examine the relationship between the wage profile and the marginal
productivity of employees, Fukao et al. (2006) estimated both the marginal
productivity and wage rate of part-time employees compared with those of
regular employees, using employer-employee matched data at the factory
level. We found that the productivity gap between part-time and regular
employees is larger than the wage gap. This finding means that firms pay
a premium to part-time workers to obtain flexibility of employment. We
also found that although there is some deferred compensation, the major

5. Suppose that Japan is on the kind of balanced growth path assumed in standard neoclas-
sical growth models. An 8 percent improvement in labor quality will raise Japan’s real GDP by
8 percent. Since the labor income share in Japan is about two-thirds, 5.3 percentage points of
the increase in GDP will be due to the increase in (quality-adjusted) labor input and 2.7 per-
centage points will be due to the increase in capital input induced by the labor input increase.

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 145

part of regular employees’ wage increase with age reflects increases in their
marginal productivity with age.6
It seems that firms are increasing the number of part-time workers
to maintain flexibility of employment in an era of declining working-age
population and economic stagnation. Most firms don’t expect their need
for employees to steadily increase, as was the case during the high-growth
era. At the same time, areas in which individual firms have a competitive
advantage over their rivals are changing quickly and Japan’s comparative
advantage as a whole is also changing over time.7
The increasing reliance of firms on part-time workers, while rational
in the context of slow economic growth and Japan’s system of high job
security, may also be imposing a huge economic loss by reducing human
capital accumulation.
One could argue that nonregular employment is increasing because
industries in which part-time work is widespread are expanding, such as
care for the elderly or eateries, and also because female labor participation
is increasing. However, according to Asano, Ito, and Kawaguchi (2013),
only one-quarter of the increase in nonregular workers can be explained
by changes in industry distribution and labor force composition. Instead,
nonregular employment is rising because new labor market entrants, male
workers of younger cohorts, and female workers of all cohorts are increas-
ingly taking up part-time employment, suggesting that long-term employ-
ment relationships have declined in importance. As shown in figure 7.9a,
the share of nonregular employees is particularly high among young male
workers, even among university graduates (Hamaaki et al. 2012). Moreover,
most female workers are nonregular employees (figure 7.9b), as are most
older male workers.
Slow economic growth and increasing international competition
are discouraging firms from employing most of their workers as regular
employees under Japan’s traditional lifetime employment system. These
workers accumulate less human capital because most of them are part-
time workers.

6. The Japanese employment system for regular workers is also changing. Using microdata
from the Basic Survey on Wage Structure, Hamaaki et al. (2012) found that the age-wage profile
has become flatter in recent years.
7. Matsuura, Sato, and Wakasugi (2011) constructed a theoretical model in which trade liber-
alization encourages firms to reduce the number of products, which raises uncertainty about
the demand firms face. This change will increase firms’ demand for temporary workers. They
empirically test their model using microdata for Japanese manufacturing plants and find
moderate support for the model’s predictions.

146 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 7.9 Labor force participation rate, by age and employment status,
2013
a. Men
percent
100

90

80

70

60

50

40

30

20

10

0
15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–
age group
b. Women
percent
80
70
60
50
40
30
20
10
0
15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 70–
age group
Regular employees
Nonregular employees
Other

Source: International Labor Organization, Labour Force Survey.

Simply prohibiting nonregular employment is not an appropriate policy

response. Such a policy would substantially misallocate workers among
firms.8 On the other hand, reducing the job security of regular workers and

8. See Fukao and Kwon (2006).

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 147

improving the social security net would also not be an appropriate policy re-
sponse, since it risks substantially slowing down skill accumulation among
workers, given that training of workers in Japan greatly depends on the life-
time employment system. Instead, Japan needs to enhance (1) labor market
flexibility and (2) human capital accumulation among workers who do not
participate in the lifetime employment system.
One option would be to increase “limited regular employment,” where
workers are employed based on job-specific labor contracts without life-
time employment guarantees but receive high compensation for their
professional skills (Tsuru 2017). A job card system, which would allow
workers to prove their skills and work experience, would also improve
reallocation of workers and accumulation of human capital. Yet another
avenue would be to replace the system of internal training of regular
employees with training and education outside firms, which would require
reforming Japan’s professional education system at universities and voca-
tional schools. Other important measures include regulations to reduce
the unfair wage gap between regular and nonregular employees.
In a survey by the Ministry of Health, Labour and Welfare (MHLW),
when asked why they worked as nonregular employees, 22.8 percent of male
nonregular workers answered that they could not find regular employ-
ment, while 5.6 percent replied that they wanted to balance work with their
family responsibilities, 15.6 percent of respondents answered that they
could not find regular employment, while 35.9 percent replied that they
wanted a job compatible with their family circumstances, such as house-
work, child care, and nursing in the case of female nonregular workers.9
Reducing nonregular employment among women would require making
regular employment compatible with workers’ family circumstances.
Life-work balance is also important to resolve Japan’s low fertility rate.
The average age at which a Japanese woman first marries has risen espe-
cially among educated full-time female workers, who tend to postpone
getting married and raising families because of the demands of regular
employment (Sakamoto and Kitamura 2008, Brinton 2015).
It is important to note that the rigidity of Japan’s labor market is
related to Japan’s low inflation problem. Reflecting the increasingly tight
labor market, wages of nonregular workers have started rising considerably
(at an annual rate of about 2 percent). However, regular workers have seen
very limited wage increases (Bank of Japan 2017, chart 28). This difference is
likely due to the following facts: First, labor unions, which are composed of
regular workers, tend to prioritize job security over wage increases. Second,

9. General Survey on Diversified Types of Employment 2014 conducted by the Ministry of Health,
Labour and Welfare (MHLW).

148 SUSTAINING ECONOMIC GROWTH IN ASIA

because of deferred compensation, workers do not change their jobs even
when their wages are temporarily lower than at other firms, Third, since
labor unions oppose wage cuts and it is difficult for firms to fire regular
workers, wage setting for regular workers is like a long-term leasing contract
for durable machines with a fixed fee. Firms’ decisions on wage rates of
regular workers crucially depend on their expectations of future inflation.
And since inflation expectations are still low in Japan, firms are reluctant to
raise the wage rates of regular workers.

Productivity and Wage Gaps between Large and Small

Firms
Japan’s market economy has been characterized by large differences in labor
productivity and wage rates between large firms and small and medium
enterprises (SMEs) since the interwar period (the so-called dual economy
[Nakamura 1983]). Oi and Idson (1999) found that firm-size wage differ-
ences in Japan are greater than in the United States. Labor productivity
differences between small and large firms in Japan are also larger than in
most other OECD countries (figure 7.10).
These differences have widened since the 1990s, especially in the manu-
facturing sector (Fukao 2013),10 where the TFP growth of large firms has
actually accelerated, while SMEs have lagged behind (Fukao and Kwon
2006). One possible explanation is that SMEs have not kept up with the
information and communication technology (ICT) revolution and interna-
tionalization (Fukao et al. 2016a; Ito, Deseatnicov, and Fukao 2016).
Another potential explanation of the slowdown of SMEs’ TFP growth is
the decline in technology spillovers from large firms (Belderbos et al. 2013).
In the manufacturing sector, large assemblers, which produce final goods,
have conducted intensive research and development (R&D) to create new
products. Meanwhile, SMEs have tended to supply parts and components
to these assemblers. Supplier relationships between large assemblers and
SMEs have usually been stable and tight, and SMEs likely benefited from
spillovers from large assemblers. Moreover, R&D intensity of SMEs is much
lower than that of larger firms in Japan. In fact, this gap is much larger in
Japan than in other OECD countries (figure 7.11). However, since the 1990s
Japan’s leading export industries, such as the electronic and automotive in-
dustries, have increasingly relocated production abroad. Because of this and
other factors such as restructuring at large assemblers, buyer-supplier rela-

10. Andrews, Criscuolo, and Gal (2015) report that TFP differences between frontier firms,
which tend to be large and internationalized, and nonfrontier firms have been widening in
many OECD countries since the 2000s. However, since their data do not cover the 1990s, one
cannot judge whether such widening in TFP differences started in the 1990s like Japan or not.

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 149

Figure 7.10 Labor productivity differences: Firms with 20–49
workers/firms with more than 250 workers, 2013

value added
90
80
70
60
50
40
30
20
10
0 O ny

Sl lan d
ov ds

Sp en
us n

Fr enia
ng nd
Tu ary

Ja ico
ec itz ree n
Un h R erla ce
ite ep nd
D ta ic
ite P nm es
d or ark

Be do l
G lgi m
er um

et o D

ed ia

A at ia
a
or ia

Sl Ita y
ov ly

nl e
to d
a
Is ey
ex l

ng ga
M rae

A ai

us vi

Fi nc
Cz Sw G pa

r n

ni
Es an
N P EC
S l

Sw ak

tr

N tral
e t
d ub

w
he la
Hu ela

rk

Ki tu

a
L
Ir

Un

OECD = Organization for Economic Cooperation and Development

Note: Value added per person employed in 2013 in firms with 20–49 workers relative
to that in firms with more than 250 workers = 100.
Source: OECD (2016b).

tionships in these industries in Japan have become more open (Paprzycki

2004, Ikeuchi et al. 2015).
Despite the importance of the widening productivity gap between
large and small firms, analysis of the “dual economy” in the nonmanufac-
turing sector has been limited. Previous studies have also not sufficiently
studied differences in labor input, such as workers’ education level, sex,
age, and employment status, between different firm-size groups. Against
this background, Fukao et al. (2014) examined these issues by splitting
KLEMS-type data of the market economy by firm size and industry. This
section summarizes the results and discusses the policy implications.
For control totals, Fukao et al. (2014) used the JIP Database (KLEMS
data on Japan). It is implicitly assumed that prices of outputs and interme-
diate inputs do not differ across different firm-size groups. To split data
by firm size, we employed the Corporate Enterprise Annual Statistics from the
Ministry of Finance. These statistics provide data on value added, capital
stock, number of workers, and total labor cost by firm size within each in-
dustry (the financial industry is not covered). Statistics by firm size are avail-
able only in terms of paid-in capital. Using the microdata underlying these
statistics, we created a matrix of the distribution of workers for each in-
dustry by amount of paid-in capital and by number of workers. Using these

150 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 7.11 Business R&D and government support for business R&D, by firm size, 2013

percent
100
SMEs share of BERD
90 SMEs share of government-funded BERD
SMEs share of government-funded BERD, 2003
80

70

60

50

40

30

20

10

s
k
a
y
n

d
d
g

ic

ia
al
ia
ia
l

ry
lia

ile

ay
es

da

nd
nd
ce

ai
ar

um
re

om

tr
ur

e
nd
an

an
an
pa

an

ak

w
en
ra
la

ug
l
Ita
at

ga
na

Ch
an

gi

t
r

la

al
m

re
l

Sp
us

n
nl

gd
Ko
Ja

bo

or
nm
i
St

c
ov

e
Po
lo

G
n
Fr y

l
A

I
or
us
ze

E
e
er

Ca
F

N
m en

Z
Be
S

S
d

Hu ce
he
P
it

Re nia
A
D
G

Ki
xe
te

et ubl
d

ew ela
Sw

ch sto
N
ni

Lu we

N
ite
U

Cz
Un

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN

SMEs = small and medium enterprises; BERD = business enterprise research and development
Source: OECD Science, Technology and Industry (STI) Scoreboard 2015, http://dx.doi.org/10.1787/sti_scoreboard-2015-en.

151
matrices, we converted statistics by amount of paid-in capital into statistics
by number of workers. Data on labor input and wage rates by firm size and
by industry are obtained from MHLW’s Basic Survey on Wage Structure, which
provides information on wage rates by age, sex, education, and employment
status and working hours by firm size within each industry. One caveat re-
garding these statistics is that they do not cover firms with fewer than 10
employees.
For each year and industry, Fukao et al. (2014) decompose labor produc-
tivity differences between firm-size group s and s’ using the following equa-
tion:
V  V  1   K   K 
ln s   ln s    ln qs   ln qs     s   s   ln s   ln s   
 s
H  s 
H  2   s s
q H  qs  H s   
 ln RTFPs , s ' 
where
Vs: Nominal value added of firm-size group s,
Hs: Total hours worked in firm-size group s,
qs: Labor quality of firm-size group s,
νs: Cost share of capital in firm-size group s,
Ks: Capital service input of firm-size group s,
RTFPs,s’: Relative TFP level of firm-size group s compared with that of firm-
size group s’.
To calculate Jorgenson-Griliches-type labor quality indices, qs, our
study uses the industry average wage premium of each category of workers.
Therefore, the study assumes that there are no differences in labor quality
among the same type of workers across different firm-size groups—for
instance, male university-educated full-time workers aged 30–34 years in
large automobile firms and their counterparts in small automobile firms.
The results of our analyses for the total market economy are summa-
rized in figure 7.12a, which shows that there are huge wage and labor
productivity differences between large (with more than 999 employees) and
small firms (with fewer than 100 employees). Differences in capital-labor
ratios mainly cause labor productivity differences. However, TFP differences
also play an important role. The contribution of labor quality differences
is declining. Figure 7.12b shows that wage and productivity differences
between medium-sized firms (with 100 to 999 employees) and small firms
(with fewer than 100 employees) are not so large. Therefore, subsequent
figures show the results for wage and productivity differences between large
firms and small firms.

152 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 7.12 Wage and productivity differences, total
market economy, 1975–2010

a. Firms with more than 999 employees/firms with

up to 99 employees
log value
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1975 1980 1990 2000 2010

b. Firms with 100 to 999 employees/firms with

up to 99 employees
log value
0.4

0.3
Total factor productivity
0.2 Capital-labor ratio
0.1 Labor quality
Labor productivity gap
0 Wage gap
1975 1980 1990 2000 2010

Source: Fukao et al. (2014).

In the manufacturing sector, differences in both labor productivity

and wages have increased (figure 7.13). Widening TFP differences substan-
tially contributed to the increase in labor productivity differences.
The results for the nonmanufacturing sector are shown in figure 7.14.
In wholesale and retail, both TFP and wage differences are declining. On
the other hand, in construction as well as transportation, communication,
utilities, and real estate, TFP differences increased from 1975 to 2010.
One of the most interesting findings of the above analysis is that while
wage differences are quite large, differences in labor quality based on the
Jorgenson-Griliches approach are not. Figure 7.15 decomposes the labor
quality gap in the total market sector between large firms (with more than
999 employees) and small firms (with fewer than 100 employees) in terms of
worker characteristics. Labor quality of large firms based on the Jorgenson-
Griliches approach is higher than that of small firms mainly because of

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 153

 Figure 7.13 Wage and productivity differences: Firms
with more than 999 employees/firms with
up to 99 employees, by manufacturing
sector, 1975–2010

a. Light industry
log value
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1975 1980 1990 2000 2010

b. Chemical industry
log value
1.2

1.0

0.8

0.6

0.4
Total factor productivity
Capital-labor ratio
0.2
Labor quality
Labor productivity gap
0 Wage gap
1975 1980 1990 2000 2010
(figure continues)

differences in education. The labor quality gap has declined over time and
was 7 percent in 2010. The decline was mainly caused by the increase in
nonregular employees in large firms (in figure 7.15, this factor is included
in “employment status”).
It is important to note that although the wage gap between the two
firm groups is more than 50 percent (figure 7.12a), measured labor quality
explains only 7 percentage points of this gap in 2010. Rebick (1993) reports
that in the United States, about one-third of firm-size wage differences are
explained by labor characteristics, such as education, experience, etc., while
154 SUSTAINING ECONOMIC GROWTH IN ASIA
Figure 7.13 Wage and productivity differences: Firms
with more than 999 employees/firms with
up to 99 employees, by manufacturing
sector, 1975–2010 (continued)

c. Machinery
log value
1.0

0.8

0.6

0.4

0.2 Total factor productivity

Capital-labor ratio
Labor quality
Labor productivity gap
0 Wage gap
1975 1980 1990 2000 2010

Source: Fukao et al. (2014).

in Japan it is only one-tenth. Our result for Japan is roughly consistent

with Rebick’s finding.
What causes firm-size wage differences that worker characteristics
cannot explain? One possible explanation is that since labor mobility across
firms is limited in Japan, workers of large firms enjoy rents as a result of
belonging to larger, more productive firms. However, since most large firms
remain large and do not go bankrupt, it is difficult to understand why em-
ployees at large firms can continue to enjoy windfalls in the form of high
wages. Two other explanations seem more plausible. The first is differences
in on-the-job and off-the-job training. As shown in figure 7.16, large firms
in Japan tend to provide much more job training to workers than SMEs.
Using microdata on labor turnover and resulting wage changes, Genda
(1996) finds that firm-size differences in job training contribute much
more to firm-size wage differences than unmeasured ability differences. The
second explanation is that, in Japan, graduates of top-ranked universities are
much more likely than other graduates to get a job at a large firm (Higuchi
1994). This suggests that there might be a large gap in innate ability across
workers in different firm-size groups that is difficult to measure using the
Jorgenson-Griliches approach.

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 155

Figure 7.14 Wage and productivity differences: Firms with more
than 999 employees/firms with up to 99 employees,
by nonmanufacturing sector, 1975–2010

a. Wholesale and retail b. Construction

log value log value
0.7 1.0
0.6
0.8
0.5

0.4 0.6

0.3 0.4
0.2
0.2
0.1

0 0
1975 1980 1990 2000 2010
–0.2
1975 1980 1990 2000 2010

c. Transportation, communication, d. Other services

utilities, and real estate log value
log value
0.9
1.4
0.8
1.2 0.7

1.0 0.6
0.5
0.8
0.4
0.6 0.3
0.4 0.2
0.1
0.2
0
0 1975 1980 1990 2000 2007
1975 1980 1990 2000 2010

Total factor productivity

Capital-labor ratio
Labor quality
Labor productivity gap
Wage gap

Source: Fukao et al. (2014).

Firm-size wage differences that are not explained by the standard

Jorgenson-Griliches approach account for about 45 percentage points
(the 52 percent in figure 7.12a minus the 7 percent in figure 7.15). If we
assume that all of this gap is due to labor quality differences, the TFP
of firms with more than 999 employees relative to firms with fewer than

156 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 7.15 Decomposition of the labor quality gap: Total market sector, firms
with more than 999 employees/firms with up to 99 employees,
1975–2010
percent
15

Contribution of labor quality

13 gap to wage gap

11

–1

–3

–5
1975 1980 1985 1990 1995 2000 2005 2010

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN

Education Employment status
Gender Age

157
Source: Fukao et al. (2014).
 Figure 7.16 Off-the-job training expenses (including opportunity
costs) by firm size, total market sector, 1975–2010
ratio
6
Firms with more than 999 employees/
5 firms with 30–99 employees
Firms with 100–999 employees/
firms with 30–99 employees
4

0
1975 1980 1990 2000 2010

Source: Fukao et al. (2014).

100 employees, when measured without taking such difficult-to-measure

labor quality differences into account, will be overestimated by this differ-
ence of 45 percentage points times the labor income share, which is around
two-thirds in Japan. This means that we could explain 30 percentage points
of firm-size TFP differences by such difficult-to-measure labor quality
differences, which is very close to the total TFP gap of 32 percent in figure
7.12a. As shown in table 7.1, small firms are much more prevalent in Japan
than in the United States. Firms with fewer than 999 employees account
for 72 percent of all employment. In the United States, such firms account
for only 55 percent. This indicates that the low productivity and low wage
rates of SMEs are a particularly pressing issue in Japan.
If we assume that difficult-to-measure labor quality differences cause all
of the firm-size wage differences not explained by the standard Jorgenson-
Griliches approach, then this factor explains almost all of the firm-size TFP
differences. Therefore, in order to understand firm-size TFP differences and
the slowdown in Japan’s TFP growth, it is important to examine such firm-
size labor quality differences in more detail.
If correct, this hypothesis implies that there might be large room to
improve Japan’s macro-level TFP by improving the quality of workers in
SMEs through education and training.

158 SUSTAINING ECONOMIC GROWTH IN ASIA

 Table 7.1 Number of employees by firm size, all industries, Japan and United States,
2001 and 2006

GRAPHICS
Japan United States

SECULAR
2001 2006 2001 2006
Firm size (number Number of Percent Number of Percent Number of Percent Number of Percent
of employees) employees of total employees of total employees of total employees of total

- CHAPTER
1 to 4 1,486,150 4 1,574,110 5 5,866,666 5 6,262,490 5
5 to 9 2,176,265 7 1,993,335 6 6,844,090 6 7,274,534 6

STAGNATION
10 to 19 2,954,728 9 2,736,690 9 8,369,988 7 8,794,210 7
20 to 49 4,442,234 13 4,188,269 13 11,767,978 10 12,260,057 10
50 to 99 3,300,383 10 3,166,835 10 8,442,216 7 8,868,873 7

7—SUSTAINING
100 to 249 4,177,981 13 4,144,598 13 9,813,665 9 10,497,066 9
250 to 499 2,832,588 9 2,794,966 9 6,258,633 5 6,762,233 6
500 to 999 2,528,727 8 2,573,958 8 5,866,407 5 6,063,319 5

AND THEECONOMIC
1,000+ 9,274,478 28 8,935,484 28 51,128,895 45 52,125,133 44
Total 33,173,534 100 32,108,245 100 114,358,538 100 118,907,915 100

LABOR MARKET
Source: Fukao et al. (2016a). Original data are from the Establishment and Enterprise Census for Japan and

GROWTH
the Business Dynamics Statistics for the United States.

IN JAPAN
IN ASIA1591
References
Andrews, Dan, Chiara Criscuolo, and Peter Gal. 2015. Frontier Firms, Technology Diffusion and
Public Policy: Micro Evidence from OECD Countries. The Future of Productivity: Main Back-
ground Papers. Paris: Organization for Economic Cooperation and Development.
Asano, Hirokatsu, Takahiro Ito, and Daiji Kawaguchi. 2013. Why Has the Fraction of Non-
standard Workers Increased? A Case Study of Japan. Scottish Journal of Political Economy 60,
no. 4: 360–89.
Asao, Yutaka. 2011. Overview of Non-Regular Employment in Japan. In Non-Regular Employ-
ment–Issues and Challenges Common to the Major Developed Countries. JILPT Report, no. 10:
1–43. 2011 JILPT Seminar on Non-regular Employment. Tokyo: Japan Institute for La-
bour Policy and Training.
Bank of Japan. 2017. Outlook for Economic Activity and Prices, July 2017. Tokyo.
Belderbos, René, Kenta Ikeuchi, Kyoji Fukao, YoungGak Kim, and Hyeog Ug Kwon. 2013.
Plant Productivity Dynamics and Private and Public R&D Spillovers: Technological, Geographic and
Relational Proximity. CEI Working Paper Series, no. 2013-05. Center for Economic Institu-
tions, Institute of Economic Research, Hitotsubashi University.
Brinton, Mary C. 2015. Japanese Low Fertility and the Low Labor Force Participation of Mar-
ried Women: The Role of Rigid Labor Markets and Workplace Norms. Perspectives from
Around the World, no. 56. Tokyo: Research Institute of Economy, Trade and Industry.
Fukao, Kyoji. 2013. Explaining Japan’s Unproductive Two Decades. Asian Economic Policy Re-
view 8, no. 2 (December): 193–213.
Fukao, Kyoji, and Hyeog Ug Kwon. 2006. Why Did Japan’s TFP Growth Slow Down in the
Lost Decade? An Empirical Analysis Based on Firm-Level Data of Manufacturing Firms.
Japanese Economic Review 57, no. 2: 195–228.
Fukao, Kyoji, Ryo Kambayashi, Daiji Kawaguchi, Hyeog Ug Kwon, YoungGak Kim, and Izumi
Yokoyama. 2006. Deferred Compensation: Evidence from Employer-Employee Matched
Data from Japan. Hi-Stat Discussion Paper Series, no. 187. Institute of Economic Research,
Hitotsubashi University.
Fukao, Kyoji, Tsutomu Miyagawa, Kentaro Mukai, Yukio Shinoda, and Konomi Tonogi.
2009. Intangible Investment in Japan: Measurement and Contribution to Economic
Growth. Review of Income and Wealth 55, no. 3: 717–36.
Fukao, Kyoji, Tatsuji Makino, Kenta Ikeuchi, Hyeog Ug Kwon, and YoungGak Kim. 2014.
Productivity and Wage Differences by Firm Size [in Japanese]. Japanese Journal of Labour
Studies, no. 649: 14–29.
Fukao, Kyoji, Kenta Ikeuchi, YoungGak Kim, and Hyeog Ug Kwon. 2016a. Why Was Japan
Left Behind in the ICT Revolution? Telecommunications Policy 40, no. 5: 432–49.
Fukao, Kyoji, Kenta Ikeuchi, HyeogUg Kwon, YoungGak Kim, Tatsuji Makino, and Miho
Takizawa. 2016b. The Structural Causes of Japan’s Lost Decades. In The World Economy,
Growth or Stagnation? ed. Dale W. Jorgenson, Kyoji Fukao, and Marcel P. Timmer. Cam-
bridge: Cambridge University Press.
Genda, Yuji. 1996. “Ability” or “Training”: Ability Difference Hypothesis on Firm-Size Wage
Differences [in Japanese]. Japanese Journal of Labour Studies, no. 430: 17–29.
Hamaaki, Junya, Masahiro Hori, Saeko Maeda, and Keiko Murata. 2012. Changes in the Japa-
nese Employment System in the Two Lost Decades. Industrial and Labor Relations Review
65, no. 4: 810–46.
Higuchi, Yoshio. 1994. University Education and Income Distribution [in Japanese]. In Income
and Wealth Distribution in Japan, ed. Tsuneo Ishikawa. Tokyo: University of Tokyo Press.

160 SUSTAINING ECONOMIC GROWTH IN ASIA

Ikeuchi, Kenta, Kyoji Fukao, Hiromichi Goko, YoungGak Kim, and Hyeog Ug Kwon. 2015.
Empirical Analysis on the Openness of Buyer-Supplier Relationships and Productivity in the Japanese
Automobile Parts Industry. RIETI Discussion Paper Series 15-J-017 [in Japanese]. Tokyo: Re-
search Institute of Economy, Trade and Industry.
Ito, Koji, Ivan Deseatnicov, and Kyoji Fukao. 2016. Do Exporters’ and Non-exporters’ Factor
Inputs Differ? A Study Based on Employer-Employee Matched Data for Japan. Presenta-
tion prepared for the joint meeting of the Working Party on Industry Analysis (WPIA) and
the Working Party on Globalisation of Industry (WPGI), OECD, Paris, October 10–11.
Kawaguchi, Daiji. 2017. Gender Gap in Skill Use in Japan: Findings from Japan-US-UK Com-
parison Based on PIAAC. In Current Trends in Economics (Gendai Keizaigaku no Choryu), ed.
Masako Ii, Chiaki Hara, Kaoru Hosono, and Hitoshi Matsushima. Tokyo: Toyo Keizai
Shinpo Sha [in Japanese].
Matsuura, Toshiyuki, Hitoshi Sato, and Ryuhei Wakasugi. 2011. Temporary Workers, Perma-
nent Workers, and International Trade: Evidence from Japanese Firm-Level Data. RIETI Discus-
sion Paper Series, no. 11-E-30. Tokyo: Research Institute of Economy, Trade and Industry.
Nakamura, Takafusa. 1983. Economic Growth in Prewar Japan. New Haven, CT: Yale University
Press. (Originally published in Japanese as Senzenki Nihon Keizai Seicho no Bunseki [An Analy-
sis of Economic Growth in Prewar Japan], Tokyo: Iwanami Shoten, 1971).
OECD (Organization for Economic Cooperation and Development). 2013. Technical Report of
the Survey of Adult Skills. Paris.
OECD (Organization for Economic Cooperation and Development). 2016a. Skills Matter: Fur-
ther Results from the Survey of Adult Skills. OECD Skills Studies. Paris: OECD Publishing.
OECD (Organization for Economic Cooperation and Development). 2016b. Entrepreneurship
at a Glance 2016. Paris: OECD Publishing. Available at http://dx.doi.org/10.1787/entrepre-
neur_aag-2016-en.
OECD (Organization for Economic Cooperation and Development). 2017. 2017 OECD Eco-
nomic Survey of Japan. Paris.
Oi, Walter, and Todd Idson. 1999. Firm Size and Wages. In Handbook of Labor Economics 3, no.
3, ed. O. Ashenfelter and D. Card. Elsevier.
Paprzycki, Ralph. 2004. Interfirm Networks in the Japanese Electronics Industry. Sheffield Centre
for Japanese Studies/Routledge Series. London: Routledge.
Rebick, Marcus E. 1993. The Persistence of Firm-Size Earnings Differences and Labor Market
Segmentation in Japan. Journal of the Japanese and International Economies 7, no. 2: 132–56.
Sakamoto, Kazuyasu, and Yukinobu Kitamura. 2008. Marriage Behavior from the Perspective
of Intergenerational Relationships. Japanese Economy 34, no. 4 (Winter): 76–122.
Tsuru, Kotaro. 2017. Reforming the Regular Employment System: Toward a New Norm of
Job-Specific Employment Contracts. Social Science Japan Journal 20, no. 1: 59–72.

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 161

Comment
Jonathan Woetzel

Kyoji Fukao makes an important contribution to our understanding of

Japan’s economic stagnation. He outlines the problem in terms of weak
labor productivity—much lower today than in other major OECD coun-
tries except South Korea—and offers a solution through his examination
of two recent labor market changes: rising nonregular employment and an
expanding wage gap between large and small firms. He builds a compelling
case that declining labor quality is a critical reason behind Japan’s wors-
ening productivity performance and argues that education and training
would go a long way in improving Japan’s economic situation. Education
and training should help, but significantly improving Japan’s productivity
performance is no small feat. The labor productivity gap between Japan
and the United States has been widening across most industries since
2000. At the current pace, it is expected to grow from 29 percent in 2010
to 37 percent by 2025, according to our research at the McKinsey Global
Institute (MGI). As a result, other measures will need to be considered in
conjunction with those Fukao recommends in his chapter.
Fukao first analyzes the rapid rise of nonregular employment in Japan
since the end of the 1980s. He defines a regular employee as anyone hired
directly by an employer, without a predetermined period of employment,
and who works for scheduled hours, while a nonregular employee is anyone
who does not meet one of these conditions. In figure 7.7, he shows that
nontraditional employment in the nonmanufacturing market economy
has grown from around 5 percent in 1980 to a little under 30 percent in
2012. But he also identifies a large wage gap between nonregular and regular
employees, especially in mid to late career years, as regular workers tend to
earn more as they gain more experience (figure 7.8). Regular workers also
receive training, which builds human capital, but nonregular workers typi-
cally don’t and therefore accumulate less human capital.
Fukao surmises that in an environment of slow economic growth
and increasing international competition, firms have become reluctant to
invest in employing more regular employees under Japan’s traditional life-
time employment system. While such firm behavior is rational in the face
of slow economic growth and Japan’s system of high job security, Fukao
shows that this behavior has resulted in a “huge economic loss by reducing
human capital accumulation.” The drop in regular employment—and con-

Jonathan Woetzel is director of the McKinsey Global Institute and a senior partner in McKinsey & Company’s
Shanghai office.

162 SUSTAINING ECONOMIC GROWTH IN ASIA

current rise in nonregular employment—has led to a decline in the quality
of Japanese labor. Fukao provides a neat calculation of this effect. Using the
Japan Industrial Productivity (JIP) Database, he calculates that if the share
of nonregular workers in total workers had not increased since 1988, the
average quality of Japanese labor would have been 8 percent higher today.
To address the labor quality problem, Fukao suggests not to do away
with labor market flexibility but to invest in the human capital of nonreg-
ular employees. Some of his options include establishing a job card system,
where workers can prove their skills and work experience, and potentially
moving training and education from within firms to universities and voca-
tional schools. On this score, more research needs to be done to uncover
the best way to train a growing nonregular workforce. One answer may be
to require firms to provide benefits to nonregular workers after a period of
time, even as they make regular workers less tenured. The ultimate solu-
tion, though, is competition to ensure that firms less willing to innovate
and thus retain talented workers no longer dominate the workplace.
On the second labor market change, Fukao argues that the widening
productivity and wage gap between large and small firms may also be a labor
quality issue. Using level accounting, Fukao finds large differences in total
factor productivity (TFP) between the two groups that cannot be explained
by standard approaches. He argues that the difference in fact may be caused
by hard-to-measure differences in labor quality. If that indeed is the case, he
believes training and education may significantly improve Japan’s macro-
level TFP and boost economic growth more broadly.
While Fukao has put the spotlight on an underappreciated aspect of
the Japanese economy, more work needs to be done on the implications of
labor quality issues for policy and companies, as well as on the right mix of
additional measures for accelerating Japanese productivity. Our research at
MGI identifies a number of important trends worth taking into account.
Japan’s advanced manufacturing industries have relied on the domestic
market while losing ground globally, which has implications for weak
productivity. Japanese companies overall have significantly underinvested
in organizational capabilities and talent for commercializing innovation
globally compared with the United States and a range of European econo-
mies. At the same time, there is a distinct lack of women in senior leader-
ship roles (a point that Fukao also makes in his chapter). Finally, Japan has
fallen behind the United States, United Kingdom, Germany, and France in
enterprise creation and growth.
These developments suggest that accelerating productivity in Japan may
require a wider array of policies and measures than those addressing labor
quality. The private sector can accelerate productivity growth by adopting

SECULAR STAGNATION AND THE LABOR MARKET IN JAPAN 163

best practices and new technologies to increase value added. Steps to adopt
best practices include going global, building best-in-class capabilities across
the entire value chain, and continuing to digitize. Steps to adopt new tech-
nologies include harnessing the power of big data, leveraging technology
to create and commercialize new products and services, and deploying ad-
vanced technologies in manufacturing processes. If Japan takes these steps,
we calculate that it could increase value added by up to 28 percent above the
current trajectory.
Government policies that encourage entrepreneurship and global com-
petitiveness are crucial to aid companies in their transformation. But as
Fukao points out in his chapter, so too are policies that unlock new talent
sources and encourage education. We find policies to unlock new talent
sources could include encouraging more women to enter the workforce,
providing seniors with flexible work arrangements to stay in the workforce,
and rethinking immigration policies. Policies to encourage education could
include building on top of the strong outcomes on hard skills (e.g., math
and science), instilling the innovative and global mindset needed in a 21st
century economy, and creating a true education-to-employment pipeline to
identify where labor will be needed and match labor supply with demand.
The payoff from these measures could be large. Accelerating productivity
growth would fundamentally change Japan’s economic trajectory over the
next decade.

164 SUSTAINING ECONOMIC GROWTH IN ASIA

 8
Declining Potential Growth
in Korea
DONGCHUL CHO AND KYOOHO KWON

Since the global financial crisis in 2008, “secular stagnation” has become a
catchword for the economic conditions of advanced countries (see Summers
2014, among many others). Recovery from the crisis has been slow, and in-
flation has remained low, despite unprecedentedly aggressive monetary
policies. Asia has maintained relatively rapid growth, but the pace has been
slower than it was before the crisis. Japan is struggling to escape from a defla-
tion trap, and China is suffering from the aftereffects of the overinvestment
it pushed to counter the negative shocks from the global financial crisis.
The economic performance of Korea, the third-largest economy in East
Asia, has been weak. The average rate of growth over the past five years was
less than 3 percent, the lowest since the 1960s. The unfavorable global envi-
ronment undoubtedly affected Korea, but its growth rate has been on a
secular decline since the 1990s, well before the global crisis. The weakening
of its growth momentum has raised concerns that Korea may be following
in Japan’s footsteps, that “lost decades” may lie ahead. What factors underlie

Dongchul Cho is a member of the Monetary Policy Board of the Bank of Korea. Kyooho Kwon is a fellow at
the Korea Development Institute. The views expressed are those of the authors and do not necessarily reflect the
views of the Bank of Korea or the Korea Development Institute. The authors are grateful to Shinwook Woo for
research assistance.
Figure 8.9 and table 8.2 in this chapter were produced using simulations of the overlapping generations model
from Kwon (2017b) published by the Korea Development Institute (KDI) Journal of Economic Policy. The au-
thors employed a Fortran code for their model that may be difficult to use to replicate the authors’ results without
their assistance. Accordingly, this chapter does not provide the Fortran code as part of the normal replication
package required by PIIE. PIIE does not assume responsibility for the calculations underlying figure 8.9 and
table 8.2 in this chapter. Anyone seeking further information should contact the authors at dccho@bok.or.kr
and kwonkh@kdi.re.kr.

165
the secular decline of Korea’s growth? What policies are needed to reverse,
or at least mitigate, the trend?
This chapter addresses these questions. The first section presents the
results of growth accounting and speculates on the likely path of the Korean
economy based on recent empirical evidence on total factor productivity
(TFP). The second section touches on the possibility of Korea’s “Japaniza-
tion,” by comparing Korea of today with Japan two decades ago. The third
section examines the decline in the natural interest rate (a result of the
decline in potential growth rate) and its implications for monetary policy.
The last section summarizes the chapter’s main results.

Korea’s Growth Trend

Growth Accounting
Figure 8.1 shows annual growth rates in Korea and the global economy for
the period between 1980 and 2017. The positive correlation between the
two series suggests that Korea’s slow growth in recent years is attributable
to some extent to the stagnation of the global economy. Korea’s growth has
been on a secular decline since the 1990s, however, whereas global growth
accelerated. The deterioration in external conditions alone can thus not
explain the recent slowdown of the Korean economy.
A standard growth accounting exercise provides a starting point for
understanding the secular decline in Korea’s growth. According to Kwon
(2017a), all three components of growth accounting (labor, capital, and
TFP) decelerated, for various reasons (table 8.1). An extremely low fertility
rate led to a continuing decrease in the absolute number of workers entering
the labor market, causing the pace of labor force expansion to decelerate
and the age profile to become older (figure 8.2).1 The capital accumulation
rate also slowed, as the capital-to-output ratio approached the ratios in
advanced countries or a steady-state level, after a decades-long capital deep-
ening process (figure 8.3). Increases in TFP also slowed, perhaps because the
technology gap between Korea and the global frontier narrowed.
The decline in Korea’s potential growth is likely to continue for the next
two decades. The size of the working-age population began to decline in
2017; given the current age structure, the pace of this decline is projected
to accelerate.2 The shrinking of the working-age population will eventually
reduce the aggregate labor force (although the rise in female participation

1. Korea’s fertility rate has been about 1.3 since 2000, lower than the rate in Japan and Germany.
2. It is sometimes argued that massive immigration or reunification with North Korea could
substantially change the demographic projections for Korea. This chapter does not consider
such scenarios.

166 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 8.1 Real GDP growth trends in Korea and the world,
1980–2017

percent, year on year

15

10

–5
Korea
World
–10
1980 1985 1990 1995 2000 2005 2010 2015

Sources: Bank of Korea; IMF Staff calculations.

Table 8.1 Growth accounting of Korea

Total factor
GDP Capital Employment productivity
Period (A+B+C) (A) (B) (C)
1981–1990 9.4 4.1 1.7 3.6
1991–2000 6.7 3.8 1.0 1.9
2001–2005 4.7 2.1 1.0 1.5
2006–2010 4.1 1.8 0.5 1.8
2011–2015 3.0 1.4 1.0 0.5
2016–2020 1.7+? 1.1 0.6 ?
2021–2025 1.3+?? 1.0 0.3 ??
2026–2030 0.7+??? 0.8 –0.1 ???
2031–2035 0.2+???? 0.6 –0.4 ????
2036–2040 0.0+????? 0.5 –0.5 ?????
Source: Kwon (2017a).

rate can mitigate the downward trend for a while).3 Inasmuch as growth in
output slows as a result of labor contraction, the pace of aggregate capital
accumulation will also decline, assuming that Korea’s capital-to-output

3. The projection in table 8.1 takes this factor into account.

DECLINING POTENTIAL GROWTH IN KOREA 167

 Figure 8.2 Fertility rates in Korea and select countries,
1980–2015

children per woman

3.0
Korea
United States
2.5 Germany
Japan

2.0

1.5

1.0

0.5
1980 1985 1990 1995 2000 2005 2010 2015

Source: Organization for Economic Cooperation and Development.

Figure 8.3 Capital-to-output ratios in Korea and selected

countries, 1970–2015

ratio
4.0

3.5

3.0

2.5

2.0 Korea
United States
1.5 Germany
Japan
1.0
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

Source: Organization for Economic Cooperation and Development.

ratio has reached a steady-state level. This argument can be concisely ex-
plained using the standard growth accounting formula:

∆Y/Y = ∆A/A +α∆L/L + (1– α)∆K/K,

where Y, A, L, K, and α denote output, technology, labor, capital, and the

labor income share, respectively. A steady-state output-to-capital ratio
implies that ∆K/K = ∆Y/Y, simplifying the formula to

168 SUSTAINING ECONOMIC GROWTH IN ASIA

∆Y/Y = (1/α) ∆A/A + ∆L/L or ∆y/y = (1/α) ∆A/A,

where y denotes income per labor. If demography exogenously drives L and

the capital-to-output ratio remains unchanged, Korea’s growth rate will
decrease in line with the contraction of the labor force, unless TFP improves
sufficiently rapidly.

Efficiency of Resource Allocation

Given demographic trends and the capital-deepening process, progress in
TFP will become the only source of growth for Korea. However, TFP prog-
ress declined until recently, and its prospects do not seem rosy. Many factors
could affect TFP progress, but few are likely to work in Korea’s favor. The lit-
erature cites various factors that can boost TFP, including the narrowing of
the technology (or income) gap with respect to the frontier, a young demo-
graphic structure, and the establishment of institutions to ensure efficient
resource allocation. Korea’s technology gap will narrow as long as Korea’s
per capita income converges to those of advanced countries.4 Korea’s popu-
lation will age over the next several decades, hampering TFP progress.5 The
improvement of institutions to achieve more efficient resource allocation
may be the key factor that Korea should rely on.
Recent empirical results on the efficiency of Korea’s resource allocation
using microdata are not encouraging. Kwon and Kim (2014) find that labor
mobility in response to industry-specific demand shocks has fallen signifi-
cantly in the manufacturing sector. The finding implies that declining
industries do not flexibly lay workers off and rising industries do not hire
sufficient numbers of workers. Instead, in response to demand shocks,
firms adjust their investment and the wages of incumbent workers. Kwon
and Kim present evidence that low mobility is attributable largely to exces-
sive protections for regular workers, who impose the burdens of employ-
ment adjustment on nonregular workers. This dual labor market structure
not only hinders resource allocation efficiency but also intensifies social
conflicts between the two groups of workers.

4. Since Barro (1991) and Mankiw, Romer, and Weil (1992), the convergence argument—that
lower-income countries tend to grow faster than higher-income countries, conditional on
their fundamentals—has become a standard notion in the empirical growth literature.
5. The negative effects of population aging on TFP have been found over a wide range of
data. For European countries, for example, Aiyar, Ebeke, and Shao (2016) project that the
current age structure could lead to a reduction in TFP growth by an average of 0.2 percentage
points every year over the next two decades. Using prefectural data from Japan for 1990–
2007, Liu and Westelius (2016) find that aging of the working-age population has had a
significant negative impact on TFP.

DECLINING POTENTIAL GROWTH IN KOREA 169

 The efficiency of financial markets has been eroded by the continued
supply of credits to “zombie” firms—firms that would be unable to survive
without preferential financial support from either the government or finan-
cial institutions. Based on their empirical results for Japan, Caballero, Hoshi,
and Kashyap (2008) argue that the increase in bank lending to zombies was
a major cause of the chronic recession in the 1990s after the bursting of the
economic bubble. Jeong (2014) finds that the portion of credit to zombies
in Korea increased after the global financial crisis, thanks to wide-ranging
public support, such as credit guarantees, loans through public financial
institutions, and other programs. Based on cross-section regression results,
he contends that zombies hinder investment and employment by sound
firms in the same industries.
Korea’s product markets are among the most heavily regulated in the
Organization for Economic Cooperation and Development (OECD)—and
entry barriers are being heightened to protect incumbent and “weak” firms.6
Applying the methodology of Hsieh and Klenow (2009) and Oberfield
(2013) to data on about 50,000 establishments in the manufacturing sector,
Oh (2017) finds that product market efficiency has been declining since the
mid-2000s. The deviation of the actual from the “optimal” level of produc-
tion for each firm indicates that distortion by firm size (overproduction in
small firms and underproduction in large firms) is especially serious.
These and other research results consistently suggest that the efficiency
of resource allocation in Korea has not been improving in recent years.
Although these findings are sobering, they may imply that there is room
for improving Korea’s TFP. The question is how to implement reforms to
enhance efficiency, which will inevitably provoke resistance from vested
interest groups.

Is Korea at Risk of “Japanization”?

The weakening dynamism of the Korean economy invokes concerns that
Korea may be heading for a prolonged recession accompanied by defla-
tion pressures, as Japan did in the 1990s. Some fundamentals in Korea
are similar to those in Japan 20 years ago. As Cho and Kwon (2014) noted,
Korea’s demography is following Japan’s in almost every dimension, with
a lag of about 20 years. The population of Korea is projected to decrease
beginning around 2030 (Japan’s population began decreasing in 2010). The
old-age dependency ratio in Korea will also follow that of Japan (figure 8.4).

6. Korea ranked 30th out of 33 member countries on the 2013 Product Market Regulation
Index of the OECD (2014).

170 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 8.4 Demographics in Korea and Japan

a. Total population growth rates

percent
2.0
Korea
1.5
Japan
1.0

0.5

–0.5

–1.0
1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Korea
1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 Japan

b. Aged dependency ratios

percent
80
70
60
50
40
30
20
Korea
10 Japan
0
1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Korea
1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 Japan

Source: Kwon (2017a).

It is not only in demography that Korea resembles Japan. As a result

of efforts to emulate its neighbor’s success, the economic structure of
Korea has become similar to that of Japan in many respects. At the heart
of both economies are export-oriented manufacturing industries such as
electronics, automobiles, and shipbuilding—all of which are included in
industry 7 of figure 8.5. Korea caught up with Japan in some of these indus-
tries over the past two to three decades but is now being fiercely pursued
by China. Jung (2017) examines the competition and catch-up processes
of Japan, Korea, and China in the export market, arguing that the pace of
China’s catching-up with Korea is accelerating.

DECLINING POTENTIAL GROWTH IN KOREA 171

 Figure 8.5 Revealed comparative advantages of Korea and
select countries (ratio)

Canada Germany
3 3

2 2

1 1

0 0
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9
Japan Korea
3 3

2 2

1 1

0 0
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9

United Kingdom United States

3 3

2 2

1 1

0 0
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9

Notes: Revealed comparative advantage is defined as the ratio of the share of the
item in the country’s exports to its share in world exports. The 1-digit Standard
International Trade Classification (SITC) codes on the horizontal axis are 0: Food
and live animals, 1: Beverages and tobacco, 2: Crude materials, inedible, except
fuels, 3: Mineral fuels, lubricants and related materials, 4: Animal and vegetable
oils, fats and waxes, 5: Chemicals and related products, n.e.s., 6: Manufactured
goods classified chiefly by material, 7: Machinery and transport equipment,
8: Miscellaneous manufactured articles, and 9: Commodities and transactions not
classified elsewhere in the SITC.
Source: UN Comtrade Database.

Korea’s per capita income also trails Japan’s by 20 years. Figure 8.6,
panel a, overlaps the trend of Korea’s nominal GDP growth rate with that of
Japan with a 20-year lag. The two series are almost indistinguishable. After a
similar pace of rapid growth for several decades, in 2010 Korea attained per
capita income of about $30,000—the level Japan reached in 1990.
Insofar as demography, industrial structure, and per capita income are
the crucial determinants of a country’s social and economic landscape, it
seems natural to study Japan’s history in order to draw a picture of Korea’s
future. An issue of interest is whether Korea will fall into chronic deflation,

172 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 8.6 Trends in nominal GDP growth rates in
Korea and Japan

a. Up to 2010 for Korea

percent, year on year
20

16

12

4 Nominal GDP growth rate (HP)—Korea

Nominal GDP growth rate (HP)—Japan
0
1985 1990 1995 2000 2005 2010 Korea
1965 1970 1975 1980 1985 1990 Japan

b. Up to 2016 for Korea

percent, year on year
20
Nominal GDP growth rate (HP)—Korea
16 Nominal GDP growth rate (HP)—Japan
12

–4
1985 1990 1995 2000 2005 2010 2015 Korea
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Japan

HP = Hodrick-Prescott filter
Sources: Bank of Korea; Cabinet Office of Japan (reproduced from
Cho and Kwon 2014).

as Japan did. Panel b of figure 8.6, which extends the Korean data to 2016,
appears to send a relieving signal. Unlike in panel a, Korea’s trend growth
in nominal GDP began deviating from that of Japan in the 2010s, implying
that Korea is not replicating Japan’s deflation. Korea’s inflation rate has
fallen to a worrisome level since 2013, but the pace of disinflation has been
slower than Japan faced after the bursting of its bubble in the 1990s.
The divergence between the two economies in terms of inflation/defla-
tion seems to stem from a difference in their real estate markets. Entering
the 1990s, Japan’s real estate market was in a bubble, the bursting of which
triggered a deflationary spiral that was too abrupt for the Bank of Japan to

DECLINING POTENTIAL GROWTH IN KOREA 173

 Figure 8.7 Real estate prices and consumer price index in
Korea and Japan

a. Japan
index (2010=100)
230
Consumer price index
Land (residential) price index
180

130

80

30
1975 1980 1985 1990 1995 2000 2005 2010 2015

b. Korea
index (2010=100)
120

110

100

90

80

70

60 Consumer price index

House price index
50
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Sources: Statistics Korea; Kookmin Bank.

counteract.7 In contrast, housing prices in Korea have risen roughly in accor-

dance with general prices over the past two decades (figure 8.7).8 Perhaps for
this reason, Korea has not experienced any major corrections in asset prices,
and the Bank of Korea has had sufficient time to adjust its policy stance to
disinflationary pressures.9

7. For the debate on Japan’s monetary policy in the 1990s, see Hayami (2000), Kuttner and
Posen (2001), and Ito and Mishkin (2004), among many others.
8. Cho (2017) extracts the market’s capital gains expectations from the data on the unique
rental system of Korea, called chonsei, which cannot be assessed as “excessive” compared with
macroeconomic variables such as inflation and interest rates.
9. Cho (2017) compares Korea’s monetary policy with Japan’s, arguing that the Bank of
Korea responded passively to inflation/deflation pressures.

174 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 8.8 Nominal and real interest rates (3-year government
bond)

percent
25
Nominal interest rate
20 Real interest rate

15

10

–5
1990 1993 1996 1999 2002 2005 2008 2011 2014

Note: The real interest rate is computed by subtracting the year-on-year

consumer price index (CPI) inflation rate from the nominal rate. For the
period of 1990Q1–1995Q4, the 3-year government bond rate is not available
and thus the rate for 3-year bonds issued by the Industrial Bank of Korea
(under government guarantee) is used.
Sources: Statistics Korea; Kookmin Bank.

Declining Natural Interest Rate and Monetary Policy

There is no a priori reason to believe that Korea should follow Japan into
deflation, but the changes in real sector fundamentals carry important im-
plications for monetary policy.10 The decline in the “natural interest rate,”
for example, could greatly complicate monetary policy by increasing the
probability that the nominal interest rate hits the zero lower bound.
In the context of the recent debate on secular stagnation, the litera-
ture on the effects of fundamental variables on the natural interest rate is
growing (see Laubach and Williams 2003 and Eggertsson, Mehrotra, and
Robbins 2017, among others). This sort of study may be more relevant to
Korea (a country experiencing drastic structural changes) than to relatively
stable advanced countries. Korea’s nominal interest rate declined from
double-digit levels in the 1990s to low single digits in the 2010s; long-term
rates are now similar to rates in the United States (figure 8.8). The real
interest rate in Korea fell by more than 5 percentage points between 1990

10. The effect of population aging on inflation is not as clear as its negative impact on
growth, because aging should reduce both aggregate demand and aggregate supply at the
same time. The empirical results are mixed. IMF researchers Yoon, Kim, and Lee (2014) and
Liu and Westelius (2016) present evidence from various data sets that aging operates as a
disinflationary factor. Using a larger data set, Juselius and Takáts (2015), of the Bank for
International Settlements, find that population aging is inflationary.

DECLINING POTENTIAL GROWTH IN KOREA 175

Table 8.2 Changes in simulated real natural interest rate (percentage
points)
1990–2015 1990–2040 2015–2040
(A) (B) (B – A)
Change in natural interest rate –4.3 –5.9 –1.6
(1) Mortality rate –1.2 –2.0 –0.8
(2) Fertility rate –0.5 –1.8 –1.3
(3) Total factor productivity progress rate –1.0 –1.2 –0.2
(4) Relative price –0.4 –0.5 –0.1
(5) Others –1.2 –0.4 +0.8
Note: The underlying simulation model is based on Kwon (2017b), which has 80 gen-
erations from ages 20 to 99. Each row measures the difference between the baseline
interest rate and the counterfactual rate that would have been realized if the respec-
tive parameter had not changed since 1990. “Others” indicates the endogenously
generated decline in the interest rate as a result of lagged effects of the changes in
parameters that took place before 1990.
Source: Kwon (2017b).

and 2015; over the same period, it dropped by about 1–2 percentage points
in advanced countries (Williams 2017).
To assess how much of the decline in (real) interest rates is attributable
to changes in real sector fundamentals, we conducted experiments similar to
those in Eggertsson, Mehrotra, and Robbins (2017) by using the overlapping
generations model of Kwon (2017b) for Korea.11 Table 8.2 summarizes the
effects of the counterfactual scenarios that the mortality rate, fertility rate,
rate of TFP progress, and relative price of capital goods have not changed
since 1990. Of the 4.3 percentage point decline in the natural interest rate
generated by the model between 1990 and 2015, 3.1 percentage points are
attributed to changes in these fundamental variables: 1.2 percentage points
to the lower mortality rate, 1.0 percentage points to the TFP slowdown, 0.5
percentage point to the reduced fertility rate, and 0.4 percentage point to
the lower price of capital goods. The remaining 1.2 percentage point decline
reflects the prolonged effects of changes in fundamentals that occurred
before 1990. It takes an extremely long time before the effects of demo-
graphic changes are fully realized (figure 8.9). The effect of the fertility rate,
for example, begins to show up after 10 years; it attains its maximum only
after almost half a century. Changes in the fertility rate since 1990 lowered
the real interest rate by 0.5 percentage point through 2015; an additional
impact of 1.3 percentage points is expected to kick in by 2040. Similarly, the

11. This model is composed of 80 overlapping generations from ages 20 to 99 in any calendar
year, reflecting all demographic changes of the past and the projected future. As there is no
money in the model, the marginal productivity of capital determines the (real) interest rate
(see Kwon 2017b for details).

176 GRAPHICS
2 - CHAPTER
SUSTAINING ECONOMIC GROWTH IN ASIA
8—SUSTAINING ECONOMIC GROWTH IN ASIA
 Figure 8.9 Effects on real natural interest rate of changes in fundamental
variables

percentage points
0.5

–0.5

–1.0

–1.5
Mortality rate
Fertility rate
–2.0
Total factor productivity
Relative price
–2.5
1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050

Note: The underlying simulation model is based on Kwon (2017b), which has 80 generations from
ages 20 to 99. Each row measures the difference between the baseline interest rate and the
counterfactual rate that would have been realized if the respective parameter had not changed
since 1990.
Source: Kwon (2017b).

DECLINING POTENTIAL GROWTH IN KOREA

177
decrease in the mortality rate can bring about another 0.8 percentage point
of decline in the real interest rate over the next two decades. The simulation
results in table 8.2 indicate that the natural interest rate will decline by 1.6
percentage points between 2015 and 2040.
A 1.6 percentage point change is sizable, considering the current level of
the natural interest rate, measured in terms of the overnight interbank call
rate—the target rate of monetary policy in Korea. Estimates differ substan-
tially depending on model specifications and econometric methodologies,
but most studies on Korea’s natural interest rate find that it has fallen
below 1 percent.12 Korea’s natural interest rate may well fall below zero in
the not so distant future, as it has in Japan. The probability of the nominal
interest rate hitting the zero lower bound will increase, squeezing the mone-
tary policy buffer to counter negative shocks.13 It is therefore increasingly
important for the monetary authority to keep inflation expectations from
falling below the target level (currently 2 percent a year). It is also advis-
able that Korea prepare emergency policy measures for cases of large nega-
tive shocks, including unconventional ones that have been experimented
with in advanced countries. However, the most desirable policy would be to
increase TFP, which would moderate the decline in the natural interest rate.

Concluding Remarks
Korea’s potential growth has declined since the 1990s. Given demographic
trends and capital deepening, it is projected to decline further for the next
two to three decades unless overall efficiency (or TFP) is substantially
enhanced. To reverse (or at least mitigate) this trend, bold structural reforms
in the labor, financial, and product markets are warranted. Successful
reforms to enhance market efficiency would boost TFP, which could raise
the natural interest rate as well as potential growth, reducing the danger of
Korea falling into the deflation trap that Japan experienced. Reforms always
confront resistance by vested interest groups. The real policy task may be to
convince Koreans of the pressing need for reforms in order to gain political
momentum for them.

12. Kim and Park (2013), among others, estimate the natural interest rate of Korea using
various versions of the Laubach-Williams (2003) methodology.
13. According to Wynne and Zhang (2017), Japan’s natural interest rate fluctuated around
zero beginning in the early 1990s, falling significantly below zero after the global financial
crisis.

178 SUSTAINING ECONOMIC GROWTH IN ASIA

References
Aiyar, Shekhar, Christian Ebeke, and Xiaobo Shao. 2016. The Impact of Workforce Aging on Eu-
ropean Productivity. IMF Working Paper WP/16/238. Washington: International Monetary
Fund.
Barro, Robert. 1991. Economic Growth in a Cross Section of Countries. Quarterly Journal of
Economics (May): 407–43.
Caballero, Ricardo J., Takeo Hoshi, and Anil K. Kashyap. 2008. Zombie Lending and De-
pressed Restructuring in Japan. American Economic Review 98, no. 5: 1943–77.
Cho, Dongchul. 2017. Is Korea’s Monetary Policy Following in the Footsteps of Japan? In
Japanization: Causes and Remedies, ed. Takatoshi Ito, Dongchul Cho, and Andrew Mason.
Cheltenham, UK: Edward Elgar Publishing.
Cho, Dongchul, and Kyooho Kwon. 2014. Japan of 20 Years Ago and Korea of Today: Aging
and Declining Dynamism [in Korean]. In Economic Dynamism of Korea: With a Focus on a
Comparison with Japan, ed. Dongchul Cho. KDI Research Monograph 2014-03. Sejong City:
Korea Development Institute.
Eggertsson, Gauti B., Neil R. Mehrotra, and Jacob A. Robbins. 2017. A Model of Secular Stag-
nation: Theory and Quantitative Evaluation. NBER Working Paper 23093. Cambridge, MA:
National Bureau of Economic Research.
Hayami, Masaru. 2000. Price Stability and Monetary Policy. Speech presented to the Research
Institute of Japan, Tokyo, March 21.
Hsieh, C., and P. Klenow. 2009. Misallocation and Manufacturing TFP in China and India.
Quarterly Journal of Economics 124, no. 4: 1403–48.
Ito, T., and F. S. Mishkin. 2004. Monetary Policy in Japan: Problems and Solutions. Photo-
copy. Graduate School of Business, Columbia University, New York.
Jeong, Daehee. 2014. Adverse Effects of Zombie Firms: The Case of Korea [In Korean]. In Eco-
nomic Dynamism of Korea: With a Focus on a Comparison with Japan, ed. Dongchul Cho. KDI
Research Monograph 2014-03. Sejong City: Korea Development Institute.
Jung, Kyu-Chul. 2017. Export Dynamics of Japan, Korea, and China. In Japanization: Causes
and Remedies, ed. Takatoshi Ito, Dongchul Cho, and Andrew Mason. Cheltenham, UK:
Edward Elgar Publishing.
Juselius, Mikael, and Előd Takáts. 2015. Can Demography Affect Inflation and Monetary Policy?
BIS Working Paper 485. Basel: Bank for International Settlements.
Kim, Minsu, and Yang Su Park. 2013. Estimating the Neutral Real Interest Rate (NRIR) and
Analyzing Factors of Its Fluctuation in Korea [in Korean]. Economic Analysis (Bank of Ko-
rea) 14: 47–86.
Kuttner, Kenneth N., and Adam S. Posen. 2001. The Great Recession: Lessons for Macroeco-
nomic Policy from Japan. Brookings Papers on Economic Activity 2001, no. 2: 93–185.
Kwon, Hyuk-Wook, and Daeil Kim. 2014. An Analysis of Korea’s Labor Market Efficiency in
Human Resource Allocations [in Korean]. In Economic Dynamism of Korea: With a Focus on a
Comparison with Japan, ed. Dongchul Cho. KDI Research Monograph 2014-03. Sejong City:
Korea Development Institute.
Kwon, Kyooho. 2017a. GDP Growth from the Perspective of Demographic Change: Will Ag-
ing Korea Become Another Japan? In Japanization: Causes and Remedies, ed. Takatoshi Ito,
Dongchul Cho, and Andrew Mason. Cheltenham, UK: Edward Elgar Publishing.
Kwon, Kyooho. 2017b. Korea’s Demographic Transition and Long-Term Growth Projection
Based on an Overlapping Generations Model. KDI Journal of Economic Policy 39, no. 2: 25–
51. Sejong City: Korea Development Institute.
Laubach, Thomas, and John C. Williams. 2003. Measuring the Natural Rate of Interest. Re-
view of Economics and Statistics 85, no. 4: 1063–70.

DECLINING POTENTIAL GROWTH IN KOREA 179

Liu, Yihan, and Niklas Westelius. 2016. The Impact of Demographics on Productivity and Inflation in
Japan. IMF Working Paper WP/16/237. Washington: International Monetary Fund.
Mankiw, N.G., David Romer, and David N. Weil. 1992. A Contribution to the Empirics of
Economic Growth. Quarterly Journal of Economics (May): 407–37.
Oberfield, Ezra. 2013. Productivity and Misallocation during a Crisis: Evidence from the Chil-
ean Crisis of 1982. Review of Economic Dynamics 16, no. 1: 100–19.
OECD (Organization for Economic Cooperation and Development). 2014. Product Market
Regulation Database. Paris. Available at www.oecd.org/economy/pmr.
Oh, Jiyoon. 2017. Misallocation in the Manufacturing Sector of Korea: A Micro Data Analysis.
In Japanization: Causes and Remedies, ed. Takatoshi Ito, Dongchul Cho, and Andrew Mason.
Cheltenham, UK: Edward Elgar Publishing.
Summers, Lawrence H. 2014. U.S. Economic Prospects: Secular Stagnation, Hysteresis, and
the Zero Lower Bound. Business Economics 49, no. 2: 65–73.
Williams, John C. 2017. Three Questions on R-Star. FRBSF Economic Letter. San Francisco:
Federal Reserve Bank of San Francisco.
Yoon, Jong-Won, Jinil Kim, and Jungjin Lee. 2014. Impact of Demographic Changes on Inflation
and the Macroeconomy. IMF Working Paper WP/14/210. Washington: International Mon-
etary Fund.
Wynne, Mark A., and Ren Zhang. 2017. Estimating the Natural Rate of Interest in an Open Economy.
Working Paper 316. Dallas: Federal Reserve Bank of Dallas, Globalization and Monetary
Policy Institute.

180 SUSTAINING ECONOMIC GROWTH IN ASIA

 9
A New Macroeconomic
Policy Framework for
Prudence and Higher-Quality
Growth in China
MA JUN

China’s economic growth potential will slow in the coming decade for struc-
tural reasons that cannot easily be reversed. Technological innovation and
reforms can cushion the deceleration, but it calls for a macroeconomic man-
agement framework that is more consistent with a slower, quality-driven
growth model.
Reining in the buildup of macroeconomic leverage (measured by the
M2/GDP ratio) will be key. There is a consensus that high and growing
macroeconomic leverage has been the main cause for rising financial sector
risks. According to the Bank for International Settlements, China’s nonfi-
nancial credit-to-GDP ratio stood at 256.8 percent in the third quarter of
2017, the highest ratio among the Group of Twenty (G-20) economies. This
ratio grew at an annual average pace of 11 percentage points over the past
decade, the highest rate among G-20 countries. China’s leadership has em-
phasized that prevention of financial sector risks is its top priority.
The targeting or pursuit of an aggressive GDP growth rate has contrib-
uted to the excessive growth of monetary aggregates and thus the rapid
buildup in leverage. In 2009–10, the period after the global financial crisis,
for example, the government targeted a GDP growth rate of 8 percent. It
achieved 9.4 percent in 2009 and 10.6 percent in 2010. M2 growth acceler-
ated to 27.7 percent in 2009, 11.2 percentage points higher than the annual

Ma Jun is director of the Center for Finance and Development, Tsinghua National Institute of Financial Re-
search, and member of the People’s Bank of China Monetary Policy Committee. The author thanks Wang Lish-
eng, Zhu Shouqing, He Xiaobei, Qi Xing, Yang Xiaohai, and Wang Tianyu for their contributions to this research
and Changyong Rhee and Alfred Schipke of the IMF for arranging the author’s presentation at the conference.

181
average M2 growth during 2000–08. During 2011–16, the M2/GDP ratio
continued to rise by 6.9 percentage points a year, as the government aimed
for GDP growth of around 7 percent.

Structural Reasons for Growth Deceleration

Three key structural factors will continue to decelerate the Chinese economy
in the coming decade or decades: demographics, environmental costs, and
a shift in consumer preferences from goods to services. All of them are diffi-
cult or impossible to reverse.

Demographics: Rapid Decline in the Labor Force

Thanks to the one-child policy introduced in the 1970s, China’s total fertility
rate dropped from 4.8 percent in the early part of that decade to 1.7 percent
in 2016 (Ministry of Health and Family Planning 2016). The policy has led
to a significant decline in the labor force (population aged 15–64) since 2013
(figure 9.1).
According to the author’s demographic model,1 which accounts for the
effect of the two-child policy introduced in 2016, the decline in China’s labor
force will accelerate from 1.8 million in 2017 to 10 million in 2028, equiva-
lent to 1.1 percent of the labor force that year. Given a baseline labor share
of 50 percent in the production function, in the production function, the
acceleration in the decline in the labor force will lead to a drop in economic
growth of about 0.5 percentage point in 2028 from the 2017 rate.2 (This
estimate does not account for the fact that the rapid aging of the popula-
tion may result in additional deceleration in growth via a lower saving rate.)

Environmental Costs
China’s traditional growth model, led by investment and heavy manufac-
turing, resulted in serious environmental degradation. This deterioration is
beginning to limit potential growth and may threaten social stability.
Most of the environmental costs China is facing arise from air and water
pollution, soil contamination, and CO2 emissions. Future generations will
bear these costs, the benefits of which were reaped by people who enjoyed
very high income and wealth growth in the past several decades. Remediation
costs in China during 2000–10 accounted for 6.5 percent of GDP for air, 2.1
percent for water, and 1.1 percent for land pollution/degradation, according

1. The author’s demographic model is available on the PIIE website at https://piie.com/

bookstore/sustaining-economic-growth-asia.
2. According to the literature, the labor share in the production function in China is about
50 percent (see Cai and Lu 2013, among many others).

182 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 9.1 Actual and projected labor force participation in China,
2000–50

population aged 15–64 (100 million) population growth (percent)

10.5 2.0

1.5
10.0
1.0

9.5
0.5

9.0 0

–0.5
8.5
–1.0

8.0 –1.5
0

12

16

20

24

28

32

36

40

44

48
0
0
0

20

20

20
20

20
20

20
20
20

20
20

20
20

Population growth (right axis)

Population aged 15–64 (left axis)

Sources: National Bureau of Statistics of China and author’s demographic model.

to estimates by the RAND Corporation (Crane and Mao 2015). Water pollu-
tion and land contamination costs are likely to rise. According to Zhuang
Guotai, deputy director general of the Ministry of Environment, the reme-
diation costs for land contamination will be far greater than the costs for air
and water pollution and could amount to many trillions of renminbi.3
These costs will significantly increase the costs of producing goods and
services. For example, carbon prices may rise by a factor of 10 to 15 in the
coming decade, according to the World Bank (2017), reaching $50 to $100
a ton by 2030.
Higher input costs should lead to lower profitability and thus less pro-
duction and lower economic growth. Our dynamic CGE model shows that
China’s annual GDP growth may slow 0.5 percentage point during the
energy transition in 2017–2030 assuming clean energy is 30 percent more
expensive than dirty energy.4 Rising water and food costs may have a similar

3. See He Gwangwei, “The Soil Pollution Crisis in China: A Cleanup Presents Daunting
Challenge,” Yale Environment 360, July 14, 2014, https://e360.yale.edu/features/the_soil_
pollution_crisis_in_china_a_cleanup_presents_daunting_challenge.
4. The author employed the GEMPACK software to calculate this estimate of GDP growth
slowdown during the energy transition period. Replicating this estimate may be difficult
without the author’s assistance. Accordingly, this chapter does not provide the GEMPACK
code as part of the normal replication package required by PIIE. PIIE does not assume

A NEW MACROECONOMIC POLICY FRAMEWORK FOR CHINA 183

impact. After accounting for these environment-related costs (essentially a
debt incurred by the past generation), it is possible that annual GDP growth
could slow by 1 percentage point in the coming decade.

Shift in Consumer Preference from Goods to Services

China’s services output-to-GDP ratio rose from 43 percent in 2008 to 52
percent in 2017 (National Bureau of Statistics of China). It will continue to
rise in the coming decade, because the penetration rates of goods (housing,
electronics, food, and clothing) are already high (close to averages in the
Organization for Economic Cooperation and Development) whereas per
capita consumption of services (health care, insurance, entertainment,
sports, and so on) remains significantly lower than in developed countries.
This structural shift in consumer preference from goods to services
will likely result in a deceleration of productivity growth. Average annual
labor productivity growth in 2007–16 was about 9.0 percent in manufactur-
ing and 5.8 percent in services.5 Given China’s current economic structure
(manufacturing accounts for 40 percent of the economy and services for
52 percent; see figure 9.2), these figures imply aggregate (nonfarm) labor
productivity growth of 7.2 percent. Based on the current trend, one can
expect a 10 percentage point rise in the share of the services sector over the
coming decade (and a decline in the share of the manufacturing sector).
This structural change implies aggregate (nonfarm) labor productivity
growth of roughly 6.9 percent in 10 years. All else held constant, a decline of
0.3 percentage point (from 7.2 to 6.9 percent) in annual labor productivity
growth reduces annual economic growth by roughly 0.3 percentage points
in a decade.

Inevitability of a Growth Slowdown

The headwinds from these structural changes are difficult to reverse. In
addition, China’s economic growth is facing other challenges, including
very high macroeconomic leverage, which means the country is no longer
positioned to borrow at the pace it had. Urbanization is also slowing, which
means that growth driven by infrastructure and property investment may
lose steam.
Fortunately, the growth slowdown may not necessarily translate into
severe unemployment, as the number of people seeking jobs will shrink as
well. According to the author’s demographic model, the labor force will

responsibility for the calculations underlying this estimate. Anyone seeking further infor-
mation should contact the author at maj@pbcsf.tsinghua.edu.cn.
5. “Manufacturing” is used as shorthand for “manufacturing, construction, and mining.”

184 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 9.2 Services and manufacturing sectors in China as percent
of GDP, 1990–2017

percent
60

50

40

30

20

10
Manufacturing, construction, and mining
Services
0
90

92

94

96

98

10

12

14

16
0

0
0
0
0

20

20
20
20
19

19
19

20
19

20
19

20
20
20

Source: National Bureau of Statistics of China.

decline by 11 million people, obviating the need to create net new jobs by
2028.6

Measures Already Taken to Offset Deceleration Pressures

The Chinese government has launched a number of initiatives to offset
some of these factors and boost growth potential. The most notable effort
is the promotion of technological innovation. Central and local govern-
ments have promoted technological innovation and the development of
new business models through measures such as tax deductions and exemp-
tions; increased government expenditure on research and development; the
establishment of more than 3,000 high-tech startup incubators; encourage-
ment of the transformation of scientific products for commercial use; and
the creation of one-stop services for patent examination, verification, and
protection. Many Chinese high-tech companies, especially internet and e-
commerce businesses, were listed on stock exchanges in the past few years.
In power, engineering, mining, high-speed rail, and construction, China has
become a technology exporter.

6. Rural labor migration is projected at about 8 million people in 2028, almost the same as
the net reduction in the urban labor force that year.

A NEW MACROECONOMIC POLICY FRAMEWORK FOR CHINA 185

 Acceleration of technology development may not be enough to offset
the economic deceleration caused by structural factors, however. According
to Gordon (2016), total factor productivity growth in the United States
averaged less than 1 percent during 1980–2010. China has many patents,
but their average quality is low, and per capita patent holding is only 10
percent that of Korea, which has not escaped economic deceleration in
the past two decades. In light of these international experiences, it seems
likely that technology may only partially offset the downward pressure on
economic growth in China.

Features of the New Macroeconomic Policy Framework

Given slowing growth and the need to curb macroeconomic leverage,
China’s macroeconomic management framework needs to shift toward one
that promotes slower but higher-quality growth. Such a new framework
should (1) replace GDP with employment as the most important macroeco-
nomic policy objective; (2) enhance the independence of monetary policy,
in order to avoid dominance by dovish tones in monetary policymaking;
and (3) make sure that the planning of fiscal and quasi-fiscal expenditures—
such as unfunded mandates for local governments on education, health-
care, environment, poverty alleviation, and so on, as well as the launching
or support of economic/high-tech development zones—is consistent with
macroeconomic and financial stability objectives, in order to avoid further
expansion of local government debt.
China’s macroeconomic management system centers on achieving
growth targets. It has the following key features:
n The central government sets 1-, 5-, and 10-year growth targets. Provin-
cial and lower-level governments also set their own targets for growth,
which are typically higher than the national target (e.g., if the central
government targets 6.5 percent, the provincial government targets 7.5
percent and the city- or county-level government possibly 8.5 percent).
Promotion of local government officials (such as party chiefs and
mayors) depends largely on achievement of local-level GDP growth
targets.
n Various ministries and bureaus (including those in charge of industrial
development, infrastructure, and social services) set targets for sectoral
growth. Once aggregated, they are typically more ambitious than the
national growth target. Based on these sectoral targets, the central
government issues mandates for local governments (such as achieving
targets for poverty reduction, environmental protection, education,
health care, elder care), many of which are not funded or only partially
funded by national budgets. Most of these mandates have been funded

186 SUSTAINING ECONOMIC GROWTH IN ASIA

 by local government financial vehicles or public-private partnership
projects that borrowed from banks or the bond market. They are a key
reason why local government debt surged over the past decade.
n China sought to spur regional development by launching many devel-
opment zones. The central and provincial governments launched about
5,000 such zones (economic development, high-tech, free trade, and so
on). City, county, and township governments also established numer-
ous development zones. No official statistics exist on their number, but
there are likely many thousands, if not tens of thousands. Infrastruc-
ture investment demand in the 5,000 development zones at or above
provincial levels could amount to nearly RMB50 trillion, or 60 percent
of China’s GDP in 2017.
n Historically, the State Council set monetary growth targets, based on
negotiations among key ministries. M2 grew at a pace that was about 3
percentage points higher than nominal GDP over the past two decades.
The main reason for the excessively high M2 growth was that most
parties influencing policymaking were beneficiaries of rapid monetary
expansion and none of them cared about its eventual macroeconomic
consequences, such as inflation and financial risks. The central bank,
where professionals are most concerned about macroeconomic stabil-
ity, has limited influence in this process.

These features suggest that the current macroeconomic management

framework may be prone to overheating, as a result of untenable promises
on deliverables (as reflected in the large unfunded mandates for local gov-
ernments), excessive incentives for achieving higher GDP growth through
overborrowing, and limited restraints on monetary expansion. When the
economy enjoyed strong underlying growth potential (of about 10.5 percent
in the 2000s) and financial leverage was still modest, such features did not
pose a major threat to macroeconomic and financial stability. In the “new
normal” phase, which may see growth slow and room for leverage shrink,
the old macroeconomic management model becomes unsustainable.
This growth-centered macroeconomic management model needs to
be replaced with a quality-centered one that focuses on macroeconomic
prudence and financial stability. The decline in growth potential and labor
force makes this transition feasible, as there is no more need to main-
tain strong growth for job creation. The need to contain macroeconomic
leverage makes the transition imperative. The new approach needs to focus
on several areas.

A NEW MACROECONOMIC POLICY FRAMEWORK FOR CHINA 187

Abandoning the GDP Growth Target
China has long used GDP growth targets to boost investment and economic
growth. In the 12th Five-Year Plan, the government set the goal of doubling
the country’s real GDP and household income by 2020 from 2010 levels.
The drawbacks of this kind of target-setting have become increasingly
evident. The main problem is that local governments top up the national
growth target by setting higher targets and then resort to heavy borrowing
to achieve them, pushing up local government debt and hence the leverage
ratio of the overall economy.
In addition to boosting local government debt and macroeconomic le-
verage, the overemphasis on GDP growth performance caused overcapac-
ity, environmental degradation, and statistical fraud. As many local gov-
ernments lack sufficient revenues to fund their infrastructure investments,
they resorted to multiple sources of debt financing, including loans, bonds,
and shadow banking products. When they failed to meet their GDP targets,
some local governments manipulated their statistics, as evidenced by the
fraud reported in Liaoning, Tianjin, and Inner Mongolia.7
China should scrap the national GDP growth target and replace it with
an unemployment rate target as the most important macroeconomic policy
objective. Doing so would reduce the political pressure on local govern-
ments to borrow. GDP growth forecasts (instead of targets) should still be
used as a guide for budgetary activities. They could be issued by the central
bank, the National Development and Reform Commission, or the Ministry
of Finance.

Enhancing Monetary Policy Independence

The State Council, rather than the central bank, makes all key monetary
policy decisions in China, including adjustments to the benchmark interest
rates and reserve requirement ratios. As the State Council makes decisions
largely by consensus, the system often reflects the collective opinions of
ministries in charge of economic development and policymaking, as well as
the views of local governments, which are less concerned than national poli-
cymakers about the macroeconomic spillover of such policies. The views
of the central bank—the agency mandated to maintain macroeconomic
and financial stability—carry limited weight. This system has a natural bias
toward excessive monetary expansion.

7. Yawen Chen and Ryan Woo, “Another Chinese city admits ‘fake’ economic data,” Reuters,
January 16, 2018, www.reuters.com/article/us-china-economy-data/another-chinese-city-
admits-fake-economic-data-idUSKBN1F60I1.

188 SUSTAINING ECONOMIC GROWTH IN ASIA

 The central bank should be given more independence in monetary policy
making, transitioning away from M2 growth targeting toward interest rate
targeting. To do so, when authorities announce the official policy rate (to
replace the benchmark lending and deposit rates) as the intermediate target
for monetary policy, the central bank should be given key decision-making
power. This arrangement would better ensure macroeconomic and finan-
cial stability as a top priority of monetary and macroprudential policies.

Creating a Macroeconomic Stability Screening Mechanism to

Cut Unrealistic Public Policy Mandates
To avoid further expansion of local government debt, especially implicit
debt, it is important to make sure that the planning of quasi-fiscal expen-
ditures (such as unfunded mandates for local governments on education,
health care, environment, poverty reduction, and so on) is consistent with
macroeconomic and financial stability objectives. A “macroeconomic sta-
bility screening mechanism” should be established for all major macro-
economic policies by quantifying their fiscal and monetary implications.
For example, many targets set for poverty reduction, pollution reduction,
and improvements in health care, education, and infrastructure require sig-
nificant fiscal resources. If not included in the official fiscal budget, these
mandates most likely result in local government quasi-fiscal debt. If they
do not pass the macroeconomic stability screening, which sets a limit on
total government borrowing (including targets of budgetary fiscal deficit
and quasi-fiscal deficit), such mandates should not be included in the gov-
ernment’s work plan.

Abolishing Some Development Zones

Some development zones have contributed to rapid regional economic
growth and technological innovation. But many are heavily indebted,
because they borrow excessively for expenditures on infrastructure invest-
ment without credible plans for repayment.
The current policy of “one county, one zone” is inappropriate, because
many counties lack the natural resources, workers, talent, and locational
comparative advantages to attract private investment and land purchase in
large-scale infrastructure development. As many as two-thirds of China’s
5,000 development zones (at and above the provincial level) lack the growth
potential initially foreseen or claimed. Continuing to advance large-scale
development zones broadly would exacerbate the local government debt
problem and increase macroeconomic leverage. The central government
should therefore consider abolishing many of these development zones and
refraining from launching new ones.

A NEW MACROECONOMIC POLICY FRAMEWORK FOR CHINA 189

Enhancing the Transparency of the Quasi-Fiscal Debt of Local
Governments
In 2013 the Third Plenary Session of the 18th Chinese Communist Party
Congress decided that central and local governments would henceforth
publish their government balance sheets, in order to enhance the transpar-
ency and responsibility of local government operations. Progress has been
slow, with many local governments compiling but not publishing their
balance sheets.
The rapid increase in local government quasi-fiscal debt in the past
five years has underscored the importance and urgency of implementing
this reform. A mandatory requirement for disclosure of local government
balance sheets, including local financing vehicle debts, which represent con-
tingent or implicit liabilities of the local governments, is critical to deter
irresponsible and excessive borrowing by local governments. Publishing
information on quasi-fiscal borrowing gives the general public, local resi-
dents, members of the local people’s congress, the media, banks, the debt
market, and third-party service providers such as rating companies impor-
tant information with which to assess local government debt risks. Public
opinion and pressure from all these parties for local governments to be
prudent could be very powerful.
The authorities could take the following steps to implement the deci-
sion:
n The Ministry of Finance should create a standardized template for local
government balance sheets, with clear definitions of contingent and
implicit liabilities of local governments. For example, debts incurred
by companies that are majority owned by local governments and for
developing infrastructure projects whose cash flows are insufficient
to cover debt repayments should be treated as implicit or contingent
government liabilities, regardless of whether the local government
issued an official guarantee.8
n The Ministry of Finance could select a few provinces and cities as pilot
programs for launching this reform.
n Once sufficient experience is gained, this practice should be quickly
replicated in the rest of the country.

8. The author was an advisor to the “balance sheeting working group” of a provincial-local
government a few years ago and was involved in many technical discussions on definitions,
scope, and valuations of local government balance sheet items. His experience suggests that
these issues are not as problematic as many people perceive.

190 SUSTAINING ECONOMIC GROWTH IN ASIA

Conclusion
Unlike countries that face secular stagnation, China is facing a slowdown in
economic growth driven by three unavoidable structural issues: demograph-
ics, environmental costs, and changing consumer preferences. Ongoing
technological innovations and reforms can partially cushion the decelera-
tion, but a further slowdown in growth is inevitable over the medium term.
This trend, together with the growing financial risks caused by a high
macroeconomic leverage ratio, calls for a new macroeconomic management
framework. Key to sustainable, high-quality, and environment-friendly
growth in China is adoption of a framework that is less growth-centric and
focuses more on macroeconomic and financial stability. To put such a new
framework in place, policymakers should consider the following actions:
n Make employment, rather than GDP, the most important macroeco-
nomic policy objective.
n Enhance the independence of monetary policy, in order to avoid the
dominance of dovish tones in monetary policymaking.
n Make sure the planning of fiscal and quasi-fiscal expenditures is
consistent with macroeconomic and financial stability objectives, in
order to avoid further expansion of local government debt, especially
implicit debt.
n Abolish unqualified development zones.
n Enhance the transparency of quasi-fiscal borrowing by local govern-
ments.

Even at somewhat lower rates, China’s growth will still be among the
highest in the world in the coming decade—and the declining size of its labor
force will allow it to tolerate a more moderate GDP growth rate without
causing serious unemployment. Hence there is no convincing reason to be
obsessed about reaching an annual GDP growth target. By reducing the
probability of financial crises and large disruptions to growth, the new mac-
roeconomic management framework described in this chapter should help
China better achieve its sustainable development objectives in the medium
and long terms.

References
Cai, Fang, and Yang Lu. 2013. Population Change and Resulting Slowdown in Potential GDP
Growth in China. China & World Economy 21, no. 2: 1–14.
Crane, Keith, and Zhimin Mao. 2015. Costs of Selected Policies to Address Air Pollution in China.
Santa Monica, CA: RAND Corporation.
Gordon, Robert J. 2016. The Rise and Fall of American Growth: The U.S. Standard of Living since the
Civil War. Princeton, NJ: Princeton University Press.

A NEW MACROECONOMIC POLICY FRAMEWORK FOR CHINA 191

Ministry of Health and Family Planning. 2016. Statistics on China’s Health and Family Planning
Development. Beijing.
World Bank. 2017. Guidance Note on Shadow Price of Carbon in Economic Analysis. Washington. Avail-
able at http://pubdocs.worldbank.org/en/911381516303509498/2017-Shadow-Price-of-
Carbon-Guidance-Note-FINAL-CLEARED.pdf.

192 SUSTAINING ECONOMIC GROWTH IN ASIA

 10
Do India’s Exports Reflect
the New Normal?
PRACHI MISHRA AND SIDDHARTHA NATH

India’s merchandise exports underwent a quiet revolution over the last two
decades, with exports of engineering goods, electronics, and pharmaceuti-
cals gradually replacing India’s traditional exports of leather, textiles, gems,
and jewelry. This chapter uses sectoral and firm-level data to examine what
drives India’s exports. It answers the following questions: How has India
fared since the global financial crisis? Is India a closed economy? Has it been
insulated from global shocks? The first two sections examine how open
India’s economy is and how the composition of its exports has changed.
The next two sections analyze the determinants of India’s exports (global
growth, the real effective exchange rate, supply bottlenecks) at the aggre-
gate and firm levels. The last section summarizes the chapter’s conclusions
and identifies their policy implications.

How Open Is India’s Economy?

Since the global financial crisis, low economic growth has been recognized
as the “new normal” or “new mediocre,” especially in advanced economies.
The new mediocre has been attributed to both demand and supply factors—
inadequate aggregate demand and a slowdown in productivity growth. Al-
though demand recently rose in advanced economies, underlying potential

Prachi Mishra is deputy division chief in the Western Hemisphere Department at the International Monetary
Fund. Siddhartha Nath is research officer at the Reserve Bank of India. They would like to thank Medha Madhu
Nair for excellent research assistance. The views expressed in this chapter are those of the authors and do not
necessarily represent the views of the IMF, its Executive Board, or its management, nor the views of the institu-
tions to which they belong.

193
 Figure 10.1a India’s exports and imports, 2000–16

percent of GDP
60

55

50

45

40

35

30

25
2000 2002 2004 2006 2008 2010 2012 2014 2016

Source: Ministry of Statistics and Program Implementation, Government

of India.

growth is significantly lower than in emerging-market economies, reflect-

ing unfavorable demographics and soft productivity growth.
The new mediocre has coincided with a slowdown in trade growth rela-
tive to overall global growth. China, India, and Indonesia have continued
to enjoy relatively higher growth since the global financial crisis, in contrast
to Japan and Korea. Their success reflects supportive macroeconomic poli-
cies and strong convergence dynamics.
India’s financial integration with the rest of the world is increasingly
appreciated and visible. Capital inflows and outflows are rising, and asset
prices in India are increasingly correlated with their global counterparts.
But India’s integration into the global real economy continues to be under-
appreciated, particularly on the export front. The conventional wisdom is
that India remains a largely closed economy, in which exports are of only
marginal significance. This line of thinking is often used to suggest that
India is insulated from global growth and/or protectionist shocks.
Over the last two decades, India’s tradable sector grew rapidly. Total
trade (the sum of exports and imports as a percent of GDP) rose from 27
percent in 2000 to 56 percent in 2012, before slowing during the deglobal-
ization of recent years (figure 10.1a). Increased integration and exposure
to global markets forced supply chain efficiencies and resulted in the tech-
nological transfer and productivity growth that comes with them. But the
growth and productivity benefits of global integration also made India’s
economy more vulnerable to global growth and protectionist shocks—
more than is commonly assumed.
Measuring India’s exposure on the export side is a function of how

194 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 10.1b Share of India’s exports in GDP, 2000–16

percent of GDP
28

25

22

19

16

13

10
2000 2002 2004 2006 2008 2010 2012 2014 2016

Source: Ministry of Statistics and Program Implementation, Government of India.

exports have grown as a share of GDP and how sensitive they are to changes
in partner-country growth. The larger the share of exports in GDP and the
more elastic/cyclical they are to changes in global economic conditions, the
more vulnerable the economy is.
India’s exports as a share of GDP more than doubled between 2000
and 2013, rising from 12 percent to 25 percent. This share declined in
recent years (figure 10.1b), broadly mirroring changes in emerging markets
(figure 10.1c). Despite the slowdown, however, India’s exports still repre-
sented almost 20 percent of GDP in 2017. This share is almost twice what
it was 15 years ago and almost as large as in Indonesia.
Complementing this openness is the fact that until recently, India’s
export basket had become progressively more sensitive to global growth.
Although the relationship between growth in the advanced economies and
exports in India has softened in recent years, elasticity to partner-country
growth remains high (Aziz and Chinoy 2012, Rangarajan and Mishra 2013).
It is the primary driver of export volumes, eclipsing the role of the exchange
rate and supply bottlenecks.
The growing role of exports in India’s growth process can be appreci-
ated by recognizing the contribution of slowing exports to slowing headline
GDP growth (figure 10.2). India’s much-celebrated 9 percent growth in the
2000s came largely as a result of surging export growth, as global growth
rose and India plugged into the global export market. Exports underpinned
India’s 8.8 percent GDP annual growth in 2003–07 (nearly 18 percent a year
in real terms).
In contrast, over 2012–16, export growth slowed to just 2.6 percent a

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 195

 Figure 10.1c Exports of goods and services (current US dollars),
2006–15

year-on-year change in percent

50

40

30

20

10

–10

–20

–30

–40
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

India
Average of emerging-market economies

Note: The average for emerging markets is an export-weighted average.

Source: IMF, Balance of Payments Statistics Database.

Figure 10.2 Role of exports in India’s growth slowdown

percent year on year

18
17.8
15
GDP growth = 8.8 percent
12

9
GDP growth = 6.9 percent
6
6.7
5.4
3
2.6
0
2003–07 2012–16

Exports
Private consumption

Source: Ministry of Statistics and Program Implementation, Government

of India.

196 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 10.3 Role of exports in slowing growth

basis points
350
300
250
200
150
100
50
0
Reduction in growth Attributable to exports
after series adjustment

Source: Authors’ calculations.

year in real terms. With exports constituting 20 percent of GDP, a 15 percent-

age point slowdown in exports—after adjusting for their import content—
would shave off more than 2 percentage points from headline GDP growth.
GDP growth in India averaged 6.9 percent over 2012–16, down from
8.8 percent in 2003–07. The slowdown in exports explains virtually all of
the slowdown in GDP growth over the two periods.
A change in the GDP methodology renders comparison between the two
periods tricky, because there appears to be a level shift up in GDP growth
under the new series. For the two years in which the series overlap, the new
GDP series is about 110 basis points higher on average than the old series. If
one bumps up the old GDP series by 110 basis points to facilitate a more ap-
ples-to-apples comparison, the slowdown in growth is closer to 3 percentage
points, with exports accounting for two-thirds of the slowdown (assuming
that the markup to GDP growth in the old series is not accompanied by a
commensurate markup in export growth). This assumption appears reason-
able, because the methodological changes to GDP appear to have affected
mainly consumption and investment, not exports. In the two years in which
the data overlap, export growth is virtually identical across the two series.
These results indicate that exports have been a critical driver of India’s
slowdown, accounting for about two-thirds of India’s GDP slowdown over
the last decade (figure 10.3). India did not escape the deglobalization blues
that afflicted the world since the global financial crisis.
If global growth remains sluggish, export growth will be tepid. Either
one needs to reevaluate potential growth in India (is 7 percent growth today
equivalent to 9 percent growth when global growth was surging in the mid-
2000s?) or India will have to increase its share of global exports, which was
flat in recent years (see figure 10.4), or India needs to find other sources of
domestic growth to offset the loss from external demand. Something will
have to give.
DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 197
 Figure 10.4 Evolution of global export share of India versus
China (share of total world goods exports)

percent percent
16 2.0
14 1.8
1.6
12
1.4
10 1.2
8 1.0
6 0.8
0.6
4
0.4
2 0.2
0 0
Ju r 2 01
Ja p 20 1
O ua il 2 02

ob 2 9
Ju 20 0
Ja p y 2 0
O ua il 2 11
ob 2 2
Ju r 2 13
Ja Ap 20 13
O ua il 2 14
ob 2 5
Ju r 2016
ly 16
17
ob 2 3
Ju r 2004
Ja Ap 20 4
O a 20 5
ob 2 6
Ju r 2 07
Ja pr 20 7
nu il 2 08
ly 00

A ly 0

ct ry 01

ct ry 01
nu ril 0
ct ry 00

ct ry 0

ct ry 0
ly 0

er 01
l 1
n r 0

20
e 0
ly 0

e 0
e 0

e 0
0

O a 0
e 0
ob 2
ct ry

r
n r
O ua

A
A

n
n
Ja

China (left axis)

India (right axis)

Source: World Trade Organization.

Changing Composition of India’s Exports

India’s export sector has undergone two revolutions. The first was the
surge in services exports. In 2003 service exports constituted 30 percent of
the total export basket. Just four years later, they jumped to 40 percent of
the basket, reflecting the global software and business process outsourcing
(BPO) revolution. Services exports then plateaued over the last decade
(figure 10.5).
The second revolution occurred in manufacturing (figure 10.6). In
2003 textiles, leather, and gems/jewelry—India’s traditional exports—consti-
tuted nearly 60 percent of the merchandise export basket (excluding petro-
leum). By 2015 they accounted for just 40 percent of the basket. In contrast,
exports of engineering goods (such as auto parts and capital goods) grew at
an average annual pace of almost 20 percent for 13 years. Their share of the
manufacturing export basket rose from 20 percent in 2003 to around 35
percent in 2015. India thus moved up the value chain, with exports increas-
ing in sophistication and value addition. By 2015, for example, engineering
goods, electronics, and pharmaceuticals/chemicals constituted almost 60
percent of the merchandise goods basket (excluding oil). (All of this growth
occurred in capital-intensive sectors, not in sectors that would have created
jobs in a country with a favorable demographic dividend.)
The changed composition of exports has a crucial bearing on the extent
to which global demand affects Indian exports, because sector elasticities

198 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 10.5 Share of services in export
basket (excluding oil),
2003–15

percent of total
45
40
35
30
25
20
15
10
5
0
2003 2007 2015

Source: Authors’ calculations.

Figure 10.6 Sectoral change in manufacturing

exports, 2003–15

change in share of indicated manufacturing products as a

percent of total merchandise exports in percentage points
15

10

–5

–10
Engineering Textiles Gems Leather
goods

Sources: Ministry of Commerce, Government of India; JP Morgan

Research.

to global growth are very heterogeneous. India’s new exports (engineering

goods, pharmaceuticals, software services) typically have far higher elastic-
ity to global growth than its traditional exports of leather, textiles, gems,
and jewelry—a hypothesis tested in this chapter. As in countries of the Asso-
ciation of Southeast Asian Nations (ASEAN), the export basket in India has
become more cyclical and more sensitive to global business cycles, surging
in good times and contracting in a down cycle. The volatility of global
growth transmits more directly and acutely into India given this changed
composition of exports.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 199

 Macroeconomic Analysis: How Sensitive Are Exports to
Changes in Trading Partners’ Growth and Exchange Rates?
This section uses time-series data to analyze the sensitivity of India’s exports
to trading partners’ growth and exchange rate movements. Appendix 10A
describes the data.
We estimate the following empirical specification:
ln�𝑋𝑋� � ��� ���� ln�𝑌𝑌� � � �� �𝐷𝐷� ∗ ln�𝑌𝑌� �� � �� ln�𝐸𝐸𝐸𝐸� � � �� �𝐷𝐷� ∗ ln�𝐸𝐸𝐸𝐸� �� � ��� � ��� � �𝐷𝐷� ��
∗ ln�𝐸𝐸𝐸𝐸� �� � ��� � ��� � �𝐷𝐷� ����� (10.1) (10.1)
where t represents time (the quarter), Xt represents the value of real exports
(in 2010 prices in millions of dollars), Yt is partner country GDP (2010 =
ln�𝐹𝐹�� � of
100), and ERt is the index ��the ln�𝐸𝐸𝐸𝐸� �country
��partner � �𝑋𝑋� � �real
� � �� ����
exchange rate (2010
���� (10.2)
= 100). The values taken by the index Yt are different for each sector. For
services, the value is world GDP at 2010 prices, in millions of dollars. Dt
is a dummy variable indicating the structural break, which we identify by
applying the Quandt-Andrews test in each regression separately. � Zt denotes
���
�
additional time-varying controls, such as the number of stalled projects (a
���
proxy for domestic supply constraint) and volatility in the rupee-to-dollar
exchange rate.
We estimate equation (10.1) separately for total merchandise exports,
manufacturing exports, exports in nine broad merchandise subsectors,
and services exports. For services, we use dollars per rupee. We examine the
differential impact of the structural break through changes in the inter-
cept (by introducing the standalone dummy variable Dt) and changes in
the slope coefficients (by interacting Dt with Yt and ERt). The regression also
controls for time-varying effects, by introducing a set of dummy variables,
Tt where each dummy variable represents a single year. The variables Xt and
Yt are seasonally adjusted for all sectors. We use Newey-West heteroskedas-
ticity and autocorrelation-adjusted standard errors in all the regressions.
The coefficient b1 measures the elasticity of India’s exports with
respect to partner country GDP in the pre–structural break sample period.
It shows the percentage by which India’s real exports change if the partner
countries grow by 1 percent. The coefficient b2 measures the differential
effect of partner country demand on Indian exports after the structural
break. b3 measures the percentage change in India’s real exports from a 1
percent appreciation in the real effective exchange rate in the pre–struc-
tural break period. b4 measures the differential effect1 of real exchange rates
on Indian exports after the structural break.
The empirical models yield several main findings:
n The elasticity of India’s exports to partner country GDP is positive and
statistically significant. A 1 percent growth in partner country GDP

200 SUSTAINING ECONOMIC GROWTH IN ASIA

Table 10.1 Determinants of India’s merchandise and services exports
(1) (2) (3)
Total
merchandise Manufactured Services
Variable exports items exports
2.8*** 2.3*** 1.6***
Log of partner GDP
(0.2) (0.1) (0.5)
Log of partner GDP * Structural break –2.5*** –1.4*** –0.8
dummy (0.2) (0.2) (0.7)
–1.3*** –0.8 2.5***
Log of partner exchange rate
(0.2) (0.9) (0.5)
Log of partner exchange rate * Structural 1.7*** 1.9 –2.7***
break dummy (0.3) (2.3) (0.6)
0.1** 0.1
Log of number of stalled projects
(0.0) (0.1)
79.1* 72.0 71.3
Volatility of dollar/rupee exchange rate
(41.4) (70.0) (59.7)
3.0** –2.3 –6.2
Structural break dummy
(1.3) (11.1) (4.8)
Model properties
Sample period 2002Q4 to 2016Q3
Number of observations 56 56 56
Structural break date 2005Q1 2005Q1 2005Q4
Max Wald-F: Quandt-Andrews test for 59.8*** 48.4*** 77.0***
structural break
Dependent variable: Log of exports (2010 prices) in US dollars.
Notes: Manufacturing exports exclude the exports of petroleum products, gems and
jewelry, and primary products like agricultural items, ores, and minerals. Numbers in
parentheses represent standard errors. *, **, and *** indicate statistical significance at
10, 5, and 1 percent, respectively.
Sources: CEIC database; World Bank, World Development Indicators; Centre for
Monitoring the Indian Economy (CMIE) Capex database.

was associated with a 2.8 percent rise in India’s merchandise exports

before 2005. The elasticity was 2.3 percent for manufacturing exports
and 1.6 percent for services exports (table 10.1).
n There is evidence of a sharp structural break in the effect of partner
country output on Indian exports around 2005, well before the global
financial crisis (see table 10.1). Partner country income elasticity falls
sharply after the structural break, declining by 2.5 percentage points,
reducing the effective elasticity to almost zero after 2005. The elastic-
ity of manufactured goods remained high, at 1 percent, and statistically
significant after 2005. In contrast, for services exports, there is no sig-
nificant evidence for changes in elasticity with respect to partner growth
after the structural break.

DO INDIA’S EXPORTS REFLECT

GRAPHICS CHAPTER 102011
THE NEW- NORMAL?
n There is significant variation in partner income elasticities across manu-
facturing sectors,1 both before and after the structural break (table 10.2).
Engineering and electronics reported the highest elasticities before
2005, followed by chemicals, pharmaceuticals, and leather and textiles.
The fall in elasticity after 2005 was very sharp for electronics and en-
gineering goods. Despite the decline, engineering and pharmaceuticals
reported the highest income elasticities after 2005. There was no change
in the elasticity of leather and textiles after the structural break.
n There is little evidence that an appreciating rupee hurts Indian exports.
The weak relationship between exports and exchange rates is not sur-
prising, as it is very hard to separate out the effect of exchange rates on
trade (Engel 2009). The estimation of exports–exchange rate equations
is plagued by concerns about reverse causality and omitted variables,
which are hard to address with macroeconomic data. Moreover, several
papers show that the aggregation of heterogeneous firms or sectors can
result in aggregation bias in the estimation of the elasticity of exports
to exchange rate changes (see, e.g., Dekle, Jeong, and Ryoo 2009; Imbs
and Méjean 2009). We examine the implications of such heterogeneity
at the firm level in the next section.
n In the leather, textiles, and chemical products subsectors, appreciation
of the rupee was associated with a fall in real exports, mostly in recent
years. In all other cases, the elasticity both before and after the struc-
tural break was either very insignificant or (counterintuitively) positive.
n The volume of stalled projects and volatility in the exchange rate do not
seem to affect Indian exports adversely (tables 10.1 and 10.2). There is
no evidence for nonlinearity in the estimates of elasticity for partner
GDP and exchange rates (see appendix table 10A.1 at the end of the
chapter). The partner income elasticity of Indian exports declined. It
remained substantial for manufactured items, however, particularly
leather, textiles, and engineering goods. In contrast, exchange rate
movements did not seem to have a significant effect on exports.

Firm-Level Analysis: How Do Exporting Firms React to

Exchange Rate Changes?
How can one explain the low aggregate elasticity of exports to exchange
rate movements in India? Is there heterogeneity across exporting firms
in the responses to exchange rate movements? If so, could it explain low
aggregate elasticities?

1. Break dates vary widely. They are 2005 for engineering and pharmaceuticals, 2008–09 for
electronics, and 2013–14 for leather, textiles, and chemicals.

202 SUSTAINING ECONOMIC GROWTH IN ASIA

2
Table 10.2 Determinants of India’s merchandise and services exports, by sector
(1) (2) (2.1) (2.2) (2.3) (2.4) (2.5) (3)
Total
merchandise Manufacturing Leather and Chemical Pharmaceutical
Variable exports exports textiles Engineering products products Electronics Services
2.8*** 2.3*** 2.0*** 5.5*** 2.3*** 2.2*** 2.8*** 1.6***
Log of partner GDP
(0.2) (0.1) (0.7) (0.3) (0.7) (0.2) (0.6) (0.5)
Log of partner GDP * Structural break –2.5*** –1.4*** –1.1 –4.2*** –1.2* –1.0*** –4.0*** –0.8
dummy (0.2) (0.2) (1.0) (0.7) (0.7) (0.0) (1.1) (0.7)

GRAPHICS - CHAPTER
–1.3*** –0.8 0.6 –0.7 0.6 5.2*** 2.2** 2.5***

10
Log of partner exchange rate
(0.2) (0.9) (0.6) (1.8) (0.8) (0.7) (1.1) (0.5)

DO INDIA’S
Log of partner exchange Rate * 1.7*** 1.9 –3.3*** 2.3 –2.7*** –6.7*** –0.1 –2.7***
Structural break dummy (0.3) (2.3) (0.6) (2.2) (0.8) (0.9) (2.1) (0.6)
0.1** 0.1 –0.2 0.0 –0.1 0.0 0.2***
Log of number of stalled projects
(0.0) (0.1) (0.1) (0.1) (0.2) (0.0) (0.0)
79.1* 72.0 –18.1 93.5 47.6 3.5 120.1* 71.3
Volatility of dollar/rupee exchange rate
(41.4) (70.0) (67.4) (118.5) (33.6) (25.0) (64.8) (59.7)
3.0** –2.3 20.7*** 7.9 18.0*** 34.9*** 19.2 –6.2
Structural break dummy
(1.3) (11.1) (4.5) (10.7) (4.6) (4.4) (11.9) (4.8)
Model properties
Sample period 2002Q4 to 2016Q3
Number of observations 56 56 56 56 56 56 56 56
Structural break date 2005Q1 2005Q1 2013Q4 2005Q1 2014Q1 2005Q1 2008Q2 2005Q4
Max Wald-F: Quandt-Andrews test for
59.8*** 48.4*** 176.0*** 45.6*** 79.4*** 91.6*** 73.7*** 77.0***
structural break

EXPORTS REFLECT THE NEW NORMAL?

Dependent variable: Log of exports (2010 prices) in US dollars.
Note: *, **, and *** indicate statistical significance at 10, 5, and 1 percent, respectively. Numbers in parentheses represent standard errors.
Source: Authors’ calculations.

203
���
��� ���
�����ln�𝑌𝑌 �����This
ln�𝑌𝑌����� �𝐷𝐷 section
ln�𝑌𝑌����
�𝐷𝐷�� ∗∗ln�𝑌𝑌 ����draws on
����ln�𝐸𝐸𝐸𝐸
ln�𝐸𝐸𝐸𝐸 the
���� �
literature
������𝐷𝐷 ln�𝐸𝐸𝐸𝐸�(see,
�𝐷𝐷�� ∗∗ln�𝐸𝐸𝐸𝐸 ���
��� e.g.,
���
�� �Berman,
�� � ��
���� � �𝐷𝐷�Martin,
��𝐷𝐷 ������� and
� ���
Mayer 2012) and granular firm-level data for India in order to answer these
questions.
We estimate the following specification:
ln�𝐹𝐹������ ���
ln�𝐹𝐹 ��� ��� ln�𝐸𝐸𝐸𝐸�����
���ln�𝐸𝐸𝐸𝐸 ��𝑋𝑋
�𝑋𝑋�� �
����� �
���
�� �
���
������ (10.2)
(10.2)
(10.2)
where firms are indexed by i; time is indexed by t; Fit is a firm-level outcome
that includes the value of exports, imported raw materials, total sales
(domestic and foreign), total costs, and profits; ERt is the average annual
��
exchange rate; and Xt are additional��� time-varying controls, which include
���

world GDP growth (to measure external demand conditions) and the
volume of stalled projects (as a proxy for domestic supply constraints). All
regressions include firm fixed effects (si) and therefore use variation within
firms over time to identify the responses to exchange rate movements. The
regressions also control for a time trend (t).2 Standard errors are clustered
at the firm level.
The main findings from the empirical analysis include the following:
n There is little evidence that appreciation of the exchange rate reduced
the value of exports. In fact, the elasticity of exports with respect to the
exchange rate was positive after 2009 and in the smaller sample of firms;
an appreciation of the exchange rate was associated with a higher value
of average exports at the firm level. The long-run effect was also positive
and stronger than the short-run effect (table 10.3).3
n The value of imported intermediates rises sharply in response to an ap-
preciation of the rupee (table 10.4). The finding is robust across several
specifications. A 1 percent increase in the real effective exchange rate
(REER) was associated with a 0.4 percent increase in the value of im-
ported intermediates. The magnitude of the effect was much larger after
2009, for the balanced panel of firms, and in the long run.
ln�𝑋𝑋� � ���
n ���
The �� � �
� ln�𝑌𝑌
effect � �𝐷𝐷
of � ∗ ln�𝑌𝑌� �� � ��of
appreciation the� �rupee
ln�𝐸𝐸𝐸𝐸 � �� �𝐷𝐷� ∗ ln�𝐸𝐸𝐸𝐸
varied � �� � ��
across � � ��� There
sectors. � �𝐷𝐷� ���
is ��
some evidence that exports of firms in sectors with high domestic value
added (top 25th percentile) declined following appreciation of the

ln�𝐹𝐹�� � ��� ��� ln�𝐸𝐸𝐸𝐸� � � �𝑋𝑋� � �� � �� � ���� (10.2)

2. Time dummies cannot be included, because the main variable of interest (the exchange
rate) varies only over time.
3. The long-run effect is estimated by introducing a lagged dependent variable in the
�
model (columns 10–15 in table 10.3). It is calculated as ��� , where b is the coefficient on
the exchange rate and q the coefficient on lagged exports. The dynamic effects shown in
table 10.1 are robust to using a generalized method of moments (GMM) estimator (which
addresses concerns about bias arising from introducing a lagged dependent variable in a
panel data model with fixed effects). The smaller sample of firms is a “balanced” panel with
the same set of firms over the entire period.

204 SUSTAINING ECONOMIC GROWTH IN ASIA

 rupee. The value of imported intermediate inputs rose more significant-
ly in sectors with low domestic value added (or high foreign value-add-
ed) content. Exports of firms in these sectors did not decline following
an appreciation of the rupee (table 10.5). This result is not surpris-
ing, as greater reliance on foreign inputs is analytically analogous to a
lower pass-through, which would dampen the competitiveness effect of
changes in the exchange rate (see, e.g., Ahmed, Appendino, and Ruta
2015).
n Consistent with the firm-level evidence, the market share of exports by
several sectors with high domestic value-added content (e.g., readymade
garments, other textiles, minerals, copper and copper products, trading)
declined with the strengthening of the rupee (figure 10.7). Sectors with
high foreign value-added content (e.g., drugs and pharmaceuticals,
other chemical products, and gems and jewelry) increased their market
shares with the strengthening of the rupee (figure 10.8).
n Total sales (domestic and foreign) increased following an appreciation
of the rupee. Total costs also increased, perhaps because of increased
spending on foreign inputs (tables 10.6 and 10.7).
n Profits showed no significant relationship with exchange rate move-
ments. There is no evidence that profitability declined at the firm level
following an appreciation of the rupee (table 10.8).
n Expenditure on wages and salaries is positively associated with a stren­
gthening of the rupee (table 10.9).
n Firms did not seem to adjust their investments in response to exchange
rate fluctuations (table 10.10).

These findings suggest that following a strengthening of the rupee,

firms import more intermediate inputs, export more, increase overall sales,
and pay higher wages and salaries. They do not adjust their investments,
and their profits do not change significantly. The adjustments on all these
margins are significantly higher for firms in which the imported interme-
diate content of exports is higher.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 205

206
Table 10.3 Firm outcomes and exchange rates: Value of exports (millions of rupees)
[1] [2] [3] [4] [5] [6]
Variable REER NEER US$/Rs REER NEER US$/Rs
Full sample Post-2009 sample
–0.349 –0.14 –0.005 1.364^ 1.131^ 0.499^
Exchange rate (one-year lag, in logs)
[0.260] [0.197] [0.13] [0.85] [0.70] [0.31]
0.020*** 0.016*** 0.014*** –0.02 –0.007 0.004
World growth (one-year lag)
[0.006] [0.004] [0.004] [0.020] [0.013] [0.007]
Volume of stalled projects (one-year –0.225*** –0.260*** –0.266*** –0.205 –0.329^ –0.306^
lag, in logs) [0.046] [0.040] [0.042] [0.171] [0.201] [0.193]
Value of exports (one-year lag, in

SUSTAINING ECONOMIC GROWTH IN ASIA

logs)
0.186*** 0.184*** 0.192*** 0.169*** 0.278*** 0.226***
Time trend
[0.01] [0.017] [0.015] [0.045] [0.088] [0.062]
Long-run effect
Number of observations 27,899 27,899 27,899 15,379 15,379 15,379
Number of firms 6,561 6,561 6,561 5,247 5,247 5,247
(continues on next page)

GRAPHICS - CHAPTER 10
3
4
Table 10.3 Firm outcomes and exchange rates: Value of exports (millions of rupees) (continued)
[7] [8] [9] [10] [11] [12] [13] [14] [15]
Variable REER NEER US$/Rs REER NEER US$/Rs REER NEER US$/Rs
Balanced panel of firms Dynamic effects Dynamic effects: Post-2009

0.863*** 0.890*** 0.674*** 0.334 2.197*** 0.869*** 1.415^ 1.173^ 0.518^

Exchange rate (one-year lag, in logs)
[0.282] [0.207] [0.131] [0.72] [0.589] [0.228] [0.91] [0.760] [0.335]

GRAPHICS - CHAPTER 10
–0.002 0.003 0.006 0.016 –0.012 0.008^ –0.012 0.002 0.013^
World growth (one-year lag)
[0.007] [0.005] [0.004] [0.015] [0.010] [0.005] [0.022] [0.014] [0.008]

Volume of stalled projects (one-year –0.253*** –0.207*** –0.163*** –0.313** –0.537*** –0.337*** –0.21 –0.339* –0.315^
lag, in logs) [0.049] [0.041] [0.041] [0.136] [0.085] [0.047] [0.169] [0.205] [0.196]

Value of exports (one-year lag, in 0.325*** 0.325*** 0.325*** 0.227*** 0.227*** 0.227***
logs) [0.019] [0.019] [0.019] [0.027] [0.027] [0.027]

0.175*** 0.220*** 0.189*** 0.161*** 0.366*** 0.219*** 0.130*** 0.243** 0.188***

Time trend
[0.015] [0.018] [0.015] [0.037] [0.061] [0.026] [0.045] [0.094] [0.065]
Long-run effect 0.504 3.255 1.287 1.831 1.517 0.670
Number of observations 10,304 10,304 10,304 20,645 20,645 20,645 13,666 13,666 13,666
Number of firms 1,288 1,288 1,288 5,486 5,486 5,486 4,743 4,743 4,743
Dependent variable: Value of exports (in logs).
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. Standard errors are clustered at the firm level. An increase in all measures of the exchange
rate denotes an appreciation. ***, **, *, and ^ denote statistical significance at 1, 5, 10, and 15 percent levels, respectively.
Source: Authors’ calculations.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL?

207
208
Table 10.4 Firm outcomes and exchange rates: Value of imported raw materials
(millions of rupees)
[1] [2] [3] [4] [5] [6]
Variable REER NEER US$/Rs REER NEER US$/Rs
Full sample Post-2009 sample
0.452 0.611*** 0.524*** 2.084* 1.727* 0.763*
Exchange rate (one-year lag, in logs)
[0.316] [0.234] [0.154] [1.099] [0.911] [0.402]
0.017** 0.018*** 0.019*** –0.02 –0.001 0.016*
World growth (one-year lag)
[0.007] [0.005] [0.005] [0.025] [0.016] [0.009]
Volume of stalled projects (one-year –0.149** –0.122** –0.083* 0.134 –0.055 –0.019
lag, in logs) [0.058] [0.049] [0.050] [0.224] [0.271] [0.260]

SUSTAINING ECONOMIC GROWTH IN ASIA

0.109*** 0.137*** 0.115*** 0.015 0.181^ 0.101
Time trend
[0.018] [0.021] [0.018] [0.058] [0.119] [0.084]
Long-run effect
Number of observations 19,427 19,427 19,427 10,777 10,777 10,777
Number of firms 4,524 4,524 4,524 3,578 3,578 3,578
(continues on next page)

GRAPHICS - CHAPTER 10
5
 Table 10.4 Firm outcomes and exchange rates: Value of imported raw materials (millions of rupees)
(continued)
[7] [8] [9] [10] [11] [12] [13] [14] [15]
Variable REER NEER US$/Rs REER NEER US$/Rs REER NEER US$/Rs
Balanced panel of firms Dynamic effects Dynamic effects: Post-2009
1.166*** 1.063*** 0.766*** 0.405 3.345*** 1.326*** 2.859** 2.369** 1.046**

GRAPHICS - CHAPTER 10
Exchange rate (one-year lag, in logs)
[0.358] [0.264] [0.171] [0.983] [0.737] [0.282] [1.112] [0.921] [0.407]
–0.003 0.005 0.008^ 0.02 –0.024** 0.006 –0.040^ –0.014 0.009
World growth (one-year lag)
[0.008] [0.006] [0.005] [0.020] [0.012] [0.006] [0.026] [0.016] [0.009]
Volume of stalled projects (one-year –0.148** –0.074 –0.02 –0.138 –0.503*** –0.199*** 0.12 –0.139 –0.091
lag, in logs) [0.071] [0.061] [0.060] [0.182] [0.107] [0.060] [0.230] [0.280] [0.268]
0.095*** 0.145*** 0.107*** 0.081^ 0.401*** 0.177*** 0.013 0.240** 0.131^
Time trend
[0.022] [0.026] [0.022] [0.050] [0.076] [0.032] [0.060] [0.122] [0.087]
Long-run effect 0.497 4.099 1.611 3.009 2.494 1.101
Number of observations 8,643 8,643 8,643 14,308 14,308 14,308 9,530 9,530 9,530
Number of firms 1,173 1,173 1,173 3,723 3,723 3,723 3,192 3,192 3,192
Dependent variable: Value of imported raw materials (in logs).
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. Standard errors are clustered at the firm level. An increase in all measures of the ex-
change rate denotes an appreciation. ***, **, *, and ^ denote statistical significance at 1, 5, 10, and 15 percent levels, respectively.
Source: Authors’ calculations.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL?

209
Table 10.5 Firm outcomes and exchange rates: By domestic value
added (millions of rupees)
Dependent variable: Value of exports (in logs)
[1] [2] [3] [4] [5] [6]
Variable REER NEER US$/Rs REER NEER US$/Rs
High domestic value added Low domestic value added
Exchange rate –0.874** –0.497* –0.227 0.575 0.696 0.578^
(one-year lag, in logs) [0.398] [0.296] [0.194] [0.714] [0.541] [0.359]
World growth 0.026*** 0.017*** 0.014** –0.01 –0.008 –0.006
(one-year lag) [0.009] [0.006] [0.006] [0.015] [0.011] [0.010]
Volume of stalled
–0.189*** –0.272*** –0.298*** –0.515*** –0.477*** –0.434***
projects (one-year lag,
[0.070] [0.060] [0.063] [0.123] [0.106] [0.111]
in logs)
0.202*** 0.187*** 0.210*** 0.261*** 0.292*** 0.266***
Time trend
[0.023] [0.025] [0.023] [0.040] [0.046] [0.040]
Number of
12,204 12,204 12,204 3,638 3,638 3,638
observations
Number of firms 2,942 2,942 2,942 788 788 788
Dependent variable: Value of imported raw
materials (in logs)
[7] [8] [9] [10] [11] [12]
Variable REER NEER US$/Rs REER NEER US$/Rs
High domestic value added Low domestic value added
Exchange rate 0.423 0.574^ 0.489** 1.015 0.939* 0.698*
(one-year lag, in logs) [0.497] [0.367] [0.239] [0.725] [0.551] [0.367]
World growth 0.012 0.013* 0.014** 0.003 0.009 0.012
(one-year lag) [0.011] [0.008] [0.007] [0.016] [0.011] [0.010]
Volume of stalled
–0.097 –0.073 –0.036 –0.109 –0.03 0.026
projects (one-year lag,
[0.088] [0.073] [0.074] [0.129] [0.109] [0.112]
in logs)
0.120*** 0.146*** 0.125*** 0.116*** 0.154*** 0.117***
Time trend
[0.027] [0.032] [0.027] [0.042] [0.048] [0.041]
Number of
8,418 8,418 8,418 3,067 3,067 3,067
observations
Number of firms 2,010 2,010 2,010 677 677 677
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. Standard errors are clustered at the
firm level. An increase in all measures of the exchange rate denotes an appreciation.
Domestic value added (DVA) content is based on an ICRIER study submitted to
the Ministry of Finance. High DVA industries are defined as those with DVA content
greater than the 25th percentile. ***, **, *, and ^ denote statistical significance at 1, 5,
10, and 15 percent levels, respectively.
Source: Authors’ calculations.

210 SUSTAINING ECONOMIC GROWTH IN ASIA GRAPHICS - CHAPTER 10 7

 Figure 10.7 Market share of exports over time: Selected sectors with high domestic
value added, 2007–14

Ready-made garments Other textiles

index percent index percent
105 1.2 105 1.6
100 1.0 100 1.4
1.2
95 0.8 95 1.0
90 0.6 90 0.8
0.4 85 0.6
85
0.4
80 0.2 80 0.2
75 0 75 0
2007 2009 2011 2013 2007 2009 2011 2013

Minerals Copper and copper products

index percent index percent
105 3.0 105 4.0
100 2.5 100 3.5
3.0
95 2.0 95 2.5
90 1.5 90 2.0
85 1.5
85 1.0 1.0
80 0.5 80 0.5
75 0 75 0
2007 2009 2011 2013 2007 2009 2011 2013

Real effective exchange rate (left axis)

Market share (percent of total exports, right axis)

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL?

Sources: Ministry of Statistics and Program Implementation, Government of India; authors’ calculations.

211
 Figure 10.8 Market share of exports over time: Selected sectors with high foreign

212
value added

Drugs and pharmaceuticals Other chemicals

index percent index percent
105 6.0 105 1.2
100 5.0 100 1.0
95 4.0 95 0.8
90 3.0 90 0.6
85 2.0 85 0.4

80 1.0 80 0.2
75 0
75 0
2007 2009 2011 2013 2007 2009 2011 2013

Gems and jewelry

index percent
105 7.0

SUSTAINING ECONOMIC GROWTH IN ASIA

100 6.0
95 5.0
4.0
90
3.0
85 2.0
80 1.0
75 0
2007 2009 2011 2013

Real effective exchange rate (left axis)

Market share (percent of total exports, right axis)

Sources: Ministry of Statistics and Program Implementation, Government of India; authors’ calculations.
8
Table 10.6 Firm outcomes and exchange rates: Total sales (millions of rupees)
[1] [2] [3] [4] [5] [6]
Variable REER NEER US$/Rs REER NEER US$/Rs

GRAPHICS - CHAPTER 10
Full sample Post-2009 sample
0.675*** 0.586*** 0.402*** 0.748** 0.620** 0.274**
Exchange rate (one-year lag, in logs)

DO INDIA’S
[0.136] [0.104] [0.069] [0.372] [0.308] [0.136]
0.002 0.006*** 0.009*** –0.006 0.001 0.007**
World growth (one-year lag)
[0.003] [0.002] [0.002] [0.009] [0.005] [0.003]
Volume of stalled projects (one-year –0.077*** –0.021 0.014 0.242*** 0.174* 0.187**
lag, in logs) [0.023] [0.020] [0.021] [0.084] [0.097] [0.094]
Total sales (one-year lag, in logs)
0.119*** 0.142*** 0.118*** 0.029 0.089** 0.060**
Time trend
[0.008] [0.009] [0.008] [0.022] [0.041] [0.030]
Number of observations 27,542 27,542 27,542 15,186 15,186 15,186
Number of firms 6,488 6,488 6,488 5,186 5,186 5,186
(continues on next page)

213 EXPORTS REFLECT THE NEW NORMAL?

214
Table 10.6 Firm outcomes and exchange rates: Total sales (millions of rupees) (continued)
[7] [8] [9] [10] [11] [12] [13] [14] [15]
Variable REER NEER US$/Rs REER NEER US$/Rs REER NEER US$/Rs
Balanced panel of firms Dynamic effects Dynamic effects: Post-2009
1.262*** 1.025*** 0.679*** 0.045 0.244 0.099 0.395 0.328 0.145
Exchange rate (one-year lag, in logs)
[0.142] [0.107] [0.068] [0.309] [0.254] [0.100] [0.381] [0.315] [0.139]
–0.013*** –0.003 0.001 0.019*** 0.017*** 0.019*** 0.01 0.014** 0.017***
World growth (one-year lag)
[0.003] [0.002] [0.002] [0.006] [0.004] [0.002] [0.009] [0.006] [0.003]
Volume of stalled projects (one-year –0.106*** –0.017 0.038** 0.056 0.032 0.054** 0.06 0.025 0.031
lag, in logs) [0.024] [0.019] [0.018] [0.058] [0.038] [0.022] [0.079] [0.093] [0.090]
0.455*** 0.454*** 0.454*** 0.416*** 0.416*** 0.416***
Total sales (one-year lag, in logs)
[0.027] [0.027] [0.027] [0.037] [0.037] [0.037]
0.119*** 0.164*** 0.124*** 0.021 0.043^ 0.027** 0.019 0.05 0.035

SUSTAINING ECONOMIC GROWTH IN ASIA

Time trend
[0.007] [0.009] [0.007] [0.017] [0.027] [0.012] [0.020] [0.041] [0.029]
Number of observations 10,242 10,242 10,242 20,375 20,375 20,375 13,493 13,493 13,493
Number of firms 1,286 1,286 1,286 5,427 5,427 5,427 4,692 4,692 4,692
Dependent variable: Total sales (in logs).
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. An increase in all measures of the exchange rate denotes an appreciation. Standard
errors are clustered at the firm level. ***, **, *, and ^ denote statistical significance at 1, 5, 10, and 15 percent levels, respectively.
Source: Authors’ calculations.

GRAPHICS - CHAPTER 10
9
10
Table 10.7 Firm outcomes and exchange rates: Total expenses (millions of rupees)
[1] [2] [3] [4] [5] [6]
Variable REER NEER US$/Rs REER NEER US$/Rs

GRAPHICS - CHAPTER
Full sample Post-2009 sample

10
0.761*** 0.685*** 0.490*** 1.287*** 1.066*** 0.471***
Exchange rate (one-year lag, in logs)
[0.114] [0.086] [0.056] [0.314] [0.260] [0.115]

DO INDIA’S
–0.001 0.004** 0.006*** –0.017** –0.005 0.005**
World growth (one-year lag)
[0.003] [0.002] [0.002] [0.007] [0.004] [0.002]
Volume of stalled projects (one-year –0.125*** –0.063*** –0.021 0.192*** 0.076 0.097
lag, in logs) [0.020] [0.016] [0.017] [0.068] [0.080] [0.077]
Total expenses (one-year lag, in
logs)
0.141*** 0.167*** 0.140*** 0.051*** 0.154*** 0.104***
Time trend
[0.006] [0.007] [0.006] [0.018] [0.034] [0.025]
Number of observations 27,891 27,891 27,891 15,375 15,375 15,375
Number of firms 6,558 6,558 6,558 5,245 5,245 5,245
(continues on next page)

215 EXPORTS REFLECT THE NEW NORMAL?

216
Table 10.7 Firm outcomes and exchange rates: Total expenses (millions of rupees) (continued)
[7] [8] [9] [10] [11] [12] [13] [14] [15]
Variable REER NEER US$/Rs REER NEER US$/Rs REER NEER US$/Rs
Balanced panel of firms Dynamic effects Dynamic effects: Post-2009
1.180*** 0.978*** 0.660*** 0.09 0.756*** 0.300*** 0.670** 0.555** 0.245**
Exchange rate (one-year lag, in logs)
[0.113] [0.086] [0.055] [0.253] [0.214] [0.084] [0.332] [0.275] [0.121]
–0.009*** 0.00 0.003** 0.019*** 0.009** 0.016*** 0.005 0.011** 0.016***
World growth (one-year lag)
[0.003] [0.002] [0.002] [0.005] [0.004] [0.002] [0.008] [0.005] [0.003]
Volume of stalled projects (one-year –0.123*** –0.041*** 0.011 0.016 –0.066** 0.003 0.056 –0.004 0.007
lag, in logs) [0.019] [0.016] [0.016] [0.047] [0.032] [0.018] [0.061] [0.073] [0.070]
Total expenses (one-year lag, in 0.471*** 0.469*** 0.469*** 0.413*** 0.413*** 0.413***
logs) [0.021] [0.021] [0.021] [0.028] [0.028] [0.028]
0.132*** 0.175*** 0.137*** 0.039*** 0.111*** 0.060*** 0.030* 0.083** 0.057**

SUSTAINING ECONOMIC GROWTH IN ASIA

Time trend
[0.006] [0.008] [0.006] [0.013] [0.023] [0.011] [0.016] [0.034] [0.023]
Number of observations 10,304 10,304 10,304 20,639 20,639 20,639 13,663 13,663 13,663
Number of firms 1,288 1,288 1,288 5,483 5,483 5,483 4,741 4,741 4,741
Dependent variable: Total expenses (in logs).
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. Standard errors are clustered at the firm level. An increase in all measures of the ex-
change rate denotes an appreciation. ***, **, *, and ^ denote statistical significance at 1, 5, 10, and 15 percent levels, respectively.
Source: Authors’ calculations.

GRAPHICS - CHAPTER 10
11
12
Table 10.8 Firm outcomes and exchange rates: Profits (millions of rupees)
REER NEER US$/Rs REER NEER US$/Rs REER NEER US$/Rs
Variable Full sample Post-2009 sample Balanced panel of firms
0.093 –0.074 –0.124 –0.703 –0.583 –0.257 0.824^ 0.427 0.187

GRAPHICS - CHAPTER
Exchange rate (one-year lag, in logs)
[0.431] [0.316] [0.209] [1.797] [1.489] [0.658] [0.531] [0.381] [0.249]

10
–0.003 –0.001 0.00 0.025 0.019 0.013 –0.021* –0.012^ –0.01
World growth (one-year lag)
[0.009] [0.006] [0.006] [0.042] [0.026] [0.014] [0.011] [0.008] [0.008]

DO INDIA’S
0.221*** 0.235*** 0.228*** –0.273 –0.209 –0.221 0.07 0.147* 0.173**
Volume of stalled projects (one-year lag, in logs)
[0.083] [0.069] [0.069] [0.347] [0.429] [0.410] [0.107] [0.088] [0.086]
0.02 0.015 0.016 0.137^ 0.081 0.108 0.051^ 0.063^ 0.042
Time trend
[0.026] [0.030] [0.026] [0.091] [0.194] [0.135] [0.033] [0.039] [0.033]
Number of observations 14,635 14,635 14,635 8,041 8,041 8,041 6,142 6,142 6,142
Number of firms 4,802 4,802 4,802 3,552 3,552 3,552 1,213 1,213 1,213
Dependent variable: Profits (in logs).
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. Standard errors are clustered at the firm level. An increase in all measures of the ex-
change rate denotes an appreciation. Firms with negative profits are dropped from the regressions. ***, **, *, and ^ denote statistical
significance at 1, 5, 10, and 15 percent levels, respectively.
Source: Authors’ calculations.

217 EXPORTS REFLECT THE NEW NORMAL?

218
Table 10.9 Firm outcomes and exchange rates: Salaries and wage expenses
(millions of rupees)
[1] [2] [3] [4] [5] [6]
Variable REER NEER US$/Rs REER NEER US$/Rs
Full sample Post-2009 sample
1.008*** 0.851*** 0.573*** 0.597** 0.495** 0.218**
Exchange rate (one-year lag, in logs)
[0.107] [0.081] [0.052] [0.288] [0.239] [0.105]
–0.014*** –0.007*** –0.004** –0.008 –0.003 0.002
World growth (one-year lag)
[0.002] [0.002] [0.002] [0.007] [0.004] [0.002]
Volume of stalled projects (one-year –0.158*** –0.074*** –0.022 0.199*** 0.145** 0.155**
lag, in logs) [0.018] [0.015] [0.016] [0.062] [0.074] [0.071]

SUSTAINING ECONOMIC GROWTH IN ASIA

Value of exports (one-year lag, in
logs)
0.170*** 0.202*** 0.167*** 0.070*** 0.117*** 0.094***
Time trend
[0.006] [0.007] [0.006] [0.016] [0.031] [0.022]
Number of observations 27,708 27,708 27,708 15,277 15,277 15,277
Number of firms 6,502 6,502 6,502 5,208 5,208 5,208
(continues on next page)

GRAPHICS - CHAPTER 10
13
 14
Table 10.9 Firm outcomes and exchange rates: Salaries and wage expenses (millions of rupees) (continued)
[7] [8] [9] [10] [11] [12] [13] [14] [15]
Variable REER NEER US$/Rs REER NEER US$/Rs REER NEER US$/Rs
Balanced panel of firms Dynamic effects Dynamic effects: post-2009
1.287*** 1.034*** 0.673*** –0.304 0.358* 0.173** 0.254 0.211 0.093
Exchange rate (one-year lag, in logs)
[0.108] [0.082] [0.052] [0.245] [0.201] [0.077] [0.310] [0.257] [0.113]

GRAPHICS - CHAPTER 10
–0.021*** –0.011*** –0.007*** 0.020*** 0.008** 0.011*** 0.003 0.006 0.008***
World growth (one-year lag)
[0.003] [0.002] [0.002] [0.005] [0.003] [0.002] [0.007] [0.005] [0.003]
Volume of stalled projects (one-year –0.127*** –0.035** 0.021 0.059 –0.042 –0.012 0.132** 0.109^ 0.114*
lag, in logs) [0.019] [0.015] [0.015] [0.046] [0.029] [0.016] [0.056] [0.069] [0.066]
Value of exports (one-year lag, in 0.553*** 0.553*** 0.553*** 0.458*** 0.458*** 0.458***
logs) [0.023] [0.023] [0.023] [0.034] [0.034] [0.034]
0.152*** 0.197*** 0.156*** 0.027** 0.078*** 0.056*** 0.018 0.038 0.028
Time trend
[0.005] [0.007] [0.005] [0.012] [0.022] [0.010] [0.014] [0.032] [0.022]
Number of observations 10,295 10,295 10,295 20,521 20,521 20,521 13,588 13,588 13,588
Number of firms 1,288 1,288 1,288 5,444 5,444 5,444 4,713 4,713 4,713
Dependent variable: Value of salaries and wage expenses (in logs).
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. Standard errors are clustered at the firm level. An increase in all measures of the ex-
change rate denotes an appreciation. ***, **, * and ^ denote statistical significance at 1, 5, 10, and 15 percent levels, respectively.
Source: Authors’ calculations.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL?

219
220
Table 10.10 Firm outcomes and exchange rates: Investment (millions of rupees)
[1] [2] [3] [4] [5] [6]
Variable REER NEER US$/Rs REER NEER US$/Rs
Full sample Post-2009 sample
–2.294 2.262 1.136 1.957 1.622 0.716
Exchange rate (one-year lag, in logs)
[2.729] [2.216] [0.847] [3.709] [3.073] [1.357]
0.080^ –0.001 0.015 –0.029 –0.011 0.005
World growth (one-year lag)
[0.056] [0.037] [0.019] [0.088] [0.055] [0.031]
Volume of stalled projects (one-year 0.59 –0.117 0.065 1.080* 0.903 0.936
lag, in logs) [0.525] [0.357] [0.195] [0.590] [0.772] [0.730]

SUSTAINING ECONOMIC GROWTH IN ASIA

Value of exports (one-year lag, in
logs)
–0.134 0.201 0.071 –0.271* –0.115 –0.19
Time trend
[0.145] [0.243] [0.104] [0.159] [0.384] [0.259]
Number of observations 17,866 17,866 17,866 12,099 12,099 12,099
Number of firms 5,182 5,182 5,182 4,420 4,420 4,420
(continues on next page)

GRAPHICS - CHAPTER 10
15
16
Table 10.10 Firm outcomes and exchange rates: Investment (millions of rupees) (continued)
[7] [8] [9] [10] [11] [12] [13] [14] [15]
Variable REER NEER US$/Rs REER NEER US$/Rs REER NEER US$/Rs
Balanced panel of firms Dynamic effects Dynamic effects: Post-2009
–3.774 1.645 1.003 –2.197 3.281 1.791* 2.489 2.063 0.911
Exchange rate (one-year lag, in logs)
[2.999] [2.630] [1.004] [2.361] [2.600] [1.020] [3.264] [2.705] [1.195]

GRAPHICS - CHAPTER 10
0.094^ –0.008 0.001 0.069^ –0.033 –0.015 –0.066 –0.043 –0.023
World growth (one-year lag)
[0.061] [0.042] [0.020] [0.048] [0.048] [0.027] [0.078] [0.049] [0.028]
Volume of stalled projects (one-year 1.118* 0.227 0.33 0.356 –0.434 –0.175 1.148** 0.923 0.965^
lag, in logs) [0.598] [0.440] [0.233] [0.442] [0.346] [0.165] [0.524] [0.657] [0.623]
Value of exports (one-year lag, in –0.225*** –0.225*** –0.225*** –0.234*** –0.234*** –0.234***
logs) [0.052] [0.052] [0.052] [0.065] [0.065] [0.065]
–0.292* 0.05 –0.025 –0.044 0.387^ 0.211** –0.237* –0.038 –0.134
Time trend
[0.166] [0.298] [0.127] [0.121] [0.268] [0.105] [0.137] [0.323] [0.216]
Number of observations 8,017 8,017 8,017 11,778 11,778 11,778 9,673 9,673 9,673
Number of firms 1,286 1,286 1,286 3,750 3,750 3,750 3,443 3,443 3,443
Dependent variable: Value of investment (in logs).
REER = real effective exchange rate; NEER = nominal effective exchange rate
Note: All regressions include firm fixed effects. Standard errors are clustered at the firm level. An increase in all measures of the
exchange rate denotes an appreciation. ***, **, * and ^ denote statistical significance at 1, 5, 10, and 15 percent levels, respectively.
Source: Authors’ calculations.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL?

221
Conclusions and Policy Implications
India’s merchandise exports underwent a quiet revolution over the last two
decades, with engineering, electronic, and pharmaceutical exports gradu-
ally replacing India’s traditional exports of leather, textiles, and gems and
jewelry. Sectoral data reveal that changes in partner country growth drive
export volumes, with new exports exhibiting the highest elasticities.
A structural break occurred around 2005—well before the global finan-
cial crisis and the subsequent deglobalization. Partner country growth elas-
ticities fell sharply after 2005, but they remained significant, especially for
engineering, textiles, and leather. In contrast, changes in the exchange rate
and supply-side constraints did not affect India’s export volumes signifi-
cantly.
Analysis based on firm-level data produces broadly similar results. Sec‑
tors with higher imported intermediate input content (such as pharmaceuti-
cals) exhibited a sharper increase in the value of imported intermediates and
a weaker response of exports to appreciation of the exchange rate. In sectors
with low foreign value added (such as textiles), the response of exports to
exchange rate movements was larger.
How can India increase its exports in the new global normal, especially
if the exchange rate is unlikely to be a helpful tool? If exports are less likely
to power growth and investment, where will growth come from? In this
new global environment—flush with manufacturing capacity—it may take
a brave entrepreneur to invest in a new manufacturing facility in India.
The answer to the first question may rest on four pillars:
n building and improving infrastructure,
n raising human capital, with emphasis on vocational and on-the-job
training,
n simplifying business regulation and taxation, and
n improving access to finance.

A related question is whether India should encourage or focus on partic-

ular industries. Creating a conducive business environment, strengthening
the four pillars, and leaving it to Indian entrepreneurs to choose the “right”
industry seems to be a better solution (Ministry of Finance 2012, Rajan
2016).
As for the new source of growth, public and private investments in phys-
ical and human infrastructure, where India still stares at a large deficit, may
be needed. Where will the fiscal space needed to finance them come from?
Asset sales remain one sustainable option in the medium term. Private
investment in infrastructure will require getting rid of the debt overhang
on public sector banks and private sector balance sheets (Acharya, Mishra,

222 SUSTAINING ECONOMIC GROWTH IN ASIA

and Prabhala 2017). Doing so will require—in one form or other—creative
destruction of capital, the political economy of which is daunting.
Boosting supply in India is critical. As the French economist Jean-
Baptiste Say once said, supply creates its own demand. Doubling down on
investment in physical and human capital would generate demand while
addressing a key bottleneck in the economy and improving productivity.
The network of infrastructure connecting many of the major industrial, ag-
ricultural, and cultural centers of India (the golden quadrilateral) boosted
asset prices; rural and urban demand; and, most of all, firm productivity
and competitiveness. This kind of policy must be the response to deglobal-
ization.
India has benefited enormously from global integration. But integra-
tion has increased its exposure to global risks. The sooner policymakers
accept and prepare for this reality, the sooner they can lay the ground for a
new set of growth drivers in the current global environment.

References
Acharya, Viral, Prachi Mishra, and N. R. Prabhala. 2017. The Anatomy of India’s Credit Cycle.
Photocopy. Reserve Bank of India.
Ahmed, Swarnali, Maximiliano Appendino, and Michele Ruta. 2015. Global Value Chains and
the Exchange Rate Elasticity of Exports. IMF Working Paper 15/252. Washington: Interna-
tional Monetary Fund.
Aziz, Jahangir, and Sajjid Chinoy. 2012. India: More Open Than You Think. JP Morgan Note.
Berman, Nicolas, Philippe Martin, and Thierry Mayer. 2012. How Do Different Exporters
React to Exchange Rate Changes? Quarterly Journal of Economics 127, no. 1: 437–92.
Dekle, Robert, Hyeok Jeong, and Heajin Ryoo. 2009. A Re-Examination of the Exchange Rate Dis-
connect Puzzle: Evidence from Firm Level Data. Los Angeles: University of Southern Califor-
nia.
Engel, Charles. 2009. Exchange Rate Policies. Staff Paper. Dallas: Federal Reserve Bank of Dallas.
Imbs, Jean, and Isabelle Méjean. 2009. Elasticity Optimism. CEPR Discussion Paper 7177. Lon-
don: Centre for Economic Policy Research.
Ministry of Finance. 2012. Seizing the Demographic Dividend. In Economy Survey. New Delhi:
Government of India.
Rajan, Raghuram. 2016. India in the Global Economy. Ramnath Goenka Lecture, March 13.
Rangarajan, C., and Prachi Mishra. 2013. India’s External Sector: Do We Need to Worry? Eco-
nomic and Political Weekly 48, no. 7: 52–59.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 223

Appendix 10A Data
Macroeconomic Data
We use quarterly data from calendar year 2002Q4 to 2016Q3 in our em-
pirical models. The data on merchandise and services exports are from the
CEIC database. They are available monthly. We derive the quarterly exports
figures from the monthly information available for October 2002–Septem-
ber 2016. For merchandise exports, disaggregated data are for the “principal
commodity groups,” which cover more than 92 percent of all merchandise
exports. Based on these groups, we identify 11 broad sectors: agriculture,
ores and minerals, chemicals, pharmaceuticals, electronics, engineering
goods, leather and textiles, metal and metal products, miscellaneous manu-
factured items, gems and jewelry, and petroleum products. We classify the
following as manufactured items: chemicals, pharmaceuticals, electronics,
engineering goods, leather and textiles, metal and metal products, and mis-
cellaneous manufactured items (products such as paper and print-related
items, ceramics and glassware, wood and wood products, plastic and rubber
products, handicrafts, and other unclassified manufactured items).
The data on merchandise and services exports are available in millions
of dollars (the value of exports). The database does not contain information
on volumes. We therefore deflate nominal exports’ to obtain real exports (in
millions of dollars). We deflate exports of chemicals, pharmaceuticals, elec-
tronics, engineering goods, leather and textiles, and miscellaneous manu-
factured items by the US Producer Price Index. We deflate services exports
by the US Consumer Price Index (CPI). We deflate the remaining sectors
(petroleum products, gems and jewelry, agricultural goods, and metal and
metal products) by the UK Brent (international crude price), international
gold prices, the Food and Agriculture Organization’s Food Price Index, and
the Commodity Research Bureau (CRB) Metal Price Index, respectively. For
ease of comparison, we normalize all price indexes to the year 2010. After
deflating sectoral exports, we sum them to derive the series on manufac-
tured items. Total merchandise exports include manufactured items as well
as exports of sectors that are not classified as manufacturing sectors.
The quarterly dataset also includes information on the following:
n GDP of India’s export partners
n the exchange rate of the rupee vis-à-vis trading partners’ currencies
n domestic supply constraints (proxied by the number of stalled projects
in a sector)
n the volatility of the dollar/rupee exchange rate.

Quarterly GDP data in current prices come from the CEIC. They cover
about 105 of India’s export partners, representing more than 92 percent of

224 SUSTAINING ECONOMIC GROWTH IN ASIA

India’s exports. We use the weighted-average index of real GDP of India’s
export partners. We first deflate the GDP series for each country by the US
CPI. We then normalize the real GDP of each country to the year 2010 and
combine the country-specific real GDP indexes (2010 = 100) to a composite
weighted-average index called Partner Country GDP Index (2010 = 100), in
which we use each country’s share in the value of India’s exports as weights.
As the export share of partner countries varies significantly across sectors,
we construct this index separately for each sector and also for total mer-
chandise and manufactured items. We create a separate export-weighted
partner country GDP series for each sector, using sector-specific export
shares. Because bilateral data on trade in services are not available, we use
the world real GDP (in millions of dollars) to estimate partner income elas-
ticity for services.
We also construct an index for the real exchange rate. Specifically, we
construct a partner country export-weighted real exchange rate for each
sector, using sector-specific export shares as weights. (By construction, a
higher value of the exchange rate indicates an appreciation of the rupee
against the partner country’s currency.) We first convert the rupee/partner
country nominal exchange rate to a real exchange rate by multiplying with
the ratio of India’s consumer price index (2010 = 100) to the partner coun-
try’s consumer price index (2010 = 100). We use annual country CPI data
from the World Bank’s World Development Indicators. We normalize the real
exchange rates of the rupee vis-à-vis each country’s domestic currency to the
year 2010. As the export shares of countries vary across sectors, we construct
this index separately for each sector and also for total manufactured items.
For total merchandise exports, we use the index of the exports’ weighted
real effective exchange rate for the rupee based on 36 partner countries,
published by the Reserve Bank of India. For the services sector, for which
export shares are not available, we use the nominal dollar/rupee exchange
rate, published by the Reserve Bank of India.
We compile quarterly series on the number of stalled projects based
on the Capex database, collected by the Centre for Monitoring the Indian
Economy (CMIE). We measure exchange rate volatility by the standard de-
viation of daily exchange rates each quarter. Daily exchange rates are from
the CEIC database.

Firm-Level Data
We compile an annual firm-level panel dataset that spans 2007–14, based
on the Prowess database, collected by the CMIE. It covers about 6,500
publicly listed exporting firms across 143 industries whose export earnings
represent roughly 40 percent of total exports (based on the 2013 balance

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 225

of payments). The dataset includes information on the value of exports of
goods and services, the value of imported raw materials, total costs, and
total sales (all in millions of rupees) at the firm level. It does not contain
information on quantities of exports or imports, unit values, or the desti-
nation of exports. The firm-level database is merged with data on average
annual exchange rates based on three measures (the real effective exchange
rate, the nominal effective exchange rate, and the dollar-to-rupee exchange
rate) and the domestic value-added content.

226 SUSTAINING ECONOMIC GROWTH IN ASIA

 Table 10A.1 Determinants of India’s manufacturing and services exports: Check for nonlinearity
(1) (2) (3) (4) (5) (6) (7) (8)
Manufacturing exports Services exports

Nonlinearity Nonlinearity Nonlinearity Nonlinearity

Linear in partner in exchange Linear in partner in exchange
Variable model GDP rate (2) + (3) model GDP rate (6) + (7)
Dependent variable: Log of exports (2010 prices) in US dollars
Log of partner GDP 2.3*** 2.3*** 2.3*** 2.3*** 1.6* 3.0*** 1.2 2.7***
Log of partner GDP * High partner growth 0.2 0.2 –0.4 –0.3
Log of partner exchange rate –0.8 –0.9 –0.8 –0.9 2.5* 3.6*** 2.3 3.4***
Log of partner exchange rate * High rupee
–0.7 –0.3 0.1 0.1
appreciation
Log of number of stalled projects 0.1 0.2 0.1 0.1
Volatility of dollar/rupee exchange rate 72.0 82.6 65.1 77.9 71.3 56.8 54.7 44.6

DO INDIA’S EXPORTS REFLECT

Structural break dummy * Log of partner
–1.4** –1.5* –1.4** –1.5** –0.8 –2.2 –0.6 –2.1*
GDP
Structural break dummy * Log of partner
0.0 0.0 0.1*** 0.1***
GDP * High partner growth

GRAPHICS
THE NEW
Structural break dummy * Log of partner
1.9 2.4 1.9 2.4 –2.7** –3.8*** –2.6 –3.7***
exchange rate
Structural break dummy * Log of partner
0.0 0.0 0.0 0.0
exchange rate * High rupee appreciation

- CHAPTER
(continues on next page)

17
NORMAL?10 227
18
Table 10A.1 Determinants of India’s manufacturing and services exports: Check for nonlinearity (continued)
(1) (2) (3) (4) (5) (6) (7) (8)
Manufacturing exports Services exports

228 GRAPHICS
Nonlinearity Nonlinearity Nonlinearity Nonlinearity
Linear in partner in exchange Linear in partner in exchange

SUSTAINING
Variable model GDP rate (2) + (3) model GDP rate (6) + (7)
Dependent variable: Log of exports (2010 prices) in US dollars
Dummy variables:

- CHAPTER
Structural break –2.3 –4.7 –2.4 –4.1 –6.2 –4.5 –7.0 –4.7

ECONOMIC
High partner growth –0.8 –0.9 1.7 1.1
High rupee appreciation 3.0 1.2 0.5 0.5
Sample period 2002Q4 to 2016Q3
Observations 56 56 56 56 56 56 56 56
Structural break date 2005Q1 2005Q4

10 GROWTH IN ASIA
Maximum Wald-F stat: Quandt-Andrews
48.4*** 97.0***
test
Durbin-Watson d-statistic 2.8 2.6 2.6 2.6 1.9 2.1 2.1 2.1
Ljung-Box Q-stat for first order
8.5*** 4.9** 5.4** 5.2** 0.0 0.2 0.2 0.1
autocorrelation
Notes: *, **, and *** indicate statistical significance at 10, 5, and 1 percent, respectively. High partner growth is a dummy variable that
assumes a value equal to 1 when the quarter-over-quarter change in partner GDP is more than 2 percent. High rupee appreciation is a
dummy variable that assumes a value equal to 1 when the quarter-over-quarter change in partner exchange rate is more than 2 percent.
Regression uses heteroskedasticity-consistent robust standard errors.
Source: Authors’ calculations.
Comment
Kenneth Kang

Prachi Mishra and Siddhartha Nath use macro-level evidence supported

by sectoral and firm-level analysis to examine the drivers of India’s exports.
They highlight how India’s exports have undergone a “quiet revolution”
over the past two decades as their composition has shifted from traditional
goods to new engineering, electronics, and pharmaceutical products.
Not surprising is the fact that external demand has been the main driver.
The negligible role of the exchange rate and relative prices is surprising. The
authors’ findings raise important implications for exports as a source of
growth should weak demand from the advanced economies persist. My
comments focus on how their findings relate with other research, including
at the International Monetary Fund (IMF), and what can be done to address
this “new mediocre” risk for India’s exports.
The IMF has analyzed the export performance of India and other emerg-
ing markets. Its results on the role of trading partner demand are similar
to those described in the chapter. It finds that weak investment accounted
for three-quarters of the global trade slowdown after the global financial
crisis (IMF 2016). IMF research using both macro and industry-level data
finds export elasticities similar to those described in their chapter, as well as
declines after 2005. However, the decline in export elasticities brings them
closer to their long-run average of 1, rather than the values of almost zero
found in the chapter. The decline in demand elasticity comes mainly from
the electronics sector, where elasticity is negative; for other sectors, they
remain significantly positive and closer to 1.
The result that exchange rate changes do not affect exports is somewhat
surprising. The chapter confirms the more standard result that appreciation
hurts exporters in higher domestic value sectors, suggesting that India’s
growing participation in global supply chains may explain the weaker link
between exports and the exchange rate. The weak link between the exchange
rate and exports suggests that relative price signals in India may be distorted
and be holding back the efficient reallocation of resources. If this is the case,
identifying the distortion and crafting policies for addressing them can
help India’s exports become more competitive and responsive to external
demand.

Kenneth Kang is the deputy director of the Asia and Pacific Department at the International Monetary Fund.
The views expressed are those of the author and do not necessarily represent the views of the IMF, its Executive
Board, or its management.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 229

 Figure 10C.1 Most-favored nation (MFN) applied tariff
rates

percent
40
35
30
25
20
15
10
5
0
a il ey and ina xico rica esia ssia nes ysia eru ore
di az rk i
In Br Tu ai
l Ch Me A
f
on Ru lipp ala
P ap
Th th Ind i M ng
u P h Si
So
2015
2000

Note: For countries where 2015 data are not available, 2013 or 2014
data are used.
Source: World Bank, World Integrated Trade Solution database.

One obvious distortion is India’s relatively high tariff barriers (figure

10C.1). Emerging-market and developing economies have made significant
progress in reducing average tariff rates, cutting them by almost two-thirds
over the past 35 years. India also made significant progress, lowering its
applied tariff rates from about 37 percent in 2000 to less than 15 percent
in 2015. India’s tariffs are still higher than most emerging-market econ-
omies and other countries in Asia, especially China, with which India is
competing. There is room to reduce tariff and nontariff barriers, especially
in the food, agriculture, and manufacturing sectors.
Another barrier is India’s limited role in global value chains. The value
of India’s intermediate goods exports increased steadily through 2008;
since then it has plateaued (figure 10C.2). India’s share of foreign value
added in total merchandising exports has stagnated at less than 25 percent,
similar to the trend in other countries but still below China and Indonesia.
Further integration into global value chains would both enhance the
responsiveness to external demand and expand access to technology and
foreign direct investment (FDI).
At the same time, given India’s large economy and population, there
is significant scope to strengthen domestic supply linkages. The July 2017
rollout of the nationwide goods and services tax (which applies a single,
uniform indirect tax on the supply of goods and services across India’s 29
states and 7 union territories) is an important step in promoting national
integration and could have important trade effects given the size of India’s

230 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 10C.2 Merchandise exports of India, 2000–14

billions of US dollars
300

250

200

150

100

50

0
2000 2002 2004 2006 2008 2010 2012 2014

Primary goods
Intermediate goods
Final goods

Note: Export values based on Broad Economic Categories classification

system in order to classify all of the trade goods by production stage.
“Primary goods” are materials to be used for food and beverages and in
industrial supplies. “Intermediate goods” are trade goods that represent the
intermediate input along the path toward becoming the final product. These
goods are manufactured goods (processed or assembled) that are produced
from primary goods but are not yet final products. “Final goods” are defined
here as goods used by the producer (as the intermediate input) and goods
consumed by households and the government.
Sources: Research Institute of Economy, Trade and Industry (REITI) database;
Yes Bank.

states (if India’s most populous state, Uttar Pradesh, were a separate country,
it would be the fifth largest country in the world by population, just after
Brazil). Leemput and Wiencek (2017) find that internal trade barriers in
India account for up to 40 percent of total trade barriers on average but vary
substantially by state. According to them, reducing trade barriers across
Indian states could lead to real GDP gains of more than 4 percent.
India’s competitiveness may also be a factor behind weak export response.
India’s nominal effective exchange rate has steadily declined since the global
financial crisis, especially against Indonesia and China. In real effective
terms, however, India’s exchange rate appreciated by more than 20 percent
since 2008, more so than that of Brazil or Russia. This appreciation largely
reflects India’s higher average rate of inflation, highlighting the need to keep
inflation under control and pursue reforms to enhance competitiveness.
The legacy of the credit bubble of the mid-2000s in the corporate and
banking sectors may be holding back credit allocation. Credit growth and
real investment are closely correlated. It is perhaps not coincidental that

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 231

Figure 10C.3 Banks’ nonperforming and restructured assets, 2012–16

percent of outstanding advances

16

14

12

10

0
2012–13

2013–14

2014–15

2015–16

2012–13

2013–14

2014–15

2015–16

2012–13

2013–14

2014–15

2015–16

2012–13

2013–14

2014–15

2015–16
Public sector Private sector Foreign banks All banks
banks banks

Gross nonperforming asset ratio

Restructured loan ratio

Sources: Reserve Bank of India; IMF Staff estimates.

the post-2005 decline in exports elasticity coincides with the peak of the
credit and investment bubble. Nonfinancial corporate debt in India is low
(about 50 percent of GDP) compared with other emerging-market countries
(among which it averages 90 percent). But the level of nonperforming loans
is high, at about 12 percent for all banks—significantly higher for public
banks—and rising (figure 10C.3). Nonperforming loans are concentrated in
metals and mining, construction and engineering, and transportation and
infrastructure—sectors closely tied with exports. High corporate leverage
and nonperforming loans may be hindering efficient resource allocation,
preventing the exit of nonviable zombie firms, the entry of new firms, and
the growth of productive ones. Forthcoming research by the IMF finds that
the misallocation of resources in India is much greater than in the United
States and that it varies widely across Indian states. Credit availability is one
of the key drivers of resource misallocation in India, suggesting that remov-
ing structural rigidities in key input markets and improving credit alloca-
tion across firms would help reduce distortions and boost competitiveness.
India is an open economy with a young and growing population
subject to spillovers from stagnation elsewhere, particularly through trade

232 SUSTAINING ECONOMIC GROWTH IN ASIA

and investment. The chapter highlights the need for India and other coun-
tries to address structural rigidities holding back export growth and diver-
sification. The priorities should be on getting prices right in order to real-
locate resources more efficiently and enhance export competitiveness.

References
IMF (International Monetary Fund). 2016. World Economic Outlook (Fall). Washington.
Van Leemput, Eva, and Ellen A. Wiencek. 2017. The Effect of the GST on Indian Growth. IFDP
Notes (March 24). Washington: Board of Governors of the Federal Reserve System.

DO INDIA’S EXPORTS REFLECT THE NEW NORMAL? 233

 11
How Has Indonesia Fared
in the Age of Secular Stagnation?
MITALI DAS

This chapter analyzes the implications of secular stagnation in advanced

economies for macroeconomic dynamics and policy tradeoffs in Indonesia,
one of the world’s largest emerging-market economies. The first section de-
scribes secular stagnation. The second section presents a retrospective anal-
ysis of the impact of secular stagnation on Indonesia. It reviews the channels
through which secular stagnation can spill over to growth, describes the evo-
lution of trade and financial exposure in Indonesia to advanced economies,
and decomposes the contributions of domestic and external factors to In-
donesia’s output dynamics. The last section presents a prospective analysis,
discussing the policies that can limit the impact of secular stagnation on po-
tential growth. It begins with a decomposition of the contributors to the po-
tential growth rate, then considers an illustrative scenario analysis by which
potential growth may evolve under policies that raise total factor productiv-
ity (TFP) growth to its precrisis trend. It concludes that, although greater
insularity has allowed Indonesia to enjoy more stable growth, to achieve its

Mitali Das is deputy division chief of the Strategy, Policy and Review Department of the International Mon-
etary Fund. She thanks Luis Breuer, Jérémie Cohen-Setton, Thomas Helbling, Ken Kang, and Adam Posen
for their overall guidance on this chapter and Agnes Isnawangsih, Yinqiu Liu, Toh Seng Guan, Medha Nair,
Ranil Salgado, Jongsoon Shin, and participants at the Peterson Institute for International Economics confer-
ence on “The New Mediocre and Asia” held in May 2017 and the conference on “Prospects and Challenges for
Sustained Growth in Asia” organized by the Bank of Korea, the Korean Ministry of Strategy and Finance, the
International Monetary Fund, and the Peterson Institute for International Economics in Seoul on September
7–8, 2017, for their helpful comments. Special thanks to M. Chatib Basri, whose insights and discussion of this
chapter improved both its contents and presentation, and to Egor Gornostay, for his outstanding assistance with
the empirical section. The views expressed in this chapter are those of the author and do not necessarily represent
the views of the IMF, its Executive Board, or its management. Any remaining errors are the author’s.

235
ambitious growth objectives and generate good-quality jobs for its expand-
ing labor force, Indonesia will require higher investment and technological
innovations that may be best facilitated by greater global integration.

Why Does Stagnation Matter?

In advanced economies, the long-lasting slowdown in output growth, along
with underemployment, disinflation dynamics, and negative market-clear-
ing real interest rates, have raised questions about whether these economies
are in secular stagnation. At the heart of this debate are questions about
whether weak growth reflects transitory factors associated with the financial
crisis, deeper structural factors, or transitory factors that have become per-
manent from the scarring effects of hysteresis. Competing explanations have
emerged, ascribing secular stagnation to deficiencies in aggregate demand;1
the debt supercycle (Rogoff 2016); a global savings glut;2 and structural sup-
ply-side weaknesses, as put forward most prominently by Gordon (2016).
There is little consensus about whether advanced economies are in
secular stagnation or what its proximate causes are. Still less is known
about secular stagnation in emerging markets. Increasingly, however,
following successive markdowns of actual and potential growth rates in
emerging markets (IMF 2015b), commentators have noted that macroeco-
nomic dynamics in these economies are unlikely to have decoupled from
those of advanced economies amid the deep trade and financial linkages
between them. This decoupling has raised questions about how strongly
secular stagnation has transmitted from advanced to emerging economies,
what the likely transmission channels are, and how quantitatively signifi-
cant the impact on output dynamics has been.
These questions are important for at least two reasons. First, emerging
markets represent a significant and growing fraction of the global economy.
In 1980 they accounted for about 36 percent of global GDP and 43 percent
of global GDP growth; by 2010–15 they accounted for 56 percent of global
GDP and 79 percent of global growth.3 Consequently, stagnation—or even
a temporary slowdown—in emerging markets presents a serious risk that

1. Lawrence Summers, “Why Stagnation Might Prove to Be the New Normal,” Financial Times,
December 15, 2013, http://larrysummers.com/commentary/financial-times-columns/why-
stagnation-might-prove-to-be-the-new-normal; Lawrence Summers, “Reflections on the
New ‘Secular Stagnation Hypothesis,’” VoxEU, October 30, 2014, https://voxeu.org/article/
larry-summers-secular-stagnation.
2. Ben Bernanke, Why Are Interest Rates So Low? Brookings Institution Blog, March 30,
2015, www.brookings.edu/blog/ben-bernanke/2015/03/30/why-are-interest-rates-so-low/.
3. IMF, World Economic Outlook June 2017 database, measured in purchasing power parity
(PPP) terms and PPP weights, respectively.

236 SUSTAINING ECONOMIC GROWTH IN ASIA

could reinforce the weak growth dynamics in advanced economies and
raise the likelihood of a global deflationary spiral.
Second, these questions are important for emerging markets them-
selves, many of which have seen a sharp deceleration in trend growth since
the global financial crisis, despite strong countercyclical stimulus and, in
some cases, favorable demographics (IMF 2015b). This deceleration has oc-
curred at much lower levels of income per capita than in richer economies
(IMF 2017). If the slowdown in advanced economies proves permanent,
emerging economies reliant on export-led growth may need significant
structural adjustments to offset the weaker growth from a slowdown in ex-
ternal demand. For economies that face significant domestic headwinds, a
durable external slowdown could be the catalyst that precipitates a crisis.
The implications of these global changes could be significant for policy
frameworks in emerging markets. Following the financial crisis, advanced
economies deployed a range of conventional and unconventional policies
to counter the protracted demand slump, with mixed results domestically
and large and often disruptive implications globally (IMF 2015b, Fischer
2015). Emerging-market policy responses to these spillovers were by and
large countercyclical, including nominal exchange rate adjustment, some
use of foreign exchange reserves, and capital flow management measures
(Sahay et al. 2014). In part, these responses reflected a perception that the
volume and volatility of capital flows spurred by unconventional mone-
tary policy were transitory ahead of an imminent normalization, requiring
effective stabilization policies but not deeper structural responses.
To the extent that the slowdown in advanced economies is now perceived
as durable, emerging markets will need to adapt policy frameworks accord-
ingly. They will need to tailor their responses to country-specific consider-
ations, making changes to traditional policy frameworks (fiscal, monetary,
and prudential) and adopting deeper reforms to institutions. Some may
need to reorient the structure of their economies so that growth is more
reliably linked to domestic factors. To limit their exposure to lower volumes
and higher volatility of capital flows, they may need to create prudential
buffers that raise resilience to external shocks and reduce their dependence
on external financing. Boosting competitiveness, limiting dependence on
external demand, and raising potential growth may require changes in the
regulation of product and factor markets, educational reforms, improve-
ment in technological know-how, and other measures. Reorienting from
external to domestic demand will require institutional changes that improve
the business climate and speed the process of structural transformation.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 237

Retrospective Analysis: Spillovers of Secular Stagnation
to Output Dynamics in Indonesia
In the years since the Asian financial crisis, Indonesia retrenched from the
global economy in terms of both trade and financial integration. This re-
trenchment reflected a complex set of factors, including a large domestic
base and favorable demographics, which enabled strong growth without
high reliance on exports, as well as a strengthening of populist forces
that explicitly favor inward-looking policies and protectionism (Basri and
Patunru 2012).
A perhaps unanticipated consequence of its rising insularity has been
the remarkable stability of output growth. Indonesia was among the few
emerging markets that did not suffer from the recessionary impact of the
global financial crisis (Blanchard, Das, and Faruqee 2010). More recently,
Indonesia’s inward-looking stance may have limited the transmission of
secular stagnation from advanced economies.
The benefits of high levels of growth and low growth volatility from
rising insularity could be ephemeral, however. Since the global financial
crisis, the potential growth rate of output in Indonesia has been falling,
driven entirely by lower TFP growth. The causes of slowing TFP growth
can be wide-ranging, including both domestic and external factors. But it
is likely that they at least partly reflect the ongoing decline in Indonesia’s
trade and financial engagement with the global economy, including a rise
in protectionism, which may have limited more efficient resource alloca-
tion, the development of the domestic financial sector, and the transfer of
technology, technological know-how, and best practices—the well-known
productivity-raising benefits of trade and financial integration.

Output Dynamics in Indonesia: Stylized Facts

After a period of high growth in 2005–08—marked by a commodity boom,
cheap global credit, and strong performance in trading partners—growth
in Indonesia decelerated (figure 11.1). Because this slowdown occurred
contemporaneously with secular stagnation in advanced economies, it is
tempting to conclude that the two are linked.
The historical evidence, however, suggests that Indonesia’s output
dynamics evolved distinctly from global output dynamics. Between 2004
and 2009, the growth rate of real GDP rose steadily in Indonesia, averaging
5.7 percent. In contrast, the growth rate declined steadily in the United
States, where it averaged 1.4 percent (figure 11.2).4 The divergence of growth

4. The growth rate in the United States is taken as a proxy for the growth rate of the global
economy.

238 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 11.1 Real GDP growth rates, 1990–2016

percent
10
9
8
7
6
5
4
3
2 Real GDP growth 2005–08
1990–97 2009–12
1
2000–04 2013–16
0
90

94

8
92

96

10

12

14

16
0

0
0
0
0

20

20
20
20
19

20
19

19

20
19

20
20
20

Sources: IMF, World Economic Outlook database; author’s calculations.

paths is reflected in a correlation of –0.73. Although the growth rates in

Indonesia and the United States have moved together more closely since the
financial crisis, the change may reflect responses to common global shocks
rather than greater synchronicity of Indonesia and global output dynamics.5
One fact that is consistent with the insularity of Indonesia’s output dy-
namics from global output dynamics is the remarkably low volatility of its
real output growth. The stability of Indonesia’s output growth in 2000–16
stands out among both its regional peers and emerging markets more gen-
erally (figures 11.3a and 11.3b). It is especially notable given the profound
changes that took place in the global economy in this period, including the
steep rise and subsequent sharp decline in commodity prices, the severity of
recessions in many of Indonesia’s trading partners following the financial
crisis, and the large swings in capital flows, including the taper tantrum,
which was especially significant for Indonesia.
This stability of output in Indonesia is unsurprising once one takes
account of the structure and evolution of its economy. Domestic demand
is higher than in most of its peers, reducing the country’s vulnerability to
the vicissitudes of external demand. Fueled by a large and growing popula-
tion, domestic demand averaged 97 percent of GDP in 2000–16, with the

5. As predicted by theory (see Cesa-Bianchi, Imbs, and Saleheen 2016), in response to common
global shocks, business cycle synchronization increases when financial integration declines.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 239

 Figure 11.2 Real GDP growth rates in Indonesia and the United States, 2000Q1–2017Q1

240
percent
10

SUSTAINING ECONOMIC GROWTH IN ASIA

–2

–4 Indonesia
United States

–6
Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Sources: IMF, World Economic Outlook database; author’s calculations.

 Figure 11.3a Real GDP growth relative to ASEAN-4, 2000–16

percent
10
8
6
4
2
0
–2
2000 2002 2004 2006 2008 2010 2012 2014 2016

Indonesia
Thailand
Malaysia
Philippines

ASEAN = Association of Southeast Asian Nations

Sources: IMF, World Economic Outlook database; author’s calculations.

Figure 11.3b Real GDP volatility, 2000–16

percent
10
9
8
7
6
5
4
3
2
1
0
ela
a
sia

a
nd

ile

sia

il

ia

ey
ne

az

in
bi

ric

in

xic
an

ss
Ch

rk
a
ne

lay

zu
nt
Ch
m

Br
pi

Af

Ru
l

ail
Me
Po

Tu
lo

ge
do

ne
ilip

Ma

Th
h
Co

Ar

Ve
In

ut
Ph
So

Note: Volatility is defined as standard deviation.

Sources: IMF, World Economic Outlook database; author’s calculations.

large domestic base accounting for two-thirds of consumption.6 External

demand is a relatively small contributor to aggregate demand, and both the
export basket and export destinations are diversified (as discussed below).7

6. IMF, World Economic Outlook June 2017 database.

7. Basri and Rahardja (2010) cite the composition of Indonesia’s export basket (particularly
its concentration in primary commodity exports) as a key reason for the stability of export
demand during the global financial crisis.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 241

The declining importance of external demand may also reflect weaknesses
in the business environment, which has long dissuaded investment, partic-
ularly in the export sector (IMF 2010, 2015a).
Policies have also played an important role in the low volatility of
output. A strong fiscal framework, supported by caps on the fiscal deficit
and public debt, has given authorities the fiscal space to maintain demand
in downturns (OECD 2016). Greater exchange rate flexibility and prudent
use of foreign exchange reserves have helped absorb large external shocks
and smooth output dynamics (IMF 2014a). In recent years the authorities
have taken steps to lower distortionary fuel subsidies, which has improved
the fiscal position. In the years since the Asian financial crisis, stronger su-
pervisory oversight in the financial sector has improved governance and
curtailed balance sheet exposures to foreign currency borrowing, limiting
the growth impact of episodic slowdowns in capital flows and currency de-
preciation (IMF 2007, 2009, 2013).
Will the stability of output dynamics continue in the new economic
environment? How will secular stagnation in advanced economies affect
the growth of output in Indonesia? What role will policies play?
The rest of this section examines whether secular stagnation in ad-
vanced economies has transmitted to Indonesia’s output dynamics. The
discussion begins with a brief review of the channels through which trans-
mission can occur.

Channels of Transmission
A large body of work has analyzed the channels through which economic
developments and policies in the North spill over to countries in the South
(IMF 2013, 2014c). Eggertsson, Mehrotra, and Summers (2016) find that
monetary and fiscal policy spillovers are stronger in the context of secular
stagnation and more likely to be persistent. The channels of transmission
do not necessarily act independently; they can be either reinforcing or
offsetting.

Trade Channel
Trade linkages are a direct source of transmission of the spillovers of secular
stagnation. Deceleration of demand in advanced economies lowers import
demand from emerging markets. Indeed, the global trade slowdown is
commonly viewed as symptomatic of secular stagnation in advanced econo-
mies (IMF 2016a). The impact of weaker external demand from advanced
economies can be amplified if it is accompanied by highly accommodative
monetary policy in advanced economies, leading to larger capital outflows
to emerging markets and an appreciation of their nominal exchange rates.

242 SUSTAINING ECONOMIC GROWTH IN ASIA

 Exposure to China may also affect trade-related spillovers of secular
stagnation to Asian economies.8 Slowing external demand from advanced
economies has affected China. Given its role as a hub in the regional supply
chain, the impact of secular stagnation in advanced economies may be
transmitted even to economies with low trade links to advanced economies,
through their links to China. China has recently begun to rebalance its
economy from investment toward consumption. Because the import inten-
sity of investment is significantly higher than that of consumption (Kang
and Liao 2016), exporters of investment goods (such as firms in the euro
area) may have been affected negatively whereas exporters of consumption
goods (many of which are in Asia) may have benefited.
The trade-related spillovers of secular stagnation will vary across coun-
tries, reflecting how high their exposure is to demand from advanced econ-
omies, how strong their trade links are with China, and how significant
external demand is in overall domestic demand.9

Financial Channel
Financial linkages are a second source of transmission, directly or indirectly.
Emerging markets’ financial linkages to advanced economies include the
net inflows of foreign capital that supply credit and finance investment in
their financial, nonfinancial, and official sectors and their stocks of external
assets and liabilities.
Financial spillovers of secular stagnation can come from the weakness
in demand in advanced economies, which generally leads to lower exchange
rates abroad as a result of weaker demand for exports (Eggertsson, Mehrotra,
and Summers 2016). Weaker exchange rates may benefit trade, but they can
have large and deleterious balance sheet effects and adverse growth conse-
quences when the foreign currency exposure of external liabilities is high
(Eichengreen, Hausmann, and Panizza 2007).10 Both the size of the external
balance sheet and its foreign currency denomination are important: For a

8. Basri and Rahardja (2010) note that the impact of the global crisis on the Indonesian
economy was relatively limited in part because of the strong demand from China and India,
in particular for primary commodity exports. As natural resources account for the majority
of Indonesia’s exports, the composition of exports stabilized external demand.
9. External demand in countries that are more exposed to the sectors slowing most sharply
in advanced economies (such as construction and manufacturing) may be more affected
than exporters of consumption goods (IMF 2016a).
10. By contrast, external assets denominated in foreign currency would improve in local
currency value. For a net creditor, the net impact of a depreciation on the external balance
may be to raise the value of the net international investment position (NIIP). For a net debtor
(like Indonesia), the negative revaluation of liabilities would likely exceed the positive revalu-
ation of assets, leading to a lower NIIP.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 243

given scale of the external balance sheet, the exchange rate impacts will be
larger the greater is foreign currency exposure. For a given foreign currency
exposure, the impact will be larger the bigger the external balance sheet.
Financial linkages also have the potential for self-fulfilling prophecies
to transmit secular stagnation from advanced to emerging-market econo-
mies, which can occur if financial markets expect a permanent slowdown in
advanced economies to reduce the growth prospects of emerging markets,
resulting in lower capital inflows to emerging markets, an increase in the
risk premium for new lending, or both. The financial spillovers of secular
stagnation may have highly heterogenous impacts across emerging markets
depending on the foreign currency exposure of external balance sheets, de-
pendence on external credit, and fundamentals that affect the risk premium
(Blanchard, Das, and Faruqee 2010).

How Exposed Is Indonesia to Secular Stagnation in Advanced

Economies?
This section compares Indonesia with a large group of emerging-market
countries, including regional peers in ASEAN-4 (Malaysia, Thailand, and
the Philippines). As historical perspective is also useful, it examines the
evolution of these exposures from the Asian financial crisis to today.

Trade-Related Exposure
Three indicators can shed light on Indonesia’s trade-based exposure to
secular stagnation: the size of the external sector in relation to the overall
economy, the diversity of exposure to exports and imports across regions,
and the contribution of net exports to output growth.
The size of Indonesia’s external sector as a share of the economy has
steadily declined since the early 2000s (figure 11.4a).11 Exports were 40
percent and imports 32 percent of GDP in 2000; both had declined to less
than 20 percent by 2016. Except for a short-lived uptick in 2003–05, these
declines were steady through the domestic business cycles, the commodity
price boom before the financial crisis, the commodity price bust starting in
2013, and the expansion of global value chains in Asia since the early 2000s.12

11. The change is not just a reflection of the rapid expansion of the economy. Growth rates
of exports and imports have been on a trend decline (in both nominal and real terms) since
the global financial crisis; in 2016 they registered negative growth rates.
12. These trends suggest that deeper structural factors lie behind the rise in insularity. By
some accounts, they reflect long-standing institutional weaknesses in the regulatory frame-
work, the legal and judiciary system, and tax administration, which have dissuaded private
investors, particularly in export sectors (IMF 2005).

244 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 11.4a Evolution of Indonesia’s external sector, 2000–16

percent of GDP
45
40
35
30
25
20
15
10
Imports
5 Exports
0
2000 2002 2004 2006 2008 2010 2012 2014 2016

Sources: IMF, World Economic Outlook database; author’s calculations.

Figure 11.4b Average openness (exports + imports), 2000–16

percent of GDP
100
90
80
70
60
50
40
30
20
10
0
o

Th nd
es

M nd
om l
a

Tu a
In key

M a

en ica

la

Po e

a
zi

bi

tin

si

si

si
l
ic

Ph zue

hi
in
ra

hi

la

la
ne

us

ay
fr
ex

C
r
en

pp
B

ai
A

al
do

e
rg
ol

ili
ut
C

V
So

Sources: IMF, World Economic Outlook database; author’s calculations.

Figure 11.4b illustrates how these trends compare with average open-
ness among emerging-market peers. The data reveal two facts. First, there
is tremendous heterogeneity in external exposure across emerging econo-
mies. Second, although Indonesia is more closed than all its regional peers,
it is more exposed than other large economies, including Brazil and Turkey.
A natural question is whether the extent of Indonesia’s declining
external exposure is heterogenous across its trading partners. To shed light
on this question, figures 11.4c and 11.4d show the evolution of Indonesia’s
export and import shares by region. All of the decline in Indonesia’s external

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 245

 Figure 11.4c Indonesia’s exports, by region, 1993–2016

percent of total exports

90

80

70

60

50

40

30

20

10

0
1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Advanced economies
Emerging-market and developing
economies, excluding China
China
Emerging-market Asia

Sources: IMF, Direction of Trade Statistics; author’s calculations.

Figure 11.4d Indonesia’s imports, by region, 1993–2016

percent of total imports

90
80
70
60
50
40
30
20
10
0
1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Advanced economies
Emerging-market and developing
economies, excluding China
China
Emerging-market Asia

Sources: IMF, Direction of Trade Statistics; author’s calculations.

246 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 11.4e Contributions to Indonesia’s real GDP growth, 2001–16

percent change, year over year

12
Domestic demand
10 Net exports
Real GDP (percent change)

–2
2001 2003 2005 2007 2009 2011 2013 2015

Sources: IMF, World Economic Outlook database; author’s calculations.

sector exposure resulted from reduced trade with advanced economies.

Between 2000 and 2016, the share of Indonesia’s exports to advanced econ-
omies declined by about 25 percentage points, and the share of imports
from advanced economies declined by 45 percentage points. The region that
absorbed those declining shares from advanced economies was predomi-
nantly emerging Asia, where exports rose by about 20 percentage points and
imports by about 30 percentage points.13
The compositional evolution of its trading partners indicates that
Indonesia is likely to have been significantly insulated from weaker export
demand. Not only did exposure to advanced economies decline, Indonesia’s
exports to these economies as a percent of GDP was modest, averaging
about 6 percent in 2000–16.14
The data in figure 11.4e indicate that Indonesia has moderate to low
exposure to external demand from advanced economies. As a share of the
economy, the external sector has declined, and its contributions to output
growth have been small (figure 11.4e).

13. China did not drive the rise in trade with emerging Asia, as shown in figures 11.4c and
11.4d.
14. IMF, World Economic Outlook June 2017 database.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 247

 Figure 11.5a Indonesia’s financial integration with the global
economy (external assets + external liabilities),
2001–16

percent of GDP
120

110

100

90

80

70

60

50

40
2001 2003 2005 2007 2009 2011 2013 2015

Sources: IMF, World Economic Outlook database; author’s calculations.

Financial Exposure
Both stock and flow indicators shed light on the channels through which
secular stagnation abroad can transmit to the domestic economy. A key
stock indicator is the size and (currency) composition of its external balance
sheet. The main flow indicators are the volume and volatility of nonofficial
international capital flows, which supply credit and finance investment to
the domestic economy.
In the years between the Asian financial crisis and the global financial
crisis, Indonesia steadily reduced its financial integration with the global
economy, as reflected in the declining sum of external assets and external
liabilities (figure 11.5a). By this measure, financial integration fell from 118
percent of GDP in 2001 to 80 percent of GDP in 2007.15 This evolution lies
in marked contrast with one of the major trends of the last quarter century
in the world economy (as well as in newly industrialized Asian economies
and developing Asia), which has seen record cross-border transactions and
a concomitant rise in financial integration (Obstfeld 2015).
One reason for the low and declining exposure to external liabilities
is a regulatory environment that has impeded foreign direct investment
(FDI). On the FDI Regulatory Restrictiveness Index of the Organization for
Economic Cooperation and Development (OECD), Indonesia has among

15. The trough of the sum of foreign assets and liabilities in ratio to GDP is 42 percent in
2008. However, the 2008 level may well be sui generis, reflecting large valuation effects from
both the steep depreciation of the rupiah and the large drops in asset prices.

248 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 11.5b Global financial integration across emerging markets,
2015 and average 2001–16

percent of GDP
300

250 2015 (external assets + liabilities/GDP)

Average 2001–16
200

150

100

50

d
a

sia

ey

zil

ia

ile
lan

an
ne
in

bi

xic
ss
a

Ch
rk
ne
nt

ail
Br

pi

Po
Ru

Me
Tu

lo
ge

do

ilip

Th
Co
Ar

In

Ph

Sources: IMF, World Economic Outlook database; author’s calculations.

the strictest limitations in the ASEAN-9 (Cambodia, Indonesia, Laos,

Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Vietnam),
including outright bans on foreign participation in certain sectors.
Another factor is the rapid decline in foreign currency liabilities by banks
and nonfinancial corporates, amid tighter supervision, intended to address
the corporate governance problems that proved so damaging in 1997. Debt
liabilities declined from 90 percent of GDP in 2000 to less than 30 percent
of GDP in 2013.16 Since the global financial crisis, Indonesia’s financial inte-
gration has reversed course, climbing to about 93 percent of GDP in 2015, a
level last seen in 2002. External assets and liabilities increased by about the
same amount.17
Nevertheless, as of 2015, Indonesia remained one of the least finan-
cially integrated economies among emerging-market peers (figure 11.5b).
External assets and liabilities each lie well in excess of 100 percent of GDP
in the Philippines and Thailand.
A few years after the Asian financial crisis, the foreign currency denom-
ination of external liabilities declined steadily in Indonesia (figure 11.5c).
On the eve of the Asian financial crisis, 83 percent of foreign liabilities
were denominated in foreign currency. By 2012 this figure had fallen to

16. IMF, External Wealth of Nations Database.

17. The increase in external assets reflects an increase in outward FDI and other investment
(including lending to foreign financial institutions and banks). Portfolio investment and
some increase in inward FDI have driven external liabilities.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 249

Figure 11.5c Foreign currency exposure of Indonesia’s external
balance sheet, 1990–2012

percent of total external liabilities

100
90
80
70
60
50
40
30
20
10
0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012

Share of external liabilities in domestic currency

Share of external liabilities in foreign currency

Source: Bénétrix, Lane, and Shambaugh (2015).

about 42 percent (Bénétrix, Lane, and Shambaugh 2015).18 The combina-

tion of a small external balance sheet and moderate foreign currency expo-
sure limits the financial channel of transmitting secular stagnation from
advanced economies to Indonesia (see figures 11.5b and 11.5c).
Some flow indicators support these conclusions. Despite some episodic
retrenchment, most notably during the taper tantrum of 2013, exposure to
capital flow volatility net and gross international capital flows to Indonesia
have been steady since the global financial crisis (IMF 2016b). Since the
early 2000s, the volatility of Indonesia’s net nonofficial inflows has been
similar to that of its emerging-market peers, and it compares favorably with
regional peers (figure 11.5d). Research indicates that the adverse impact of
global financing shocks is stronger for emerging markets prone to higher
capital flow volatility (IMF 2014c). This finding suggests that vicissitudes in
the supply of foreign capital are not a strong source of growth spillovers for
Indonesia. There is also no evidence that the cost of external financing has
changed (figure 11.5e). There is thus little evidence of financial market self-
fulfilling prophecies. With low to moderate financial linkages with the rest
of the world and steadily falling foreign currency denomination of external
liabilities, Indonesia is not heavily exposed to financial-related spillovers
from advanced economies.

18. Indonesia is not unique in this respect. The decline in “original sin” (as Eichengreen,
Hausmann, and Panizza [2007] call the foreign currency exposure of liabilities) has been
observed in emerging markets across the world.

250 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 11.5d Volatility of the supply of foreign capital (net nonofficial
inflows), 2000–15

percent of GDP
7
6
5
4
3
2
1
0
a

na

nd

es

le

ia

nd

a
zi
di

si

si
e
s
hi
ra
in

rk
hi

la

la
ne

us

ay
In

C
pp

B
C

Po

ai
Tu
R

al
do

Th
ili

M
In
Ph

Sources: IMF, Balance of Payments database; author’s calculations.

Decomposition of Domestic and External Contributions to

Growth
The analysis in this section is based on a vector auto-regression (VAR)
analysis, using quarterly data from the first quarter of 1999 to the fourth
quarter of 2016. The VAR is limited to 10 variables (4 external and 6
domestic factors), in order to limit the number of estimable parameters
relative to the number of observations. All variables enter the VAR with
four lags.
External factors include US real GDP growth (a proxy for advanced
economy demand shocks); US inflation (a proxy for advanced economy
supply shocks, once US growth is controlled for); and the one-year US Trea-
sury bond rate (to capture the advanced economy monetary policy stance).
The VAR includes the real growth rate in China, given China’s growing sig-
nificance in the region. Domestic factors include the real output growth
rate in Indonesia, domestic consumer price inflation, the rate of real ex-
change rate appreciation versus the US dollar, the Emerging Market Bond
Index (EMBI) spread (to proxy external financing conditions), the percent-
age change in terms of trade (capturing factors other than changes in ex-
ternal demand or financing conditions), and the short-term interest rate.
Terms of trade are arguably either a domestic or an external factor; results
were not sensitive to this choice.19 All growth rates and inflation rates are
measured in year-over-year terms.

19. In the estimation of the VAR, the key restriction is that shocks to the external block
are assumed to be exogenous to shocks to the internal block. The external variables thus

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 251

252
Figure 11.5e Indonesia’s Emerging Market Bond Index (EMBI) spread, 1998–2016

basis points
1,200
Average 1998–2001
1,000 Average 2002–08
Average 2009–16

800

600

400

SUSTAINING ECONOMIC GROWTH IN ASIA

200

0
1 1 1 12 13 13 14 5 6 16
98 99 00 0 0 02 03 04 04 05 06 07 07 08 09 10 10 01 1 1
r 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 r 2 20 20 20 20 20 20 20
e e r e h e r r e r r e h r r e h e e h r
b b n rc b e n c h e e n c e e n rc b
r e ne ch er er
b u r b b n rc be
e m em Ju Ma em mb Ju ar mb mb Ju ar mb mb Ju Ma em em J Ma em em Ju Ma em
t e M ce te M e te t
ec p ec t e p ec p ec p ec pt ec
D Se D Sep D Se D Se D Se D Se D

Sources: JP Morgan; author’s calculations.

 Table 11.1 Correlation between domestic real GDP growth in
Indonesia and domestic and global factors
Variable 2000–16 2009–16
US real GDP growth –0.06 0.41
US inflation 0.36 0.72
One-year US Treasury bond rate –0.05 –0.36
China’s real GDP growth 0.29 0.51
Inflation in Indonesia –0.07 –0.07
Terms-of-trade growth 0.07 0.28
Emerging Market Bond Index spread –0.42 –0.31
Real effective exchange rate change (increase is –0.27 0.35
depreciation)
Monetary policy in Indonesia –0.40 –0.45
Sources: Author’s calculations using data from IMF World Economic Out-
look database, JP Morgan, and US Treasury.

Table 11.1 displays the correlations between domestic real GDP growth
in Indonesia and external and domestic factors over the entire period (2000–
16) and the period after the global financial crisis (2009–16). The correla-
tion between real GDP growth in the United States and Indonesia was negli-
gible over the entire period (consistent with figure 11.2) but strengthened
considerably after the global financial crisis. This finding provides prima
facie evidence that secular stagnation in advanced economies may have been
transmitted to domestic output dynamics since the global financial crisis.
The VAR estimates are used to compute what fraction of Indonesia’s
real GDP growth (relative to its estimated average growth over the sample
period) can be attributed to external versus domestic factors. The results
show that domestic factors explain more than three-fourths of the deviation
of Indonesia’s growth from the estimated sample mean between 1999 and
the end of 2002 (figure 11.6). The contribution of external factors to devia-
tions from average growth started rising intermittently in 2003, making
positive contributions for most quarters in 2006 through the first quarter
of 2009. External factors strongly dominated growth dynamics during the

do not respond to the internal variables contemporaneously. Within the external factors,
identification of the shocks is based on a recursive scheme: US growth affects all variables
within a quarter, whereas shocks to other variables can affect US growth only with a lag of at
least one quarter. US inflation shocks may affect all variables other than US growth within
a quarter; the one-year US Treasury rate is placed last in the recursive ordering, implying
that it responds contemporaneously to all external factors but not to any of the domestic
shocks. Within the internal block, shocks are not explicitly ordered. These assumptions
follow closely the VAR exercise in IMF (2014a).

HOWGRAPHICS
HAS INDONESIA FARED
- CHAPTER IN THE AGE OF
11—SUSTAINING SECULARGROWTH
ECONOMIC IN ASIA2531
STAGNATION?
254
Figure 11.6 Contribution of domestic and global factors to Indonesia’s real GDP growth deviations,
1999Q1–2016Q4

percent
3

SUSTAINING ECONOMIC GROWTH IN ASIA

–1

Domestic
–2
Global
Deviation of growth from
estimated mean
–3
Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Sources: IMF, World Economic Outlook database; author’s calculations.

global financial crisis, with domestic factors reemerging as important late
in the second quarter of 2012.
Both domestic and external factors played roles in output dynamics
after 2012. The contribution of external factors was predominantly nega-
tive; domestic factors somewhat offset their impact.
The VAR results are consistent with the hypothesis that secular stag-
nation in advanced economies was transmitted to Indonesia. However,
the quantitative impact of the drag from external factors was not histori-
cally large. After late 2013, it was strong enough to partially or even fully
offset the strength of domestic factors. That it did so despite fairly low and
declining linkages between Indonesia and the global economy points to the
complexity of spillovers, including possibly reinforcing channels and indi-
rect transmission through regional trade partners.
It is impossible to know how secular stagnation would have affected
Indonesia in the counterfactual scenario in which its international trade
and financial linkages were strong and rising. A reasonable (albeit quali-
fied) conclusion from the results presented here is that the impact would
have been larger.

Prospective Analysis: Countering the Headwinds from

Secular Stagnation
Assessing the medium-term path of output is critical for the conduct of
monetary, fiscal, and structural policies. To the extent that demand weak-
ness in advanced economies is temporary, countercyclical stabilization
policies may suffice in addressing the short-term deceleration of domestic
growth. If the decline represents a longer-term structural change in the
growth rate of advanced economies, different policies may be needed in
Indonesia.

Possible Output Dynamics in Indonesia: Stylized Facts

Potential growth in Indonesia has been on a downward trend since the global
financial crisis, a period that broadly coincides with the slowdown in ad-
vanced economies (figure 11.7). The potential growth rate rose from about
4.0 percent in 2000 to 5.7 percent in 2008. It edged down to 5.4 percent in
2009–16 and is projected to trend down farther over the medium term.20
This trend is not unique to Indonesia. The IMF (2015b) finds that the
potential growth rate declined by about 2 percentage points in emerging
markets after the global financial crisis. From that perspective, the declining
trend in Indonesia is mild.

20. IMF, World Economic Outlook June 2017 database.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 255

Figure 11.7 Potential output growth rate in Indonesia, actual and forecast,
2000–22

percent
6.5
6.0
5.5
5.0
4.5
4.0
Potential output growth rate
3.5 Average 2000–04
3.0 Average 2005–08
Average 2009–16
2.5 Projected 2017–22
2.0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022

Sources: IMF, World Economic Outlook database; author’s calculations.

Identifying the sources of lower potential growth in Indonesia is a first

step in identifying policy implications. To the extent that lower potential
growth rates emerged from lower factor accumulation—including human
and physical capital—policy measures may need to target raising the supply
of labor, reducing rigidity in hiring and firing policies, and addressing
other domestic impediments to investment, including regulation, red tape,
and the business environment. If they arose from declining TFP, deeper
structural issues may be at play.
Decomposition of the growth rate of potential output into growth
rates of factor inputs and growth rates of TFP reveals the importance of
TFP (figure 11.8).21 Before the global financial crisis, acceleration of TFP lay
behind the increase in potential growth rates. More than half of the increase
in the growth rate of potential output—from about 5 percent in 2001–04
to 6 percent in 2006–08—reflected the increase in employment, possibly
in part due to weak employment growth in preceding years due to a base
effect. Weak employment growth in 2001–04 (reflected in the low contri-
bution of labor input in 2001–04) in turn reflects the very slow decline in
unemployment after the Asian financial crisis, in part the result of weak
investment and a poor investment climate (IMF 2005). The contribution of
capital accumulation also increased between 2001–04 and 2006–08.

21. The decomposition is of potential growth rate into the actual capital growth rate and
the potential employment growth rate reflecting the working-age population, as described
in IMF (2014c).

256 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 11.8 Decomposition of Indonesia’s potential growth rate,
2001–14

percent
7

0
2001–04 2006–08 2009–14

Total factor productivity

Employment
Capital stock
Potential output

Sources: World Bank, World Development Indicators database; IMF, World

Economic Outlook database; author’s calculations.

The decline in potential output growth rate since the global financial
crisis is attributable largely to the decline in the growth of TFP. The contri-
bution of employment and capital growth helped offset some of the decline
in TFP growth, thanks partly to the modest stimulus Indonesia imple-
mented as external demand softened in 2009. For emerging markets as a
whole, TFP has been found to account for the entire postcrisis decline in
potential growth rates (see Cubeddu et al. 2014, IMF 2015b).

Sources of Changes in Total Factor Productivity

A prominent supply-side explanation for secular stagnation, associated with
Gordon (2016), is that the slowdown in technological innovation accounts
for low growth in advanced economies. Because advanced economies are
at the technological frontier, and technological spillovers across borders
have been found to raise TFP and growth (see, e.g., Coe and Helpman 1995,
Amman and Virmani 2015), a slowdown in TFP growth in advanced econo-
mies could transmit to lower TFP growth and lower potential growth in
emerging markets. Lower trade and financial spillovers, for instance, could
limit the diffusion of technology, technological know-how, and best prac-
tices.
HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 257
 Declining TFP growth may also be a result of convergence to the tech-
nological frontier. After more than a decade of rapid factor accumulation
during the process of catch-up, it may have been inevitable that the growth
rate of factor utilization and human capital growth—an important compo-
nent of TFP—would slow (IMF 2015b).22
These arguments imply a role only for domestic factors. Stylized evi-
dence suggests that both domestic and external factors could have played a
role in the deceleration of TFP growth in Indonesia.

Human Capital Growth

Human capital accumulation—distinct from labor input—affects TFP
(Manuelli and Seshadri 2014). The human capital component of TFP can
decline during downturns because of lower learning by doing in recessions
(Martin and Rogers 1997). In addition, uncertain future growth prospects
may temporarily or permanently lower the desire for higher educational
attainment.
Figure 11.9 illustrates the growth rate and level of high-skill human
capital accumulation (proxied by completion of tertiary education) in In-
donesia. The growth rate of human capital accumulation was gradually
rising until the global financial crisis; since the crisis, it has declined (figure
11.9a).23 The decline is not severe, but given the low levels of tertiary educa-
tion in Indonesia (figure 11.9b), a slowdown in human capital accumula-
tion could present bottlenecks for high-value-added employment.

Trade Restrictions
Restrictions on trade (such as import tariffs) result in inefficient alloca-
tion of the factors of trade, reducing TFP. A large body of empirical work
corroborates this theoretical prediction (e.g., Caselli, Esquivel, and Lefort
1996; Hall and Jones 1999).
Data from the World Bank’s Temporary Trade Barriers database show
a steady increase in protectionism in Indonesia. The rise in protectionism
is evident in both average tariff rates and nontariff barriers (Basri and
Patunru 2012). Protectionist measures were on a downward trend before
the global financial crisis but rose sharply thereafter (figure 11.10). They
have since edged down but remain high relative to the precrisis years.

22. In traditional growth decomposition, TFP traditionally accounts for factor utilization—
such as hours worked, capacity utilization, and the quality of labor and capital inputs, as
distinct from the volume of labor and capital inputs—rather than labor or capital inputs.
23. Data on human capital accumulation are from the Barro-Lee dataset, which go only
through 2010.

258 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 11.9a Indonesian students with higher education, 2002–16

percent change, year over year

100
80
60
40
20
0
–20
–40
–60
–80
2002 2004 2006 2008 2010 2012 2014 2016

Higher education
Of which: Higher education, with diploma
Of which: Higher education, with postgraduate
or higher degree

Sources: Central Statistics Bureau; author’s calculations.

Figure 11.9b Share of Indonesia’s population with tertiary

education, 1988–2010

percent of population
8
7
6
5
4
3
2
1
0
1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Source: Barro and Lee database, www.barrolee.com.

Institutions
The quality of institutions—such as labor regulations, the legal and judi-
cial system, and accountability—can play a key role in a country’s ability to
adopt superior technologies, and thus raise TFP, income, and living stan-
dards (McGuiness 2007, Acemoglu 2008).
Indonesia ranks poorly on several measures of the ease of doing busi-
ness (World Bank 2017). International financial institutions have noted
structural impediments, including a weak investment climate, complex reg-
ulations, and shallow financial markets (IMF 2016b). After improvements
between the Asian financial crisis and global financial crisis, the regulato-
HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 259
 Figure 11.10 Temporary trade restrictions on Indonesian imports,
1996–2015
percent of total imports
1.2

1.0

0.8

0.6

0.4

0.2

0
1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

Sources: World Bank; author’s calculations.

Figure 11.11 Regulatory quality in Indonesia, Malaysia, the

Philippines, and Thailand, 1996–2015
index (1 = highest; 0 = weakest)
0.9
0.8
0.7
0.6
0.5
0.4
Indonesia
0.3 Malaysia
0.2 Philippines
0.1 Thailand
0
1996 2000 2003 2005 2007 2009 2011 2013 2015

Source: International Country Risk Guide, 2015.

ry environment weakened in Indonesia (figure 11.11). In 2015 the ease of

doing business was somewhat lower than in regional peers. The declining
path of regulatory quality may have contributed to lower TFP growth.

Baseline and Alternative Paths for Potential Output Growth

Rates
To identify the role for policies if the deceleration in TFP continues to
present headwinds to the growth rate of potential output, this section
considers the evolution of potential growth through a scenario analysis,
assuming a path for each of its components (labor, capital accumulation,
and TFP). The scenarios are illustrative only, given the high uncertainty of
the projections on which they are based.

260 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 11.12a Growth rate of the working-age population
in Indonesia, 1955–2065

percent, 5-year intervals

16
14
12
10
8
6
4
2 Actual
0 Estimated
–2
1955 1965 1975 1985 1995 2005 2015 2025 2035 2045 2055 2065

Sources: UN Population Division; author’s calculations.

Figure 11.12b Labor force participation rates, Indonesia vs. EU-28

and OECD, 2006–13

percent
73

72

71

70

69

68
Indonesia
67 EU-28
OECD
66
2006 2007 2008 2009 2010 2011 2012 2013

EU-28 = 28 member countries of the European Union; OECD = Organization for

Economic Cooperation and Development
Sources: World Bank, World Development Indicators; IMF, World Economic
Outlook database; OECD Statistics; author’s calculations.

The future paths for labor are derived from demographic projections and
assumptions about labor force participation rates. Indonesia has the oppor-
tunity to reap large demographic dividends, given the projected increase in
the working-age population through 2060 (figure 11.12a). However, labor
force participation rates, which were rising steadily before the financial
crisis, have been volatile since 2010, registering negative average growth rate
in 2010–13 (figure 11.12b). Taking the working-age projections as given, the
scenario analysis assumes that the labor force participation rate reverts to

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 261

 Figure 11.13 Illustrative scenario analysis

percent
8
7
6
5
4
3
2
1
0
2001–08 2009–14 2015–18 2019–2022

Capital stock
Employment
Total factor productivity
Potential output

Sources: World Bank, World Development Indicators; IMF, World

Economic Outlook database; author’s calculations.

its precrisis growth rate and stabilizes at the level in the OECD and EU-28
(as shown in figure 11.12b).
The capital stock is assumed to continue to grow at its postcrisis
average rate. This assumption is more optimistic than the scenario analysis
in IMF (2015b), which notes that less favorable external financing condi-
tions, infrastructure bottlenecks, and softer or flat commodity prices will
likely lead to a decline in emerging-market capital growth over the medium
term.
TFP growth is assumed to rise to its precrisis mean, thanks to an
increase in human capital accumulation, the removal of trade restrictions,
greater foreign participation in industry through FDI, and a better busi-
ness climate (achieved by simplifying regulations and increasing financial
depth).
Under this scenario, medium-term potential growth in Indonesia in-
creases from about 6 percent under the baseline to 7 percent (figure 11.13).
Potential growth could evolve differently for several reasons, such as an
upward revision to the forecast of commodity prices (which could spur in-
vestment and capital growth), a more rapid easing of barriers to FDI inflows
(which could raise TFP), or a downward revision to global growth.

262 SUSTAINING ECONOMIC GROWTH IN ASIA

Conclusion
In the years after the Asian financial crisis, Indonesia’s engagement with the
global economy weakened in terms of both trade and financial integration.
This retrenchment was extremely unusual, among both regional peers and
emerging markets more generally. The low exposure to global economic and
financial developments, a large and strengthening domestic base, a strong
fiscal framework, and prudence in managing international capital move-
ments insulated Indonesia significantly from the vicissitudes of global devel-
opments, reflected in remarkably stable output growth. Its inward-looking
stance limited the transmission of secular stagnation from advanced econo-
mies to the domestic economy.
Policies will need to calibrate the tradeoff between the potential stability
gains of retrenching from a slowing world economy and the long-term costs
that an increasingly inward-looking stance may entail, however.
Low exposure to global developments has also limited the diffusion
of technological advances and productivity-enhancing spillovers of global
economic integration. Potential output growth declined in recent years,
despite strong demographic tailwinds and steady capital accumulation, as
a result of lower TFP growth. The slowdown in human capital accumula-
tion, the rise in protectionism, and some weakening of the regulatory envi-
ronment may have contributed to the TFP growth slowdown. Structural
supply-side policies that enhance productivity may help Indonesia raise
potential growth and weather the negative impulse from a slowing world
economy.

References
Acemoglu, Daron. 2008. Introduction to Modern Economic Growth. Princeton, NJ: Princeton Uni-
versity Press.
Amman, E., and S. Virmani. 2015. Foreign Direct Investment and Reverse Technology Spill-
overs: The Effect on Total Factor Productivity. OECD Journal: Economic Studies. Available at
www.oecd.org/eco/growth/Foreign-direct-investment-and-reverse-technology-spillovers-
The-effect-on-total-factor-productivity-OECD-Journal-Economic-Studies-2014.pdf.
Basri, M. Chatib, and A. Patunru. 2012. How to Keep Trade Policy Open: The Case of Indone-
sia. Bulletin of Indonesian Economic Studies 48, no. 2.
Basri, M. Chatib, and S. Rahardja. 2010. The Indonesian Economy amidst the Global Crisis:
Good Policy and Good Luck. ASEAN Economic Bulletin 27, no. 1: 77–97.
Bénétrix, Agustin S., Philip R. Lane, and Jay C. Shambaugh. 2015. International Currency
Exposures, Valuation Effects and the Global Financial Crisis. Journal of International Eco-
nomics 96: 98–109.
Blanchard, Olivier J., Mitali Das, and Hamid Faruqee. 2010. The Initial Impact of the Crisis on
Emerging Market Countries. Brookings Papers on Economic Activity (Spring): 263–307. Avail-
able at www.brookings.edu/wpcontent/uploads/2010/03/2010a_bpea_blanchard.pdf.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 263

Caselli, F., G. Esquivel, and F. Lefort. 1996. Reopening the Convergence Debate: A New Look
at Cross-Country Growth Empirics. Journal of Economic Growth 1, no. 3: 363–89. Available
at https://link.springer.com/article/10.1007/BF00141044.
Cesa-Bianchi, A., J. Imbs, and J. Saleheen. 2016. Finance and Synchronization. London: Centre
for Economic Policy Research. Available at www.cepr.org/active/publications/discussion_
papers/dp.php?dpno=11037.
Coe, David T., and Elhanan Helpman. 1995. International R&D Spillovers. European Economic
Review 39, no. 5: 859–87.
Cubeddu, L., A. Culiuc, G. Fayad, Y. Gao, K. Kochhar, A. Kyobe, C. Oner, R. Perrelli, S. Sanya,
E. Tsouta, and Z. Zhang. 2014. Emerging Markets in Transition. IMF Staff Discussion Note
14/06. Washington: International Monetary Fund. Available at www.imf.org/external/
pubs/ft/sdn/2014/sdn1406.pdf.
Eggertsson, G., N. Mehrotra, and L. Summers. 2016. Secular Stagnation in the Open Econo-
my. American Economic Review 106, no. 5: 503–507.
Eichengreen, B., R. Hausmann, and U. Panizza. 2007. Currency Mismatches, Debt Intoler-
ance and Original Sin: Why They Are Not the Same and Why It Matters. In Capital Controls
and Capital Flows in Emerging Economies: Policies, Practices and Consequences, ed. Sebastian Ed-
wards. Chicago: University of Chicago.
Fischer, S. 2015. Monetary Policy in the United States and in Developing Countries. Speech
at the Crockett Governors’ Roundtable 2015 for African Central Bankers, University of
Oxford, Oxford, UK, June 30.
Gordon, Robert. 2016. The Rise and Fall of American Growth: The U.S. Standard of Living Since the
Civil War. Princeton, NJ: Princeton University Press.
Hall, Robert E., and C. I. Jones. 1999. Why Do Some Countries Produce So Much More Out-
put Per Worker than Others? Quarterly Journal of Economics 114, no. 1: 3–116.
IMF (International Monetary Fund). 2005. Indonesia: 2005 Article IV Consultation and Third Post-
Program Monitoring Discussions. IMF Country Report 05/326. Washington. Available at
www.imf.org/external/pubs/ft/scr/2005/cr05326.pdf.
IMF (International Monetary Fund). 2007. Indonesia: 2007 Article IV Consultation and Third Post-
Program Monitoring Discussions. IMF Country Report No. 07/272. Washington. Available at
www.imf.org/external/pubs/ft/scr/2007/cr07272.pdf.
IMF (International Monetary Fund). 2009. Indonesia: 2009 Article IV Consultation and Third Post-
Program Monitoring Discussions. IMF Country Report No. 09/230. Washington. Available at
www.imf.org/external/pubs/ft/scr/2009/cr09230.pdf.
IMF (International Monetary Fund). 2010. Indonesia: 2010 Article IV Consultation. IMF Coun-
try Report 10/284. Washington. Available at www.imf.org/external/pubs/ft/scr/2010/
cr10284.pdf.
IMF (International Monetary Fund). 2013. 2013 Spillover Report. IMF Multilateral Policy Issues
Report. Washington. Available at www.imf.org/external/np/pp/eng/2013/070213.pdf.
IMF (International Monetary Fund). 2014a. On the Receiving End? External Conditions and
Emerging Market Growth Before, During and After the Global Financial Crisis. In World
Economic Outlook, April. Washington. Available at www.imf.org/~/media/Websites/IMF/
imported-flagship-issues/external/pubs/ft/weo/2014/01/pdf/_c4pdf.ashx.
IMF (International Monetary Fund). 2014b. 2014 Pilot External Sector Report. Multilateral Policy
Issues Report. Washington. Available at www.imf.org/external/np/pp/eng/2014/062614.
pdf.
IMF (International Monetary Fund). 2014c. 2014 Spillover Report. IMF Multilateral Policy Is-
sues Report. Washington. Available at www.imf.org/external/np/pp/eng/2014/062514.
pdf.

264 SUSTAINING ECONOMIC GROWTH IN ASIA

IMF (International Monetary Fund). 2015a. Indonesia: 2015 Article IV Consultation. IMF Coun-
try Report 16/81. Washington. Available at www.imf.org/external/pubs/ft/scr/2016/
cr1681.pdf.
IMF (International Monetary Fund). 2015b. Where Are We Headed? Perspectives on Potential
Output. In World Economic Outlook, April. Washington. Available at www.imf.org/~/me-
dia/Websites/IMF/imported-flagship-issues/external/pubs/ft/weo/2015/01/pdf/_c3pdf.
ashx.
IMF (International Monetary Fund). 2016a. Global Trade: What’s Behind the Slowdown?
In World Economic Outlook, October. Washington. Available at www.imf.org/~/media/
Websites/IMF/imported-flagship-issues/external/pubs/ft/weo/2016/02/pdf/_c2pdf.
ashx+&cd=1&hl=en&ct=clnk&gl=us.
IMF (International Monetary Fund). 2016b. Indonesia: 2016 Article IV Consultation. IMF Coun-
try Report 17/37. Washington. Available at www.imf.org/~/media/Files/Publications/
CR/2017/cr1737.ashx.
IMF (International Monetary Fund). 2017. Indonesia: 2017 Article IV Consultation. IMF Country
Report No. 18/32. Washington. Available at www.imf.org/~/media/Files/Publications/
CR/2018/cr1832.ashx.
Kang, J., and C. Liao. 2016. Chinese Imports: What’s behind the Slowdown? IMF Working Paper
WP/16/106. Washington. Available at www.imf.org/external/pubs/ft/wp/2016/wp16106.
pdf.
Manuelli, R., and A. Seshadri. 2014. Human Capital and the Wealth of Nations. American
Economic Review 104, no. 9: 2736–62.
Martin, Phillip, and Carol A. Rogers. 1997. Stabilization Policy, Learning by Doing and Eco-
nomic Growth. Oxford Economic Papers 49, no. 2: 152–66.
McGuiness, Anne. 2007. Institutions and Total Factor Productivity Convergence. Working Paper 9/
RT/07. Dublin: Bank of Ireland.
Obstfeld, M. 2015. Trilemmas and Tradeoffs: Living with Financial Globalization. BIS Working
Paper 480. Basel: Bank for International Settlements. Available at www.bis.org/publ/
work480.htm.
OECD (Organization for Economic Cooperation and Development). 2016. Economic Survey of
Indonesia 2016. Paris. Available at www.oecd.org/eco/surveys/economic-survey-indonesia.
htm.
Rogoff, Kenneth. 2016. Debt Supercycle, Not Secular Stagnation. In Progress and Confusion: The
State of Macroeconomic Policy, ed. Olivier Blanchard, Raghuram Rajan, Kenneth Rogoff, and
Lawrence Summers. Cambridge, MA: MIT Press.
Sahay, R., V. Arora, T. Arvanitis, H. Faruqee, P. N’Diaye, and T. Mancini-Griffoli. 2014.
Emerging Market Volatility: Lessons from the Taper Tantrum. Staff Discussion Note
SDN/14/09. Washington: International Monetary Fund.
World Bank. 2017. Doing Business 2017: Indonesia. Washington. Available at www.doingbusi-
ness.org/~/media/wbg/doingbusiness/documents/profiles/country/idn.pdf.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 265

Comment
Muhamad Chatib Basri

Mitali Das’ excellent chapter dissects Indonesia’s economic growth since

the Asian financial crisis. Her analysis is commendable, and I agree with
her findings. To sharpen the analysis, I offer several comments.
First, Basri and Rahardja (2011) indicate that an outward-looking strat-
egy has been the source of Indonesia’s economic growth. Exports have a
large effect in supporting economic growth, despite being less stable than
domestic demand. Therefore, a strategy that safeguards a balance between
the domestic economy and an outward-looking strategy must be part of
the development strategy. An outward-looking strategy brings the risk of
growth volatility, however. An important question in the chapter is how to
adopt an outward-looking strategy without destabilizing economic growth.
Basri and Rahardja (2011) show a positive correlation between the
concentration of exports in products and markets with export volatility:
The higher the export concentration of markets and products, the higher
the level of export volatility. This finding suggests that export diversifica-
tion helps reduce export volatility. Medium- and high-tech export manu-
factures have a negative relationship with export volatility. The higher
concentration in medium- and high-tech manufacturing makes export
volatility lower. This finding supports the argument that overdependence
on a single export has a risk of higher export volatility. This argument is
particularly true for Indonesia. Export diversification is as important as
human capital accumulation and open trade regime.
Second, the chapter does not discuss the terms of trade. Indonesia is a
producer of energy and commodities such as coal and palm oil. The prices
of oil and these commodities are strongly correlated (World Bank 2015),
because when oil prices rise, the demand for nonoil energy substitutes rises.
Deterioration in the terms of trade will have a negative impact on household
consumption, exports, and economic growth. A decrease in commodity
and energy prices will also lead to a reduction in government revenue.
The government’s fiscal expansion capacity is thus limited, especially in
Indonesia, where the budget deficit is legally limited to 3 percent of GDP.
Improvement in the terms of trade has had a positive impact on
Indonesia’s economy. One of the explanations for the relatively strong-
growth of the economy between 2002 and 2012 was the commodity super-

Muhamad Chatib Basri, former minister of finance of Indonesia, is chairman of the Advisory Board of the Man-
diri Institute and chairman of Indonesia Infrastructure Finance. He teaches at the Department of Economics,
University of Indonesia, and is cofounder of Creco Research Institute, a Jakarta-based economic consulting firm.

266 SUSTAINING ECONOMIC GROWTH IN ASIA

cycle. The reverse was true after 2013, when the energy price began to
collapse.
If Indonesia wants to boost economic growth and reduce its boom-bust
cycle, the growth strategy should target manufacturing, especially exports.
Services exports also play an important role, because exports of tourism,
design, and workers’ remittances are likely to have direct links with private
consumption.
Third, I agree with Das that structural supply-side policies will help
Indonesia increase productivity. But one cannot ignore the importance
of demand-side issues, especially in the short term. Private consumption
accounts for about 50 percent of GDP in Indonesia. Improving purchasing
power is an important tool for boosting economic growth. The admin-
istration of President Joko Widodo has emphasized supply-side policies
by building infrastructure. It also engaged in deregulation. Despite these
measures, economic growth hovered around 5 percent in the last three
years and may stay there for a few years. The Bank of Indonesia cut interest
rates by 300 basis points between January 2016 and September 2017, but
annual credit growth remained stuck at 7 to 8 percent, much lower than
the 18–20 percent Indonesia enjoyed in the past. Undisbursed loans are
steadily rising. This suggests that in the short term, boosting demand is
very important.
If demand for goods and services is weak, what can the government do
to encourage firms to resume borrowing and expand production? If there
is no demand, what is the point in expanding?
Basri, Fitrania, and Zahro (2016) show that consumption encourages
investment after one quarter but that an increase in investment does not
significantly increase consumption. Supply-side policies and interest rate
cuts do not necessarily produce demand, at least in the short term.
What are the implications for policy? The government needs to launch
a fiscal stimulus that targets the lower middle class. It is very important
to stimulate domestic consumption through direct cash and conditional
cash transfer programs, as Indonesia did during the global financial crisis
and taper tantrum, because doing so boosts aggregate demand in the short
term.
Fourth, I support Das’ suggestion that Indonesia focus on increasing
productivity by improving the quality of human capital and opening up the
economy. In the short term, it can do so by attracting foreign direct invest-
ment and incentivizing foreign investors to provide training or build R&D
departments by offering tax deductions. The Indonesian government’s
policy of providing scholarship and sending the best students to the best
universities in the world should be continued.

HOW HAS INDONESIA FARED IN THE AGE OF SECULAR STAGNATION? 267

References
Basri, M. Chatib, Namira Fitrania, and Shirin Zahro. 2016. Causality between Investment and
Private Consumption in Indonesia. Jakarta: Creco Research.
Basri, M. Chatib, and Sjamsu Rahardja. 2011. Should Indonesia Say Goodbye to Strategy Fa-
cilitating Export? In Managing Openness: Trade and Outward-Oriented Growth After the Crisis,
ed. Mona Haddad and Ben Shepherd. Washington: World Bank.
World Bank. 2015. Commodity Markets Outlook, July. Washington.

268 SUSTAINING ECONOMIC GROWTH IN ASIA

 12
Twenty-Five Years of Global
Imbalances
MAURICE OBSTFELD

The modern floating exchange rate era that began in 1973 falls into two
stages. The first, which ended in the mid-1990s, was a period of adjustment
to the new international monetary regime. Financial markets underwent lib-
eralization, internationally as well as domestically, and international trade
expanded, while central bankers learned how to manage inflation in a world
of national fiat currencies. The period’s end is marked by the foundation
of the World Trade Organization (WTO) and the 1994–95 Mexican crisis,
which Michel Camdessus, the managing director of the International Mon-
etary Fund (IMF) at the time, labeled “the first 21st century crisis.”
The second 21st century crisis arrived even before the 21st century did,
in the form of the 1997–98 Asian crisis. The Asian crisis was notable in
that several of its victims succumbed despite the absence of garden-variety
macroeconomic imbalances—such as big public deficits—that IMF econo-
mists and others viewed as red flags. This facet of the crisis certainly influ-
enced academic thinking (witness the celebrated Kaminsky and Reinhart
1999 analysis of twin banking and currency crises), but it also pointed to
the dawn of a new, second stage of the post-1973 era.

Maurice Obstfeld is director of the Research Department at the International Monetary Fund. This chapter is
based on remarks made at the conference on “Prospects and Challenges for Sustained Growth in Asia” orga-
nized by the Bank of Korea, the Korean Ministry of Strategy and Finance, the International Monetary Fund,
and the Peterson Institute for International Economics in Seoul on September 7–8, 2017. The author is grate-
ful for suggestions and assistance from Gustavo Adler, Eugenio Cerutti, Luis Cubeddu, Thomas Helbling, and
Haonan Zhou. The views expressed in this chapter are those of the author and do not necessarily represent the
views of the IMF, its Executive Board, or its management. All errors and interpretations are the author’s alone.

269
 That stage, covering roughly the last 25 years, is characterized by hyper-
financialization, greater exchange rate flexibility on the part of many emerg-
ing-market economies, and a decisive shift of Asian growth leadership from
Japan to China, especially after China’s accession to the WTO. Subrama-
nian and Kessler (2013) characterize developments in international trade,
including rapidly expanding global value chains, as “hyperglobalization.”
The forces unleashed after the mid-1990s led to the global financial crisis of
2008–09, a crisis with long-lived repercussions that are still being felt.
One notable feature of the second period of floating was a significant
widening of global current account imbalances, which roughly tripled in
the decade after 1995 and remained higher, albeit at lower levels than the
peak they reached before the global financial crisis (figure 12.1). That global
imbalances had the potential to widen is not surprising, given financial
market development, including further financial opening, after the mid-
1990s. Controversy remains, however, over both the causes of these imbal-
ances and their potential causal role in the global financial crisis. On one
side is the view, expressed by US Treasury Secretary Henry Paulson as he was
leaving office in 2009, that global imbalances originating in the emerging
markets were the root cause of the global crisis.1 Others (such as Obstfeld
and Rogoff 2009) have argued that this view deflects too much blame from
other critical factors.
The debate raises at least four questions, which this chapter tries to
answer:
n Why did global imbalances expand after the mid-1990s?
n What circumstances and concomitant factors provide clues about the
origins of the global financial crisis?
n If one accepts that a monocausal story about the global financial crisis
based on global imbalances is inaccurate, how should one view the
potential threats from excessive global imbalances today?
n What policy implications follow?

Rise of Global Imbalances

A prominent feature of the expansion of global imbalances after the mid-
1990s was the growing US deficit (see figure 12.1). In a justly famous
speech, Bernanke (2005) argued for “locating the principal causes of the
US current account deficit outside the country’s borders.” In his telling, in

1. See Krishna Guha, “Paulson says crisis sown by imbalance,” Financial Times, January 1,
2009, www.ft.com/content/ff671f66-d838-11dd-bcc0-000077b07658.

270 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 12.1 Global current account balances and reserve purchases, 1990–2017

percent of world GDP

2.5
2.0
1.5
1.0
0.5
0
–0.5
–1.0
–1.5
–2.0
1990 1993 1996 1999 2002 2005 2008 2011 2014 2017

United States Other advanced economies

China Other EMDEs
Oil exporters

EMDEs = emerging-market and developing economies

Note: “Oil exporters” follows WEO classification and includes Norway. Bars represent regional current account balances,
the dotted line total reserve purchases.
Source: IMF, World Economic Outlook.

TWENTY-FIVE YEARS OF GLOBAL IMBALANCES

271
 Figure 12.2 Real 10-year bond yield in advanced economies,
1990–2018

percent
8
United States
6 Japan
United Kingdom
Canada
4 Euro area

–2
Asian crisis Global financial crisis

1990 1993 1996 1999 2002 2005 2008 2011 2014 2017

Sources: Consensus Forecasts; Haver Analytics. For the euro area, the period
before 2003 is based on German data.

a global capital market equilibrium, higher net saving by emerging-market

and developing economies (EMDEs)—precautionary saving by Asian emerg-
ing economies after their crisis, bigger surpluses of oil exporters as the price
of oil rose—needed to be matched by bigger US deficits. What mechanism
induced the United States to save less and invest more? Initially responsible
was a run-up in equity prices. After they crashed, the main driver became a
fall in real interest rates that, among other effects, fueled a housing price
and residential investment boom. Real 10-year Treasury rates rose after the
Asian crisis, returning to their long-term decline as recession set in during
2001 (figure 12.2). The pattern was similar for the short-term “natural”
policy rates (r*), as calculated by Holston, Laubach, and Williams (2017)
(see figure 12.3).
Figure 12.4 is another way to visualize the dramatic widening of the US
external deficit after the mid-1990s. Where are the counterpart widening
surpluses? The top panel shows that deficits of non-oil-exporting EMDEs
(including China) did start to rise as the Asian crisis erupted, but the extent
of the increase was dwarfed by the rise in the US deficit. More important was
the increase in oil exporters’ surpluses. In Asia (bottom panel), surpluses
of East Asian economies other than China and Japan rose after the Asian
crisis. China’s surplus began to rise later, in the mid-2000s, reaching about
0.6 percent of global GDP in 2008.
There were multiple counterpart surpluses to the US deficit; one of
the biggest came from oil exporters. Their surpluses were driven by steeply

272 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 12.3 Natural rate of interest (r*) in advanced economies,
1990–2016

percent
4

–1
Asian crisis Global financial crisis

1990 1993 1996 1999 2002 2005 2008 2011 2014

United States
United Kingdom
Canada
Euro area

Source: Holston, Laubach, and Williams (2017).

rising global oil prices, however (figure 12.5).2 As US real activity, financial
conditions, and domestic oil production are critical determinants of con-
ditions in the world oil market, it seems implausible that changes in oil
prices emanated entirely from outside US borders to influence its current
account. Instead, the US deficit reflected domestic as well as global forces—
global forces coming from other advanced economies, not just EMDEs—all
of which helped set the stage for the global financial crisis.

Background to the Global Financial Crisis

Common factors drove oil prices, oil surpluses, and bigger current account
deficits in several advanced economies. They included very loose global fi-
nancial conditions, enabled by generally accommodative monetary policies,
as well as financial deregulation and innovation, and a global reach for yield
and safety that contributed to widespread housing market booms.
One component of this constellation, though likely not the most im-
portant one, was the accumulation of foreign reserves by EMDEs, which
Bernanke (2005) noted (figure 12.1). During 1998–2008, intervention

2. In the early 2000s, measured world surpluses surged above world deficits; a substantial
global discrepancy—a “missing deficit”—emerged. Missing deficits tend to be positively corre-
lated with oil prices. The discrepancy could be related to some countries understating the cost
of oil imports.

TWENTY-FIVE YEARS OF GLOBAL IMBALANCES 273

 Figure 12.4 Current account balances, 1991–2017

a. World
percent of world GDP
1.0

0.5

–0.5

–1.0 United States

Other advanced
–1.5 economies
Oil exporters
–2.0 Other EMDEs
Asian crisis Global financial crisis

1991 1995 1999 2003 2007 2011 2015

b. East Asia
percent of world GDP
1.0

0.6

0.2

–0.2 China
Japan
–0.6 Other East Asian
advanced economies
Other East Asian
–1.0 EMDEs
Asian crisis Global financial crisis

1991 1995 1999 2003 2007 2011 2015

EMDEs = emerging-market and developing economies

Source: IMF, World Economic Outlook.

tended to be associated with wider surpluses for countries purchasing

foreign exchange. But the US housing bubble also drew fuel from European
banks’ purchases of asset-backed securities, as Bernanke et al. (2011) and
Bayoumi (2017) document. Perceiving low sovereign and no currency risk
within a permissive regulatory environment, banks in France, Germany,
and other European countries recycled global funding into peripheral econ-
omies in the euro area, such as Spain, Ireland, Portugal, and Greece, financ-
ing housing or sovereign debt bubbles and big external deficits (Hale and
Obstfeld 2016). Similar dynamics played out in the Baltics.

274 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 12.5 World oil price and oil exporters' current account,
1991–2017

current account balance Brent crude oil

(percent of GDP) (dollars per barrel)
1.0 120
Oil exporters’ current
0.8 account 100
Brent crude oil price
0.6 (right axis) 80
0.4
60
0.2
40
0

–0.2 20

–0.4 0
97

11

13

15
91

93

95

99

17
0

0
0
0

0
20

20

20

20
19

20
19
19

19

19

20
20
20

20

Source: IMF, World Economic Outlook; Haver Analytics; IMF Staff

calculations.

A symptom of global financial ease during the 2000s, coupled with

ongoing deregulation and innovation, was an explosion of two-way capital
flows across borders. Figure 12.6 shows the pattern of gross international
financing corresponding to the net international financing needs illustrated
in figure 12.1. Starting in the mid-1990s, gross flows began to expand mark-
edly, reaching unprecedented proportions on the eve of the global financial
crisis. Tellingly, the vertical axis scale needed to chart gross flows in figure
12.6 is a full order of magnitude greater than the scale needed in figure 12.1:
The bare minimum capital flows needed to finance current account imbal-
ances would have been only a tenth of the flows that actually took place.
In theory, gross two-way flows fulfill economic functions beyond the fi-
nancing of current account imbalances, notably, asset swaps that enable in-
ternational portfolio diversification and risk sharing. No known economic
or financial model can easily rationalize gross flows that were persistently
such a big share of global GDP in the run-up to the crisis, however.
Much of the asset churning was likely driven by tax avoidance strategies
or regulatory arbitrage, exemplified by the surge in European purchases
of US mortgage-related assets (see, e.g., Acharya and Schnabl 2010). These
developments were a symptom of the financial distortions that ultimately
helped power the US current account deficit.
Following their sharp expansion starting in the mid-1990s, gross ex-
ternal positions leveled off after the finanical crisis, at least in the aggre-
gate (see figure 12.7). An increasing fraction of gross external positions now
seems to represent foreign direct investment (FDI), although as Lane and

TWENTY-FIVE YEARS OF GLOBAL IMBALANCES 275

276
Figure 12.6 Global gross financial flows, 1990–2017

percent of world GDP

25

20

15

10

–5

–10

–15

SUSTAINING ECONOMIC GROWTH IN ASIA

–20

–25
1990 1993 1996 1999 2002 2005 2008 2011 2014 2017

Other investment inflow Portfolio investment outflow

Other investment outflow FDI inflow
Portfolio investment inflow FDI outflow

FDI = foreign direct investment

Source: IMF, Balance of Payments Statistics.
 Figure 12.7 Global investment liabilities stock, 1990–2016

percent of world GDP

250
FDI
200 Other investment
Portfolio debt
150 Portfolio equity

100

50

0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

FDI = foreign direct investment

Source: IMF, International Investment Position, World Economic Outlook.

TWENTY-FIVE YEARS OF GLOBAL IMBALANCES

277
Miles-Ferretti (2018) document, much of this so-called FDI reflects not risk
sharing or greenfield investment but claims on offshore financial centers,
driven by tax minimization strategies.

Risks from Excess Global Imbalances

Twenty-first century crises have largely been balance sheet crises, driven
more by predetermined vulnerabilities (currency and maturity risk) than
by the risk that the flow of financing for the current account deficit dries
up. The very large gross flows in figure 12.6 represent potentially fragile
balance sheet positions that could threaten financial stability, with severe
macroeconomic consequences.
Why, then, worry about even outsized current account imbalances? One
reason is that an excessive flow deficit may be one symptom of an under-
lying buildup of stock vulnerabilities. Generally easy financial conditions
(including lax or badly designed regulation) will tend to promote leverage,
which in turn facilitates divergence between spending and income. The
eventual adjustment process can be costlier when it follows bigger imbal-
ances. Lane and Milesi-Ferretti (2015), for example, show how countries
with bigger current account deficits before the global financial crisis tended
to suffer greater demand compression when the crisis struck.
Another reason to worry about excess imbalances, especially if they are
persistent, is that they could imply widening international disparities in
net external wealth. Although the growth of gross external positions has
leveled off for now, the net international investment positions of inter-
national creditors and debtors have grown increasingly divergent (figure
12.8). Absent future changes in asset valuations, the IMF projects that this
divergence trend will continue (IMF 2017).
The tendency for debtors to go ever farther into debt (while credi-
tors accumulate those debts as assets) has been a salient phenomenon,
as figure 12.9 illustrates for the sample of countries covered in the IMF’s
annual External Sector Reports. Since 2010 countries’ cumulative current
account balances have been strongly positively correlated with initial net
foreign assets. That propensity, however, leads to a sustainability problem.
Eventually, debtor countries will have to reduce spending to reflect their
intertemporal budget constraints; the longer the adjustment is postponed,
the more likely it is that the process will be abrupt and disruptive, creating
a global deflationary impulse unless creditor countries just as abruptly
decide to spend more.
Low real interest rates encourage debtors to continue borrowing and
creditors to accumulate wealth more assiduously as they seek to ensure
adequate resources for the future. Higher equilibrium interest rates could

278 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 12.8 Net international investment positions, 2002–17

percent of world GDP

20

10

–10

–20

–30
2002 2005 2008 2011 2014 2017

Oil exporters United Kingdom

Other surplus Deficit emerging-market economies
Surplus advanced economies Advanced-economy commodity
Japan exporters
Germany/Netherlands Other deficit countries
China East Asia (other)
United States Discrepancy

Surplus advanced economies = Korea, Hong Kong, Singapore, Sweden, Switzerland,

and Taiwan; Advanced-economy commodity exporters: Australia, Canada, and New
Zealand; Deficit emerging-market economies = Brazil, India, Indonesia, Mexico,
South Africa, and Turkey; Oil exporters = World Economic Outlook definition plus
Norway.
Source: IMF, World Economic Outlook; IMF Staff calculations.

TWENTY-FIVE YEARS OF GLOBAL IMBALANCES 279

create problems if they rise more than underlying economic growth rates.
In integrated capital markets, real interest rates depend on global, not
just national, growth. IMF projections of growth prospects for advanced
economies are relatively subdued. As the External Sector Report (IMF 2017)
points out, however, big deficits (and surpluses) have also increasingly
become concentrated in precisely these slower-growing advanced debtors
(and creditors). Down the road, therefore, higher real interest rates could
spell trouble for some advanced debtors if the levels of those rates reflect
primarily the better growth prospects in lower-income economies.
A possible mitigating factor could be the effect identified by Gourinchas
and Rey (2007) for the United States, according to which an unexpected fall
in net exports triggers an unexpected capital gain on the net international
investment position, thereby mitigating the negative effect of the net export
innovation. When the World Economic Outlook examined the generality of
this effect more than a decade ago (IMF 2005), it detected some evidence of
the Gourinchas-Rey effect for advanced economies but not for EMDEs. The
finding was unsurprising in view of the pervasive foreign currency denomi-
nation of EMDEs’ foreign liabilities and the relatively early stage of their
integration into world capital markets.
Currency mismatch has declined since the early 2000s, however (Béné-
trix, Lane, and Shambaugh 2015), and it appears that the Gourinchas-Rey
effect now applies more broadly. This stabilizing force may owe more to do-
mestic asset price developments than exchange rates. Whatever its source,
estimates suggest that it provides only a partial offset to trade balance de-
velopments (Adler and Garcia-Macia 2018). Figures 12.9 and 12.10 (which
can be compared) illustrate how the Gourinchas-Rey effect has muted but
not eliminated the divergent behavior of net international investment posi-
tions. The tendency of current accounts to drive divergence in net foreign
assets shown in figure 12.10 remains problematic.
A third danger from big and persistent global imbalances, one that is
currently prominent, is that they may spark trade warfare as deficit coun-
tries vainly try to counter macroeconomic forces with import barriers. From
the Asian crisis through the early 2010s, global current account surpluses
were closely aligned with global foreign exchange purchases (figure 12.1).
Indeed, several economies, notably in emerging East Asia, intervened to
maintain their currencies at undervalued levels, as did some important oil
exporters. Charges of currency manipulation and trade tensions resulted.3

3. In the cases of oil exporters with exchange rate pegs (e.g., Saudi Arabia), causality clearly
ran from the price of oil through the current account to reserve accumulation. In Asian econ-
omies—notably China, with its capital control regime—intervention likely had some causal
impact on the current account, to different degrees in different economies.

280 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 12.9 Current account behavior has led net international
investment positions to diverge

change in international investment positions (2017–2010)

excluding valuation changes, percent of GDP
100
NOR
80 CHE
NLD
60 DNK
DEU

MYS
SWE
40 KORRUS
THA y = 0.4108x + 8.2883
ISR JPN R2 = 0.4173
20 HUN
IRL
PHL
AUT CHN

ESP ITA BEL

0 CZE PAK GTMFRA FIN
ARG
IDN
MEX IND
CHL
PRT POL USA
NZL PERCAN
–20 AUS BRACRI
ZAF
COL
GBR
MAR
TUR
GRC
–40
TUN
–60

–80
–150 –100 –50 0 50 100 150 200
net international investment positions
(2010, percent of GDP)

Note: Vertical axis measures cumulated nominal current accounts between 2010 and 2017,
adjusted to reflect the path of nominal income growth.
Source: IMF, World Economic Outlook; IMF Staff calculations.

In more recent years, however, global imbalances owe little to foreign

exchange intervention by EMDEs, as IMF (2017) observes and figure 12.1
shows. This change is consistent with the greater concentration of imbal-
ances within the advanced economy group noted above. Although discus-
sion of currency manipulation is much more muted than it was a few years
ago, it can be expected to return to the fore if the US dollar strengthens
in response to Federal Reserve tightening and procyclical fiscal expansion
in the United States. Because of those US macroeconomic prospects, the
IMF now projects a bigger US current account deficit in coming years,
notwithstanding the US administration’s avowed aim of trimming deficits
through trade policies. That tension will doubtless be a source of future
trade frictions between the United States and the rest of the world. Those
frictions may in turn take a toll on global economic growth.

TWENTY-FIVE YEARS OF GLOBAL IMBALANCES 281

Figure 12.10 Valuation effects have moderated but not eliminated
the tendency for net international investment positions
to diverge

change in net international investment positions (2017–2010)

including valuation changes, percent of GDP
150
NOR

100
NLD
DNK y = 0.1697x + 8.2267
50 HUN DEU R2 = 0.0666
TUN CZE ZAF KORRUS ISR
ESP POL ITA AUT
PRT GRC BRA THA SWE
NZL AUS IDN ARG CHN
MEX PHLGBR
FRA
IND
MYS JPN CHE
0 TUR PAK GTM CHL
MAR PER FIN BEL
COL
CRI USA

–50

–100 IRL

–150
–150 –100 –50 0 50 100 150 200
net international investment positions
(2010, percent of GDP)

Source: IMF, World Economic Outlook; IMF Staff calculations.

Policy Implications
This analysis yields four main implications for policy:
n Excess global imbalances usually reflect global forces and multiple
distortions in many countries, including diverse financial sector dis-
tortions. Monocausal explanations rarely apply. Reducing global im-
balances should therefore be a collective effort based on a shared appre-
ciation of the roles individual countries need to play. To promote that
objective, the IMF monitors external excess imbalances annually and
makes its findings public through the External Sector Report.
n Notwithstanding the need for collective action, excess surplus countries
still face little that would force them to adjust (outside of the threat
of protectionist responses). In contrast, most deficit countries face the
risk that lenders will withdraw. The trend of diverging net international
investment positions thus looks unsustainable. Its end is likely to origi-
nate with external debtors.
n That endgame could be messy. Gross capital flows far exceed the
minimal needs of current account financing. These flows can give rise

282 SUSTAINING ECONOMIC GROWTH IN ASIA

 to balance sheet vulnerabilities, underlining the need for continuing
international financial cooperation (within the Basel Committee, the
Financial Stability Board, and other international forums) to address
cross-border financial stability risks. Similar cooperation to address so-
cially costly tax minimization strategies and money laundering is richly
warranted.
n Economic theory suggests that trade restrictions will do little to alter
global current account imbalances, which are primarily macroeconom-
ic phenomena. But protectionism can do great harm to global growth.
Stronger multilateral dispute resolution within the rules-based system
is the right way to deal with the range of policies that distort trade and
tilt the playing field.

References
Acharya, V., and P. Schnabl. 2010. Do Global Banks Spread Global Imbalances? Asset-Backed
Commercial Paper During the Financial Crisis of 2007–09. IMF Economic Review 58, no.
1: 37–73.
Adler, G., and D. Garcia-Macia. 2018. The Stabilizing Role of Net Foreign Asset Returns. IMF Work-
ing Paper WP/18/79 (March). Washington: International Monetary Fund.
Bayoumi, T. 2017. Unfinished Business: The Unexplored Causes of the Financial Crisis and the Lessons
Yet to Be Learned. New Haven, CT: Yale University Press.
Bénétrix, A., P. Lane, and J. Shambaugh. 2015. International Currency Exposures, Valuation
Effects, and the Global Financial Crisis. Journal of International Economics 96, no. S1: S98–
S109.
Bernanke, B. 2005. The Global Saving Glut and the U.S. Current Account Deficit. Sandridge
Lecture to the Virginia Association of Economists, Richmond, VA, March 10.
Bernanke, B., C. Bertaut, L. DeMarco, and S. Kamin. 2011. International Capital Flows and
the Returns to Safe Assets in the United States, 2003–2007. Financial Stability Review
(Banque de France) 15: 13–26.
Gourinchas, P.-O., and H. Rey. 2007. International Financial Adjustment. Journal of Political
Economy 114, no. 4: 665–703.
Hale, G., and M. Obstfeld. 2016. The Euro and the Geography of International Debt Flows.
Journal of the European Economic Association 14, no. 1: 115–44.
Holston, K., T. Laubach, and J. Williams. 2017. Measuring the Natural Rate of Interest: Inter-
national Trends and Determinants. Journal of International Economics 108, no. S1: S59–S75.
IMF (International Monetary Fund). 2005. Globalization and External Imbalances. In World
Economic Outlook (April). Washington.
IMF (International Monetary Fund). 2017. External Sector Report (July). Washington.
Kaminsky, G., and C. Reinhart. 1999. The Twin Crises: The Causes of Banking and Balance-
of-Payments Problems. American Economic Review 89, no. 3: 473–500.
Lane, P. and G. Milesi-Ferretti. 2015. Global Imbalances and External Adjustment After the
Crisis. In Global Liquidity, Spillovers to Emerging Markets and Policy Responses, ed. C. Raddatz,
D. Saravia, and J. Ventura. Santiago: Central Bank of Chile.
Lane, P., and G. Milesi-Ferretti. 2018. The External Wealth of Nations Revisited: International
Financial Integration in the Aftermath of the Global Financial Crisis. IMF Economic Review
66, no. 1: 189–222.

TWENTY-FIVE YEARS OF GLOBAL IMBALANCES 283

Obstfeld, M., and K. Rogoff. 2009. Global Imbalances and the Financial Crisis: Products of
Common Causes. In Asia and the Global Financial Crisis, ed. R. Glick and M. Spiegel, 131–
172. Federal Reserve Bank of San Francisco.
Subramanian, Arvind, and Martin Kessler. 2013. The Hyperglobalization of Trade and Its Future.
PIIE Working Paper 13-6 (July). Washington: Peterson Institute for International Eco-
nomics.

284 SUSTAINING ECONOMIC GROWTH IN ASIA

 13
Monetary and Exchange Rate
Policies for Sustained Growth
in Asia
JOSEPH E. GAGNON AND PHILIP TURNER

The more advanced economies in Asia are experiencing slower growth rates.
Structural reforms are the most important policies for keeping growth rates
up, but this chapter takes the growth slowdown as given and focuses on
implications for monetary policy. The key policy implication is the impor-
tance of keeping core inflation at or above 2 percent to avoid prolonged
periods of economic slack.
Flexible inflation targeting has provided a successful and adaptable
framework for monetary policy worldwide.1 It is hard to overstate the im-
portance for monetary policy of keeping inflation within the central bank’s
policy mandate. Such mandates typically specify some target for average
inflation in the medium term, either a single number or some range. In
our view, such a target should be no less than, and possibly greater than, 2
percent.
Forecasts at the time of writing (January 2018) suggest strong global
growth in 2018. The threat of deflation has faded in most countries. But
with inflation expectations still comparatively low, monetary policy should
react promptly to any significant negative shocks to growth or to inter-

Joseph E. Gagnon is senior fellow at the Peterson Institute for International Economics. Philip Turner is visiting
lecturer at the University of Basel and former deputy head of the Monetary and Economic Department at the
Bank for International Settlements. This chapter reflects the views of the authors and should not be interpreted
as reflecting the views of the Institute or the Bank or any members of the boards or staffs of those institutions.
1. Graeme Wheeler, “Reflections on 25 years of inflation targeting,” speech at Reserve Bank
of New Zealand and International Journal of Central Banking conference, Wellington,
December 1, 2014.

285
national financial markets (e.g., a new taper tantrum). In Thailand, core
inflation remains well below 2 percent and there is a case for additional
monetary ease already.
Japan failed to keep inflation above zero after a severe financial crisis
and suffered two decades of excess unemployment and forgone output.
The longer inflation is allowed to remain below target, the harder it is to
raise inflation to target. When inflation expectations settle at low levels,
central banks have less scope to use conventional monetary policy to stabi-
lize cyclical fluctuations. Central banks, however, can still expand their
balance sheets (so-called unconventional policy) when the conventional
policy rate is near zero.
We rebut three possible criticisms of our advice.
First, it is argued that monetary policy has only a weak impact on infla-
tion as reflected in declining estimates of the slope of the Phillips curve. We
suggest that the Phillips curve slope is nonlinear in both the output gap and
the level of inflation. When inflation is close to zero, a negative output gap
has very little effect on inflation because of downward rigidities in nominal
wages and prices. But a positive output gap is expected to have a significant
effect, and this effect is likely to grow as the gap becomes larger.
Second, it is argued that central banks should stick to setting the over-
night rate and should avoid the so-called unconventional balance sheet
policies of the kind implemented by the Federal Reserve, European Central
Bank (ECB), Bank of Japan (BOJ), and Bank of England. This view is unhis-
torical. Central banks have used their balance sheets to advance their objec-
tives since their inception more than 300 years ago. In Asia, the accumula-
tion of foreign exchange reserves and related policies to stabilize financial
markets and control any excess liquidity in domestic banks were major
planks of monetary policy in the years during and after the Asian financial
crisis. What is new is that the substantial development of domestic financial
markets in emerging-market economies has widened the possibilities for
balance sheet policies. Because bond markets have become more important
in monetary policy transmission in Asia, and because the liquidity of such
bond markets can be especially fragile when global markets are disturbed,
balance sheet policies should be on the policy agenda. This would be rein-
forced if weak growth and low inflation were to push the policy rate to zero.
Third, it is argued that easy monetary policy encourages risky behavior
in financial markets. We argue that the evidence for such an effect is very
weak. Moreover, ultra-low inflation and persistent negative output gaps
themselves raise risks to financial stability. Prudential regulatory policies
are far more potent at preserving financial stability than monetary policy.
Regulatory policy includes tools such as bank capital and liquidity require-
ments; rules on currency and maturity mismatches in banks; limits to

286 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 13.1 Growth rate of real per capita GDP, 1985–2016,
five-year moving average (percent)

Australia Bangladesh China

10
5
0
−5
Hong Kong India Indonesia
10
5
0
−5
Japan Korea Malaysia
10
5
0
−5

New Zealand Philippines Singapore

10
5
0
−5

Taiwan Thailand Vietnam

10
5
0
−5
1985 1995 2005 2015 1985 1995 2005 2015 1985 1995 2005 2015

Source: IMF World Economic Outlook database.

interest rate exposures; and enhanced stress tests to make sure the balance
sheets of financial intermediaries are resilient to any eventual tightening of
monetary policy. New macroprudential policy tools (such as loan-to-value
and debt-to-income ratios for house mortgages) give the central bank new
ways of limiting risks to financial stability arising from low interest rates.

Macroeconomic Developments in Asia

Figure 13.1 displays five-year moving averages of the growth rate of real per
capita GDP for the 15 largest economies (based on 2016 GDP at market
exchange rates) in the Asia-Pacific region. Many economies appear to be
growing more slowly over time. However, for some economies there is no
clear trend in the growth rate, and for a few economies growth seems to be
increasing.
MONETARY AND EXCHANGE RATE POLICIES 287
 Figure 13.2 CPI inflation rates in Asia, 1985–2016
(percent per year)

Australia Bangladesh China

10
10
5 10
5
5
0 0 0

Hong Kong India Indonesia

10
10 10
5
5 5
0
0 0
Japan Korea Malaysia
10 10 10

5 5 5
0
0 0

New Zealand Philippines Singapore

10
10 5
10
5 5
0
0 0

Taiwan Thailand Vietnam

10 10

5 5 10
5
0 0 0

1985 1995 2005 2015 1985 1995 2005 2015 1985 1995 2005 2015

CPI = consumer price index

Note: Observations above 20 percent have been dropped.
Source: IMF, World Economic Outlook database.

The main policy option to raise an economy’s growth rate regardless

of its relative income level is structural reform that opens up protected
sectors to competition and encourages investments in human and physical
capital and research and development (R&D). However, structural reform
is the topic of other chapters in this volume. This chapter focuses on impli-
cations for monetary policy, which can help to avoid prolonged underem-
ployment of resources and to sustain investment. Monetary policy can
thus help an economy achieve its long-run potential growth rate.
Figure 13.2 displays inflation rates in the Asia-15 economies. In every
case, inflation in 2016 was below its historic average, often by a consider-
able amount. In 2016, inflation was below 5 percent in 14 of the 15 econo-

288 SUSTAINING ECONOMIC GROWTH IN ASIA

mies; below 3 percent in 12 economies; below 2 percent in 8 economies;
below 1 percent in 5 economies; and below 0 percent in 2 economies.
In most of the economies with inflation below 1 percent in 2016—
Singapore, Japan, Thailand, New Zealand, and Korea—GDP growth slowed
markedly over time. In these economies, there was probably a gap between
actual and potential GDP growth at some point either before or when
inflation was declining. As we discuss in the next section, a key priority for
monetary policy in Asia should be keeping inflation from falling persis-
tently below 2 percent and possibly even targeting a rate slightly higher
than 2 percent.

Dangers of Ultra-Low Inflation

It is a widely acknowledged human failing that we learn more readily from
our own mistakes than from the mistakes of others. A similar truth holds
at the level of national economic policies. The United States and the euro
area did not fully absorb the lessons of real estate bubbles and banking
crises in the Nordic countries, the United Kingdom, and Japan in the early
1990s and in developing Asia in the late 1990s. This failure doomed them
to suffer their own crises in 2008.
In our view, many Asian policymakers are not sufficiently concerned
about the likely persistence and economic costs of ultra-low inflation. All
their lives, inflation has been a problem only when it was too high, and they
take satisfaction in having conquered it. Yet both economic research and
the experiences of some advanced economies suggest that inflation can get
stuck below target for extended periods and that allowing ultra-low infla-
tion to persist has serious costs.

Lessons from Japan

The bursting of the equity and real estate price bubbles in 1990–91 devas-
tated the Japanese financial system but it was not until 2000 that the
authorities developed systematic policies to deal with insolvent banks.2
Since 1993, nominal GDP growth and core consumer price index (CPI)
inflation in Japan have fluctuated around zero3 (see figure 13.3). Real
GDP growth slowed after 1993. The severe impairment of financial inter-

2. Nakaso (2001) provides an authoritative account of how it took years for the authorities
in Japan to develop policies to deal with this financial crisis.
3. In the four quarters following each increase in the consumption tax, the inflation rate
shown in figure 13.3 was adjusted downward by the change in inflation in the first quarter
minus the change in inflation in the fifth quarter divided by two. The consumption tax was
increased in 1997Q2 and 2014Q2.

MONETARY AND EXCHANGE RATE POLICIES 289

Figure 13.3 Macroeconomic developments in Japan, 1954Q1–2017Q1

Growth rate of nominal GDP Inflation rate, CPI excluding food and energy
4-quarter change, percent 4-quarter change, percent
20
20
15
10 10
5
0
0
−10 −5
1955 1975 1995 2015 1955 1975 1995 2015

Unemployment rate Output gap (Bank of Japan)

percent of labor force percent of potential GDP

5 5.0
4 2.5
3 0
2 −2.5
1
−5.0
0
1955 1975 1995 2015 1955 1975 1995 2015

CPI = consumer price index

Note: Inflation has been adjusted to remove the effects of increases in the consumption tax.
Sources: Bank of Japan, Haver Analytics, and authors’ calculations.

mediation during much of the 1990s, slower growth of the working-age

population, and convergence toward per capita income levels of the most
advanced economies would have slowed real growth regardless of mone-
tary policy or inflation.
However, ultra-low inflation makes it harder to reduce the real value
of debts, and banks find it difficult to improve their balance sheets when
nominal GDP is stagnant. In any event, the level of output does appear to
have been below potential on average since the advent of ultra-low infla-
tion. The BOJ’s estimate of the output gap has fallen from an average of
1.7 percent before 1993 to –1.0 percent since 1993 (see figure 13.3). This
estimate probably understates the economy’s true underperformance. The
unemployment rate has risen from an average of 1.9 percent before 1993
to 4.1 percent since then. Even if the natural rate of unemployment is now
close to 3 percent, as appears likely, Okun’s law suggests that an average
excess unemployment rate of 1 percent implies a shortfall in GDP of 2
percent.4 Cumulated over more than 20 years, this shortfall represents an

4. Note that the BOJ estimates that output is slightly above potential in 2017 with an unem-
ployment rate slightly below 3 percent.

290 SUSTAINING ECONOMIC GROWTH IN ASIA

enormous loss of goods and services that could have been consumed or
invested in Japan.
In 1993—well after the bubbles had burst—the BOJ’s policy interest
rate was above 3 percent. Adam Posen (1998) provided one of the earliest
critiques of Japanese macroeconomic policy after the bubbles. He argued
for coordinated monetary and fiscal expansion to return output to poten-
tial and avoid deflation. Soon thereafter, the BOJ began to take more
aggressive monetary policy measures (the zero interest rate policy in 1999
and quantitative targeting in 2001) but it reversed direction before defla-
tion was fully conquered. A faster, stronger, and more sustained response
to deflation in the early 1990s would have been warranted and might have
maintained inflation near 2 percent in the subsequent decades (Ahearne et
al. 2002). Twenty years of zero inflation, however, have changed the expecta-
tions of firms and workers in Japan. Raising inflation back to 2 percent is
much harder now, as evidenced by the limited success of the BOJ’s massive
quantitative easing policy since 2013 (Ball et al. 2016, 48–50). Surveys of
professional forecasters reveal that long-term inflation expectations in
Japan were very slow to decline and remained above actual inflation for
most of the 1990s and 2000s.5 Getting expectations to rise will also take
time and is likely to require a sustained increase in actual inflation. A key
lesson for Asian economies is not to allow inflation and inflation expecta-
tions to become entrenched below 2 percent.

Economic Costs of Ultra-low Inflation

One important cost of ultra-low inflation is that relative prices become
more difficult to change. When inflation is positive and prices and wages
are rising on average, firms can adjust relative prices in response to shifts in
tastes, technology, and competitive conditions by increasing some prices at
a faster rate and others at a slower rate. When overall inflation is zero, ad-
justments in relative prices require firms to reduce some prices in nominal
terms. If there is resistance to cutting prices, the economy needs to run below
its full-employment level to force some wages and prices down to keep the
average inflation rate near zero (Akerlof, Dickens, and Perry 1996; Benigno
and Ricci 2011). Studies find evidence of such downward rigidities in wages
in many countries (Dickens et al. 2007; Fallick, Lettau, and Wascher 2016).
Once nominal wages begin to decline, the fear of a deflationary spiral can
lead households and firms to cut spending, adding further downward pres-
sure to an already weak economy.
Akerlof, Dickens, and Perry estimated that an overall inflation rate of

5. Consensus Forecasts, various issues.

MONETARY AND EXCHANGE RATE POLICIES 291

at least 2 percent is needed to avoid the bulk of the economic cost of down-
ward nominal wage rigidity.6 In comments published with the Akerlof,
Dickens, and Perry paper, Robert Gordon and Greg Mankiw suggested that
downward nominal wage rigidity might become less apparent as people
become used to low inflation or in the event of severe economic distress.
However, Fallick, Lettau, and Wascher show that these conjectures did not
prove correct in the aftermath of the Great Recession in the United States.
Another reason to prefer a positive inflation rate is that price indexes
do not fully control for quality improvements and the welfare benefits of
new goods. These omissions bias published inflation measures up by as
much as 1 percent per year, so that a reported inflation rate of 1 percent
may reflect constant true prices (Bank of Canada 2013).

Ultra-low Inflation, Interest Rates, and Monetary Policy

Because economic equilibrium depends on the real rate of interest over
both short and long horizons, an environment of low expected inflation
must be accompanied by low nominal rates of interest. Moreover, recent
studies document a decline in the equilibrium real rate of interest in many
advanced economies (Holston, Laubach, and Williams 2016; Williams
2017). Figure 13.4 shows that the real short-term interest rate has trended
down in many of the Asia-15 economies. In 2016 it was below its historical
average in these economies except India and Vietnam.
In most economies in the Asia-Pacific region with active and open
bond markets, long-term interest rates have declined since 2000, a period in
which long-term inflation expectations are likely to have been fairly stable.
As figure 13.5 shows, this has mirrored the movement in average long-term
rates in advanced economies. Since the mid-2000s local-currency bond
markets of many Asian emerging-market economies have thus become
part of this expanding global market (Obstfeld 2015). But note that the
long-term rates of emerging Asian economies on average rose more sharply
than rates in advanced economies in the two periods of bond market turbu-
lence—in 2008 and during and after the 2013 taper tantrum.
As King and Low (2014) have concluded, given the high correlation
between bond yields of different countries (emerging as well as advanced
economies), “it therefore is quite reasonable to talk about a ‘world’ interest
rate.” The real long-term rate has been declining for about 30 years.
Observations for the most recent years using a principal-components esti-
mate based on 10-year government bonds of three major markets show
that the world real long-term interest rate has been hovering around zero

6. Wyplosz (2001) also argues for optimal rates of inflation above 2 percent in major Euro-
pean economies.

292 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 13.4 Real short-term interest rates in Asia, 1985–2016
(percent)

Australia Bangladesh China

10 10 10

0 0 0

−10
−10 −10
Hong Kong India Indonesia
10 10
10
0 0
0
−10 −10 −10
Japan Korea Malaysia
10 10 10

0 0 0

−10 −10 −10

New Zealand Philippines Singapore

10 10
10
0 0 0

−10 −10 −10

Taiwan Thailand Vietnam

10 10 10
0 0 0
−10
−10 −10
1985 1995 2005 2015 1985 1995 2005 2015 1985 1995 2005 2015

Note: Real interest rate is three-month Treasury bill rate (or closest equ-
ivalent) minus current consumer price index (CPI) inflation rate.
Sources: Haver Analytics and IMF International Financial Statistics and
World Economic Outlook databases.

since mid-2011 (graph 2 in Hördahl, Sobrun, and Turner 2016). Rachel and
Smith (2015) attribute about two-thirds of the long-run decline to secular
factors shaping desired saving and investment rates in the global economy.
They argue that the likely persistence of these factors suggests that the
underlying global neutral real rate will settle at around 1 percent in the
medium to long run.7 If this prognosis is correct, central banks will again
grapple with the zero lower bound for the policy rate.

7. Laubach and Williams (2015) also estimate a long-run equilibrium real rate of around 1
percent for the United States. They attribute the decline in the long-run equilibrium real rate
mainly to the decline in the growth rate of potential output.

MONETARY AND EXCHANGE RATE POLICIES 293

 Figure 13.5 Long-term bond yields: Emerging Asia and selected
advanced economies, January 2001 through July 2017

yield

Emerging Asia
2 United States
Advanced economies
excluding United States
and Japan
0

2002m1 2007m1 2012m1 2017m1

Note: Vertical lines denote the collapse of Lehman Brothers in September 2008
and the taper tantrum in July 2013. Emerging Asia is the unweighted average of
Indonesia, Korea, Malaysia, Philippines, Taiwan, and Thailand. Advanced
economies are the unweighted average of Australia, Canada, euro area, New
Zealand, Sweden, and United Kingdom.
Sources: Bank for International Settlements and Haver Analytics.

The decline in equilibrium rates of interest took both markets and

central banks by surprise. Because policy rates have lagged the equilibrium
rate in coming down, there has been a secular downward pressure on infla-
tion. To some degree this downward trend in inflation has been welcome.
However, in a few economies it has gone too far. Asian economies with
slowing trend growth rates are at risk of getting trapped in harmful defla-
tion, as in the case of Japan since the early 1990s.
With a low inflation rate and a low equilibrium real interest rate, the
nominal policy interest rate will be close to zero in the future. The diffi-
culty of setting the policy rate much below zero greatly reduces the scope
for countercyclical monetary policy, at least using the conventional policy
tool. Ball et al. (2016) show for the United States that the zero bound is
likely to constrain conventional monetary policy in all but the mildest of
recessions as long as inflation and inflation expectations remain near 2
percent. Although fiscal policy can, in principle, play an important role in
macroeconomic stabilization when monetary policy faces the zero bound,
the experience of the major advanced economies in the aftermath of the
global financial crisis demonstrates that political and institutional barriers
to effective fiscal policy can be substantial.

294 SUSTAINING ECONOMIC GROWTH IN ASIA

 The case of Korea is instructive (see figure 13.6). From 2000 through
2008, Korea’s short-term interest rate averaged 4.3 percent and core infla-
tion averaged 2.8 percent, implying a real interest rate of 1.5 percent. To
stabilize the Korean economy during the global financial crisis, the Bank
of Korea cut short-term interest rates by more than 3 percentage points.
Since 2009, the short-term interest rate has averaged 2.2 percent and
the core inflation rate also averaged 2.2 percent, implying a real interest
rate of 0.0 percent. Currently, core inflation is 1.4 percent. If inflation
were to settle in at 1 percent and the equilibrium real interest rate is now
0 percent, then the “new normal” policy rate would be 1 percent. In the
event of a negative shock to the Korean economy, the Bank of Korea would
not be able to lower the policy rate by as much as it did in 2009. Without
the help of fiscal policy or unconventional monetary policy (discussed
below) Korea would be subject to longer recessions and slower recoveries.
To reduce this risk, the Bank of Korea should set its policy stance to ensure
that inflation returns at least to its target rate of 2 percent and seriously
consider a slightly higher target, say 3 percent, which had been the target
only two years ago.
Figure 13.7 shows that Thailand is at risk of falling into sustained defla-
tion. With the policy rate at 1.5 percent, the Bank of Thailand would not
be able to deliver the 2 percentage point easing of conventional policy that
it did during the global financial crisis. Moreover, policy seems to be too
tight as the real interest rate is higher than in Korea and core inflation is
falling further below target. In the latest Article IV consultation, IMF staff
recommended further monetary ease. Thai authorities preferred to preserve
space for future policy action, arguing that inflation expectations are well
anchored and lower interest rates could raise risks to financial stability.
However, as seen in Japan in the 1990s, measures of expectations typically
lag rather than lead actual inflation trends. In addition, research suggests
that preserving policy space is the wrong strategy for an economy at risk of
deflation. In such circumstances, an inflation surprise to the downside is
harder to deal with than a surprise to the upside. Therefore, a central bank
should be more aggressive than otherwise in easing policy as it approaches
the zero lower bound to avoid the danger of the liquidity trap (Reifschneider
and Williams 2000).
We also note that Singapore and Taiwan are currently at the zero lower
bound. Any additional monetary ease in these economies must come in the
form of unconventional monetary policies, which we discuss in the next
section.

MONETARY AND EXCHANGE RATE POLICIES 295

 Figure 13.6 Korean monetary policy since the Asian financial
crisis, 2000Q1–2017Q2

percent

1
Short-term interest rate
Core CPI inflation
0

2000Q1 2005Q1 2010Q1 2015Q1

CPI = consumer price index

Source: Haver Analytics. Interest rate is the call money rate. Inflation is the
four-quarter change in the CPI excluding agriculture and oil.

Figure 13.7 Thai monetary policy since the Asian financial

crisis, 2000Q1–2017Q2

percent

Short-term interest rate

Core CPI inflation
−2

2000Q1 2005Q1 2010Q1 2015Q1

CPI = consumer price index

Source: Haver Analytics. Interest rate is the call money rate. Inflation is the
four-quarter change in the CPI excluding food and energy.

296 SUSTAINING ECONOMIC GROWTH IN ASIA

Risks from Global Bond Markets: The “Taper Tantrum”
Central banks in Asia facing subpar growth and below-target inflation are
vulnerable to shocks from global bond markets that could suddenly tighten
domestic financing conditions. The increased importance of domestic
bond markets in monetary policy transmission (Mohanty 2014) means that
central banks in Asia may have to make greater use of their balance sheets
than when credit was supplied exclusively by banks at rates linked to the
short-term interest rate set by the central bank.
Central banks need to be ready for external shocks to the world real
interest rate. As described above, there are sound reasons to believe that the
equilibrium real interest rate has declined secularly. Yet part of the decline
observed since 2011 is cyclical and reflects central bank purchase of bonds.
The unexpected depth and length of the recession in advanced economies
after the global financial crisis (and the associated pessimism about the
future) has depressed the world long-term rate. If growth-friendly policies
succeed in closing the global output gap and end the trend decline in infla-
tion, investment rates would probably rise and precautionary savings fall.
We cannot know how suddenly global long-term rates would rise. The taper
tantrum of 2013 showed that expectations of monetary policy tightening
in the United States could have a large effect on Asian bond markets even
when domestic conditions in Asia do not change (see figure 13.5).
The taper tantrum demonstrated how externally driven swings in local
bond market liquidity in emerging markets can affect local financial con-
ditions in a dramatic way. Monetary policy, notably central bank balance
sheet policies, may need to offset these shocks. During the taper tantrum,
the average of yields on local-currency government bonds in the more open
Asian markets rose sharply—from 3.2 percent in April 2013 to 4.4 percent
by January 2014—and market volatility spiked higher. In some emerging
markets, currencies fell sharply just as bond prices declined, although ex-
change rate movements in the Asia-15 economies were mixed and mostly
rather small.
The bond markets in most emerging markets are of recent birth, and
market liquidity is vulnerable to swings in foreign investor sentiment. In
many countries, the domestic investor base is narrow, dominated by banks
or state-run pension funds. Because the intrinsic liquidity of the markets
for government bonds in many emerging markets is still comparatively
low, some foreign investors tend to rely on intermediary instruments (bond
funds, synthetic exchange traded funds [ETFs], etc.) that promise daily li-
quidity. When market sentiment changes, this liquidity illusion can be shat-
tered, leading to very heavy sales: Shek, Shim, and Shin (2015) have shown
that investor flows into and out of emerging-market funds tend to cluster
much more than for advanced-economy bond flows.
MONETARY AND EXCHANGE RATE POLICIES 297
 As discussed further later in the chapter, the central bank can use its
balance sheet to keep the markets for local financial assets operating in the
face of a market liquidity shock. This can forestall any self-feeding price
movements that could produce a sharp and unwarranted tightening of
financial conditions. A particularly bold policy move along these lines was
the decision of the Hong Kong Monetary Authority in 1998 to buy 7 percent
of domestic equities to thwart a joint speculative attack on its currency
peg and stock market (Bayoumi and Gagnon 2018). This policy worked
because of the credibility of the central bank’s commitment to free finan-
cial markets in normal times. In the aftermath of the global financial crisis,
several central banks in the emerging markets undertook to lend against
(or even buy) financial assets, private as well as public (BIS 2009). Some
offered to indemnify asset holders for any eventual losses from continuing
to hold government bonds or other paper. Such policies aimed at coun-
tering temporary bouts of extreme market illiquidity.

Is Low Inflation beyond the Control of Central Banks?

The Phillips Curve Is Dormant, Not Dead
Many observers have noted that very large increases in unemployment rates
during the Great Recession had only small effects on inflation in advanced
economies. However, it does not follow that inflation is beyond the control
of central banks. Rather, the very low trend rates of inflation coupled with
downward nominal wage and price rigidity have put economies in a region
where the Phillips curve is flat. But the slope is likely to increase as econo-
mies exceed potential by a significant amount. And the slope is likely to be
higher in general when inflation is significantly above zero.
This subsection examines the evidence on inflation and the output gap
in the United States, which has the longest available series of these data.
The first column in table 13.1 presents an estimate of an expectations-
augmented Phillips curve (equation 13.1). The dependent variable is the
four-quarter percent change in the GDP deflator minus the value of infla-
tion that had been predicted eight quarters ago for the following four quar-
ters.8 The gap is the difference between the Congressional Budget Office’s
estimate of the natural rate and the actual unemployment rate. The infla-

8. The predicted value is from a survey of professional economic forecasters and is provided
by Haver Analytics. The inflation forecast of four quarters earlier should have incorporated
the effects of the gap of four quarters earlier, leaving no systematic prediction error if fore-
casters are efficient. An alternative specification based on the contemporaneous output gap
and forecasted inflation with a four-quarter lag had a lower R2 (0.12) and a much smaller
coefficient on the gap of 0.46.

298 SUSTAINING ECONOMIC GROWTH IN ASIA

Table 13.1 Phillips curves regressions on US GDP deflator
Inflation – Expected Inflation = α + β Gap + γ Gap*(Inflation Dummy)
Expectations
measure Survey Survey Lag Lag Lag
1.16*** 1.00*** 1.15***
Emp Gap
(0.25) (0.12) (0.13)
0.53*** 0.45***
Cap Util Gap
(0.08) (0.05)
–0.96*** –0.98*** –1.01***
Emp Gap (Inf<3)
(0.22) (0.16) (0.14)
–0.56*** –0.45***
Cap Util Gap (Inf<3)
(0.09) (0.07)
0.32 –0.38** –0.02 0.16 –0.41***
Constant
(0.29) (0.17) (0.17) (0.17) (0.14)
R-squared 0.58 0.58 0.47 0.56 0.60
Observations 182 182 271 199 199
1972Q2– 1972Q2– 1950Q1– 1968Q1– 1968Q1–
Sample
2017Q3 2017Q3 2017Q3 2017Q3 2017Q3
Note: ***, **, and * denote 1, 5, and 10 percent significance levels, respectively. Newey-
West standard errors with three lags are in parentheses.
Sources: Haver Analytics and authors’ calculations. See text for description of vari-
ables.

tion dummy equals zero when inflation is above 3 percent and one when
inflation is below 3 percent.

Inflation(t) – Expected Inflation(t–8) = (13.1)

α + β Gap(t–4) + γ Gap(t–4)*(Inflation Dummy(t–4))

The results show that the gap has a large and strongly significant effect
when inflation is above 3 percent, but the effect largely disappears when
inflation is below 3 percent. This simple model can explain nearly 60 percent
of the overall variance of inflation, as shown by the R2 statistic. The second
column displays results using an alternative measure of the output gap: the
Federal Reserve’s index of capacity utilization in manufacturing, mining,
and utilities minus its average value since 1967. The effect on inflation of a
1 percentage point gap in capacity utilization is about half as large as that of
a 1 percentage point employment gap, but the explanatory power is essen-
tially identical. As with the employment gap, the effect of the capacity utili-
zation gap declines sharply when inflation is very low.
The remaining columns of table 13.1 display results using a lagged
three-year moving average of inflation as a measure of inflation expecta-
tions (equation 13.2). The regression shown in column 3 has a much longer
sample, back to 1950, and a somewhat lower R2 than column 1. But the
coefficients on the employment gap are reasonably close to those in column

GRAPHICS - CHAPTERMONETARY AND EXCHANGE

13—SUSTAINING ECONOMIC RATE IN ASIA2991
POLICIES
GROWTH
1. Column 4 displays the same regression starting in 1968. The coefficients
are almost identical to those in column 1. The final column displays a re-
gression using lagged inflation and capacity utilization. It obtains results
similar to those of column 2.

Inflation(t) – 3-year Ave. Inflation(t–4) = (13.2)

α + β Gap(t–4) + γ Gap(t–4)*(Inflation Dummy(t–4))

To check whether the Phillips curve slope may have changed over time,
we also ran the regressions of table 13.1 starting in 1992Q1, just after US
inflation fell below 3 percent on a sustained basis. There are only nine quar-
ters with inflation above 3 percent in this subsample (2004Q4–2006Q3 and
2007Q1), yet we obtain estimates of the gap coefficients that are almost
identical to those shown in columns 1 and 2 and moderately smaller than
those in columns 3 to 5.
Figure 13.8 displays the inflation surprises (left-hand sides of equa-
tions 13.1 and 13.2) and employment gaps (right-hand sides of both equa-
tions), where the sample is split between lagged inflation above and below 3
percent. Similar results (not shown) are obtained using capacity utilization.
The greater slope in the high inflation regime is apparent for both measures
of expectations. It also appears that the slope may steepen for the most
positive values of the output gap. However, adding the interacted value of
the output gap and a dummy when the output gap is above its mean value
yields a significant coefficient (at the 10 percent level) in only the first of the
five regressions shown in table 13.1.

Scope for Unconventional Monetary Policy

Central banks can ease policy and achieve their objectives even at the zero
bound. It would be absurd to assume that—irrespective of circumstances—
the only legitimate policy tool for a central bank is the overnight rate in
interbank markets. The analysis by Reddy (2017) supports this policy con-
clusion. The following paragraphs suggest some possibilities. What would
work best will depend on country circumstances (including political con-
straints) and on macroeconomic conditions.
To a small extent, central banks can reduce policy rates below zero. Swit-
zerland has pushed short-term interest rates more deeply negative than any
other economy, at –0.75 percent. It may be possible to go more negative, but
there is a risk that at some point banks and firms might begin to store large
volumes of paper currency. In addition, banks in any economy with negative
policy rates have not passed the negative rates through to household depos-
its. This lack of pass-through to household deposits limits the effectiveness
of negative policy rates and hurts bank profitability.

300 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 13.8 US Phillips curves for GDP inflation and employment gap,
1955Q1–2017Q1

Inflation < 3 percent Inflation > 3 percent

surprise in inflation surprise in inflation

8 8
6 6
4 4
2 2
0 0
−2 −2
−4 −4
−6 −6
−5 −4 −3 −2 −1 0 1 2 3 −5 −4 −3 −2 −1 0 1 2 3
employment gap employment gap

Inflation < 3 percent Inflation > 3 percent

change in inflation change in inflation

8 8
6 6
4 4
2 2
0 0
−2 −2
−4 −4
−6 −6
−5 −4 −3 −2 −1 0 1 2 3 −5 −4 −3 −2 −1 0 1 2 3
employment gap employment gap

Source: Haver Analytics and authors’ calculations.

Another channel for easing policy at the zero bound is to provide

forward guidance to markets that the policy rate will remain near zero for
several years. The credibility of such a commitment almost surely declines
with the horizon of the commitment, as central bank governors and policy
board members cannot legally restrict their own future actions, let alone
those of their successors. But forward guidance does appear to have worked
over horizons of two to three years (Campbell et al. 2012).
Probably the most general avenue for easing policy at the zero bound
is the active use of the central bank’s own balance sheet. Table 13.2 repre-
sents a stylized central bank balance sheet to show just how many tools a
central bank has at its disposal. In almost all emerging-market economies
in Asia, the central bank balance sheet is very large. A major driver of this
expansion was the huge accumulation of foreign exchange reserves after
the Asian financial crisis (discussed in the next section). One consequence
of foreign exchange accumulation for the domestic financial system was an

MONETARY AND EXCHANGE RATE POLICIES 301

 Table 13.2 A central bank balance sheet
Assets Liabilities
Foreign assets Cash
Government bills Required bank reserves
Government bonds Excess bank reserves
Loans to domestic banks Government deposits
Other local financial assets Equity

increase in local bank deposits, usually raising commercial bank reserves

held with the central bank. When the central bank wanted to prevent this
accumulation leading to an increase in bank lending, it typically raised
required reserve ratios. For much of this period, however, central banks
welcomed the stimulus from bank lending expansion that foreign reserve
accumulation supported. Buying domestic financial assets or lending to
domestic banks would similarly stimulate aggregate demand even if policy
interest rates do not change.
Historically, central banks in the advanced economies have used their
balance sheets extensively for macroeconomic purposes. Tobin’s (1969)
classic work on portfolio rebalancing mechanisms in the transmission of
monetary policy (changes in the term premium and other risk spreads)
found a recent echo in Gertler and Karadi (2013). Ben Friedman (2014) has
argued that the central bank’s balance sheet is likely to become a part of
the standard toolkit of monetary policy in the years ahead. Farmer (2017)
shows how official purchases of equities can counter too-pessimistic animal
spirits in markets, and so sustain business investment.
Table 13.3 shows the scope for quantitative easing (QE) in the Asia-15
economies. Among these economies, only Japan is currently engaged in
QE, as reflected in the very large size of the BOJ’s balance sheet. Many other
Asian economies have large central bank balance sheets, primarily reflecting
large stockpiles of foreign exchange reserves. We discuss exchange rate and
intervention policy in the final section below.
All of these economies have at least some scope for central banks to
conduct QE through purchases of government bonds. Three advanced
economies (the euro area, Japan, and the United Kingdom) have also
subsidized lending to the banking system. The Bank of England estimates
that its Funding for Lending scheme has had a macroeconomic impact
equivalent to a reduction in the policy rate of 0.75 to 1.50 percent (Bank
of England 2014). The column labeled broad money in table 13.3 gives an
approximate size of the domestic banking system through which subsi-
dized central bank credit could operate.

302GRAPHICS
2 - CHAPTER
SUSTAINING ECONOMIC GROWTH IN ASIA
13—SUSTAINING ECONOMIC GROWTH IN ASIA
Table 13.3 Scope for quantitative easing, December 2016 (percent of
2016Q4 GDP, seasonally adjusted annual rate)
General
Central bank government Equity market
Country liabilities gross debt Broad moneya capitalization
Australia 10 41 112 100
Bangladesh b
n.a. 33 55 16
China 45 44 202 66
Hong Kongc 60 0 232 560
Indiad 21 70 82 69
Indonesia 8 28 38 45
Japan 88 239 236 104
Korea 28 38 201 92
Malaysia 35 56 128 131
New Zealande 10 29 99 n.a.
Philippines 30 35 62 80
Singapore 37 112 137 221
Taiwan 87 36 238 157
Thailand 46 42 124 103
Vietnamb n.a. 61 n.a. 3
n.a. = not available
a. M3 where available, otherwise M2.
b. Based on 2016 annual GDP.
c. Data exclude assets denominated in foreign currency and shares of mainland com-
panies.
d. Broad money based on April 2017 to avoid effects of demonetization in late 2016.
e. Central government debt.
Sources: Haver Analytics, Hong Kong Monetary Authority, and International Monetary
Fund World Economic Outlook database.

Perhaps the most untapped channel for QE is equity purchases. The

BOJ is buying about 1 percent of domestic equities per year, but this pace
could be increased considerably. In some economies, equities represent a
much larger potential market for central banks than government bonds.
A vast literature documents the powerful effects of QE on long-term
bond yields in the euro area, Japan, the United Kingdom, and the United
States (Gagnon 2016). Though more difficult to prove, there is evidence
that QE has stimulated economic activity and inflation. The Federal Reserve
purchased long-term bonds equivalent to nearly 25 percent of GDP in suc-
cessive rounds from 2008 through 2014. Staff estimate that these purchases
had a macroeconomic effect roughly equal to that of a 250 basis point cut
in the federal funds rate (Engen, Laubach, and Reifschneider 2015). Wu
and Xia (2015) estimate a shadow federal funds rate to capture the mac-
roeconomic impact of QE. They find that as of 2014, QE had reduced the

GRAPHICS - CHAPTER MONETARY AND ECONOMIC

13—SUSTAINING EXCHANGE GROWTH
RATE POLICIES 3
IN ASIA303
shadow federal funds rate by 200 to 300 basis points. A similar estimate
by Lombardi and Zhu (2014) places more weight on the Federal Reserve’s
balance sheet and implies a shadow federal funds rate of minus 400 basis
points by 2011.
In some circumstances, however, governance considerations may in
practice limit the large-scale use of the central bank’s balance sheet. This
can be the case in jurisdictions where full instrument independence of the
central bank is not securely established. Central banks must avoid the traps
of fiscal and financial dominance; they need to be sure they are free to decide
to sell the assets they have purchased if monetary policy so requires. Govern-
ments with large debts to refinance may resist higher bond yields. While no
central bank will want to provoke financial market volatility, worries about
destabilizing bond or equity markets should not prevent central banks from
gradually tightening monetary policy when inflation is expected to remain
above their targets. The warning of Shirakawa (2015) that markets must
not be misled into believing the policy regime has become a “put-option
type of monetary policy” is well taken.

Should Risks to Financial Stability Constrain Monetary

Policy?9
Some central banks feel in a quandary. They worry that a prolonged period
of very low interest rates could create risks for financial stability—a reason-
able worry since monetary policy works in part by changing financial risk
exposures. Lower interest rates reduce the debt service burdens of borrow-
ers and may help keep them solvent. And lower rates typically increase asset
prices, raising the value of collateral held by firms and households, thus
making them seem better credit risks in the eyes of potential lenders. Debt-
to-income ratios can be expected to rise if the decline in interest rates per-
sists, as some recent research suggests (Laubach and Williams 2015, Rachel
and Smith 2015). Higher debt and asset prices can be regarded as natural
equilibrating mechanisms to a move to a low interest rate environment. Yet
it is possible to overshoot the new equilibrium, and any sudden correction
would be disruptive. Because no one can know how such worries might ma-
terialize, regulatory policy needs to be prepared.
Keeping policy interest rates higher than warranted by macroeconomic
conditions would not solve this quandary. This is because a prolonged
period of subpar growth and high unemployment also creates financial
stability risks. Such risks would be all the greater if prices are falling. The
question whether a central bank should keep the policy rate higher than

9. This section draws on Turner (2017), which provides fuller details and references.

304 SUSTAINING ECONOMIC GROWTH IN ASIA

that needed on macroeconomic grounds to counter financial stability risks
is not new. Dennis Robertson (1928/1966) answered this question with a
clear “no” when he took the (young) Federal Reserve to task in 1928 for
focusing its interest rate policy on limiting speculative lending of commer-
cial banks. At that time, the Federal Reserve was guided by what it called the
Principle of Productive Credit. Underlining the danger at that time of an
undesirable fall in the general level of prices, Robertson proposed instead
what he termed the Principle of Price Stabilization, “the stabilization of the
price level as the sole and sufficient objective of (central) banking policy.”
The subsequent history surely vindicated his view. The Federal Reserve’s
acquiescence in the massive collapse of the money supply and a 25 percent
decline in the price level after the 1929 crash turned an ordinary recession
into the Great Depression.
As the head of economic research at the Reserve Bank of Australia (RBA)
has documented, there is little historical evidence that low interest rate
environments are inherently unstable—either in creating macroeconomic
instability or in destabilizing the financial system (Simon 2015). The main
common sense argument for not allowing financial stability worries to
override the macroeconomic considerations driving monetary policy is that
interest rates high enough to counter some potential financial threat would
cripple the rest of the economy. In addition, expectations that determine
asset prices or lending expansions are not as stable or predictable functions
of policy variables as are macroeconomic variables (BIS 1998). The most
general analysis of the issue to date shows that the marginal cost of keeping
the policy rate high and accepting higher unemployment outweighs the
marginal benefit from the lower probability of a crisis under a wide range of
assumptions about the economy (Svensson 2016).
Recent history fully supports this conclusion. From mid-2004 to mid-
2006, a substantial rise in policy rates worldwide, which bond markets
expected to be sustained, went together with increased risk-taking in the
global financial system on all the standard metrics (Turner 2017). In his
press conferences as chairman of the bimonthly global economy meetings
of central bank governors at the Bank for International Settlements (BIS),
Jean-Claude Trichet repeatedly during 2006 and early 2007 underlined the
concerns of the governors about overextended financial markets. He ex-
plained that central banks had prepared the ground by raising interest rates
substantially as economies neared full employment and that the financial
industry should prepare for a significant correction. But banks and markets
remained entirely complacent.
One telling international comparison between the Bank of England and
the Bank of Canada throws some useful light on what would have happened

MONETARY AND EXCHANGE RATE POLICIES 305

had short-term rates been kept higher before 2004. The Bank of England,
worried about strong domestic demand as well as continued rises in house
prices and expecting a return of core inflation to around 2 percent from a
lower level, did not follow the sharp cuts in the US federal funds rate in 2001.
By mid-2004, the bank rate had been raised to 4¾ percent—even though
core inflation was below 1½ percent during almost all of 2003 and 2004.
The bank was concerned about “financial imbalances creating problems
beyond the two-year horizon of our inflation target.” Yet tighter monetary
policy did not prevent the buildup of financial imbalances in the United
Kingdom. And this policy did contribute to an overvalued currency, which
created its own financial risks.
The Bank of Canada, by contrast, cut interest rates aggressively. But
lower rates did not induce Canadian banks to become overextended because
of much stricter regulation (notably the existence of a leverage ratio and
limits to banks’ off-balance sheet exposures to securitized products) and
because a less contestable domestic banking market allowed fatter margins.
The major policy shortcomings that aggravated the 2008–09 financial
crisis were not related to monetary policy. They were rather the failures of
domestic supervisors to address the new risks that innovation in the finan-
cial industry had created (Ramaswamy 2017).
The BIS has challenged the Svensson (2016) analysis. In its 2016 Annual
Report, the BIS put forward an alternative path for the federal funds rate
from 2002 (BIS 2016). The new policy rule guiding this path was a Taylor
rule augmented by a financial cycle proxy. Had the Federal Reserve followed
this rule, the BIS argues, the financial crisis would have been avoided and
there would have been a gain of about 1 percent a year in real US GDP over
a decade or so, or 12 percent cumulatively.
As Turner (2017) documents, however, the methodology underlying
this calculation raises many questions. The federal funds rate implied by the
financial cycle–augmented Taylor rule rises earlier but by much less than
the actual funds rate over the 2003–06 period. As noted above, substan-
tial rises in the Bank of England’s policy rate and in the federal funds rate
to over 5 percent failed to curb financial market risk-taking—much to the
chagrin of Trichet and the other governors. Why then would a more modest
rise started a little earlier have worked? It is implausible that a new mone-
tary policy rule would have significantly reined in the housing bubble and
added so much to US GDP. We are skeptical that a Taylor rule augmented
by any financial cycle proxy would be a useful guide to policy. As Federal
Reserve chair Janet Yellen put it shortly after the publication of the 2014
BIS Annual Report, which had urged central banks to more quickly return
interest rates to normal levels because of financial stability worries, “there

306 SUSTAINING ECONOMIC GROWTH IN ASIA

is no simple rule that can prescribe, even in a general sense, how monetary
policy should adjust to shifts in the outlook for financial stability.”10
The implication from this new rule for monetary policy that was under-
lined by the BIS in June 2017 was that central banks “may have to tolerate
longer periods of inflation below target, and tighten monetary policy if
demand is strong, even if inflation is weak, so as not to fall behind the curve
with respect to the financial cycle.”11 Certainly, strong demand growth
especially when the economy is near full employment justifies a tightening
in monetary policy. But we would not agree that central banks should
keep interest rates up in the face of prolonged periods of inflation below
their targets. Such a policy would run counter to the inflation targeting
mandates of many central banks and aggravate the risks of recession.
Almost everywhere, the postcrisis policy response focused primarily on
tightening regulation and developing new macroprudential tools. Monetary
policy was progressively eased to counter a deep and prolonged weakness in
aggregate demand. Although such a recession was perhaps inevitable given
the severity of the global financial crisis, its persistence was a surprise. Few
(if any) expected interest rates to remain low for so long. The United States
both tightened regulations (notably forcing the banks to recapitalize) more
rigorously and uniformly and eased monetary policy more promptly than
was the case in the euro area. This difference, as well as the fragmented
policy response to the euro area’s existential crisis, likely explains why the
United States was more successful is ending its recession.
The implications of unusually low interest rates globally for the balance
sheets of households, companies, and financial institutions are going to
be much larger than in the past because rates have been low for so long
(Hannoun and Dittus 2017). Those responsible for prudential regulation
need to pay particular attention to two important classes of risk. The first
is the risk associated with borrowers becoming more highly leveraged. The
second is the interest rate risks on the balance sheets of financial intermedi-
aries. Near-zero or negative interest rates on shorter maturities have induced
banks and other investors to seek yield by lengthening the maturity of the
bonds they hold as assets. The profitability of interest rate carry-trades for
many years has led many financial firms to lengthen the maturity of their
debt instruments, which has lowered long-term rates. Falling long-term
interest rates for some years have given large capital gains to financial firms
holding bonds on the asset side of their balance sheets.

10. Janet Yellen, “Monetary policy and financial stability,” 2014 Michel Camdessus Central
Banking Lecture, International Monetary Fund, Washington, DC, July 2, 2014.
11. “Central banks warned on inflation risk: BIS annual report focuses on danger of interest
rates staying low for too long,” Financial Times, June 26, 2017.

MONETARY AND EXCHANGE RATE POLICIES 307

 At the same time, this lengthening in duration has made the market
value of portfolios of debt securities more sensitive to changes in bench-
mark long-term rates. Interest rate risk exposures have therefore risen.
Even in normal times, regulatory and accounting rules do not treat interest
rate risk well. Some recent regulations (e.g., the international banking rules
of Basel 3 and the Solvency 2 regulations for European insurance compa-
nies) have inadvertently magnified interest rate exposures. When global
interest rates are lower and more stable than they have been historically,
those supervising banks and institutional investors need to look especially
hard at how their current rules encourage greater interest rate risk expo-
sures. And they need to redouble efforts to better manage such exposures.
Interest rate developments should also influence the design of macro-
prudential instruments. Consider the evolution of rules on household
property mortgages. One way of protecting households from borrowing too
much when interest rates are unusually low is to impose debt-to-income
ratios. After successfully using loan-to-value (LTV) ratios, for example, the
Reserve Bank of New Zealand (RBNZ) recently proposed that it be given
powers to use debt-to-income ratios. It argues that such ratios would help
to constrain the credit/asset price cycle in a manner most other macropru-
dential ratios would not (Reserve Bank of New Zealand 2017).
To those who argue that macroprudential tools are not perfect and
occasionally may need to be reinforced by monetary policy, we counter
that all policy tools are imperfect. It is better to develop new macropru-
dential tools—or improved techniques for existing tools—than to sacrifice
the important objectives of monetary policy for a goal that it is ill suited
to achieve. In many cases, the financial risks that cause most concern are
sector-specific (e.g., increased mortgage debt in many Asian countries) and
require a tailored policy response. A recent speech by the vice president of
the ECB exemplified the approach we support.12 QE has produced stronger
growth, which has helped make banks in the euro area stronger. But he also
reported ECB data on how very low interest rates were creating larger matu-
rity mismatch risks in nonbank financial institutions and then explained
the need to expand the macroprudential toolkit to address these new risks.
So our answer to the question of this section is: No, financial stability
considerations should not in general constrain monetary policy. But pru-
dential policies may need to be adapted to curb risks created by higher
levels of debt and by the maturity mismatches/interest rate exposures as-
sociated with a long period of very low rates. A possible rejoinder is that this
answer amounts to advocating that one arm of policy (regulation) undo

12. Vítor Constâncio, “The evolving risk landscape in the euro area,” speech at a Banco do
Portugal conference on financial stability, October 17, 2017.

308 SUSTAINING ECONOMIC GROWTH IN ASIA

the inevitable consequences of another arm of policy (monetary expansion).
This rejoinder is not convincing. Expansionary monetary policy in a de-
pressed economy can also improve financial stability (higher incomes from
stronger growth lower debt-to-income ratios of many borrowers, lower in-
terest charges help liquidity-constrained but viable debtors avoid default,
encouraging investors to buy risky assets that are typically undervalued in
a recession, and so on). It deserves emphasis that there is no logical pre-
sumption that monetary tightening needs to complement macroprudential
tightening measures.
In many circumstances the central bank will want to ease monetary
policy but tighten macroprudential policies. The recent policies of the
RBNZ illustrate this well. As the governor noted, the introduction of
macroprudential speed limits on high loan-to-value lending for mortgages
“moderated excesses in the housing market, thereby enabling the Bank to
delay the tightening of interest rates, and reducing the incentive for further
capital inflows into the New Zealand dollar.”13

Exchange Rate Policy

If Asian economies were suffering a loss of external demand, it might be
argued that officials should seek a more competitive exchange rate by selling
domestic currency for foreign currencies. However, in economies where the
growth slowdown is most pronounced (Korea, Japan, Thailand, Hong Kong,
Singapore, and Taiwan) the cause is domestic not external. These economies
have current account surpluses, and in all but Hong Kong, these surpluses
have been rising in recent years (see figure 13.9). Thus, the external sector
has on net been supporting growth in most of the Asian economies expe-
riencing slower growth. Among the large emerging Asian economies, only
India and Indonesia have current account deficits, which in both cases are
quite modest at around 2 percent of GDP.
Figure 13.10 shows that many Asian economies have piled up unprec-
edented levels of foreign exchange reserves and paid down official external
debts in some cases. In many cases, official foreign assets far exceed rea-
sonable precautionary needs (Bergsten and Gagnon 2017). Moreover, com-
batting currency mismatches in the private sector is much more important
than accumulating reserves for preserving financial stability and prevent-
ing future balance of payments crises (Gagnon 2013). For instance, the
currency mismatch data reported in Chui, Kuruc, and Turner (2016) show
that foreign-currency debts in the corporate sector in some Asian econo-
mies grew too rapidly because foreign lenders took too much comfort from

13. Wheeler, “Reflections on 25 years of inflation targeting.”

MONETARY AND EXCHANGE RATE POLICIES 309

 Figure 13.9 Current account balances in Asia, 1985–2016
(percent of GDP)

Australia Bangladesh China

0 2 10
−2
0
−4 5
−6 −2
−8 −4 0
Hong Kong India Indonesia
15 2 5
10 0
5 0
0 −2
−5 −4 −5
Japan Korea Malaysia
5 10 20
4
3 5 10
2 0 0
1
0 −5 −10
New Zealand Philippines Singapore
0 5 30
−2 20
−4 0
−6 10
−8 −5 0

Taiwan Thailand Vietnam

20 10 5
15 0
10 0 −5
5 −10
0 −10 −15
1985 1995 2005 2015 1985 1995 2005 2015 1985 1995 2005 2015

Source: IMF World Economic Outlook database.

very high levels of official foreign exchange reserves. In some cases, quasi-
fixed exchange rate regimes encourage foreign-currency borrowing. Neither
macroeconomic nor precautionary needs justify continued accumulation
of foreign exchange reserves by most large Asian economies. The IMF esti-
mates that the aggregate savings ratio of developing Asia has exceeded 40
percent of GDP for many years—far above that prevailing elsewhere. In such
circumstances, and given the chronic shortfall of aggregate demand at the
global level since the global financial crisis, excess reserve accumulation that
supports a current account surplus exerts a powerful negative externality on
the rest of the world.
In some cases, the stock of foreign exchange reserves may fall as offi-
cials seek to prevent unwanted currency depreciation. For example, China’s
reserves fell by roughly $1 trillion over the past three years. The central bank

310 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 13.10 Foreign exchange reserves and net official
assets in Asia, 1985–2016 (percent of GDP)

Australia Bangladesh China

6 10 60
4 0 40
−10 20
2 −20
0 −30 0

Hong Kong India Indonesia

120 20 20
100 15 0
80 10
60 5 −20
40 0 −40
Japan Korea Malaysia
30 30 50
20 20
0
10 10
0 0 −50

New Zealand Philippines Singapore

20 50 200
15 150
10 0
100
5 −50 50

Taiwan Thailand Vietnam

100 60 50
80 40 0
60
40 20 −50
20 0 −100
1985 1995 2005 2015 1985 1995 2005 2015 1985 1995 2005 2015

Net official assets

Foreign exchange reserves

Note: Net official assets are foreign exchange reserves plus other official
assets (including sovereign wealth funds) minus official borrowing in
foreign currencies.
Source: Bergsten and Gagnon (2017).

may want to prevent a credit-depressing shrinkage of its balance sheet. In

the case of China, a rise in loans to domestic banks offsets the decline in
foreign-currency assets on the central bank’s balance sheet.

Conclusion
Growth in several Asian economies remains disappointing, and there are
downside risks. With inflation declining to very low levels, central banks in
Asia should be ready to use the policy tools at their disposal to sustain aggre-
gate demand to meet medium-term inflation targets. Indeed, for Thailand
at least, conditions already support further monetary ease.

MONETARY AND EXCHANGE RATE POLICIES 311

 Financial stability worries do not in general justify keeping the policy
rate higher than warranted by macroeconomic conditions. Indeed, such
a policy would be fraught with risks: For instance, when discussing the
Swedish Riksbank’s decision to raise rates in 2010 to counter a property
price bubble, Brunnermeier and Schnabel (2016) pointed out that increasing
rates when banks are vulnerable and leverage in the economy high might
not be the best option. At the top of the policy agenda should be measures
to adapt both regulatory and macroprudential policies and the focus of
supervision should be on the new (or accentuated) financial risks created by
a very long period of exceptionally low interest rates, long as well as short.
Although globalization has weakened some of the channels of mone-
tary policy transmission, Asian central banks have not lost their monetary
autonomy.14 The volatility of capital flows and vulnerability of domestic
financial conditions to sudden shifts in investor risk preferences create
difficult monetary policy dilemmas. As Obstfeld (2015) put it, “financial
globalization has worsened the trade-offs monetary policy faces in navi-
gating between multiple domestic objectives.” The most reliable compass
remains flexible inflation targeting. Even if the policy rate gets stuck at the
zero lower bound, central banks still have tools to keep inflation within
their policy mandate. In particular, they have the scope to expand their
holdings of domestic assets (financial securities and loans to banks).

References
Ahearne, Alan, Joseph Gagnon, Jane Haltmaier, and Steven Kamin. 2002. Preventing Deflation:
Lessons from Japan’s Experience in the 1990s. International Finance Discussion Paper 729.
Washington: Board of Governors of the Federal Reserve System.
Akerlof, George, William Dickens, and George Perry. 1996. The Macroeconomics of Low In-
flation. Brookings Papers on Economic Activity 1: 1–59. Washington: Brookings Institution.
Ball, Laurence, Joseph Gagnon, Patrick Honohan, and Signe Krogstrup. 2016. What Else Can
Central Banks Do? Geneva Reports on the World Economy 18. Geneva: International Center
for Monetary and Banking Studies.
Bank of Canada. 2013. Backgrounders: Measurement Bias in the Canadian CPI. Available at www.
bankofcanada.ca/wp-content/uploads/2010/11/measurement_bias_canadian_cpi.pdf
(accessed July 13, 2017).
Bank of England. 2014. Inflation Report (November). London.
Bayoumi, Tamim, and Joseph Gagnon. 2018. Unconventional Monetary Policy in the Asian
Financial Crisis. Pacific Economic Review 23, no. 1: 80–94.
Benigno, P., and Luca Ricci. 2011. The Inflation-Output Tradeoff with Downward Wage Ri-
gidities. American Economic Review 101, no. 4: 1436–66.

14. Veerathai Santiprabhob, panel remarks at the 2017 Andrew Crockett Lecture: Boom-Bust
Cycles, Interest Rates and the Global Financial System, Bank for International Settlements,
Basel, June 25, 2017.

312 SUSTAINING ECONOMIC GROWTH IN ASIA

Bergsten, C. Fred, and Joseph Gagnon. 2017. Currency Conflict and Trade Policy: A New Strategy
for the United States. Washington: Peterson Institute for International Economics.
BIS (Bank for International Settlements). 1998. The Role of Asset Prices in the Formulation of Mon-
etary Policy. BIS Conference Paper 5 (March). Basel.
BIS (Bank for International Settlements). 2009. Capital Flows and Emerging Economies. CGFS
Paper 33 (January). Basel: BIS Committee on the Global Financial System.
BIS (Bank for International Settlements). 2016. The Financial Cycle, the Natural Rate of Interest
and Monetary Policy. 86th Annual Report (June). Basel.
Brunnermeier, M. K., and I. Schnabel. 2016. Bubbles and Central Banks: Historical Perspec-
tives. In Central Banks at a Crossroads: What Can We Learn from History? ed. Michael D. Bordo,
Øyvind Eitrheim, Marc Flandreau, Jan F. Qvigstad. Cambridge: Cambridge University
Press.
Campbell, Jeffrey, Charles Evans, Jonas Fisher, and Alejandro Justiniano. 2012. Macroeco-
nomic Effects of Federal Reserve Forward Guidance. Brookings Papers on Economics Activity
(Spring): 1–80. Washington: Brookings Institution.
Chui, Michael, Emese Kuruc, and Philip Turner. 2016. A New Dimension to Currency Mismatches
in the Emerging Markets: Non-Financial Companies. BIS Working Paper 550 (March). Basel:
Bank for International Settlements.
Dickens, W., L. Goette, E. Groshen, S. Holden, J. Messina, M. Schweitzer, J. Turunen, and M.
Ward. 2007. How Wages Change: Micro Evidence from the International Wage Flexibility
Project. Journal of Economic Perspectives 21, no. 2: 195–214.
Engen, Eric, Thomas Laubach, and David Reifschneider. 2015. The Macroeconomic Effects of the
Federal Reserve’s Unconventional Monetary Policies. Finance and Economics Discussion Series
2015-005. Washington: Board of Governors of the Federal Reserve System.
Fallick, Bruce, Michael Lettau, and William Wascher. 2016. Downward Nominal Wage Rigidity
in the United States during and after the Great Recession. Finance and Economics Discussion
Series 2016-001. Washington: Board of Governors of the Federal Reserve System.
Farmer, Roger. 2017. Prosperity for All. Oxford: Oxford University Press.
Friedman, B. M. 2014. Has the Financial Crisis Permanently Changed the Practice of Monetary Policy?
Has It Changed the Theory of Monetary Policy? NBER Working Paper 20128 (May). Cam-
bridge, MA: National Bureau of Economic Research.
Gagnon, Joseph. 2013. Currency and Maturity Mismatches in Asia. In Asian Capital Market
Development and Integration, ed. Iwan Azis and Hyoung-Tae Kim. New Delhi: Oxford Uni-
versity Press.
Gagnon, Joseph E. 2016. Quantitative Easing: An Underappreciated Success. PIIE Policy Brief 16-4.
Washington: Peterson Institute for International Economics.
Gertler, M., and P. Karadi. 2013. Monetary Policy Surprises, Credit Costs and Economic Activ-
ity. Paper presented at NBER conference on Lessons from the Crisis for Monetary Policy,
October 18–19, Boston.
Hannoun, Hervé, and Peter Dittus. 2017. Revolution Required: The Ticking Time Bombs of the G7
Model. Kindle edition.
Holston, Kathryn, Tomas Laubach, and John Williams. 2016. Measuring the Natural Rate of In-
terest: International Trends and Determinants. Federal Reserve Bank of San Francisco Work-
ing Paper 2016-11. Federal Reserve Bank of San Francisco.
Hördahl, P., J. Sobrun, and P. Turner. 2016. Low Long-Term Interest Rates as a Global Phenomenon.
BIS Working Paper 574 (August). Basel: Bank for International Settlements.
King, M., and D. Low. 2014. Measuring the “World” Real Interest Rate. NBER Working Paper
19887 (February). Cambridge, MA: National Bureau of Economic Research.

MONETARY AND EXCHANGE RATE POLICIES 313

Laubach, T., and J. C. Williams. 2015. Measuring the Natural Rate of Interest Redux. Hutchins
Center Working Paper 15. Washington: Hutchins Center of Fiscal and Monetary Policy at
the Brookings Institution.
Lombardi, M., and F. Zhu. 2014. A Shadow Policy Rate to Calibrate US Monetary Policy at the Zero
Lower Bound. BIS Working Paper 452 (June). Basel: Bank for International Settlements.
Mohanty, M. S. 2014. The Transmission of Unconventional Monetary Policy to the Emerging Markets.
BIS Papers 78. Basel: Bank for International Settlements.
Nakaso. H. 2001. The Financial Crisis in Japan during the 1990s: How the Bank of Japan Responded
and the Lessons Learnt. BIS Papers 6 (October). Basel: Bank for International Settlements.
Obstfeld, M. 2015. Trilemmas and Tradeoffs: Living with Financial Globalisation. BIS Working Pa-
per 480 (February). Basel: Bank for International Settlements.
Posen, Adam S. 1998. Restoring Japan’s Economic Growth. Washington: Institute for Interna-
tional Economics.
Rachel, L., and T. D. Smith. 2015. Secular Drivers of the Global Real Interest Rate. Bank of England
Staff Working Paper 571 (December). London: Bank of England.
Ramaswamy, S. 2017. Financial Crisis and Regulatory Reforms. Basel: IngramSpark.
Reddy, Y. V. 2017. Advice and Dissent: My Life in Public Service. New York: HarperCollins.
Reifschneider, David, and John Williams. 2000. Three Lessons for Monetary Policy in a Low-
Inflation Era. Journal of Money, Credit, and Banking 32: 936–78.
Reserve Bank of New Zealand. 2017. Serviceability Restrictions as a Potential Macroprudential Tool
in New Zealand. RBNZ Consultation Paper (June). Wellington.
Robertson, D. H. 1928/1966. Theories of Banking Policy. Lecture at the London School of
Economics, February 13, 1928. Reprinted in Essays in Money and Interest (1966, Collins,
Fontana Library).
Shek, J., I. Shim, and H. S. Shin. 2015. Investor Redemptions and Fund Manager Sales of Emerging
Market Bonds: How Are They Related? BIS Working Paper 509 (August). Basel: Bank for In-
ternational Settlements.
Shirakawa, M. 2015. Excessive Debt and the Monetary Policy Regime. BIS Papers 80 (January). Ba-
sel: Bank for International Settlements.
Simon, J. 2015. Low Interest Rate Environments and Risk. Paper presented at the Paul Wool-
ley Centre for the Study of Capital Market Dysfunctionality Conference, October 8, Syd-
ney.
Svensson, L. E. O. 2016. Cost-Benefit Analysis of Leaning Against the Wind: Are Costs Larger Also with
Less Effective Macroprudential Policy? IMF Working Paper WP/16/13 (January). Washington:
International Monetary Fund.
Tobin, James. 1969. A General Equilibrium Approach to Monetary Theory. Journal of Money,
Credit, and Banking 1: 15–29.
Turner, Philip. 2017. Did Central Banks Cause the Last Crisis? Will They Cause the Next? Special
Paper 249 (November). London School of Economics, Financial Markets Group.
Williams, Rebecca. 2017. Characterizing the Current Economic Expansion: 2009 to Present
Day. Reserve Bank of New Zealand Bulletin 80, no. 3: 3–22.
Wu, Cynthia, and Dora Xia. 2015. Measuring the Macroeconomic Impact of Monetary Policy at the
Zero Lower Bound. Chicago Booth Research Paper 13-77. University of Chicago, Booth
School of Business (revised 2015).
Wyplosz, Charles. 2001. Do We Know How Low Inflation Should Be? In Why Price Stability?
Proceedings of the First ECB Central Banking Conference, ed. A. García-Herrero, V. Gaspar, L.
Hoogduin, J. Morgan, and B. Winkler. Frankfurt: European Central Bank.

314 SUSTAINING ECONOMIC GROWTH IN ASIA

 14
Monetary Policy in the New
Mediocre
RANIA AL-MASHAT, KEVIN CLINTON, BENJAMIN HUNT,
ZOLTAN JAKAB, DOUGLAS LAXTON, HOU WANG, AND
JIAXIONG YAO

The global economy has been underperforming for several years. The perfor-
mance of advanced economies has been lackluster since the 2008–09 global
financial crisis. Strong expansion in emerging-market economies boosted
global growth, but growth in these economies has also decelerated since
2011, largely because of weak commodity prices and the transition in China
to a more balanced pattern of growth. Although global growth picked up in
2017, output gaps remain in many economies, and most countries’ inflation
rates were still below official central bank target rates. Interest rates have
been historically low—but evidently not low enough to fuel a normal recov-
ery and strong expansion. Fiscal policies have been inconsistent, with priori-
ties switching between stimulus and consolidation, often without a long-
term strategy, sometimes adding to cyclical instability. Political opposition
has caused long delays to structural reforms that would raise trend growth.
With a still weak global economy and excess capacity, the possibility of
further large negative shocks remains a concern. Financial vulnerabilities
remain evident in various parts of the world. Credit-fueled booms have in-
flated real estate values in many cities. Corporate balance sheets have weak-

Rania Al-Mashat is an advisor at the Research Department of the International Monetary Fund (IMF) and
former subgovernor for monetary policy at the Central Bank of Egypt. Kevin Clinton is a former research ad-
viser at the Bank of Canada and a consultant to the IMF. Benjamin Hunt is assistant director at the Research
Department of the IMF. Zoltan Jakab is a senior economist at the Research Department of the IMF. Douglas
Laxton is the chief of the Economic Modeling Division of the Research Department of the IMF. Hou Wang is an
economist at the Research Department of the IMF. Jiaxiong Yao is an economist at the Research Department of
the IMF. The views expressed in this chapter are those of the authors and do not necessarily represent the views of
the IMF, its Executive Board, or its management.

315
ened, with leverage rebounding to high levels. Restructuring in China could
bring disruptive adjustments and financial stress. Governments have been
raising trade barriers, and protectionist sentiment is on the rise.
In Asia demographics are another potential source of chronic demand
shortfalls. Some analysis suggests that a decline in population growth could
reduce aggregate demand, with few or no self-correcting forces. Several Asian
economies, advanced and emerging alike, are in or will soon enter the phase
of the demographic transition in which fertility declines and lower mortality
results in rapidly aging populations and sharply lower population growth.
This demographic change could mean that “the new mediocre” could
become a problem in Asia—a concern that in part motivated this volume.
Gaspar et al. (2016) argue for assertive policy actions following persis-
tent demand weakness. For countries with wide output gaps, low levels of
employment, below-target rates of inflation, and extremely low long-term
rates of interest, the argument in favor of such a response extends beyond
regular cyclical stabilization. Assertive policy action is also called for to avoid
risk, because a negative shock to the global economy could push several
countries into a low-inflation trap, with central banks unable to respond
strongly, as interest rates are already near their effective lower bound.
Another recession could set off a downslide in expectations of inflation,
even a deflation mentality, undermining the nominal anchor that central
bank inflation targets are supposed to provide. The outcome would be a
bad quasi-equilibrium (a “dark corner”): a low-inflation or deflation trap in
which resources lie idle, with no end in sight.1
A comprehensive, consistent, and coordinated approach (which
Gaspar et al. 2016 refer to as “the 3Cs”) is needed across all macroeconomic
instruments. This chapter focuses on monetary policy, with country case
studies of Australia, India, and Japan.
Assertive monetary policy responses to negative output shocks may
lead central banks to overshoot the official inflation target in the medium
term; they can also cause it to undershoot the target, as in India, where the
response to a jump in food prices may mean that a year of above-target
headline inflation is followed by a short period below target. Central banks
should not be averse to such outcomes when a bad equilibrium lies nearby.
In these situations, the prudent policy is to get away from the bad equilib-
rium as quickly as possible. A prolonged period of below- or above-target
inflation is more likely to destabilize the nominal anchor than a cycle that
sees inflation on both sides of the target rate.

1. The term low-inflation trap refers to a situation in which inflation expectations are stuck
significantly below the inflation target and there are significant risks of deflation.

316 SUSTAINING ECONOMIC GROWTH IN ASIA

 In India the central bank is attempting to reduce long-term expecta-
tions of inflation. Shocks (e.g., to food prices) could unhinge inflation ex-
pectations on the upside (another dark corner). Prompt action is needed,
especially in view of India’s history of high and volatile inflation and there-
fore imperfect credibility.
This chapter is organized as follows. The first section discusses a risk-
avoidance approach to monetary policy in a world with chronic underper-
formance and downside risks to growth. Section two illustrates the argu-
ment with country studies on Australia, India, and Japan. Section three
describes the role of macroprudential policies, stressing the importance of
sound banking systems and orderly markets for sustainable growth. The
last section makes the case that risk avoidance may involve assertive policy
actions when public confidence in policy objectives is low.

Risk Avoidance Strategy for Monetary Policy

From Inflation Targeting to Inflation-Forecast Targeting
In 1977 Kydland and Prescott reported that discretionary monetary policy
has a bias toward inflation. Following the inflation of the 1970s, their finding
made a strong impression on central banks and led to a search for a more
robust framework for monetary stability. The solution proposed by Barro
and Gordon (1983) was for the central bank to commit to a time-consistent
price stability policy, in order to prevent policymakers from yielding to the
temptation to give output a short-run boost through a surprise spurt of
monetary stimulus.
During the 1980s, inflation in most economies was moderate but per-
sistent. Some central bankers saw a parallel with the chronic inflation pre-
dicted for discretionary monetary policy by the time-consistency theory.
Early practitioners of inflation targeting saw hitting their targets, short
run as well as long run, as important in establishing their commitment to
price stability. The first country to adopt inflation targeting, New Zealand,
started out with a fairly rigid approach. It produced a rapid decline in infla-
tion to the long-run target, along with instrument instability.
During the 1990s, as inflation stabilized at low rates, in line with the
announced objective, it became clearer that the key to establishing confi-
dence in inflation targeting was not rigid adherence to targets but a trans-
parent strategy to eliminate over time any deviations that arose or were
expected to arise. Announcing an explicit numerical target was in itself a
major step toward clarifying what monetary policy was aiming to achieve.
Following the introduction of inflation targeting, central banks took
additional steps to improve their monetary policy communications, through
regular monetary policy reports (sometimes called inflation reports), speech-

MONETARY POLICY IN THE NEW MEDIOCRE 317

es by senior officials on strategy, media briefings after interest rate decision
meetings, and so on. By the turn of the 21st century, one could argue that the
transparent pursuit of a low-inflation objective by a politically accountable
central bank provided a solution to the Kydland-Prescott time-consistency
problem. Inflation targeting had put a constraint on discretion, removing
the inflation bias.
Another interpretation of the evidence would be that the success of
inflation targeting refuted the Kydland-Prescott theory. Central banks
showed no sign of reneging on inflation control in pursuit of short-run
output goals; in inflation-targeting countries, at least, the authorities did
not display the short-sighted bias at the heart of the argument. On the
contrary, governments left and right of center supported the low-inflation
objective—by a formal instruction where the central bank does not have
goal independence (e.g., the United Kingdom); by an endorsement where
it does (e.g., the Czech Republic); or by a statement of agreement where the
government and central bank jointly assume responsibility for the goal (e.g.,
Canada and New Zealand).
Under these arrangements, the central bank is typically accountable for
the conduct of monetary policy, to government or parliament and implicitly
to the general public. In large part because of the clear delegation of respon-
sibility, implementation of inflation targeting has been accompanied by a
vast increase in the transparency of the conduct of monetary policy.
During the 1990s, central bankers came to realize that the better their
policies were understood, the more effective they were—a remarkable turn-
around within one generation for a profession formerly known (not entirely
fairly) for its secrecy. With respect to numerical variables, the debate has
been about what to disclose above and beyond the target for the rate of
inflation and the current setting of the policy rate—in particular, which ele-
ments of the quarterly macroeconomic forecast of the central bank should
be released. Publishing the forecast for inflation and output has not been
controversial, because policymakers had to show the public that they had a
plan for keeping inflation on target and that the plan recognized the poten-
tial short-run implications for output. Svensson (1997) pointed out that the
central bank’s inflation forecast represents an ideal conditional intermedi-
ate target, because it takes into account all available information, including
the preferences of policymakers and their view of how the economy works.
The flexible inflation targeting regime now in place at many central
banks can be described as inflation-forecast targeting. It implies a balancing
between the deviations of inflation from its target and the deviations of
output from its potential. Under a flexible inflation-forecast targeting
regime, the central bank has a dual mandate (either explicit or implicit): It
recognizes that there is a short-run tradeoff between output and inflation.

318 SUSTAINING ECONOMIC GROWTH IN ASIA

 The history of inflation targeting and the transition to full-fledged
inflation-forecast targeting provides a line of openness, or accountability.
Milestones along the way include the following:
n announcement of targets with a multiyear horizon (clarity of target),
n precision on policy interest rate setting (clarity of instrument),
n transparent communications on policy implementation,
n publication of a complete macro forecast (including inflation) (clarity
of the intermediate target [inflation-forecast targeting]), and
n publication of a conditional forecast path, alternative scenarios, and
confidence bands for the short-term interest rate (full-fledged infla-
tion-forecast targeting).

Unlike pioneers, such as New Zealand, newcomers to inflation targeting

do not have to reach each of these milestones. The road has already been
tested, over a few decades.2 Depending on its technical capacities, a central
bank can take to the road at any point.
Numerous countries have built durable inflation targeting regimes from
unpromising starting conditions, as the survey of central banks by Batini
and Laxton (2007) shows. None had a reputation for stable low inflation.
Many were emerging from a financial or exchange market crisis that had
shaken confidence in the monetary authorities (e.g., the United Kingdom,
Sweden, and the Czech Republic). Some were in the midst of economywide
structural changes that would completely alter the transmission of mon-
etary policy (e.g., New Zealand and the Czech Republic). Special problems
enfeebled the monetary transmission mechanism in certain countries (e.g.,
dollarization in Peru and severe financial fragility in the Czech Republic).
Of the early adopters of inflation targeting, only the Bank of Canada
had anything close to a forecasting and policy analysis system matched to
the task. A common omission was the lack of an appropriate policy model.
No country had the external communications program required to explain
how the monetary policy objective was to be achieved and maintained. The
list of prerequisites is not demanding; if a central bank can conduct mone-
tary policy, it can engage in inflation targeting.
Central banks that adopted inflation targeting put in place suitable
frameworks for making it effective. These frameworks are still evolving, as
central banks learn by doing.

2. For a discussion of the timeline for several inflation-targeting countries, see Clinton et al.
(2015).

MONETARY POLICY IN THE NEW MEDIOCRE 319

 Loss-Minimization Approach to Policy Formulation
The model or models used by central banks to conduct policy under infla-
tion-forecast targeting should incorporate an endogenous interest rate. A
model in which the interest rate is exogenous has no nominal anchor; the
inflation rate drifts indeterminately following disturbances. For inflation
targeting to be logically consistent, the interest rate must adjust to the
requirements of the target. Many central banks incorporate this principle
into their forecasting models and thus produce an endogenous path for
the interest rate.3
Endogenous monetary policy in this type of model is often represented
by a linear inflation forecast–based policy reaction function for the short-
term interest rate, with equal weighting of marginal changes in devia-
tions from target for any starting point. For inflation-forecast targeting,
the current setting of the policy rate (i in the example below) takes into
account the contemporaneous output gap (y), the model’s forecast of year-
on-year inflation (π4), the equilibrium real interest rate (𝑟𝑟̅ ), the inflation
𝑟𝑟̅ policy rate (e ). In the example below,
i
target (π*), and possible shocks to the
the forecast horizon is set at three quarters: ���� ��������
𝑖𝑖� � ���𝑖𝑖��� � �� � �����𝑟𝑟̅� � ����� � ��������� � � ∗ � � ����� � �
𝑖𝑖� � ���𝑖𝑖��� � �� � �����𝑟𝑟̅� � ������ ��������
��� � ��������� � � ∗ � � ����� � � ��� .

In contrast to the Taylor rule, in which the policy rate reacts � to the
contemporaneous output gap and
�
� � ∑��� 𝛽𝛽
𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿inflation �
������������
deviation��� from � the �
� ∗ � target,
� ���� � ����𝑖𝑖��� � 𝑖𝑖����
the
. � � �������� ∗ � � � �. that
�
inflation 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿� � ∑��� 𝛽𝛽 ����
forecast–based reaction
��� � � � �ignores
function ���� � ����𝑖𝑖 ��� � 𝑖𝑖�����
inflation shocks
. are expected to reverse within the three-quarter policy horizon. Although
central banks should not care about small temporary deviations, they
should care a lot about large, long-lasting deviations. A quadratic loss func-
tion, in which the marginal cost of deviations of inflation and output from
desired values increases exponentially as they grow, better captures such
preferences. It is particularly appropriate when a bad equilibrium threatens
to take hold of the economy. Historically, two types of bad equilibrium (in
both of which inflation expectations go adrift) have preoccupied central
banks: a low-inflation trap and an inflation spiral. As a quadratic loss func-
tion would imply aggressive policy actions when confronted with a mate-

3. Indeed, policy modelers design these models to make it impossible to simulate with an
exogenous path (constant or based on market expectations) for the policy rate. Sometimes
these models are simulated by assuming that the interest rate is exogenous over some short-
term horizon but that agents believe that at some point in the future the monetary authority
will start to adjust its instruments sufficiently aggressively to anchor inflation and infla-
tion expectations. In these forward-looking models, this assumption helps anchor inflation
expectations in the short run. Policymakers should be skeptical of forecasts based on exog-
enous paths for the policy rate.

320 SUSTAINING ECONOMIC GROWTH IN ASIA

 rial probability of such situations, it provides the basis for a risk-avoidance
strategy.
An appropriate quadratic loss function for inflation-forecast targeting
includes the discounted sum of squared expected deviations of headline
inflation from the target, the squared expected output gap, and (with a
smaller weight) the squared𝑟𝑟̅ change in the policy rate. The central bank sets
the path of the policy interest rate to minimize this loss function, which
heavily penalizes large deviations of the target variables, output, and infla-
𝑖𝑖� � ���𝑖𝑖��� � ��
tion � �����𝑟𝑟̅
from � ������
��� �
their� preferred ��������
������The
values. � � ∗ � � ����
��� discounting � � �(β
factor ��,� .a value between
0 and 1) means that a given deviation from target imposes a lower cost the
farther it is in the future:
�������� �
𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿� � ∑� �
��� 𝛽𝛽 �������
�
� � ∗ � � ���� � ����𝑖𝑖��� � 𝑖𝑖����� �� �.

The rationale for an aversion to large changes in the interest rate,

which is not in itself a target variable, is to avoid overreacting to small tran-
sitory disturbances and the associated problem of instrument instability
and to ensure that the central bank provides clear signals for the intent
of its actions (Woodford 2005). In effect, inclusion of this term in the loss
function smooths the response of the interest rate following shocks to the
system. It has strong empirical support in the actual, incremental, rate-
setting behavior of central banks.
A procedure that minimizes a loss function can best illustrate the prin-
ciples of robust monetary policymaking under uncertainty. The principal
arguments of the loss function are the deviation of inflation from its target
and the deviation of output from potential (i.e., the output gap). Given the
dual-mandate preferences of policymakers, central banks engaging in flex-
ible inflation-forecast targeting in effect choose a policy rate path such that
the forecasts for inflation and output “look good,” where “looking good”
means finding the best policy interest rate path to stabilize the inflation
forecast around its target and output around its long-run sustainable path.
Linear reaction functions may give reasonable results in normal times,
but they perform poorly in abnormal times. When policy interest rates are
at or very near the effective lower bound, the loss-minimization approach
produces better results, because its response to disinflationary conditions
involves an extended commitment to keep the rate at the floor. As the effec-
tive lower bound approaches, the loss-minimization approach provides ever
stronger policy reactions to contractionary shocks, to keep the economy
away from the deflation dark corner. At the effective lower bound, this risk-
avoidance principle is reflected in a commitment to hold the short-term
interest rate low long enough that inflation will rise, perhaps above the long-
run target rate for a while. The boost to inflation expectations reduces the

MONETARY POLICY IN THE NEW MEDIOCRE 321

real medium-term interest rate and depreciates the real exchange rate, even
though the nominal short rate can go no lower.4 Where there is very little
risk of sustained inflationary pressure but a high risk of getting stuck in a
low-inflation trap, such a policy reaction represents prudent policymaking.

Conventional Forward Guidance for Transparent

Communications
The Bank of Canada, the US Federal Reserve, the Bank of England, the Bank
of Japan, and other central banks provided forward guidance on the policy
interest rate after the global financial crisis. Such guidance “talks down” the
expected policy rate path and term premium, thereby reducing longer-term
interest rates. In this respect, forward guidance succeeded (Charbonneau
and Rennison 2015; Engen, Laubach, and Reifschneider 2015). This type
of communication represents unconventional forward guidance, because
it was introduced along with other unconventional measures as an ad hoc
tool when the effective lower bound put constraints on reductions in the
policy rate.
Another form of forward guidance is more robust. Conventional for­
ward guidance, as practiced by the Czech National Bank and the Reserve
Bank of New Zealand, is a systematic part of the policy framework. It in-
volves the publication of a complete central bank macroeconomic forecast,
with an endogenous interest rate path and confidence bands around key
variables. The endogenous policy rate moves to achieve the announced
inflation target over a medium-term horizon in a way that reflects policy-
makers’ preferences with respect to the short-run tradeoffs between output,
inflation, and interest rate variability. The policy rate path is clearly condi-
tional on a range of assumptions and subject to a range of uncertainty, as
indicated by the confidence bands around it.
Publication of the forecast policy rate path increases the effective-
ness of the interest rate instrument in good times as well as bad. In and of
itself, setting a short-term interest rate for the following six weeks (a typical
interval between policy meetings) has no material impact on inflation or
output. The policy rate has an effect only as far as it moves the longer-term
interest rates at which households and firms borrow and invest. In effect,
the central bank must ensure that public expectations of the future over-
night rate move in line with the current setting. If market participants have
similar forecasts as the central bank, longer-term interest rates, the exchange

4. See annex II of Clinton et al. (2015) for the analytics.

322 SUSTAINING ECONOMIC GROWTH IN ASIA

rate, and asset prices generally are likely to move in support of the objectives
of monetary policy.5
That full disclosure helps lead expectations has long been accepted
with respect to the inflation rate. Expected inflation and the actual nominal
interest rate are the two components of the real interest rate. The published
inflation rate forecast influences the expected real interest rate, generally in
support of monetary policy. When nominal interest rates are, or are likely
to be, constrained by the effective lower bound, disclosure is particularly
important. A central bank fighting deflation risks might well envisage a
strategy in which it temporarily overshoots the inflation target (as in policy
simulations reported later in this chapter). Overshooting would reduce the
real interest rate and help move the economy away from a deflation, or low-
inflation, dark corner. With full publication of its forecast, the central bank
can communicate the whole story underlying its strategy, allaying any risk
to the credibility of the target that the planned overshoot might otherwise
create.
Many Asian countries have the opportunity to leapfrog some of the
countries that have well-established inflation-forecast targeting regimes by
adopting full-fledged inflation-forecast targeting. Figure 14.1 shows the
path of the Dincer-Eichengreen index of transparency for four Asian coun-
tries (Japan, Australia, Korea, and India); the United States; and the three
most transparent central banks in the index (the Sveriges Riksbank, the
Czech National Bank, and the Reserve Bank of New Zealand). All institu-
tions improved over time, but there is a sizable gap between the top central
banks and the rest.
The main objection to publication of the interest rate forecast is its con-
ditionality. Monetary policy must allow the interest rate to vary to offset
shocks. It cannot commit unconditionally to a forecast path for the rate.
Central bankers have worried that their credibility might be impaired if it
becomes necessary to deviate from a previous forecast. With effective com-
munications, this issue need not arise: Markets have readily adjusted in
countries where the central bank publishes its interest rate. Indeed, with a
deeper understanding of the intentions of policymakers, markets are more
likely to perform a strong buffering role against shocks. Model-derived
confidence bands, and alternative forecasts based on shocks to the baseline
forecast, are useful tools for communicating the conditionality of the pro-
jection and the implications of disturbances for which the probability is not
calculable from historical data.

5. For a discussion of the Czech experience with publishing forecasts, see Clinton et al.
(2017).

MONETARY POLICY IN THE NEW MEDIOCRE 323

 Figure 14.1 Central bank transparency index, 1998–2014

index
15

13

11

1
1998 2000 2002 2004 2006 2008 2010 2012 2014

Top three (Sweden, Czech Republic, New Zealand)

United States
Japan
Australia
Korea
India

Source: Dincer and Eichengreen (2014).

Country Examples
Three country examples illustrate how a risk-avoidance strategy for mone-
tary policy can be implemented. Australia is a large commodity-exporting
economy closely linked to the rest of the world. Its monetary policy faces
several challenges going forward—challenges that are similar to those that
central banks in other advanced economies have faced. With the comple-
tion of the institutional foundation of a flexible inflation-targeting regime,
India has seen Consumer Price Index (CPI) inflation decline markedly, but
it will take time to establish credibility in the new regime. After years of low
growth and near deflation, Japan needs comprehensive policies to achieve a
higher sustainable growth path and anchor inflation to the 2 percent target.

Australia: Managing Downside Risks and Avoiding a Deflation

Trap
The Reserve Bank of Australia practices flexible inflation targeting, with a
target range of 2–3 percent. As its mandate calls for the maintenance of
full employment, its monetary policy must consider objectives for output
as well as inflation. The inflation-forecast targeting model is squarely in line
with this policy framework (see appendix 14A for a summary of the model).

324 SUSTAINING ECONOMIC GROWTH IN ASIA

Baseline Scenario with Alternative Policy Strategies
Figure 14.2 shows a baseline scenario for a hypothetical forecast for
Australia for the period between the third quarter of 2016 and the fourth
quarter of 2019. The assumptions are that the global economy gradually
recovers and commodity prices rise smoothly. The policy rate is constrained
by the effective lower bound, which is assumed to be 0.75 percent for the
Reserve Bank of Australia.
The black line shows a forecast based on the linear inflation forecast–
based reaction function. The policy rate declines a little in the short run
and then rises toward its long-run equilibrium value as the output gap
approaches zero and inflation approaches 2.5 percent, the midpoint of
the target range. The process of adjustment gets an early boost from an
increase in the exchange rate (a depreciation of the Australian dollar) of
about 8 percent.
The light grey line shows results for the alternative, risk-avoidance,
strategy. Here the central bank cuts the interest rate to the lower bound
of 0.75 percent and holds it there for four quarters. The lower-for-longer
interest rate implies a considerably greater depreciation in the exchange
rate (of about 13 percent from the peak). These changes boost output,
closing the gap more rapidly. Headline inflation quickly returns to the
target range, by virtue of the pass-through to consumer prices from the
depreciation of the Australian dollar. In all, the risk-avoidance strategy
results in better outcomes than a strategy based on delaying rate cuts and
“keeping the powder dry.”
An early fiscal stimulus (the dotted line in figure 14.2) could further
improve outcomes. With the direct addition to aggregate demand, the
central bank keeps the interest rate at the floor for only one quarter. This
policy experiment illustrates how fiscal policy can assist monetary policy
in providing support to the economy when there are significant risks that
monetary policy could be constrained by the effective lower bound.

Downside Scenario with Alternative Policy Strategies

The downside scenario includes a large and unexpected weakening of eco-
nomic growth in China, such that the Chinese output gap is 3.5 percent
wider than in the baseline case. Weakening in China has significant spill-
overs to the rest of the world. For Australia it reduces export incomes by a
few percentage points and leads to about a 5 percent decline in the prices of
Australian commodity exports. Both changes reduce output in Australia.
Figure 14.3 shows the simulated results from the model. In response
to the shocks, the central bank cuts the interest rate in a series of steps,
reaching the effective lower bound after about three quarters. Together with

MONETARY POLICY IN THE NEW MEDIOCRE 325

 Figure 14.2 Baseline scenario with alternative policy strategies for Australia
a. Policy rate b. Output gap
percent per annum percent

326
3.0 0.5

2.5 0

2.0 –0.5

1.5 –1.0

–1.5
1.0
–2.0
0.5 1 1 1
1 1 1 1 6: 3 :1 3 8: 3 9: :3
3 7: 3 8: 3 3 1 1 6: 17 1 7: 1 18: 1 19
16: 6: 1 1 7: 8: 19: 9: 20 20 20 20
01 20 201 01 01 20 20 20 20
20 2 20 2 20 2
c. Exchange rate d. Headline inflation
Australian dollar per US dollar percent, year over year
1.6 4.0

3.5

SUSTAINING ECONOMIC GROWTH IN ASIA

1.5
3.0

1.4 2.5

2.0
1.3 1.5

1.0
1.2 :1 :3 :1 :3 :1 :3 :1 :3
:1 :1 1 1 16 16 17 17 18 18 19 19
16 6 :3 17 7 :3 8: 8:3 1 9: 9:3 20 20 20 20 20
01 20 20 20
20 201 20 201 2 2 01 20 2 01

Linear inflation forecast–based reaction function

Risk-avoidance strategy based on loss-minimization
Risk-avoidance strategy based on loss-minimization with fiscal support

Source: Helbling et al. (2017).

 Figure 14.3 Alternative policy strategies for Australia following a negative external shock
a. Policy rate b. Output gap
percent per year percent
3.0 0.5

2.5 0

2.0 –0.5

1.5 –1.0

1.0 –1.5

–2.0
0.5
:1 3 :1 1 3 1 3 :1 :3 :1 :3 :1 :3 :1 :3
16 6: 17 :3 18: 8: 19: 9: 16 16 17 17 18 18 19 19
1 17 1 1 20 20 20 20 20 20 20 20
20 20 20 20 20 20 20 20
c. Exchange rate d. Headline inflation
Australian dollar per US dollar percent, year over year
1.6 4.0

3.5
1.5
3.0

1.4 2.5

2.0
1.3
1.5

1.0
1.2
1 3 :1 3 :1 3 1 :3
:1 :3 :1 :3 :1 :3 :1 :3 1 6: 1 6: 17 1 7: 18 1 8: 1 9: 19
16 16 17 7 8 8 9 9 20 20 20 20
20 01 01 01 01 01 20 20 20 20
20 20 2 2 2 2 2

MONETARY POLICY IN THE NEW MEDIOCRE

Risk-avoidance strategy based on loss-minimization
Risk-avoidance strategy based on loss-minimization

327
with fiscal support and more aggressive monetary policy

Source: Helbling et al. (2017).

the worsening in the terms of trade, the cut causes a jump in the exchange
rate. As a result, despite the widening of the output gap, the inflation rate
rises, reaching the center of the target range within a year. The gap in output
performance remains below –1 percent for two years.
Given these results, one might question the appropriateness of the
strategy. In view of the size and prominence of the shock, the rationale
for smoothing the adjustment of the interest rate disappears. It would be
more prudent to cut the policy rate quickly to the effective lower bound. In
addition, as the effective lower bound becomes effective and constrains the
amount of stimulus monetary policy can deliver, the government might
enact a fiscal stimulus. The dotted line in figure 14.3 shows the results for
a package that includes a more aggressive rate cut and a fiscal stimulus.
In these simulations, fiscal policy complements monetary policy, which,
if left to itself, would achieve less satisfactory outcomes, especially with
respect to output. With a large negative shock, prudent policy thus involves
a sharp cut in the rate to the effective lower bound, in order to reduce the
risk of a bad (deflation) equilibrium. If doing so implies a modest overshoot
of the announced inflation target, there is no cause for concern. The greater
risk by far is that expectations of inflation will become unanchored on the
downside.

India: Building Credibility and Mitigating Supply Shocks

Before 2014 India’s central bank had no clear price stability mandate. The
history of unstable, moderate inflation doubtless weighs heavily in the
public mind, despite the announced change in regime. Credibility is earned,
over time, by achieving announced objectives and by effective, transparent
communications. It can be lost through policy actions that are inconsistent
with stated objectives.
Where monetary policy credibility is low, persistent food and fuel
price shocks can result in upward ratcheting in inflation expectations and
contribute to higher levels of inflation persistence. The example presented
here—based on technical collaboration between the Reserve Bank of India
and the International Monetary Fund to develop a quarterly projection
model for India (Benes et al. 2017a, 2017b)—has an endogenous credibility-
building process built into it (see appendix 14A for a summary of the model).

Building Monetary Policy Credibility Amid Frequent Supply

Shocks
Demand shocks are relatively easy for central banks to deal with, because
the policy response to the impacts on both output and inflation point in
the same direction (a negative demand shock reduces both output and

328 SUSTAINING ECONOMIC GROWTH IN ASIA

inflation, unambiguously calling for a cut in the policy rate). In contrast,
supply shocks present a dilemma, because the effects on output (decrease)
and inflation (increase) move in different directions.
The policy conflict is stark in the context of repeated supply shocks,
such as a drought lasting several years, which would cause repeated upward
shocks to the headline rate of inflation. Core inflation is also affected to
some extent, through the pass-through of higher costs of intermediate agri-
cultural goods. Although output is depressed at the same time as inflation
is rising, the central bank must respond, at least in the short run, with an
interest rate hike, because the primary objective of Indian monetary policy
is to establish firm expectations that inflation will, over time, return to the
announced target rate and the Reserve Bank of India is still in a phase in
which it needs to bolster the credibility of the inflation target.
Figure 14.4 shows the results of simulations of the model. The simu-
lated, loss-minimizing policy response is severe: a sharp rise in the interest
rate that provokes a steep decline in the foreign exchange rate and widens
the output gap. Even so, over the medium term, inflation increases and
the short-run policy tradeoff looks bad (“stagflation”). Monetary policy
credibility unavoidably takes a hit for a while, because of the nature of the
shock, not because of any failure of policy. Under the risk management
strategy, policy eventually restores inflation to 4 percent and prevents
long-term inflation expectations from ratcheting upward. These simula-
tions illustrate that, even with bad luck, a committed central bank can
anchor long-term inflation expectations.

Importance of Prompt Action

In the previous experiment, prompt effective action helped reduce the losses
of output and credibility in monetary policy that may follow supply shocks.
The importance of timely policy can be shown with experiments in which
policymakers wait before responding to a large, nasty supply shock. The
eventual interest rate hike has to be much greater than in the baseline (no
delay) case, and the cumulative output loss is much greater (figure 14.5).
Inflation is also higher over the medium term, and it stays above 4 percent
longer. Delay thus causes a substantial deterioration in the medium-term
tradeoff between output and inflation, as well as long-lasting damage to the
credibility of the 4 percent target.
Contrary to popular misunderstanding, the historical dominance of the
food component in major cycles in the inflation rate is not a valid argument
against an inflation target–based monetary policy. It underlines instead the
major weakness in the old monetary policy regime, which did not provide a
firm enough anchor to resist pass-through of food price shocks.

MONETARY POLICY IN THE NEW MEDIOCRE 329

 Figure 14.4 Sequence of adverse supply shocks in India

Output gap Inflation Policy rate

percent percent, year over year percent per year
0.5 6.0 13.0
0 5.5 11.5
–0.5 5.0 10.0
–1.0 4.5 8.5
–1.5 4.0 7.0
–2.0 3.5 5.5
0 8 16 24 32 40 0 8 16 24 32 40 0 8 16 24 32 40

Log real exchange rate Credibility stock Credibility signal

index index index
1 1.0 1.0
0
0.8 0.8
–1
–2 0.6 0.6
–3
0.4 0.4
–4
–5 0.2 0.2
–6
0 0
0 8 16 24 32 40
0 8 16 24 32 40 0 8 16 24 32 40

Cumulated output gap Coefficient on lagged Bias in inflation

percent inflation in expectations expectations
index percent
0
–1 0.8 0.20
–2 0.6 0.15
–3
–4 0.4 0.10
–5
0.2 0.05
–6
–7 0 0
0 8 16 24 32 40 0 8 16 24 32 40 0 8 16 24 32 40
One quarter
Two consecutive quarters
Three consecutive quarters

Source: Benes et al. (2017a).

Japan: Comprehensive Policies to Exit Deflation

Since the early 1990s, Japan has experienced weak nominal and real GDP
growth and repeated deflationary episodes. The effective lower bound has
constrained the Bank of Japan.
A 2013 policy package, which included fiscal stimulus, monetary easing,
and structural reforms, aimed at reflating the economy. It helped narrow
the wide output gap, reversed an appreciation of the yen, boosted corpo-

330 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 14.5 Delay of policy responses in India

Output gap Inflation Policy rate

percent percent, year over year percent, per year
0.5 6.5 14.5
0 6.0 13.0
–0.5 5.5 11.5
–1.0 5.0 10.0
–1.5 4.5 8.5
–2.0 4.0 7.0
5.5
–2.5 3.5
0 8 16 24 32 40
0 8 16 24 32 40 0 8 16 24 32 40

Log real exchange rate Credibility stock Credibility signal

index index index
2 1.0 1.0
0 0.8 0.8
–2 0.6 0.6
–4 0.4 0.4
–6 0.2 0.2
–8
0 0
0 8 16 24 32 40
0 8 16 24 32 40 0 8 16 24 32 40

Cumulated output gap Coefficient on lagged Bias in inflation

percent inflation in expectations expectations
index percent
0
0.8 0.20
–2
0.6 0.15
–4
0.4 0.10
–6 0.2 0.05

–8 0 0
0 8 16 24 32 40 0 8 16 24 32 40 0 8 16 24 32 40

Baseline
Nastier supply shock
Nastier supply shock with delayed policy

Source: Benes et al. (2017a).

rate profits, and lifted actual and expected inflation into positive territory.
The economy reached close to full employment, and base wages increased
modestly.
Despite this progress, the economy was still in a dark corner in early
2017, with interest rates at their floor and long-term inflation expectations
well below 2 percent. Adoption of inflation-forecast targeting should have
provided the basis for more consistent communications to explain how

MONETARY POLICY IN THE NEW MEDIOCRE 331

the Bank of Japan was adjusting its instruments to achieve its objectives.
However, as Gaspar et al. (2016) argue, monetary policy on its own is un-
likely to jolt the economy out of the trap it is in. It will require support from
fiscal policy, incomes policy, and structural reforms.
This section examines what might be achieved with a comprehensive
set of policies against deflation that Gaspar et al. (2016) call Three Arrows
Plus (Prime Minister Abe’s Three Arrows of fiscal stimulus, monetary
expansion, and structural reform plus an incomes policy).6 It emphasizes
the need for credible and transparent monetary and fiscal frameworks that
reinforce one another. Over the longer run, structural reforms to increase
labor market participation and mobility and reduce government protection
of certain sectors of the economy support faster growth of real output. An
unorthodox component of the package is an incomes policy aimed directly
at sluggish wage–price dynamics, through public sector wage settlements
and a “comply or explain” requirement for the private sector. This policy
would build on recent measures taken by the authorities, including higher
minimum wage increases, tax incentives for firms that raise wages, and
moral suasion to encourage wage growth.
The simulations are presented as deviations from the baseline projec-
tions for Japan included in the Staff Report for the 2016 Article IV
Consultation for Japan (see appendix 14A for a summary of the model).
According to the simulations, over the medium term the Three Arrows
Plus package (excluding structural reforms) would result in higher growth
and inflation than the baseline (figure 14.6). On average, real GDP growth
would be 0.4 percentage point higher every year over the forecast horizon,
and CPI inflation (excluding the effects of the value-added tax [VAT]
increase) would overshoot the Bank of Japan target of 2 percent by 2019.7
The combination of higher real GDP growth, gradual VAT increases, and
inflation would put the net-debt-to-GDP ratio on a downward trajectory,
which would marginally reduce the term premium, as financial market
participants viewed the debt as being more sustainable. Overall, the
proposed policy would end deflation and moderately improve real GDP
growth and debt sustainability over the medium term.

6. This section is based largely on Arbatli et al. (2016) and Gaspar et al. (2016).
7. Monetary policy in the simulations targets inflation excluding the direct impact of the
VAT increase. As a result, the relevant inflation comparison between the scenarios is the one
excluding the direct impact of the VAT increase.

332 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 14.6 Japan's Three Arrows Plus policy package

Real GDP growth Output gap

percent percent
1.4 0.5
1.2
0
1.0
0.8 –0.5

0.6 –1.0
0.4
–1.5
0.2
0 –2.0
2015 2016 2017 2018 2019 2020 2015 2016 2017 2018 2019 2020

Long-term and policy Core CPI inflation

(dotted line) interest rate percent, year over year, (excluding VAT impact)
percent, per annum
2.5
1.2
1.0 2.0
0.8
1.5
0.6
0.4 1.0
0.2
0.0 0.5
–0.2
0
–0.4
1 2 3 4 5 6
2015 2016 2017 2018 2019 2020

Net government debt Real effective exchange rate

percent of GDP (+ = appreciation)
percent deviation from baseline
136
1
134
0
132 –1

130 –2

–3
128
–4
126
–5
2015 2016 2017 2018 2019 2020
2015 2016 2017 2018 2019 2020
Baseline
Three Arrows Plus policy package

CPI = consumer price index; VAT = value added tax

Note: Baseline reflects the projections in the IMF Staff Report for the 2016 Article IV Consultation
for Japan. In the baseline scenario, the direct VAT impact is 0.3 percentage point in 2019 and
1.0 percentage point in 2020. In the other scenario, the direct VAT impact is 0.4 percentage
point in 2017 and 0.5 percentage point thereafter.
Source: Arbatli et al. (2016).

MONETARY POLICY IN THE NEW MEDIOCRE 333

Financial Stability Considerations
The effects of monetary policy on the real economy are transmitted through
various channels, which require a healthy financial system. Well-functioning
financial markets allow monetary policy to affect a wide array of market
prices. Institutions, especially banks, play a critical role in extending credit
to households and firms. A weak banking sector can be a significant hurdle
to monetary policy transmission; overstretched balance sheets can limit the
effects of monetary policy on the real economy.
With the aggressive monetary policy easing required to address chronic
demand shortfalls and other problems related to the new mediocre, a fre-
quently voiced concern is that such a policy could lead to a buildup of
financial imbalances in some sectors. In many cities, for instance, low in-
terest rates, in combination with the limited elasticity of the housing supply,
have contributed to steep increases in house prices and a run-up in house-
hold debt and leverage.
Macroprudential policy can help address building vulnerabilities from
accommodative monetary policy. Experience suggests that a broad range
of tools may be needed to address potential vulnerabilities in both the time
and structural dimensions. In the time dimension, prudential tools can be
used to build up additional buffers as systemic risk accumulates, so that
they are adequate in times of stress. Doing so helps constrain the risks to
the financial system from the boom and dampen the negative impact on
credit provision in the bust. In the structural dimension, various pruden-
tial tools can be used to address the externalities that systematically impor-
tant financial institutions can impose on the financial system. Capital
surcharges can enhance the resilience of these institutions to shocks.
Additional loss absorbency requirements can facilitate their orderly reso-
lution and reduce spillover risks within the financial system (IMF/FSB/
BIS 2016).
The financial stability from the very accommodative monetary policy
in the major advanced economies after the global financial crisis has been
widely discussed. The 2017 Annual Report of the Bank for International
Settlements notes that “even if inflation does not rise, keeping interest
rates too low for long could raise financial stability and macroeconomic
risks further down the road, as debt continues to pile up and risk-taking in
financial markets gathers steam” (BIS 2017, 68).
In response to this argument, one might point out that there is little
evidence that the very low interest rates of recent years weakened finan-
cial systems. Indeed, it is more likely that monetary ease reduced the risks
of banking system loan portfolios, by boosting economic activity, which
strengthened the incomes of households and firms, and raising the values

334 SUSTAINING ECONOMIC GROWTH IN ASIA

of assets. After the global crisis, the United States, which applied monetary
stimulus early and aggressively, also reestablished financial stability rela-
tively quickly.
The greater risk is that low rates may lead central banks to hold back
in their response to a negative output shock. When the economy is at risk
of slipping into a dark corner, with expectations of inflation falling below
target rates, and the policy rate is near the effective lower bound, prompt
corrective action is a risk-avoiding strategy for the central bank. Delay would
shake confidence in the nominal anchor and jeopardize financial stability.

Conclusion
The global economy has underperformed for a decade. There is strong evi-
dence that the real equilibrium level of the interest rate—the rate consis-
tent with the maintenance of full employment and stable inflation—may
be close to zero. Bond yields have fallen to extremely low levels, and nega-
tive output gaps and below-target inflation rates persist, despite near-zero
central bank policy rates. The effective lower bound on rates suggests that
conventional monetary policy may not be able to put up strong resistance
to a future economic downturn.
Current uncertainties suggest that policymakers should prepare for
nasty surprises from a variety of sources, including structural rebalancing
in China; financial sector vulnerabilities in several countries; weakened cor-
porate balance sheets; an asymmetric tightening of monetary conditions;
runaway real estate booms in cities, allied to high levels of household debt;
a low but rising tide of protectionism; and the deleterious effects of de-
mographics. The effect of a large shock from one or more of these sources
could be worse than a short-lived global recession. It could push the world
economy to the dark corner of a low-inflation or deflation trap, in which
resources lie idle with no end in sight. The nominal anchor for the economy
could become unhinged, as the public comes to expect that official inflation
targets will not be met and large-scale unemployment becomes chronic.
Weak growth in nominal incomes would worsen prospects for stabilizing
the uptrend in government debt-to-GDP ratios.
Governments and central banks should have contingency plans for
dealing with such dismal prospects. Despite widespread perception that
monetary and fiscal policies have run out of ammunition, there is room for
maneuver. Transparent, assertive policies must be conducted within a cred-
ible framework for long-term stability. A prudent approach to a recession-
ary shock from a weak starting point may be to take an aggressive stance,
committing to holding the policy rate at the effective lower bound for an
extended period. Such an approach may well lead to a short-run overshoot-

MONETARY POLICY IN THE NEW MEDIOCRE 335

ing of the inflation rate—but under the circumstances, such overshooting
would not be a cause for alarm. In countries in which there is a history of
high and unstable inflation and the central bank is still in the middle of a
credibility-building exercise, prompt reactions to shocks are the key to es-
tablishing credibility.

336 SUSTAINING ECONOMIC GROWTH IN ASIA

Appendix 14A Summary of Models
Open-Economy New Keynesian Model for Australia
The open-economy New Keynesian model for Australia bears many simi-
larities to the models many central banks use for forecasting and policy
analysis. It has a standard core structure, with equations for the output gap,
core inflation, the policy interest rate, and the exchange rate. Expectations
are forward looking, consistent with the projections of the model itself,
but the behavioral equations also embody significant lags and rigidities.
In addition, the model includes equations for headline, food, and energy
inflation; commodity terms of trade; trade and financial linkages with the
rest of the world; and bond yields of various maturities. It exhibits some
nonlinearities, primarily as a result of the effective lower bound constraint
on the policy interest rate and its implications for monetary policy.
The model is calibrated with reference to both historical data and the
theoretical literature, to ensure that it plausibly replicates historical data
and generates sensible projections. In practice, real-time forecasts also
involve a substantial deal of judgment and data analysis, given the well-
known problems with simple model-driven projections in the short term.
That caveat notwithstanding, under given forecasts and preferences, the
model simulations can demonstrate the key features of an optimal mone-
tary policy strategy.

Model of Endogenous Policy Credibility for India

The model of endogenous policy credibility is a stripped-down version of
the practical forecasting and policy analysis models many central banks
use. The simple core structure allows nonlinearities in the policy reac-
tion function, the Phillips curve, and a credibility-building process, along
with other technical modifications, which enable the exploration of more
complex problems and policy options than linear models can examine. The
simple core structure includes equations for the output gap, core inflation,
the policy interest rate, and the exchange rate. The Phillips curve contains
a nonlinear output gap, which implies an increasing marginal effect on the
inflation rate as the gap increases. At wide negative output gaps, the curve
becomes quite flat.
In a context of low monetary policy credibility, persistent food and fuel
price shocks can result in upward ratcheting in inflation expectations and
contribute to higher levels of inflation persistence. Inflation expectations are
modeled as a linear combination of backward-looking expectations (lags)
and forward-looking, model-consistent expectations (leads). The weight on
the former is decreasing with the stock of credibility. When central bank
credibility is imperfect, inflation expectations are biased upward relative to

MONETARY POLICY IN THE NEW MEDIOCRE 337

the linear combination of backward-looking and forward-looking compo-
nents. This property strengthens the propagation mechanism from supply
shocks and therefore requires a more aggressive tightening in monetary
conditions, to contain inflationary pressures and anchor long-term infla-
tion expectations. Over time, an established inflation forecast–targeting
regime provides an effective strategy for dealing with the second-round
effects of supply shocks.
Credibility is earned, over time, by achieving announced objectives and
by providing effective, transparent communications. It can be lost through
policy actions that are inconsistent with stated objectives. Technically, the
modeling is done by introducing a credibility signal that takes a higher
value if policy outcome is closer to announced objectives. High values of
the credibility signals increase the stock of credibility. However, because of
the built-in high persistence in the credibility stock, it takes repeated good
signals to improve credibility significantly.
Monetary policy minimizes a quadratic loss function, which penalizes
squared deviations from output and inflation objectives as well as squared
deviations of changes in the policy interest rate. This approach penalizes
dark corners (deflation or high-inflation traps). For India the relevant dark
corner could be a situation in which expectations of high inflation become
so entrenched that their elimination would require huge costs in lost output
and employment.

The International Monetary Fund’s G20MOD Module of Flexible

System of Global Models (FSGM)
G20MOD is an annual, forward-looking, multieconomy model of the global
economy that combines both micro-founded and reduced-form formula-
tions of economic sectors. This model contains individual blocks for the
G-20 countries and five additional regions to cover the remaining coun-
tries in the world. The key features of a typical G20MOD country model
are outlined below. More detailed description of the model can be found in
Andrle et al. (2015).
Consumption and investment have microeconomic foundations. Con-
sumption features overlapping-generations households, which can save
and smooth consumption, and liquidity-constrained households, which
must consume all of their current income every period. Firms’ investment is
determined by a Tobin’s Q combined with the Bernanke-Gertler-Gilchrist
model. Firms are net borrowers. Their risk premiums rise during periods
of excess capacity, when the output gap is negative, and fall during booms,
when the output gap is positive. This pattern mimics the effect of falling/
rising real debt burdens.

338 SUSTAINING ECONOMIC GROWTH IN ASIA

 Trade is modeled by reduced-form equations. They are a function of a
competitiveness indicator and domestic or foreign demand. The competi-
tiveness indicator improves one-for-one with domestic prices; there is no
local market pricing.
Potential output is endogenous. It is modeled by a Cobb-Douglas
production function with endogenous capital and labor. Total factor pro-
ductivity (TFP) is also endogenous and affected by public investments and
commodity prices.
Consumer price and wage inflation are modeled by reduced-form
Phillips’ curves. They include weights on both lagged and future inflation
and on the output gap as well. Consumer price inflation also includes a
weight on the real effective exchange rate and second-round effects from
commodity prices.
Monetary policy is governed by an interest rate reaction function. For
most countries, it is an inflation forecast–based rule that seeks to achieve
a long-run inflation target through uncovered risk-adjusted interest rate
parity.
Fiscal policy is governed by long-run fiscal policy targets that seek to
maintain a stable debt level. Short-run fiscal policy can result in significant
deviations away from this target; automatic stabilizers also operate. The
model includes four types of expenditures: public consumption, public in-
vestments, targeted transfers, and general transfers. Public capital is produc-
tive and has an impact on potential output. Major taxes are consumption
taxes, personal income taxes, and capital taxes.
Inclusion of three commodities (oil, metals, and food) allows for a dis-
tinction between headline and core consumer price inflation and provides
richer analysis of the macroeconomic differences between commodity-
exporting and commodity-importing regions. Demand for commodities
is driven by world demand; it is relatively price inelastic in the short run,
because of limited substitutability of the commodity classes considered.
The supply of commodities is also price inelastic in the short run. Countries
can trade in commodities, and households consume food and oil explicitly,
allowing for the distinction between headline and core CPI inflation. The
global output gap (a short-run effect), the overall level of global demand,
and global production of the commodity in question determine global real
prices.
Commodities can moderate business cycle fluctuations. In times of
excess aggregate demand, upward pressure on commodity prices from slug-
gish adjustment in commodity supply relative to demand will put some
downward pressure on demand. If there is excess supply, falling commodi-
ties prices will ameliorate the deterioration.

MONETARY POLICY IN THE NEW MEDIOCRE 339

 Countries are largely distinguished from one another in the G20MOD
by their unique parameterizations. Each economy in the model is structur-
ally identical (except for commodities) but has different key steady-state
ratios and behavioral parameters.

References
Andrle, M., P. Blagrave, P. Espaillat, K. Honjo, B. Hunt, M. Kortelainen, R. Lalonde, D. Laxton,
E. Mavroeidi, D. Muir, S. Mursula, and S. Snudden. 2015. The Flexible System of Global Mod-
els: FSGM. IMF Working Paper 15/64. Washington: International Monetary Fund.
Arbatli, E., D. Botman, K. Clinton, P. Cova, V. Gaspar, Z. Jakab, D. Laxton, C. Ngouana,
J. Mongardini, and H. Wang. 2016. Reflating Japan: Time to Get Unconventional? IMF Work-
ing Paper 16/157. Washington: International Monetary Fund.
Barro, R., and D. Gordon. 1983. Rules, Discretion, and Reputation in a Model of Monetary
Policy. Journal of Monetary Economics 12, no. 1: 101–21.
Batini, N., and D. Laxton. 2007. Under What Conditions Can Inflation Targeting Be Adopt-
ed? The Experience of Emerging Markets. In Monetary Policy Under Inflation Targeting, ed. F.
S. Mishkin and K. Schmidt-Hebbel. Santiago: Banco Central de Chile.
Benes, J., K. Clinton, A. George, J. John, O. Kamenik, D. Laxton, P. Mitra, G.V. Nadhanael,
H. Wang, and F. Zhang. 2017a. Inflation-Forecast Targeting for India: An Outline of the Ana-
lytical Framework. IMF Working Paper 17/32. Washington: International Monetary Fund.
Benes, J., K. Clinton, A. George, P. Gupta, J. John, O. Kamenik, D. Laxton, P. Mitra, G.V. Nad-
hanael, R. Portillo, H. Wang, and F. Zhang. 2017b. Quarterly Projection Model for India: Key
Elements and Properties. IMF Working Paper 17/33. Washington: International Monetary
Fund.
BIS (Bank for International Settlements). 2017. 87th Annual Report (June). Basel.
Charbonneau, K., and L. Rennison. 2015. Forward Guidance at the Effective Lower Bound: Inter-
national Experience. Bank of Canada Staff Discussion Paper 2015–15. Ottawa: Bank of
Canada.
Clinton, K., C. Freedman, M. Juillard, O. Kamenik, D. Laxton, and H. Wang. 2015. Inflation-
Forecast Targeting: Applying the Principle of Transparency. IMF Working Paper 15/132. Wash-
ington: International Monetary Fund.
Clinton, K., T. Hlédik, T. Holub, D. Laxton, and H. Wang. 2017. Czech Magic: Implementing In-
flation-Forecast Targeting at the CNB. IMF Working Paper 17/21. Washington: International
Monetary Fund.
Dincer, N., and B. Eichengreen. 2014. Central Bank Transparency and Independence: Updates
and New Measures. International Journal of Central Banking 10, no. 1: 189–259.
Engen, E., T. Laubach, and D. Reifschneider. 2015. The Macroeconomic Effects of the Federal Re-
serve’s Unconventional Monetary Policies. Finance and Economic Discussion Series 2015–005.
Washington: Federal Reserve Board.
Gaspar, V., M. Obstfeld, R. Sahay, D. Laxton, D. Botman, K. Clinton, R. Duval, K. Ishi, Z. Jakab,
L. Jaramillo Mayor, C. Lonkeng Ngouana, T. Mancini Griffoli, J. Mongardini, S. Mursula,
E. Nier, Y. Ustyugova, H. Wang, and O. Wuensch. 2016. Macroeconomic Management When
Policy Space Is Constrained: A Comprehensive, Consistent and Coordinated Approach to Economic
Policy. IMF Staff Discussion Note 16/09. Washington: International Monetary Fund.
Helbling, T., P. Karam, O. Kamenik, D. Laxton, H. Wang, and H. Yao. 2017. Inflation Targeting
in Australia: Performance, Challenges, and Strategy Going Forward. IMF Country Report 17/43.
Washington: International Monetary Fund.

340 SUSTAINING ECONOMIC GROWTH IN ASIA

IMF/FSB/BIS (International Monetary Fund/Financial Stability Board/Bank for Internation-
al Settlements). 2016. Elements of Effective Macroprudential Policies: Lessons from International
Experience (August). Washington: International Monetary Fund.
Kydland, F., and E. Prescott. 1977. Rules Rather than Discretion: The Inconsistency of Opti-
mal Plans. Journal of Political Economy 85, no. 3: 473–92.
Svensson, L. 1997. Inflation Forecast Targeting: Implementing and Monitoring Inflation Tar-
gets. European Economic Review 41, no. 6: 1111–46.
Woodford, M. 2005. Central-Bank Communication and Policy Effectiveness. Paper presented
at the Federal Reserve Bank of Kansas City Symposium, Jackson Hole, WY, August 25–27.

MONETARY POLICY IN THE NEW MEDIOCRE 341

 15
Avoiding a New Mediocre in
Asia: What Can Fiscal Policy
Do?
ANA CORBACHO, DIRK MUIR, MASAHIRO NOZAKI,
AND EDDA ZOLI

Several countries in Asia face the risk of settling into a new mediocre of
lower growth. Rapid population aging and weak productivity are increas-
ingly taking a toll on medium- to long-term economic potential. There is
a lot fiscal policy can do to address the risk of a new mediocre. Investing in
productive public infrastructure, lowering distortionary taxes, and reform-
ing labor markets are prominent avenues through which fiscal policy can
improve productivity and growth paths. These fiscal initiatives can be costly,
however. Countries already experiencing symptoms of the new mediocre
face the double challenge of absorbing higher levels of age-related spending
and coping with low productivity growth.
This chapter analyzes the implications of the new mediocre for
medium- to long-term fiscal sustainability and quantifies the potential of
fiscal reforms to improve growth. It is organized as follows. The first section
examines the effects of the new mediocre on long-term projections of public
debt. The second section examines whether current public spending and
revenue policies are aligned with the challenges of the new mediocre. The
third section illustrates the growth payoff of fiscal reform packages in three
areas: public infrastructure, tax rebalancing, and labor markets. The last

Ana Corbacho is division chief at the Intenational Monetary Fund (IMF). Dirk Muir, Masahiro Nozaki, and
Edda Zoli are senior economists at the IMF. The authors thank Medha Madhu Nair for valuable research as-
sistance; Francis Landicho for coordination of the editing and production; and Thomas Helbling, Kenneth Kang,
Junghun Kim, Changyong Rhee, Markus Rodlauer, and IMF country teams for useful comments and feedback.
The views expressed in this chapter are those of the authors and do not necessarily represent the views of the IMF,
its Executive Board, or its management.

343
section discusses the design of fiscal institutions and key reforms to address
the new mediocre and boost growth in Asia.

Public Debt Dynamics under the New Mediocre

Asia is expected to have generally favorable public debt dynamics in the
near to medium term, for two reasons. First, public debt is lower than in
other regions. In 2016 the median public debt-to-GDP ratio was 40 percent
in advanced Asia and 42 percent in developing Asia—lower than in other
advanced economies (66 percent), Latin America (50 percent), and emerging
Europe (44 percent) (figure 15.1). Japan, with a debt ratio of 240 percent in
2016, is an outlier in Asia. Second, with a few exceptions (including China),
public debt-to-GDP ratios in Asia are projected to decrease or remain stable
over the medium term (figure 15.2).1
Population aging in the new mediocre can pose a challenge for public
debt sustainability over the long term. The speed of aging in Asia is remark-
ably high compared with the historical experience of Europe and the United
States. It took 26 years in Europe and more than 50 years in the United
States for the old-age dependency ratio to increase from 15 to 20 percent;
it will take China, Korea, and Thailand less than a decade to complete the
same demographic transition (IMF 2017a). Moreover, several countries
in Asia are likely to face the fiscal costs of aging at relatively low levels of
income per capita. Age-related public spending on pensions and health care
is thus projected to increase substantially in several Asian economies. If the
spending increase is financed by borrowing, public debt will rise.
Prospects for age-related spending differ significantly across countries.
Japan, which already has the oldest population in the world, spent 20 percent
of GDP on pensions and health care in 2015 (figure 15.3). In China, Korea,
and Thailand, age-related spending is projected to rise over the next several
decades, approaching Japan’s current level by 2050. India, Indonesia, and
the Philippines will see their working-age populations grow substantially
in coming decades. As they will continue to reap a demographic dividend,
their age-related spending should remain contained over the long term.
Aside from the fiscal cost, aging will slow potential growth, through its
negative impact on the labor force and on trend total factor productivity
(TFP). The impact of aging on the labor force could reduce annual growth
by about 0.5 to 0.75 percentage points in China, Korea, and Thailand. For
the region as a whole, aging could slow annual trend TFP growth by 0.1 to
0.3 percentage point (IMF 2017a).

1. Projections are based on the World Economic Outlook of April 2017 (IMF 2017c).

344 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 15.1 Public debt as percent of GDP, by region, 2016

percent of GDP
250
10–90th percentile band
200 Median

150

100

50

0
Advanced Other Developing Emerging Latin Sub-Saharan Middle East
Asia advanced Asia Europe America Africa and
North Africa
Sources: IMF, World Economic Outlook April 2017 database; IMF Staff calculations.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 345
 Figure 15.2 Actual and projected public debt in selected
Asian economies

percent of GDP
120
2016
100 2022

80

60

40

20

0
ia a h s r a d a ia s ia a e
es di es ine ma ore ilan hin ys nam Lao Ind ank por
n bo lad ipp a n K a C la iet L a
do m g il y Th M
a
V i
Sr Sin
g
In Ca Ban Ph M

Sources: IMF, World Economic Outlook April 2017 database; IMF Staff
calculations.

Figure 15.3 Actual and projected public spending on pensions

and health care in selected Asian economies

percent of GDP

25
2015
2030
20 2050

15

10

0
a ia s
di es ne sia ka in
a
am re
a nd pa
n
In n pi ay an Ch tn Ko ila Ja
do p al L e a
In ili M Sri Vi Th
Ph

Sources: United Nations (2015); Amaglobeli and Shi (2016); IMF Staff calculations.

The slowdown in productivity growth also has a bearing on public debt

dynamics. As in other regions, productivity growth in many parts of Asia
has slowed since the global financial crisis, as a result of sluggish invest-

346 SUSTAINING ECONOMIC GROWTH IN ASIA

 Box 15.1 Country groupings
This chapter examines the implications of population aging and
weak productivity for fiscal policy in Asia. In doing so, it recog-
nizes that these features of the new mediocre vary across coun-
tries. Broadly speaking, countries in Asia can be cast into three
main groups, depending on the prominence of these features of
the new mediocre:1

n Group 1 includes Japan, the most affected country.

n Group 2 includes economies in which the demographic shift has

already taken place and productivity growth is slowing. This
group comprises China, Hong Kong, Korea, Singapore, and Thai-
land.2

n Group 3 includes developing economies that are reaping a

demographic dividend but are subject to potentially negative
spillovers from a slowdown elsewhere in the region. This group
comprises Bangladesh, Cambodia, India, Indonesia, Laos, Ma-
laysia, Maldives, Mongolia, Myanmar, the Philippines, Sri Lanka,
Vietnam, and the Pacific Island countries.

1. The number of countries analyzed in each section of this chapter is based

on data availability.
2. For an overview of demographic and productivity trends across Asia,
see IMF (2017a, 2017b).

ment, little impetus from trade, weaker human capital formation, and the
reallocation of resources to less productive sectors of the economy. The pro-
ductivity slowdown has been most pronounced in the advanced economies
of the region and China (IMF 2017b). The decline in real interest rates ob-
served since the 1980s can be expected to reverse only moderately over the
medium term (IMF 2014c).
This section presents long-term projections for public debt reflecting
the impact of the new mediocre. The analysis covers 10 Asian economies
(chosen based on data availability): Japan (Group 1); China, Korea, and
Thailand (Group 2); and India, Indonesia, Malaysia, the Philippines, Sri
Lanka, and Vietnam (Group 3). (See box 15.1.)

Methodology
The ratio of public debt to GDP(dt) evolves as follows:
𝑖𝑖��� � ����
𝑑𝑑��� � 𝑑𝑑� � 𝑑𝑑 � �����,,
� � ���� �

CHAPTER
AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL 15—GRAPHICS
POLICY DO? 3471
 𝑖𝑖��� � �
𝑑𝑑��� � 𝑑𝑑� �
� � ���
where it, gt, and pbt denote the nominal interest rate on public debt (com-
puted as interest payments divided by the previous year’s debt outstanding),
nominal GDP growth, and the primary balance as percent of GDP, respec-
tively. The debt ratio increases at the rate of ���� �����
������ (the interest rate growth
differential [IRGD] divided by (1 + gt+1)) and decreases with the primary
surplus.
Four scenarios are considered, based on extrapolation of the projec-
tions in the IMF’s World Economic Outlook April 2017 database (IMF 2017c)
for 2022–50:2
n Scenario 1: Status quo. The primary balance, nominal GDP growth, and
nominal interest rate for 2023–50 are assumed to remain at 2022 levels.
The IRGD is negative for all countries. The primary balance is in surplus
in Korea, Malaysia, and Sri Lanka and in deficit for the remaining coun-
tries (table 15.1).
n Scenario 2: Historical primary balance. The primary balance for 2023–50
is assumed to reflect its historical average over 2000–15. This scenario
examines cases in which the projected primary balance for 2022 appears
optimistic compared with the historical track.3 GDP growth and the
interest rate for 2023–50 are unchanged at their 2022 levels.
n Scenario 3: Demographics. The primary balance is assumed to worsen
from the 2022 level by the projected increase in pension and healthcare
spending.4 GDP growth from Scenario 1 is adjusted by the impact of
aging on the underlying labor force.5 The interest rate is left unchanged
at the 2022 level.
n Scenario 4: Productivity and interest rates. Productivity growth could con-
tinue to slow. Together with the demographic shift, it could cause the

2. These scenarios correspond to stylized simulations under certain assumptions for the key
parameters that underlie the evolution of the public debt-to-GDP ratio and should not be
considered the baseline projections of the IMF. For the latter, refer to the respective Article
IV Staff Reports.
3. In some countries (e.g., India), the lower primary balance projected by 2022 is supported
by structural reforms driving a decline in primary fiscal spending.
4. Pension and healthcare spending projections are based on United Nations (2015) and
Amaglobeli and Shi (2016). These projections incorporate mainly the impact of demographics
on age-related spending, holding constant the coverage as well as replacement ratios, except
in countries with ongoing parametric reforms. Reforms that improve coverage or benefits
would translate into higher spending. The IMF Fiscal Monitor (www.imf.org/en/publications/
fm) publishes estimates of the net present value of the projected increase in pension and
healthcare spending over 2015–50.
5. The labor force is assumed to grow at the rate of the working-age population, computed
from United Nations (2015).

348 SUSTAINING ECONOMIC GROWTH IN ASIA

2 AVOIDING
Table 15.1 Key parameters for long-term public debt projections in selected Asian economies

CHAPTER
Projection for 2022 Historical average for 2000-15
Real Real
interest Interest interest Interest
Primary rate on rate Primary rate on rate
balance Real GDP public growth balance Real GDP public growth
(percent growth debt differential (percent growth debt differential
Country Group of GDP) (percent) (percent) (percent) of GDP) (percent) (percent) (percent)

A 15—GRAPHICS
Japan 1 –2.0 0.6 –0.6 –1.2 –5.8 0.9 1.9 1.0
China 2 –2.1 5.7 –0.3 –6.0 –0.8 9.6 –2.0 –11.6
Korea 2 1.7 3.1 1.9 –1.2 1.5 4.3 2.3 –2.0
Thailand 2 –0.6 3.0 1.7 –1.3 0.4 4.1 0.1 –4.0
India 3 –1.3 8.2 3.4 –4.8 –3.7 7.1 2.2 –4.8
Indonesia 3 –0.8 5.5 2.3 –3.2 1.1 5.4 –3.4 –8.8
Malaysia 3 0.9 4.8 2.2 –2.6 –2.5 5.1 2.0 –3.1
Philippines 3 –0.4 7.0 1.5 –5.5 2.2 5.1 3.7 –1.4
Vietnam 3 –2.2 6.2 –0.7 –6.9 –2.0 6.5 –6.7 –13.2
Sri Lanka 3 1.2 5.3 2.2 –3.1 –1.7 5.7 –1.5 –7.2
Sources: IMF World Economic Outlook April 2017 database; IMF Staff calculations.

NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 349

 natural interest rate to fall. The net effect on the IRGD for debt dy-
namics is ambiguous. Two scenarios are therefore considered, both of
which assume the primary balance is as in Scenario 3. In Scenario 4a,
the IRGD decreases gradually to 1 percentage point below the 2022
level by 2030, as interest rates charged on public debt fall by more than
growth. In Scenario 4b, the IRGD rises gradually, reaching 2 percentage
points above the 2022 level by 2030, as the sovereign risk premium in-
creases with rising public debt and/or conditions underlying negative
IRGDs (e.g., financial repression) reverse.

Some caveats are in order. First, Scenarios 3, 4a, and 4b assume that the
increase in age-related spending will be fully financed by government bor-
rowing. But governments would likely react by rationalizing spending, mo-
bilizing revenue, or both. Second, an acceleration of problems in later years
makes the response of governments even more uncertain and difficult to
assess. Third, when considering the IRGD, whether sovereign risk premia can
remain contained in countries with a rapid increase in public debt is ques-
tionable and can modify the results. Fourth, the methodology abstracts from
the impact of exchange rate movements on the debt ratio or any fiscal risks.

Results
Countries’ public debt-to-GDP ratios develop as follows under each scenario
(figure 15.4):
n Under Scenario 1, the debt ratio decreases or stabilizes in all econo-
mies, driven by the negative IRGD. The debt ratio is contained well
below 50 percent of GDP in the long term for all economies except
Japan, where the debt ratio is projected to decrease but remain above
200 percent of GDP.
n Under Scenario 2, the debt ratio increases in Japan, India, and Malaysia
and remains flat in Sri Lanka, where the historical primary deficit was
substantially higher than the level projected for 2022.
n Under Scenario 3, pension and healthcare spending impose a signifi-
cant fiscal burden on the rapidly aging Asian economies. In China,
Japan, and Thailand, the debt ratio rises shortly after 2022. In China
and Thailand it exceeds 100 percent by 2050. In Korea and Vietnam, the
debt ratio hits its nadir around 2030 and then rises to 110 percent in
Korea and 115 percent in Vietnam by 2050. The debt paths for Malaysia
and Sri Lanka are well above those of Scenario 1. For India, Indonesia,
and the Philippines—the youngest economies in the sample—the impact
of aging on the debt path is much more muted, with the debt ratio for
2050 exceeding that in Scenario 1 by only about 15 percentage points.

350 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 15.4 Actual and projected public debt in selected Asian
economies, 2000–50 (percent of GDP)

Japan China
450 160
400 140
350 120
300 100
250 80
200
60
150
40
100
20
50
0
0
2000 2010 2020 2030 2040 2050
2000 2010 2020 2030 2040 2050

Korea Thailand
160 250
140
120 200
100
150
80
60 100
40
20 50
0
–20 0
–40 2000 2010 2020 2030 2040 2050
2000 2010 2020 2030 2040 2050

India Indonesia
90 100
80
80
70
60 60
50
40
40
30 20
20
0
10
0 –20
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050

Scenario 1: status quo (g, i, pb at 2022 levels)

Scenario 2: historical average (pb at 2000–15 average, g and i
at 2022 levels)
Scenario 3: aging impact (high pension and healthcare spending
and low growth)
Scenario 4a: aging impact plus lower IRGD
Scenario 4b: aging impact plus higher IRGD
(figure continues)

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 351
Figure 15.4 Actual and projected public debt in selected Asian
economies, 2000–50 (percent of GDP) (continued)

Malaysia Philippines
90 80
80
70 60
60
40
50
40 20
30
20 0
10
0 –20
2000 2010 2020 2030 2040 2050
–40
2000 2010 2020 2030 2040 2050

Vietnam Sri Lanka

160 100
140 90
80
120
70
100 60
80 50
40
60
30
40 20
20 10
0 0
2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050

Scenario 1: status quo (g, i, pb at 2022 levels)

Scenario 2: historical average (pb at 2000–15 average, g and i
at 2022 levels)
Scenario 3: aging impact (high pension and healthcare spending
and low growth)
Scenario 4a: aging impact plus lower IRGD
Scenario 4b: aging impact plus higher IRGD

i, g, pb = nominal interest rate on public debt (computed as interest payments

divided by the previous year’s debt outstanding), nominal GDP growth, and the
primary balance as percent of GDP, respectively; IRGD = interest rate growth
differential
Sources: IMF World Economic Outlook April 2017 database; IMF Staff calculations.

n Under Scenario 4a, the gradual decline in the IRGD does not have a
large impact on the debt path, except in Japan, where the debt ratio in
2050 falls by 50 percentage points compared with Scenario 3, reflecting
its elevated level.
n Under Scenario 4b, the gradual but significant increase in the IRGD
raises the debt-to-GDP ratio in 2050, particularly in Japan (by 140 per-
centage points compared with Scenario 3) and Thailand (40 percentage

352 SUSTAINING ECONOMIC GROWTH IN ASIA

 points). A potential reversal of the negative IRGD constitutes a signifi-
cant upside risk for the debt path.6

Figure 15.5 decomposes the change in the debt ratio over 2022–50 under
Scenario 4b into three factors: the status quo primary balance as of 2022,
the increase in age-related spending, and the IRGD. In Japan, where the debt
level is already very high, the stability of the debt path hinges on maintain-
ing a favorable IRGD and a primary balance in the face of still rising age-re-
lated spending. China, Korea, and Thailand (Group 2) face significant fiscal
pressures from aging in the long term, but Korea has kept its debt level low
and will have more time to address rising age-related spending than China
and Thailand. Group 3 countries are more diverse. Indonesia and the Philip-
pines have built fiscal buffers, and demographic prospects do not appear to
pose a threat to fiscal sustainability. Stabilizing debt paths in India, Malay-
sia, and Sri Lanka hinges on a negative IRGD (India) or a behavioral shift
from the past to maintain a primary surplus over the long term (Malaysia
and Sri Lanka). Vietnam’s debt prospects look similar to China’s, with de-
mographics negatively affecting debt dynamics relatively early.

Are Public Spending and Revenue Policies Aligned with

the Challenges of the New Mediocre?
Fiscal policy can play an important role in supporting sustained and inclu-
sive growth (IMF 2014c, 2015a). Aside from its overarching role in ensuring
macroeconomic stability, fiscal policy can boost potential growth through
the following channels:
n physical capital accumulation, through efficient public investment, es-
pecially in infrastructure (Buffie et al. 2012; IMF 2014c, 2015b), and
taxation of capital and other incentives;
n labor supply, through provision of social benefits for specific groups
(e.g., public spending for child care and active labor market policies);
taxation of labor income to encourage labor force participation (IMF
2014a); and spending on education and health care to boost human
capital accumulation;
n trend TFP, through spending on research and development (R&D); tax
incentives to encourage private R&D or new business ventures that spur

6. Theoretically, a negative IRGD implies that the economy is dynamically inefficient (i.e.,
reducing the saving rates would leave all generations better off). Mankiw (2015) argues that
such a Pareto-improving opportunity is unlikely to be sustained. Escolano, Shabunina, and
Woo (2017) attribute negative IRGDs in developing economies to real interest rates well
below market equilibrium, reflecting financial repression and captive and distorted markets.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 353
354
Figure 15.5 Decomposition of change in debt ratio under Scenario 4b

percent of GDP

250
200
150
100
50
0
–50
–100

SUSTAINING ECONOMIC GROWTH IN ASIA

–150
Japan China Korea Thailand India Indonesia Malaysia Philippines Vietnam Sri Lanka

Primary deficit, status quo in 2022

Age-related spending increase
Interest rate growth differential
Change in debt, 2022–50, Scenario 4b

Sources: IMF, World Economic Outlook April 2017 database; IMF Staff calculations.
 innovation (IMF 2016); and fiscal policies for labor and product market
reforms that increase productivity (IMF 2014b, Banerji et al. 2017).

This section presents stylized facts on the level, composition, and effi-
ciency of public spending and tax systems in Asian economies, benchmarked
against countries from other regions at similar income levels.

Public Spending Level and Efficiency

Infrastructure. In many Asian economies, there is scope to increase the
level and quality of infrastructure spending, through public investment and
public-private partnerships (PPPs). In Japan (Group 1) and the advanced
economies in Group 2, the quality of infrastructure exceeds that in peer
economies. In some developing economies in Group 2 and Group 3, it has
room to catch up to levels in emerging markets in Europe and Latin America
(figure 15.6).7 Most developing economies in Asia in Group 2 and Group 3
lie within the public investment efficiency frontier (figure 15.7).8
Health Care. Despite increases in per capita income since the early 2000s,
public spending on healthcare remains modest in Asia (except in Japan)
(figure 15.8), although private healthcare expenditure is broadly at similar
levels.9 Public health care spending in Hong Kong, Korea, and Singapore
reached at most 4 percent of GDP—less than half the average of 8.5 percent
of GDP in advanced economies outside of Asia. In most Asian developing
countries, public healthcare spending is below the average in emerging
Europe and Latin America.10
The efficiency of public healthcare spending is also low in parts of Asia:
Group 3 countries fall well within the frontier (figure 15.9), indicating that
better health outcomes could be achieved with the same level of healthcare
spending.
Education. The level and efficiency of public spending on education is
low in some parts of Asia (figure 15.10). Public spending on education in
advanced economies in Group 2 is somewhat lower than in other countries

7. Emerging Europe includes countries belonging to the Commonwealth of Independent

States.
8. The efficiency frontier is calculated using Data Envelopment Analysis techniques, which
use linear programming methods. See appendix 15A.
9. In Group 2 and Group 3 countries, private healthcare spending amounted to 2.3 and
2.6 percent of GDP, respectively, in 2015, compared with 1.8 percent of GDP in advanced
economies outside Asia and about 3 percent of GDP in emerging markets in other regions.
10. In Indonesia, Korea, and Thailand, the gap between actual and expected public spending
on health care was 0.8–1.7 percentage points of GDP in 2015 (Phillips et al. 2013).

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 355
356
Figure 15.6 Index of quality of infrastructure

index (1 is worst, 7 is best)

SUSTAINING ECONOMIC GROWTH IN ASIA

0
n a g a e d sh dia sia dia ka lia sia es am ed pe ica
pa in on re or an
nc uro er
Ja Ch K Ko gap hail la de bo ne In Lan ngo lay pin etn
o i a m
ng n T Sr M
o Ma ilip Vi dv g E A
Ho Si B a ng Cam Ind
Ph
e r a gin atin
r
th e L
O Em

Sources: IMF, World Economic Outlook April 2017 database; World Economic Forum; IMF Staff calculations.
 Figure 15.7 Efficiency frontier of public investment

indicator of quality of infrastructure

(100 = average in advanced economies)
140

120

100

80

60

40
Frontier Group 1
Group 2 Group 3
20 Emerging Europe Latin America
Other advanced Other
0
0 10,000 20,000 30,000 40,000 50,000 60,000
public capital stock per capita (dollars)

Note: The efficiency frontier is calculated using Data Envelopment Analysis

(DEA) techniques (described in appendix 15A). The distance to the efficiency
frontier Indicates the size of the loss in infrastructure quality as a result of
inefficiencies in public investment (IMF 2015b).
Sources: IMF, World Economic Outlook April 2017 database; World Economic
Forum; IMF Staff calculations.

at similar income levels. For developing countries in Group 2 and Group

3, public spending on education is generally lower than in peers in Latin
America. Based on secondary school enrollment (figure 15.11, panel a) and
PISA scores (figure 15.11, panel b), most Group 2 and 3 countries lie within
the efficiency frontier (Korea is on the frontier for PISA scores).
R&D spending. Data on R&D spending are limited. They seem to indicate
that Group 2 economies except Hong Kong are roughly comparable to their
peers in other regions but that Group 3 countries (most notably Indonesia,
the Philippines, and Sri Lanka) lag their peers (figure 15.12).
Some public R&D spending is not effective. In Korea, for instance, high
R&D expenditure has not been accompanied by sustained gains in produc-
tivity because of weak links between government research institutes, univer-
sities, and industry that hinder technology transfers and commercialization
(Jones and Kim 2014).

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 357
Figure 15.8 Public spending on health care as percent of GDP in
selected economies, 2001 and 2014

percent of GDP

9
8
7
6
5
4
3
2
1
0

pa
n a g a e
in on re or an
d sh dia sia dia os ka lia sia es am ed pe ica
Ja Ch K Ko gap hail a de bo ne In La Lan ngo lay pin etn a nc uro er
l o i a i p i v E m
ng n T ng am Ind
o
Sr M M hil V ad ing in A
Ho Si Ba C P r
e g at
th er L
O Em
2014
2001

Sources: IMF, World Economic Outlook April 2017 database; World Economic Forum;
IMF Staff calculations.

Figure 15.9 Efficiency frontier of public spending on

health care, 2015

healthy life expectancy (years)

80

75

70

65

60

55

50

45 Frontier Group 1
Group 2 Group 3
40 Other advanced Latin America
Emerging Europe Other
35
0 1,000 2,000 3,000 4,000 5,000
healthcare expenditure per capita
(purchasing power parity dollars)

Sources: IMF, World Economic Outlook April 2017 database; World

Economic Forum; IMF Staff calculations.

358 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 15.10 Public spending on education as percent of GDP in
selected economies, 2001 and 2014

percent of GDP
8
7 2014
2001
6
5
4
3
2
1
0

pa
n a g a e d
in on re or an sh dia sia dia os ka sia es d e
ce rop eric
a
Ja Ch K Ko gap hail a de bo ne In La Lan lay pin a n u
l o i a ip m
ng n T ng Cam Ind Sr M hil dv g E A
Ho Si Ba P r a gin atin
e
th er L
O Em

Sources: IMF, World Economic Outlook April 2017 database; World Economic
Forum; IMF Staff calculations.

Revenue Collection and Composition

Group 2 economies will face increasing spending pressures from demo-
graphic trends, and Group 3 economies need to step up public spending
on infrastructure, health care, and education. Is there scope to expand the
revenue intake?
In several of these economies, general government revenue ratios are
low compared with other economies at similar income levels (figure 15.13).
In the advanced economies of Group 2, revenue-to-GDP ratios are almost
20 percentage points lower than in advanced economies in other regions.
In emerging economies from Group 2 and Group 3, revenue-to-GDP ratios
average 7 and 13 percentage points less than in peers from Latin America
and Europe, respectively.
The relatively low revenue intake partly reflects an explicit policy choice
in favor of a low tax environment. However, tax yields, which are indicators
of the efficiency of the tax system, also tend to be low in several Asian econo-
mies. Figure 15.14 shows two measures of personal income tax (PIT) yields:
PIT revenues as a percent of GDP divided by the middle and the maximum
statutory PIT.
Three factors contribute to low tax yields: narrow tax bases, reflecting
widespread exemptions, which are often inefficient (IMF 2012, World Bank
2012); lingering weaknesses in tax administration, despite reform efforts
(Araki and Claus 2014); and extensive informality.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 359
360
Figure 15.11 Efficiency frontier of public spending on education

a. Secondary school enrollment

net secondary school enrollment (percent)
120

100

80

60

40

SUSTAINING ECONOMIC GROWTH IN ASIA

Frontier Group 1
20 Group 2 Group 3
Emerging Europe Latin America
Other advanced Other
0
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000
education spending per secondary school student (purchasing power parity dollars)

Sources: IMF, World Economic Outlook April 2017 database; World Economic Forum; IMF Staff
calculations.
 Figure 15.11 Efficiency frontier of public spending on education

b. PISA scores
PISA score
600

500

400

300

200 Frontier Group 1

Group 2 Group 3
Other advanced Emerging Europe
Latin America Other
100
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000
education spending per secondary school student (purchasing power parity dollars)

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO?

PISA = Programme for International Student Assessment
Sources: IMF, World Economic Outlook April 2017 database; World Economic Forum; IMF Staff

361
calculations.
Figure 15.12 Spending on research and development in selected
economies, 2015

percent of GDP

0
a a a a a sia es m
pa
n e
in ong re or and sia di nk oli n a ed pe ica
Ja Ch K Ko gap ail o ne In La ng alay ppi etn a nc uro er
i ili Vi m
ng n Th d Sr M
o M dv g E A
Ho Si In Ph r a gin atin
e r
th e L
O Em

Sources: IMF, World Economic Outlook April 2017 database; World Economic Forum;
IMF Staff calculations.

Figure 15.13 Correlation between GDP per capita and

government revenue as percent of GDP, 2015

government revenue as percent of GDP

60

50

40

30

20

10

0
5 6 7 8 9 10 11 12
GDP per capita (US dollars in logs)
Group 1 Group 2
Group 3 Other advanced (average)
Emerging Europe Latin America (average)
(average) Other

Sources: IMF, World Economic Outlook April 2017 database; World Economic
Forum; IMF Staff calculations.

There is scope to enhance the efficiency of individual taxes and tax

collection in Asia; there is also room to improve the tax structure to support
growth. Consumption and property taxes are less harmful to growth than

362 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 15.14 Yields from personal income tax (PIT) in selected
economies, 2015

percent

40

35 Yields based on middle PIT

Yields based on maximum PIT
30

25

20

15

10

0
a h ia ia ia s a ia s
pa
n a e
in ong re or and es od es Ind Lao ank ys ine ed pe ica
Ja Ch K Ko gap ail l ad b on L ala ipp a nc uro er
h i m
n g n T ng Cam Ind Sr M hil d gE A
v
Ho Si Ba P r a gin atin
e
th er L
O Em

Note: Tax yields are defined as income tax revenues in percent of GDP divided by
the middle and the maximum statutory tax rates.
Sources: IMF, World Economic Outlook April 2017 database; IMF, Government
Finance Statistics; IMF Staff calculations.

taxes on factors of production, such as personal or corporate income tax

(Arnold et al. 2011). A broad-based, single-rate VAT is regarded as relatively
efficient, because it does not introduce large distortions in relative prices or
saving decisions (Ebrill, Keen, and Summers 2001).
Relative to advanced and developing economies outside Asia, many
Asian countries rely more on personal and corporate income tax than on
general consumption taxes (figure 15.15). Tax rebalancing away from direct
taxes on labor and capital toward indirect taxes on consumption could
boost economic growth.

Quantifying the Payoff of Fiscal Policy Reforms

This section quantifies the growth payoff of fiscal reforms that reallocate
expenditure and revenue composition toward growth- and productivity-
enhancing items. The fiscal packages cover three areas: public infrastruc-
ture, tax rebalancing, and labor markets. The purpose of these packages is
to raise long-term growth, through higher TFP, labor productivity, or labor
force participation. A more flexible labor force, better infrastructure, and
a more favorable tax environment should also allow economies to adjust
more rapidly to shocks, increasing resilience and stability.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 363
364
Figure 15.15 Sources of tax revenue in selected economies, 2015

percent of total tax revenue

100
90
80
70
60
50
40
30
20
10
0

SUSTAINING ECONOMIC GROWTH IN ASIA

n a a e h ia ia ia s a ia ia s ed pe ica
pa in ong re or and es od es nd Lao ank gol ys ine nam
nc uro er
Ja Ch K Ko gap ail
g h lad b on I i L on ala ipp iet v a E m
n n T ng am Ind Sr M M hil V d g A
Ho Si Ba C P
e r a gin atin
th er L
Corporate income tax O Em
Personal income tax
Property tax
Goods and services taxes
Other

Sources: IMF, World Economic Outlook April 2017 database; IMF, Government Finance Statistics;
IMF Staff calculations.
 The simulations are based on the IMF’s APDMOD, a semi-structural
model of the global economy that includes blocks for 16 Asian countries
and 8 regions representing the rest of the world.11 The model’s fiscal sector
includes spending on consumption; infrastructure; transfers; lump-sum
taxation on households; and distortionary taxation on consumption, labor,
and capital. Only some households (“saving” households) hold debt as a
source of wealth, which allows them to smooth consumption in the face of
shocks or policy changes. Other households cannot save effectively; these
liquidity-constrained households live off their current income.12 These non-
Ricardian properties allow a powerful role for fiscal policy in both the short
and long term. For most countries, monetary policy is assumed to follow an
inflation-targeting regime.
By implementing fiscal policy reforms, Asian countries can achieve
stronger growth and productivity, although some measures impose short-
term costs on households and firms. These costs can be mitigated by ad-
dressing adverse effects on income equality and by phasing implementation
over time. Better results may be achievable by taking advantage of synergies
with monetary policy in countries facing low inflation.13

Illustrative Fiscal Reform Packages

Most of the simulations of the three packages are implemented in a budget-
neutral manner. The size of the measures is illustrative and not necessarily
optimal. Policies account for idiosyncratic country features and tackle the
main gaps in the composition of tax and expenditure policies identified in
the previous section. Measures are notionally carried out relative to a base-
line scenario consistent with the IMF’s World Economic Outlook (IMF 2017c).
Appendix 15B provides details on the fiscal packages in each economy.
The three packages include the following:
1. Public infrastructure push. Infrastructure investment increases by 1 percent
of GDP for five years in developing countries in Groups 2 and 3, driving

11. APDMOD is part of the Flexible System of Global Models (FSGM). See Andrle et al.
(2015) for a complete treatment of the model.
12. APDMOD assumes that 60 percent of households in Asian countries are liquidity
constrained, except in Hong Kong, Japan, Korea, and Singapore, where liquidity-constrained
households represent just 35 percent of the population. Liquidity-constrained households
account for 20 to 30 percent of total consumption in the advanced Asian economies and 45
to 55 percent in the others.
13. For a discussion of the concepts of coordination and comprehensiveness, particularly
around the use of spending, tax, and labor measures, see Gaspar, Obstfeld, and Sahay (2016).

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 365
 up the size of the public capital stock by 5 percentage points.14 Invest-
ment then increases permanently by 0.2 percent of GDP to replace the
loss from an annual depreciation rate of 4 percent. Economies near the
frontier or in which spending is already high (the advanced economies
and China) do not participate.
Financing varies across the region. In many of the economies
analyzed, half of it is in the form of PPP, accounted for as private invest-
ment and debt and recouped by user fees.15 The other half is traditional
public investment, financed by a temporary reduction in general trans-
fers. In two countries (Cambodia and the Philippines), the fiscal deficit
is allowed to increase.
2. Rebalancing toward indirect taxes. VAT collection increases by 1.5 percent
of GDP. It is offset by cuts of 0.75 percent of GDP to both the personal
and the corporate income tax.16 The increase excludes additional tax
changes needed to finance labor sector reforms (see below). Some econ-
omies do not participate fully or at all, either because doing so would
conflict with other goals (Japan and India, with baseline scenarios that
already factor in VAT increases) or because direct taxes already repre-
sent a small share of tax revenues (Bangladesh, China, Hong Kong, and
Sri Lanka).
For countries in Group 3, liquidity-constrained households receive
a 0.5 percent of GDP rebate, to address the adverse impact of tax reform
on income inequality. To achieve a net VAT increase of 1.5 percent of
GDP, VAT collection is increased by 2 percent of GDP. In addition,
the cut in personal income tax is progressive, benefiting only liquidity-
constrained households.
3. Labor sector reforms. All countries participate in the following reforms:
n The retirement age is increased by two years. Average retirement
ages are currently 57 in Asia, 61 in emerging-market economies in
Europe, 62 in Latin America, and 64 years in advanced economies

14. The exception in Group 3 is India, which has limited scope to reallocate spending within
the budget envelope.
15. The exception is Bangladesh, whose PPP framework needs further development, where
the assumption is that 100 percent of infrastructure is built through public procurement
and accounted for as public investment.
16. In APDMOD the personal income tax is combined with social security taxes. Simulations
here consider only changes to the personal income tax.

366 SUSTAINING ECONOMIC GROWTH IN ASIA

 outside Asia. The longer working life raises the labor force participa-
tion rate of people 55–64 (and hence the overall participation rate).17
n Childcare spending is permanently increased by 0.25 percent of
GDP a year, leading to higher female labor force participation.18
n Active labor market policies permanently increase government
spending by 0.25 percent of GDP a year, lowering long-term un-
employment and consequently raising labor productivity.19
The financing of labor sector reforms varies across economies. It
includes a slight increase in VAT in Bangladesh and Japan, an increase
in the fiscal deficit in Cambodia and the Philippines, an increase in
excise taxes in India (where the personal income tax base is small),
and an increase in personal income tax elsewhere. In China, Thailand,
and most Group 3 countries, the increase in personal income tax is
progressive, exempting liquidity-constrained households.
Other avenues to improve productivity and long-term growth include
public spending on R&D, education, and health care. The model does not
capture these components directly, but some of their effects are implied. For
example, some infrastructure investment, such as building high-tech facili-
ties, is related to R&D; spending on schools and medical facilities supports
human capital accumulation. Cuts in corporate income tax could be in
the form of tax credits for innovation or R&D. Active labor market poli-
cies generally include adult education, which contributes to human capital
accumulation.

Effects of the Reforms across Asia

If all countries in Asia implemented these fiscal reforms, growth in the
region would be about 0.3 percentage point higher during the first three
years and about 0.1 to 0.2 percentage point a year higher from then on once
reforms were completed. In the short term, more than two-thirds of the
increase in growth would come from labor sector reforms. In the long term,
investment in infrastructure is the largest contributor to growth.

17. Figures are based on estimates in Bouis and Duval (2011), originally calculated for the
OECD.
18. Figures are based on estimates in Bassanini and Duval (2006), originally calculated for
the OECD.
19. Figures are based on estimates in Bouis and Duval (2011), originally calculated for the
OECD.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 367
 Figure 15.16 Projected effects of fiscal measures on real
GDP in selected Asian economies

a. Asia
percentage point difference in
real GDP associated with fiscal measure
2.4
2.0
1.6
1.2
0.8
0.4
0
–0.4
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027

b. Group 1 (Japan)
percentage point difference in
real GDP associated with fiscal measure
2.4
2.0
1.6
1.2
0.8
0.4
0
–0.4
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027

Infrastructure investment
Tax reform
Labor sector reform (figure continues)

Panel a of figure 15.16 shows the effects on real GDP. By the end of the
first decade, labor reforms contribute about half of the increase in real GDP;
infrastructure investment and tax reform each contribute about a quarter.
Infrastructure investment has the longest-lasting impact on growth,
through higher productivity. The larger stock of infrastructure lowers
transaction and transportation costs. Healthcare- and education-related
infrastructure can also increase the quality and productivity of the labor
force in the long term.
Rebalancing toward indirect taxes provides most of the benefits to
growth in the short term. A hike in VAT discourages consumption, but a
cut in personal and corporate income tax more than compensates by stim-

368 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 15.16 Projected effects of fiscal measures on real
GDP in selected Asian economies (continued)

c. Group 2
percentage point difference in
real GDP associated with fiscal measure

2.4
2.0
1.6
1.2
0.8
0.4
0
–0.4
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027

d. Group 3
percentage point difference in
real GDP associated with fiscal measure
2.4

2.0

1.6

1.2

0.8

0.4

–0.4
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027

Infrastructure investment
Tax reform
Labor sector reform

Source: IMF Staff projections.

ulating investment and encouraging exports, so that real GDP is higher

overall (the light grey line minus the dark grey line in figure 15.16). Cutting
the personal income tax not only increases the labor supply, it can lower
wage costs (the pre-tax wage) faced by firms. Therefore, all else equal, firms
pay a higher post-tax wage to households. Similarly, a cut in the corporate
income tax encourages investment and capital formation, expanding poten-
tial output, lowering the cost of consumer goods, and increasing firms’
demand for labor to complement their new capital.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 369
 Labor reforms also have a long-term impact (the black line minus
the light grey line in figure 15.16). Raising the retirement age shrinks the
cohort that relies on government pensions and increases the labor supply.
It allows governments to increase the size of pensions per capita without in-
creasing overall spending, which may be crucial in countries where replace-
ment ratios are too low. Provision of child care expands the labor supply by
making it easier for women to work, which can be especially useful in coun-
tries facing downward labor supply pressures from demographics. Active
labor market policies can help lower the long-term unemployment rate. All
three measures expand the labor force and increase tax collection (especially
personal income tax).
The effect of reforms varies across the three country groups. In Group
1 (Japan) benefits accrue only from increasing childcare spending, which
would offset some of the significant impact of demographics; the remaining
impacts are spillovers from other countries’ measures. Group 2 countries
benefit most strongly from labor sector reforms. Outside of China, there is
also strong support from infrastructure investment and tax reform. Group
3 countries gain from both their own reforms and the positive spillovers
from improvements in major countries in the region with which they trade.
Gains are higher than in Group 2: Real GDP is 1.3 percent higher after 10
years for Group 3 countries and 1.2 percent higher for Group 2 countries.
Group 3 has the most balanced gains from each of the reform packages.

Payoff to Fiscal Reforms in a Low-Inflation Environment

In the short term, most of the fiscal package is inflationary, as it stimulates
demand, although there is also counteracting disinflationary pressure from
the expansion of potential output. Interest rates increase somewhat over the
first five years, peaking at about 50 basis points to contain inflation.
A feature of the new mediocre in many countries has been subdued
inflation (figure 15.17).20 Many central banks still have room for inflation to
rise before breaching their targets. They could follow a policy of monetary
accommodation in the short term, which would lead to lower real interest
rates than if monetary policy worked to offset the inflationary impact. Such
a strategy could be justified in countries facing persistently low inflation,
allowing temporary inflation overshooting and convergence to the target
from above.21

20. See Garcia Morales et al. (2018) for a discussion of the Association of Southeast Asian
Nations (ASEAN) countries.
21. Similar results are obtained in a scenario in which, rather than adopting a standard linear
Taylor rule, monetary policy follows a policy reaction function that minimizes a quadratic
loss function of inflation and output gaps.

370 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 15.17 Average number of months inflation rate was below
inflation target in selected Asian economies

Japan

Korea
Thailand

Bangladesh
India
Indonesia
Mongolia
Philippines

0 12 24 36 48 60
number of months below target

Note: Figure excludes effects of changes in indirect taxes on Consumer Price Index
inflation. Data cover January 2013–December 2017.
Sources: IMF, International Financial Statistics; IMF Staff calculations.

Infrastructure spending in Thailand—which is financed by lowering

general transfers, while the central bank conducts policy as usual—illustrates
this point (figure 15.18). If instead the central bank chose to accommodate
the stimulus for five years, real GDP would increase by 1.2 percentage points.
Fiscal policy reinforces this effect, provided that higher debt does not
increase the sovereign risk premium. If the government were to use debt
financing instead of cutting general transfers, it would achieve the same
level of growth with the normal conduct of monetary policy, as consump-
tion would be stronger (figure 15.18, solid light grey line minus solid dark
grey line). Monetary accommodation with debt financing (figure 15.18,
dashed light grey line) would lead to higher real GDP, for two reasons. First,
the lower real interest rate provides more stimulus to private investment.
Second, higher inflation would further increase nominal GDP, reducing the
impact on the debt-to-GDP ratio to 0.8 percent of GDP. Overall, monetary
accommodation with debt financing would lead to real GDP that would be
3.2 percent higher by the fifth year.

Need for Supporting Fiscal Institutions and Structural

Reforms
Given the fiscal pressures exerted by rapid population aging in many coun-
tries, ensuring medium- to long-term sustainability calls for upgrading

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 371
 Figure 15.18 Projected effect of accommodative monetary policy on real GDP, government

372
debt, consumption, and current account balance in Thailand

a. Real GDP b. Government debt

percent difference percentage point of GDP difference
3.5 6.0
3.0 4.0
2.5
2.0 2.0
1.5 0
1.0 –2.0
0.5
0 –4.0
2017 2018 2019 2020 2021 2022 2017 2018 2019 2020 2021 2022

c. Consumption d. Current account

percent difference percentage point of GDP difference
3.5
–0.1
2.5

SUSTAINING ECONOMIC GROWTH IN ASIA

1.5 –0.6

0.5 –1.1
–0.5
–1.6
–1.5 2017 2018 2019 2020 2021 2022
2017 2018 2019 2020 2021 2022

Fiscal package funded by cuts in transfers; Fiscal package funded by deficit financing;
monetary policy conducted normally monetary policy conducted normally
Fiscal package funded by cuts in transfers; Fiscal package funded by deficit financing;
interest rates fixed for five years interest rates fixed for five years

Source: Saenz et al. (2018); IMF Staff calculations.

fiscal institutions and delivering structural reforms. This section focuses on
four issues:
n fiscal rules to provide a credible medium-term anchor,
n structural reforms to ensure the viability of pension and healthcare
systems,
n efficient revenue mobilization and energy pricing reforms to create
fiscal space for spending needs, and
n reforms to state-owned enterprises (SOEs) to unlock growth potential
while managing fiscal risks.

Fiscal Rules
A fiscal rule imposes a long-lasting constraint on fiscal policy by setting
numerical limits on budgetary aggregates, in order to establish a credible
medium- and long-term anchor for fiscal sustainability while mitigating
any resulting short-term economic volatility. There are four types of fiscal
rules: debt rules, fiscal balance rules, expenditure rules, and revenue rules
(Schaechter et al. 2012).
Figure 15.19 shows the share of countries with one of the four types of
fiscal rules in each country group over the 30-year period between 1985 and
2015. Most advanced economies, including Japan and Singapore, adopted
fiscal rules. In contrast, only about a quarter of Asia’s developing econo-
mies (including Indonesia, Malaysia, Maldives, Mongolia, Sri Lanka, and
Thailand) did so—a significantly smaller share than in emerging Europe,
Latin America, or Sub-Saharan Africa.
The design of fiscal rules should be guided by the following principles
and practices (Schaechter et al. 2012), while taking country circumstances
into account:
n Provide broad coverage and ensure strong legislative support. Fiscal rules
should also cover all revenue and expenditure items. Excluding public
investment from a fiscal balance rule weakens the rule’s link to public
debt and can provide incentives for creative accounting. Rules enshrined
in high-level legislation or the constitution can instill greater commit-
ment and policy credibility.
n Be neutral or countercyclical. Rules governing the (unadjusted) fiscal bal­
ance can lead to procyclical fiscal policy. To prevent such an effect, rules
can set limits on the cyclically adjusted or structural fiscal balance (al-
though there are technical challenges in estimating these measures).
An expenditure rule that limits expenditure growth to potential GDP
growth is one way to reduce procyclicality.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 373
 Figure 15.19 Share of countries with fiscal rules, by region,
1985–2015

percent of all countries in region

100

90

80

70

60

50

40

30

20

10

0
85

87

89

91

93

95

97

99

01

05
03

07

09

11

13

15
20
19

20

20

20
19
19

19
19

19
19

19

20

20
20

20
Advanced (34)
Developing Asia (19)
Emerging Europe (23)
Latin America (30)
Middle East and North Africa (22)
Sub-Saharan Africa (46)

Note: Figure excludes supranational rules for countries belonging to

currency unions. The number of countries in each group is shown in
parentheses in legend.
Source: IMF Fiscal Rules Dataset, www.imf.org/external/datamapper/
fiscalrules/map/map.htm; IMF Staff calculations.

n Allow flexibility in the face of large shocks. An escape clause can provide flex-
ibility in dealing with a rare but devastating shock (such as a natural
disaster). A defined trigger and exit strategy should be articulated.
n Create supportive arrangements. Effective implementation of fiscal rules
calls for sound public financial management institutions, including
medium-term budget frameworks, procedural and transparency re-
quirements for budget processes and fiscal authorities, and indepen-
dent fiscal councils tasked with assessing compliance with fiscal rules
(Corbacho and Ter-Minassian 2013).

Reform of Pension and Healthcare Systems

Rising age-related spending in aging Asian economies may result in unfund-
ed liabilities over the long term. This risk can be mitigated by parametric
reforms on pension systems and structural reforms in the healthcare sector.
374 SUSTAINING ECONOMIC GROWTH IN ASIA
 Aging reduces the sustainability of defined-benefit pension systems.
Pensions in China, Japan, Korea, Thailand, and Vietnam—where pension
spending is projected to rise faster than in the rest of the region—are at
least partially defined-benefit systems (Arbatli et al. 2016). A defined-ben-
efit system with unfunded liabilities requires parametric reforms: raising
retirement ages, cutting pension benefits, and/or increasing contribution
rates. Among these options, raising retirement ages is generally perceived as
preferable, given increases in life expectancy in recent decades and the posi-
tive effects of older workers on labor force participation (Clements, Eich,
and Gupta 2014). Many Asian economies have scope to increase retirement
ages (Arbatli et al. 2016). Reforms could accompany enhanced antipoverty
programs for older people and a larger role for private retirement saving
schemes to help offset the potential decline in lifetime retirement income
(Clements et al. 2015).
Such reforms should start now and proceed gradually, in order to
spread the burden across generations. They can be accompanied by auto-
matic adjustment mechanisms that respond to demographic, macroeco-
nomic, and financial developments in a predetermined fashion and without
the need for additional policy intervention. Automatic adjustment mech-
anisms have been embedded in the pension systems of many advanced
economies, including Japan, but not in the rest of Asia (Arbatli et al. 2016).
Another option is to adopt a notional defined contribution pension scheme,
in which contributions are accumulated in an individual account with a
notional rate of return and the pension benefit is calculated by dividing the
pension wealth by the life expectancy at retirement (Arbatli et al. 2016).
Healthcare costs tend to grow faster than GDP, because of population
aging and technological improvements that result in better but less afford-
able services (Clements et al. 2015). The experience in advanced economies
suggests that structural reforms to slow the increase in healthcare costs
include promoting some degree of competition among insurers and service
providers, supporting primary and preventive care, and improving provider
payment systems (Clements et al. 2015).
Given population pressures, policymakers in developing Asia need to
consider how to manage the process of expanding the coverage of pensions
and healthcare systems. In economies with low labor market informality
and efficient revenue administration, it is easier to expand a social insurance
scheme. In economies with high labor informality, general taxes may need
to finance such protection (Jenkner, Clements, and Shang 2012).

Efficient Revenue Mobilization and Energy Pricing Reforms

Many Asian economies will need to mobilize revenue mobilization to in-
crease spending on physical and human capital, health care, and pensions.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 375
Increasing efficiency calls for a larger role for indirect taxes. Closing VAT
collection gaps arising from exemptions and low compliance can enhance
the productivity of the tax without increasing rates. Raising excise taxes on
alcohol, tobacco, and even sugar can be justified on health grounds (IMF
2015a). The regressive impact of higher indirect taxes can be compensat-
ed for by expanding programs that disproportionately benefit the poor.
Rationalizing tax exemptions and preferential tax treatment could mobilize
revenue and promote equity. Revenue administration should be improved
through risk-based compliance strategies, proper segmentation of taxpay-
ers, and simplification of laws and procedures (IMF 2015a).
Continuing to eliminate energy subsidies and introducing environmen-
tal taxes can create fiscal space efficiently and equitably. Energy subsidies
distort consumption and production decisions and are often very poorly
targeted (Coady, Flamini, and Sears 2015). Replacing these subsidies with
targeted social assistance can free up resources for productive spending
while enhancing the redistributive role of fiscal policies. Environmental
taxes such as carbon taxes can create fiscal space while addressing negative
environmental externalities.

Reform of State-Owned Enterprises

Reforming SOEs can boost productivity growth and create new jobs by
increasing market orientation. SOEs have large footprints in emerging
Asia. A composite measure of SOEs’ share in sales, assets, and market values
exceeds 50 percent in China, India, Indonesia, and Malaysia (Kowalski et
al. 2013). Unlocking growth potential requires establishing a level playing
field for SOEs and privately owned enterprises operating on a commercial
basis (OECD 2015). It can be accomplished by increasing the share of SOEs’
profits that go to the government budget, eliminating government subsi-
dies, strengthening governance, and improving SOEs’ commercial orienta-
tion. Ultimately, reform must also include greater tolerance of bankruptcy
or exit of SOEs (Lipton 2016).
At the same time, contingent liabilities emanating from SOEs need to
be managed prudently. Bailouts of troubled SOEs cost governments about 3
percent of GDP on average—15 percent of GDP in the most extreme cases—
between 1990 and 2014 (Bova et al. 2016). Governments can reduce expo-
sure to SOE-related fiscal risks by strengthening governance arrangements,
holding SOE management accountable for performance against financial
and operational targets, mandating financial reporting and audits based
on International Financial Reporting Standards, and legislating explicit
no-bailout clauses for SOEs (IMF 2016, Allen and Alves 2016).

376 SUSTAINING ECONOMIC GROWTH IN ASIA

Concluding Remarks
This chapter analyzed the implications of rapid population aging and weak
productivity for medium- to long-term fiscal sustainability and quantified
the potential of fiscal reforms to improve future growth paths.
Left unchecked, rapid population aging and the related increase in
public pension and healthcare spending may threaten long-term fiscal sus-
tainability in many Asian economies. The scenario analysis conducted in
this chapter shows that the impact of demographics will drive public debt
up over the long term, even under the benign assumption that interest rates
remain low enough and growth high enough. A reversal of the favorable dif-
ferential between interest rates and growth that helped keep debt ratios in
check in recent times would put debt dynamics at further risk. In a highly
uncertain global setting, higher sovereign risk premia or tighter financing
conditions could trigger this risk, particularly for countries with weaker
fundamentals.
Fiscal policy can, and must, address the long-term implications of the
new mediocre. Many countries have scope to spur growth by redirecting
budgets to infrastructure, human capital, and R&D and increasing the effi-
ciency of spending and tax systems. Model-based simulations calibrated
for Asia illustrate the growth payoff of fiscal reform packages that increase
infrastructure spending, rebalance taxes toward VAT, and implement
labor sector reforms. Even budget-neutral packages have the power to raise
growth, in both the short and the long term. Dividends are higher when
reforms exploit synergies, mitigate any adverse on impact inequality, and
generate positive regional spillovers.
The time for action is now. After the global financial crisis, inflation
and interest rates remained low by historical standards but are expected
to rise over the medium term. In this environment, a case can be made for
financing fiscal reforms through debt and for combining fiscal and mon-
etary stimulus. In the example in the chapter, the impact on growth of an
infrastructure stimulus financed with debt under monetary accommoda-
tion could be 2.5 times larger than the same budget-neutral reform without
monetary accommodation. Critical to this result is that sovereign risk
premia be contained. Given pressures from rapid aging in many countries,
credible fiscal anchors and broader institutional and fiscal reforms should
be immediate priorities to secure long-term fiscal sustainability and reap
the growth benefits of fiscal stimulus.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 377
Appendix 15A Public Spending Efficiency Frontiers
Data Envelopment Analysis (DEA) applies linear programming methods
to build an efficiency frontier by connecting the bundles of units (coun-
tries) for which no other unit (country) produces the same or more output
with a given amount of input. For each economy analyzed in this chapter,
table 15A.1 shows the distance from the efficiency frontier, calculated as the
percentage difference in the output variable (e.g., quality of infrastructure,
secondary school enrollment, scores on the Programme for International
Student Assessment (PISA), and healthy life expectancy relative to the fron-
tier at the corresponding level of the input variables (public capital stock,
public spending on education and health care).

Table 15A.1 Distance of selected Asian economies from public

spending efficiency frontier (percent)
Secondary
Group/ Quality of school Healthy life
economy infrastructure enrollment PISA score expectancy
Group 1
Japan –13.7 –2.1 –4.1 0
Group 2
Korea –10.0 –5.2 –5.8 0
Singapore 0 n.a. 0 –1.0
China –29.2 n.a. –14.7 –4.7
Thailand –18.9 –19.9 n.a. –6.4
Group 3
Bangladesh –31.0 –36.9 n.a. –2.3
Cambodia –13.3 n.a. n.a. –7.1
India –30.4 –32.9 n.a. –6.5
Indonesia –31.4 –18.2 –4.7 –3.8
Laos –28.8 n.a. n.a. –6.7
Malaysia –9.6 –30.9 –9.6 –6.7
Mongolia –58.1 –13.2 n.a. –6.8
Philippines –36.2 –19.1 n.a. –7.5
Sri Lanka –17.1 –7.2 n.a. –1.8
Vietnam –36.6 n.a. n.a. –2.2
n.a. = not available; PISA = Programme for International Student Assessment
Sources: Organization for Economic Cooperation and Development; World
Economic Forum; World Health Organization.

378 SUSTAINING ECONOMIC GROWTH IN ASIA

Appendix 15B Composition and Cost of Fiscal Packages
Tables 15B.1 and 15B.2 show the composition and cost of the fiscal pack-
ages examined in this chapter. Spending and tax measures are permanent
(annual) unless stated otherwise; spending measures are phased in over
three years. For spending, a positive number indicates higher spending; for
taxes, a positive number indicates higher collection. For the decomposition
of the personal income tax household shares in table 15B.2, the share of
liquidity-constrained households is approximate.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 379
380
Table 15B.1 Composition of fiscal packages in selected Asian economies
Infrastructure Tax reform (increase in VAT, cut in Labor reforms (increase in retirement age, extension
Group/economy spending personal and corporate income tax) of child care, active labor market policies)
Group 1

4 CHAPTER
Personal income tax only (1 percent

SUSTAINING
Japan None of GDP to liquidity-constrained Yes, financed by higher VAT
households from saving households)
Group 2
Yes, financed by additional personal income tax (not
China None None
imposed on liquidity-constrained households)

15—GRAPHICS
Cut to taxes other than personal
Hong Kong None Yes, financed by additional personal income tax
income tax
Korea None Yes Yes, financed by additional personal income tax
Singapore None Yes Yes, financed by additional personal income tax
Public–private Liquidity-constrained households
partnership (PPP), receive all of the personal income tax Yes, financed by additional personal income tax (not

ECONOMIC GROWTH IN ASIA

Thailand
financed by five-year cut and a VAT rebate of 0.5 percent imposed on liquidity-constrained households)
cut in transfers of GDP
Group 3
VAT increase only (not imposed on
Public only, financed
Bangladesh liquidity-constrained households); no Yes, financed by tax reform
by tax reform
personal or corporate income tax
Liquidity-constrained households
PPP, funded by receive all of the personal income tax
Cambodia Yes, funded by deficit finance
deficit finance cut and a VAT rebate of 0.5 percent
of GDP
India None None Yes, financed by additional excise taxes
(table continues)
 Table 15B.1 Composition of fiscal packages in selected Asian economies (continued)
Group 3
Liquidity-constrained households
PPP, financed by
receive all of the personal income tax Yes, financed by additional personal income tax (not
Indonesia five-year cut in
cut and a VAT rebate of 0.5 percent imposed on liquidity-constrained households)
transfers
of GDP
Liquidity-constrained households
PPP, financed by
receive all of the personal income tax Yes, financed by additional personal income tax (not
Malaysia five-year cut in
cut and a VAT rebate of 0.5 percent imposed on liquidity-constrained households)
transfers
of GDP
Liquidity-constrained households
PPP, financed by
receive all of the personal income tax Yes, financed by additional personal income tax (not
Mongolia five-year cut in
cut and a VAT rebate of 0.5 percent imposed on liquidity-constrained households)
transfers
of GDP
Liquidity-constrained households
PPP, financed by
receive all of the personal income tax
Philippines five-year cut in Yes, deficit finance
cut and a VAT rebate of 0.5 percent
transfers
of GDP
PPP, financed by
Yes, financed by additional personal income tax (not
Sri Lanka five-year additional None
imposed on liquidity-constrained households)
VAT

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT

Liquidity-constrained households
PPP, financed by
receive all of the personal income tax Yes, financed by additional personal income tax (not
Vietnam five-year cut in

CHAPTER
cut and a VAT rebate of 0.5 percent imposed on liquidity-constrained households)
transfers
of GDP
Liquidity-constrained households

CAN FISCAL
PPP, financed by
receive all of the personal income tax Yes, financed by additional personal income tax (not
Pacific Island countries five-year cut in
cut and a VAT rebate of 0.5 percent imposed on liquidity-constrained households)
transfers
of GDP
VAT = value-added tax; PPP = purchasing power parity

15—GRAPHICS
Source: IMF Staff assumptions for illustrative purposes.

POLICY DO?
5 381
382
Table 15B.2 Cost of fiscal packages in selected Asian economies (percent of GDP change in fiscal instruments relative
to baseline)
Transfers
Share of to liquidity- Value- Corporate
Group/ Government Government public General constrained added income

6 SUSTAINING
economy deficit consumption investment transfers households tax Personal income tax tax
Group 1
–1.00 for liquidity-constrained
Japan None 0.25 None 0.25 None 0.5 households, None
+1.00 for saving households

CHAPTER 15—GRAPHICS
Group 2
China None 0.25 None 0.25 None None +0.50 for saving households None
–0.25 for –0.40 for liquidity-constrained
five years, households,
Hong Kong None 0.25 None None 1.5 –0.75
+0.25
thereafter +0.15 for saving households

ECONOMIC GROWTH IN ASIA

–0.40 for liquidity-constrained
Korea None 0.25 None 0.25 None 1.5 households, –0.75
+0.15 for saving households
–0.25 for –0.40 for liquidity-constrained
five years, households,
Singapore None 0.25 None None 1.5 –0.75
+0.25
thereafter +0.15 for saving households

–0.25 for –0.59 for liquidity-constrained

+0.50 for five years, households,
Thailand None 0.25 0.5 2 –0.75
five years +0.25
thereafter +0.34 for saving households

(table continues)
 Table 15B.2 Cost of fiscal packages in selected Asian economies (percent of GDP change in fiscal instruments relative
to baseline) (continued)
Group 3
–0.80 after
+1.00 for
Bangladesh year five 0.25 +0.25 0.5 2 None None
five years
years
+0.50 for –0.75 for liquidity-constrained
Cambodia 0.5 0.25 +0.25 0.5 2 –0.75
five years households
None, but
+0.5 on
India None 0.25 None 0.25 None None None
excise
tax
–0.25 for –0.59 for liquidity-constrained
+0.50 for five years, households,
Indonesia None 0.25 0.5 2 –0.75
five years +0.25
thereafter +0.34 for saving households

–0.25 for –0.59 for liquidity-constrained

+0.50 for five years, households,
Malaysia None 0.25 0.5 2 –0.75
five years +0.25
thereafter +0.34 for saving households

–0.25 for –0.59 for liquidity-constrained

+0.50 for five years, households,
Mongolia None 0.25 0.5 2 –0.75
five years +0.25

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT

thereafter +0.34 for saving households

+0.50 for –0.75 for liquidity-constrained

Philippines 0.5 0.25 +0.25 0.5 2 –0.75
five years households

CHAPTER
+0.5 five
+0.50 for
Sri Lanka None 0.25 +0.25 0.5 years +0.50 for saving households None
five years

CAN FISCAL
only
–0.25 for –0.59 for liquidity-constrained
+0.50 for five years, households,
Vietnam None 0.25 +0.5 2 –0.75
five years +0.25
thereafter +0.34 for saving households

15—GRAPHICS
–0.25 for –0.59 for liquidity-constrained
Pacific
+1.00 for five years,

POLICY DO? 7
Island None 0.25 0.5 2 households, –0.75
five years +0.25
Countries

383
thereafter +0.34 for saving households

Source: IMF Staff assumptions for illustrative purposes.

References
Allen, R., and M. Alves. 2016. How to Improve the Financial Oversight of Public Corporations. IMF
Fiscal Affairs Department How-to-Note 5/2016. Washington: International Monetary
Fund.
Amaglobeli, D., and W. Shi. 2016. How to Assess Fiscal Implications of Demographic Shifts: A Granu-
lar Approach. IMF Fiscal Affairs Department How-to-Note 2/2016. Washington: Interna-
tional Monetary Fund.
Andrle, M., P. Blagrave, P. Espaillat, K. Honjo, B. Hunt, M. Kortelainen, R. Lalonde, D. Lax-
ton, E. Mavroeidi, D. Muir, S. Mursula, and S. Snudden. 2015. The Flexible System of Global
Models (FSGM). IMF Working Paper 15/64. Washington: International Monetary Fund.
Araki, S., and I. Claus. 2014. A Comparative Analysis of Tax Administration in Asia and the Pacific.
Manila: Asian Development Bank.
Arbatli, E., C. Feher, J. Ree, I. Saito, and M. Soto. 2016. Automatic Adjustment Mechanisms in Asian
Pension Systems? IMF Working Paper 16/242. Washington: International Monetary Fund.
Arnold J., B. Brys, C. Heady, A. Johansson, C. Schwellnus, and L. Vartia. 2011. Tax Policy for
Economic Recovery and Growth. Economic Journal 121: 1006–1036.
Banerji, A., V. Crispolti, E. Dabla-Norris, R. Duval, C. Ebeke, D. Furceri, T. Komatsuzaki, and
T. Poghosyan. 2017. Labor and Product Market Reforms in Advanced Economies: Fiscal Costs,
Gains, and Support. IMF Staff Discussion Note 17/03. Washington: International Mon-
etary Fund.
Bassanini, A., and R. Duval. 2006. Employment Patterns in OECD Countries: Reassessing the Role of
Policies and Institutions. OECD Economics Department Working Paper 486. Paris: OECD
Publishing.
Bouis, R., and R. Duval. 2011. Raising Potential Growth After the Crisis: A Quantitative Assessment of
the Potential Gains from Various Structural Reforms in the OECD Area and Beyond. OECD Eco-
nomics Department Working Paper 835. Paris: OECD Publishing.
Bova, E., M. Ruiz-Arranz, F. Toscana, and H. Elif Ture. 2016. The Fiscal Costs of Contingent Li-
abilities: A New Dataset. IMF Working Paper WP/16/14. Washington: International Mon-
etary Fund.
Buffie, E., A. Berg, C. Pattillo, R. Portillo, and L. Zanna. 2012. Public Investment, Growth, and
Debt Sustainability: Putting Together the Pieces. IMF Working Paper WP/12/144. Washington:
International Monetary Fund.
Clements, B., K. Dybczak, V. Gaspar, S. Gupta, and M. Soto. 2015. The Fiscal Consequences of
Shrinking Populations. IMF Staff Discussion Note 15/21. Washington: International Mon-
etary Fund.
Clements, B., F. Eich, and G. Gupta, eds. 2014. Equitable and Sustainable Pensions: Challenges and
Experience. Washington: International Monetary Fund.
Coady, D., V. Flamini, and L. Sears. 2015. The Unequal Benefits of Fuel Subsidies Revisited: Evidence
for Developing Countries. IMF Working Paper WP/15/250. Washington: International Mon-
etary Fund.
Corbacho, A., and T. Ter-Minassian. 2013. Public Financial Management Requirements for
Effective Implementation of Fiscal Rules. In The International Handbook of Public Financial
Management, ed. R. Allen, R. Hemming, and B.H. Potter. London: Palgrave Macmillan.
Ebrill, L. P., M. Keen, V. P. Summers, eds. 2001. The Modern VAT. Washington: International
Monetary Fund.
Escolano, J., A. Shabunina, and J. Woo. 2017. The Puzzle of Persistently Negative Interest-
Rate-Growth Differentials: Financial Repression or Income Catch-Up? Fiscal Studies 38,
no. 2: 179–217.

384 SUSTAINING ECONOMIC GROWTH IN ASIA

Garcia Morales, J.A. et al. 2018. Monetary Policy in the New Normal. In The ASEAN Way: Sus-
taining Growth and Stability, ed. A. Corbacho and S. J. Peiris. Washington: International
Monetary Fund.
Gaspar, V., M. Obstfeld, and R. Sahay. 2016. Macroeconomic Management When Policy Space Is
Constrained: A Comprehensive, Consistent, and Coordinated Approach to Economic Policy. IMF
Staff Discussion Note 16/09. Washington: International Monetary Fund.
Global Infrastructure Hub and Oxford Economics. 2017. Global Infrastructure Outlook: Infra-
structure Investment Needs: 50 Countries, 7 Sectors to 2040. https://www.gihub.org/.
IMF (International Monetary Fund). 2012. Philippines: Technical Assistance Report on Road Map
for pro-Growth and Equitable Tax System. IMF Country Report 12/60. Washington
IMF (International Monetary Fund). 2014a. Can Fiscal Policy Do More for Jobs? Fiscal Moni-
tor, October. Washington.
IMF (International Monetary Fund). 2014b. Is It Time for an Infrastructure Push? The Mac-
roeconomic Effects of Public Investment. World Economic Outlook, October. Washington.
IMF (International Monetary Fund). 2014c. Perspectives on Global Real Interest Rates. World
Economic Outlook, April. Washington.
IMF (International Monetary Fund). 2015a. Fiscal Policy and Long-Term Growth. IMF Policy Pa-
per, June. Washington.
IMF (International Monetary Fund). 2015b. Making Public Investment More Efficient. IMF Policy
Paper, June. Washington.
IMF (International Monetary Fund). 2016. Analyzing and Managing Fiscal Risks: Best Practices.
IMF Policy Paper, May. Washington.
IMF (International Monetary Fund). 2017a. Asia: At Risk of Growing Old before Becoming
Rich? In Regional Economic Outlook: Asia and Pacific, April. Washington.
IMF (International Monetary Fund). 2017b. The New Mediocre. and the Outlook for Produc-
tivity in Asia. In Regional Economic Outlook: Asia and Pacific, April. Washington.
IMF (International Monetary Fund). 2017c. World Economic Outlook, April. Washington.
Jenkner, E., B. Clements, and B. Shang. 2012. Health Reform Lessons from Experiences of
Emerging Economies. In The Economics of Public Health Care Reform in Advanced and Emerging
Economies, ed. B. Clements, D. Coady, and S. Gupta. Washington: International Monetary
Fund.
Jones, R., and M. Kim. 2014. Fostering a Creative Economy to Drive Korean Growth. OECD Eco-
nomics Department Working Paper 1152. Paris: OECD Publishing.
Kowalski, P., M. Büge, M. Sztajerowska, and M. Egeland. 2013. State-Owned Enterprises: Trade
Effects and Policy Implications. OECD Working Paper of the Trade Committee, TAD/TC/
WP(2012)10/FINAL. Paris: Organization for Economic Cooperation and Development.
Lipton, D. 2016. Preparing the Ground: China’s Quest for Sustainable Growth Calls for Bold
Fiscal Reforms. Finance & Development, March.
Mankiw, N. G. 2015. Yes, r > g. So What? American Economic Review: Papers and Proceedings 105,
no. 5: 43–47.
OECD (Organization for Economic Cooperation and Development). 2015. Guidelines on Cor-
porate Governance of State-Owned Enterprises. Paris.
Phillips, Steven T., Luis Catão, Luca A. Ricci, Rudolfs Bems, Mitali Das, Julian Di Giovanni,
Filiz D. Unsal, Marola Castillo, Jungjin Lee, Jair Rodriguez, Mauricio Vargas. 2013. The
External Balance Assessment (EBA) Methodology. IMF Working Paper 13/272. Washington:
International Monetary Fund.

AVOIDING A NEW MEDIOCRE IN ASIA: WHAT CAN FISCAL POLICY DO? 385
Saenz, M., A. Corbacho, I. Fukunaga, D. Muir, S.J. Peiris, and S. Rafiq. 2018. Macroeconomic
Policy Synergies for Sustained Growth. In The ASEAN Way: Sustaining Growth and Stability,
ed. A. Corbacho and S.J. Peiris. Washington: International Monetary Fund.
Schaechter, A., N. Budina, T. Kinda, and A. Weber. 2012. Fiscal Rules in Response to the Crisis:
Toward the “Next-Generation” Rules. A Dataset. IMF Working Paper 12/187. Washington:
International Monetary Fund.
United Nations. 2015. World Population Prospects: The 2015 Revision. New York: Population Divi-
sion, Department of Economic and Social Affairs.
World Bank. 2012. Creating Fiscal Space through Revenue Mobilization. South Asia Economic Fo-
cus, June. Washington.

386 SUSTAINING ECONOMIC GROWTH IN ASIA

 16
Global Imbalances and the
Trade Slowdown: Implications
for Asia
CAROLINE FREUND

The current global trade slowdown is unprecedented in recent history.

Global trade volumes plummeted 13 percent in 2009, many times the 2
percent decline in real GDP growth experienced in the depths of the Great
Recession. While the trade collapse shocked economists, the slowdown in
trade growth since 2011 has been an even bigger surprise. Real trade grew
more than twice as fast as real GDP from 1990 to 2007, and more than 1.5
times as fast even before 1990, but since 2011 trade has grown only slightly
faster than GDP.
Researchers have explored a number of potential explanations for the
recent change in the relationship between income growth and trade growth.
Most research points to a decline in demand, especially for investment
goods that weigh heavily in trade flows, as the main factor.
Overlooked in the debate is how greater capital mobility and widening
global imbalances may have enhanced the effects of demand on trade in the
1990s and early 2000s. In the period when trade surged, global imbalances
also ballooned. If the excess savings in some countries financed more con-
sumption and investment in other countries, then trade and trade imbal-
ances would logically move together. Put differently, the ability to borrow
from abroad allowed deficit countries to import more than they would have
if they had been constrained by their exports, thus stimulating trade growth.

Caroline Freund, senior fellow at the Peterson Institute for International Economics since May 2013, is on leave
for public service as director of trade, competition and investment climate at the World Bank. She is grateful to
Olivier Blanchard, Jérémie Cohen-Setton, Davin Chor, Joseph Gagnon, Thomas Helbling, Maurice Obstfeld,
Adam Posen, and conference participants for discussions and comments on an earlier draft.

387
 Figure 16.1 Global imbalances and global trade since 1982

trade/GDP current account balance/GDP

.30 .06

.05
.25

.04

.20
.03

Trade/GDP
Imbalances/GDP .02
.15

1980 1990 2000 2010 2020

Note: Data are for a balanced sample of 76 countries, taken in US dollars.

Sources: World Bank, World Development Indicators; author’s calculations.

Figure 16.1 shows global imbalances across 76 countries with data

going back to 1982. Global imbalances are calculated as the sum of the
absolute values of the countries’ current account balances relative to the
sum of their incomes. When savings and investment in the large coun-
tries are equal, global imbalances will be close to zero. When some coun-
tries, such as China, Japan, Korea, and the Gulf countries, expanded their
surpluses, and others, like the United States and several Southern European
countries, expanded their deficits, the measure of global imbalances grew.
The figure also shows the ratio of trade to GDP, measured as total
imports to total GDP for the same group of countries. Since about 1995, as
global imbalances surged, so too did trade growth. The correlation between
the ratio of trade to GDP and the ratio of global imbalances to GDP since
1995 is 0.80; the correlation before 1995 was –0.61.
Figure 16.2 shows the relationship in growth rates. Again, the correla-
tion expands over time. The correlation was zero prior to 1995 and is 0.66
after 1995.
Further evidence of the relationship between imbalances and trade is
apparent when countries are split according to the magnitude of their current
account balances. For countries with current account balances exceeding 2.5
percent of GDP, the correlation between trade growth and growth in global
imbalances since 1995 is 0.67, while for the group with more balanced trade
it is 0.25. The high correlation is not just a US and China effect. Excluding
China and the United States, the correlation between global imbalances and
trade growth for large-imbalance countries is 0.55 and the correlation for
balanced-trade countries remains 0.25.
388 SUSTAINING ECONOMIC GROWTH IN ASIA
 Figure 16.2 Growth in ratios of trade/GDP and imbalances/GDP

growth in openness growth in balances

.1 .2

0 0

−.1 −.2

dTrade/GDP (left axis)

dCA/GDP (right axis)
−.2 −.4

1980 1990 2000 2010 2020

Note: Data are for a balanced sample of 76 countries, taken in US dollars.

Sources: World Bank, World Development Indicators; author’s calculations.

Relationship between Imbalances and Trade

There is no reason that growing imbalances must be associated with rapid
trade growth, but there is a potential mechanical relationship if trade
grows faster than GDP. This section shows that rapid trade growth was not
sufficient to explain widening trade balances in the late 1990s and 2000s.
Assume imports and exports both grow at a constant rate, higher than
GDP growth. Starting from a position of unbalanced trade, imbalances as a
share of income will expand over time as trade grows. However, in this case,
global imbalances relative to global trade should remain constant. This is
a conservative assumption in the sense that it implies that for deficit coun-
tries international borrowing expands faster than income as trade grows.
For example, assume exports and imports are growing twice as fast
as GDP (as happened in the 1990s), but trade growth is constant across
exports and imports, then we have:
Global imbalances/GDP at time

t����
∑ |������∗� �������∗� |
∑�|������∗��� �������∗���| ������
� ������ ∗∗ ∑∑��|�������|
∑ |� �� | (16.1)
� t����
���� � ���� �
� � ∑∑��������� � ����� ����
∑ � ���
���������� � ����� � �

Global imbalances/total trade at time

∑ |������∗� �������∗� | ∑ |� �� |
t���� � t������ ∑��|������∗�
∑
�� �������∗���| ∑ �|� ��� �|
� ∑∑������ ���� � (16.2)
∑��������� ��� ��
��������� ��
� � � �� �� �
� � �

where x denotes GDP growth and i is the index for individual countries.
As���� trade
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 grows
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇����
����
faster than income, imbalances
𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴��
��
as a����
𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 share of GDP in-
����
crease, � � � � � � � fact, �over
� �����the period
𝐺𝐺𝐺𝐺𝐺𝐺�� but 𝐺𝐺𝐺𝐺𝐺𝐺
��
imbalances
����
����
as a share of
�
� total
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇
����
trade do𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇
not. In ����
����

GLOBAL IMBALANCES AND THE TRADE SLOWDOWN 389

 Figure 16.3 Global imbalances/GDP versus global imbalances/trade

ratio

3.0
Global imbalances/GDP
Global imbalances/trade

2.5

2.0

1.5

1.0

1995 2000 2005 2010 2015

Note: Data are for a balanced sample of 76 countries, taken in US dollars.

Sources: World Bank, World Development Indicators; author’s calculations.

in question both increased (figure 16.3), indicating that the growth in im-
balances was associated with relatively high import growth in the importer/
deficit countries and relatively high export growth in the exporter/surplus
countries.

Decomposing Trade Growth

If trade imbalances boosted trade growth in the 1990s and 2000s, then
exports from surplus countries and imports by deficit countries would be
expected to drive a large share of global trade. In fact, this is exactly what
happened.
Figure 16.4 divides countries into three groups: deficit countries (those
with average deficit above 2.5 percent of GDP), surplus countries (those with
average surplus above 2.5 percent of GDP), and countries with balanced
trade. It shows each group’s contribution to global real trade growth during
the rapid growth period and during the slowdown. The period of rapid
trade growth is 1998–2008 (1998 is the first year because data on real trade
growth for China begin in this year). As shown in the left panel, during the
period of rapid trade growth, strong import growth in deficit countries and
strong export growth in surplus countries were important contributors to
trade growth. In the recent slowdown (2012–15), trade growth has been
both lower and more balanced.

390 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 16.4 Contribution to real trade growth, by type, before
and after the financial crisis

1998−2008 2012−2015
percentage points percentage points

4 4

3 3

2 2

1 1

0 0
Balance Deficit Surplus Balance Deficit Surplus

Import growth
Export growth

Sources: International Monetary Fund; author’s calculations.

Figure 16.5 shows contributions by region. The rapid trade growth

period was associated with rapid export growth in Asia and rapid import
growth in the Americas. During the slowdown, trade in Europe also slowed
markedly, but it was more balanced, linked to slowing growth associated
with the euro crisis.
Figure 16.6 shows the contribution to trade growth by countries in
East Asia and the United States, before and after the financial crisis. China
recorded especially fast-growing exports and the United States recorded
especially fast-growing imports in the rapid growth period. Both moder-
ated in the slowdown period, the gaps between export and import growth
closed, and in the case of China reversed. Among other East Asian coun-
tries, Japan and Korea also experienced widening trade imbalances in the
early period, which have disappeared or reversed since 2011.

Imbalances and Trade Growth at the Country Level

If imbalances are associated with more trade, this link should be apparent
at the country level as well, at least for the large countries driving aggre-
gate trade and imbalances. This section examines whether trade grew faster
than GDP in periods when imbalances were expanding, within countries

GLOBAL IMBALANCES AND THE TRADE SLOWDOWN 391

 Figure 16.5 Contribution to global real trade growth by region
.and period

1998−2008 2012−2015
percentage points percentage points

3 3

2 2

1 1

0 0
c a as sia p e i a ca as sia pe ia
fri ric A ro an fri er
ic A ro an
A
m
e Eu ce A
m Eu ce
A O A O

Import growth
Export growth

Sources: International Monetary Fund; author’s calculations.

over time, controlling for global growth. In the basic regression, the depen-
dent variable is the change in the ratio of trade to GDP and the indepen-
dent variable is the change in the absolute volume of imbalances to GDP.
Year fixed effects are included in all regressions to pick up global trade
growth. ∑ |(1+2𝑥𝑥𝑥𝑥)∗𝑋𝑋𝑋𝑋𝑖𝑖𝑖𝑖 −(1+2𝑥𝑥𝑥𝑥)∗𝑀𝑀𝑀𝑀𝑖𝑖𝑖𝑖 | (1+2𝑥𝑥𝑥𝑥) ∑ |𝑋𝑋𝑋𝑋𝑖𝑖𝑖𝑖 −𝑀𝑀𝑀𝑀𝑖𝑖𝑖𝑖 |
t + 1 = t + 1 = 𝑖𝑖𝑖𝑖 ∑ (1+𝑥𝑥𝑥𝑥)𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺 = (1+𝑥𝑥𝑥𝑥) ∗ ∑𝑖𝑖𝑖𝑖 𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺
One concern with 𝑖𝑖𝑖𝑖 this specification
𝑖𝑖𝑖𝑖 is that 𝑖𝑖𝑖𝑖 if 𝑖𝑖𝑖𝑖exports and imports are

growing at a constant rate, then trade imbalances could widen because of

the mechanical∑𝑖𝑖𝑖𝑖|(1+2𝑥𝑥𝑥𝑥)∗𝑋𝑋𝑋𝑋
relationship described
𝑖𝑖𝑖𝑖 −(1+2𝑥𝑥𝑥𝑥)∗𝑀𝑀𝑀𝑀𝑖𝑖𝑖𝑖 | ∑ |𝑋𝑋𝑋𝑋earlier. −𝑀𝑀𝑀𝑀 | To account for this possi-
t+1=t+1= = ∑ 𝑖𝑖𝑖𝑖(𝑋𝑋𝑋𝑋𝑖𝑖𝑖𝑖 +𝑀𝑀𝑀𝑀𝑖𝑖𝑖𝑖 )
bility, the growth in ∑𝑖𝑖𝑖𝑖 absolute
(1+2𝑥𝑥𝑥𝑥)(𝑋𝑋𝑋𝑋 𝑖𝑖𝑖𝑖 +𝑀𝑀𝑀𝑀 𝑖𝑖𝑖𝑖 balance
) 𝑖𝑖𝑖𝑖 relative
𝑖𝑖𝑖𝑖 𝑖𝑖𝑖𝑖 to trade is, conservatively,
used as an alternative independent variable:
𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖−1 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑇𝑇𝑇𝑇𝐴𝐴𝐴𝐴𝑇𝑇𝑇𝑇𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑇𝑇𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑇𝑇𝑇𝑇𝐴𝐴𝐴𝐴𝑇𝑇𝑇𝑇𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝑇𝑇𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖−1
− = 𝛾𝛾𝛾𝛾𝑖𝑖𝑖𝑖 + 𝛽𝛽𝛽𝛽 � − � + 𝜀𝜀𝜀𝜀𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖−1 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖−1
The variable of interest, the growth in the absolute balance relative to
trade, will increase only if imbalances expand faster than total trade. This
regression tests whether periods when trade is growing faster than income
are also periods when imbalances are growing faster than total trade. The
year fixed effect, γt, controls for increasing openness over the period and

392 SUSTAINING ECONOMIC GROWTH IN ASIA

Figure 16.6 Contribution to global real trade growth, by country
and period

a. United States and China

1998−2008 2012−2015
percentage points percentage points

1.0 1.0

.8 .8

.6 .6

.4 .4

.2 .2

0 0
China United States China United States

b. East Asia
1998−2008 2012−2015
percentage points percentage points

.4 .4

.3 .3

.2 .2

.1
.1

0
0

sia es am sia nd ore ng rea an sia es am sia nd ore ng rea an

o ne pin etn lay aila ap Ko Ko Jap o ne pin etn lay aila ap Ko Ko Jap
d lip Vi Ma Th ing ng d lip Vi Ma Th ing ng
In Phi S Ho In Phi S Ho

Import growth
Export growth

Sources: International Monetary Fund; author’s calculations.

GLOBAL IMBALANCES AND THE TRADE SLOWDOWN 393

global shocks.1 The variable of interest is the increase in the level of imbal-
ances to total trade. A positive coefficient on imbalances relative to trade
would imply that periods of high trade growth are accompanied by rela-
tively faster import growth in a deficit country and relatively faster export
growth in a surplus country.
The results, reported in table 16.1, show that over the whole period
there is a positive correlation between growth in relative imbalances and
growth in openness (column 1). The results for the period before 1995 are
not significantly different from the results after, suggesting that the higher
correlation in the recent period in the aggregate data is due to larger swings,
and not a different relationship between imbalances and trade. The final
three columns consider large economies (with average GDP over the period
greater than $200 billion), countries larger than Denmark. The idea is to see
whether balances in systemic countries were more important in the second
period. The results are robust for all countries and for large countries post-
1995, irrespective of which specification is used—the more liberal one,
comparing imbalances with income, or the conservative one, comparing
imbalances with trade (column 6).
Overall, the results offer evidence that within countries, imbalances
and trade have tended to move together, especially for large countries in
the period of hyperglobalization. The results from the liberal method in
the upper panel suggest that a one percentage point increase in the change
in trade imbalances to GDP is associated with a 0.6 to 0.8 percentage point
increase in the change in trade to GDP.
The magnitudes are economically meaningful. For the United States,
the absolute balance relative to GDP increased by 4 percentage points
and the absolute balance relative to trade increased by 13 percentage
points between 1995 and 2007. The trade-to-GDP ratio increased by 5.5
percentage points. The liberal method in the upper panel suggests that for
the United States, about half of the faster growth in trade compared with
GDP was driven by widening imbalances (0.6 * 0.04 = 0.024). The conser-
vative method in the lower panel also suggests that for the United States,
about half of the faster growth in trade compared with GDP was driven
by widening imbalances (0.19 * 0.13 = 0.025). In fact, the exports-to-GDP
ratio for the United States rose by less than a percentage point over this
period, so faster trade growth for the United States was almost entirely a
result of rapidly growing imports, and hence the imbalance.
For China, the role of imbalances is less pronounced. The absolute
balance relative to GDP increased by 7 percentage points and the abso-

1. Results are almost identical if country fixed effects are included (not reported). Since the
equation is in first differences, country effects control for constant growth in openness.

394 SUSTAINING ECONOMIC GROWTH IN ASIA

 Table 16.1 Panel regressions on trade and absolute balance
Dependent variable: d(trade/GDP)
Pre-1995 Post-1995
All Pre-1995 Post-1995 Large large large

GRAPHICS - CHAPTER
Variable (1) (2) (3) (4) (5) (6)
0.756*** 0.813*** 0.730*** 0.669*** 0.829 0.560***
d(absolute value of imbalance/GDP)
(0.0550) (0.0654) (0.0765) (0.245) (0.553) (0.163)
Observations 5,870 2,013 3,857 922 356 566
R-squared 0.317 0.281 0.337 0.270 0.175 0.356

GLOBAL IMBALANCES
16—SUSTAINING
AND
0.123* 0.155** 0.1000 0.0916 0.0342 0.194**
d(absolute value of imbalance/trade)
(0.0641) (0.0654) (0.0999) (0.0608) (0.0736) (0.0913)
Observations 5,870 2,013 3,857 922 356 566
R-squared 0.065 0.034 0.082 0.203 0.082 0.316

ECONOMIC
THE TRADE
Notes: All regressions include year fixed effects. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
Source: Author’s calculations.

GROWTH
SLOWDOWN
IN ASIA3951
lute balance relative to trade increased by 9 percentage points between
1995 and 2007, implying that imbalances contributed to an expansion
of openness in China of only 2 to 4 percentage points. For China, the
trade-to-GDP ratio increased by 28 percentage points, mainly driven by
the exports-to-GDP ratio, which increased by 18 percentage points. But
import growth was also vibrant: The imports-to-GDP ratio was up 10
percentage points.

What Would Trade Growth Have Looked Like under More

Balanced Global Trade?
An alternative way of investigating the potential effect of global imbal-
ances on trade growth is to assume exports constrain current account defi-
cits. The focus is on import constraint in deficit countries because global
imbalances after the financial crisis contracted almost entirely because of a
reduction in demand in deficit countries (IMF 2014). In addition, widening
deficits—not surpluses—are a systemic risk.
For this exercise, nominal data are used because current accounts are
measured in current dollars. The year 2015 is excluded because nominal
trade declined sharply that year owing to the 50 percent drop in oil prices.
Two series are created: one restricts countries to trade balance and
the second allows countries to run deficits similar to historical norms.
Specifically, under the latter case, for each deficit country, imports are
assumed to be constrained to ensure that the trade deficit to GDP does
not exceed
1. 2.5 percent of GDP and
2. the average value plus 2 standard deviations, during the period 1980–
95.

The first criterion ensures that the deficit is reasonably large. The
second ensures that it is larger than its historical norm. The second condi-
tion is important because a number of small developing countries had
large imbalances in the 1990s, and the rule allows them to maintain these
imbalances.
For countries where the deficit reaches the limit, imports are assumed
to decline to the level that would allow the current account deficit to be
within one standard deviation of its historical value or to trade balance,
whichever is larger. This case maintains global imbalances near an average
of 2.5 percent of GDP.
For both specifications, feedback effects to surplus countries are incor-
porated because exports use imports in their production. The feedback

396 SUSTAINING ECONOMIC GROWTH IN ASIA

 Figure 16.7 Trade/GDP ratio without growing imbalances

trade/GDP ratio

.30

.25

Observed
.20 Trade balance
Constrained

1995 2000 2005 2010 2015

Sources: International Monetary Fund; Organization for Economic Cooperation and

Development; author’s calculations.

effect is based on the import content of exports in each year.2 Specifically,

for global trade to be balanced, exports above balance from surplus coun-
tries must decline by the same amount as the decline in excess imports
by deficit countries. The export reduction is then translated to an import
reduction in surplus countries, using the annual average import content of
exports. No further feedback effects from the decline in imports by surplus
countries are assumed.
Figure 16.7 shows observed exports and predicted exports assuming
current account balances were restricted to be balanced or to prevent imbal-
ances from exceeding historical norms. In both cases, the ramp-up in trade
would have been somewhat slower than was realized. In the case where trade
deficits are constrained, but not limited to balanced trade, the 2012–14
growth slowdown is less dramatic because the 2010–11 rebound in trade
happens more gradually than actually occurred.
Table 16.2 records average trade growth for the periods 1995–2008 and
2012–14, for observed flows and the series where current account balances
were restricted. It also shows average trade growth relative to average income
growth. The evidence is consistent with growing imbalances contributing
to more rapid trade growth during the first period and moderating trade

2. Data on exports, imports, GDP, and the import share of exports are available for 57 coun-
tries from the OECD Trade in Value Added (TiVA) database.

GLOBAL IMBALANCES AND THE TRADE SLOWDOWN 397

Table 16.2 Responsiveness of trade to income and global imbalances
Trade growth (percent) Elasticity
Period Actual Balanced Restricted Actual Balanced Restricted
1995–2008 8.5 8.1 8.1 1.60 1.53 1.53
2012–2014 2.2 2.3 2.7 0.98 1.03 1.21
Share of decline explained by imbalances (percent) 19 48
Source: Author’s calculations.

imbalances contributing to slower growth in recent years. Overall, about

20 to 50 percent of the drop in trade growth relative to GDP growth can be
explained by the rise and fall in imbalances.

Transmission of US Shock to East Asian Economies

The previous section assumed that the large deficit countries drove trade
growth. How feasible is this? The United States was the most important
deficit country. To the extent that the period of rapid trade growth was
driven by US demand fueling export growth in Asia, trade in the Asian
economies would tend to move with US total imports. This section explores
whether US imports are positively correlated with the total exports and
imports of East Asian surplus countries.
Table 16.3 shows results from regressing real import and export
growth in the East Asian countries on real US import growth. East Asian
imports may also be correlated with US imports, because a high share of
the imports are inputs used in exports. For example, in Japan the import
share of exports is 20 percent, in China it is 30 percent, and in Korea it is
40 percent.
Table 16.3 shows that, for surplus Asian countries, both import growth
and export growth are highly correlated with US import growth. In contrast,
for the East Asian countries with roughly balanced trade or trade deficits,
the correlations are much lower and not significant. While the surplus
economies and the United States could be responding to global growth or
other excluded variables, the fact that only the surplus countries show a
strong correlation with US import growth is consistent with movements in
imbalances enhancing the transmission from slow US growth to East Asian
exports and imports.

Is This Explanation Consistent with the Existing Literature?

The main explanations put forward for the recent trade slowdown are
(1) weak demand, especially for investment goods that are a big part of
trade flows, (2) a slowdown in the development of global supply chains,
and (3) protectionism.

398GRAPHICS
2 - CHAPTER
SUSTAINING ECONOMIC GROWTH IN ASIA
16—SUSTAINING ECONOMIC GROWTH IN ASIA
 Table 16.3 Trade growth in surplus Asian countries is highly correlated with US import growth
Deficit Balance Surplus
Hong
Variable Laos Cambodia Vietnam Indonesia Philippines Japan Korea China Thailand Kong Taiwan Malaysia Singapore
Dependent variable: Import growth
US import 0.0223 0.66 –0.186 0.505 0.0767 0.636*** 0.695 0.596* 1.110** 0.625** 1.237*** 1.173*** 0.825***

GRAPHICS - CHAPTER
growth (0.414) (0.489) (0.418) (0.523) (0.375) (0.174) (0.400) (0.282) (0.417) (0.234) (0.206) (0.323) (0.269)
Observations 18 18 18 18 18 18 18 18 18 18 18 18 18
R-squared 0.00 0.10 0.01 0.06 0.00 0.46 0.16 0.22 0.31 0.31 0.69 0.45 0.37
Average CA/
–16.26 –5.82 –1.14 1.12 1.83 2.72 3.28 3.84 3.86 7.22 7.66 10.28 18.89
GDP

GLOBAL IMBALANCES
Dependent variable: Export growth

16—SUSTAINING
AND
US import 0.275 0.783* –0.108 0.133 0.434 1.189*** 0.718*** 1.051** 0.931*** 0.720*** 1.013*** 0.918*** 0.740***
growth (0.343) (0.381) (0.238) (0.371) (0.464) (0.241) (0.177) (0.364) (0.131) (0.222) (0.206) (0.206) (0.212)
Observations 18 18 18 18 18 18 18 18 18 18 18 18 18
R-squared 0.039 0.209 0.013 0.008 0.052 0.603 0.508 0.342 0.761 0.397 0.602 0.553 0.432

ECONOMIC
THE TRADE
Average CA/
–16.26 –5.82 –1.14 1.12 1.83 2.72 3.28 3.84 3.86 7.22 7.66 10.28 18.89
GDP
Notes: * indicates significance at the 10 percent level, ** 5 percent level, and *** 1 percent level. Data are real import and export growth from the

GROWTH
International Monetary Fund.
Source: Author’s calculations.

SLOWDOWN 3
IN ASIA 399
 The IMF World Economic Outlook (2016) did the most extensive study
to date and found that demand is largely to blame for the trade slowdown,
accounting for 50 to 80 percent, with supply chains and protectionism
each explaining at most 5 percent. The results are consistent with research
showing that investment tends to drive trade movements (Bussière et al.
2013). While trade imbalances are, of course, related to demand, neither
study considered whether increasing global imbalances in the mid-1990s
may have affected the relationship between trade and growth.
A number of studies focus on the role of China in the trade slowdown.
Constantinescu, Mattoo, and Ruta (2015) find that the deceleration in
vertical integration, particularly in China, is important. Gaulier et al. (2015)
focus on China’s rise as a manufacturing center and its shift to domestic
demand as important in explaining changes in global trade growth over
time. Consideration of the role of widening global imbalances in fueling
trade growth is complementary to these studies in that it helps to explain
how China’s exports could grow so rapidly and the timing of the shift to
domestic demand-driven growth.
The importance of borrowing in the period of hyperglobalization
before the financial crisis is also consistent with work by Shin (2011) and
Borio, James, and Shin (2014) on the surge in credit in the period leading
up to the financial crisis. A banking glut led to an expansion of gross and
net capital flows in the United States. On the real side, that was reflected in
an import surge in the United States and the credit expansion also fueled
greater unbundling of production.
While there is no reason trade could not surge in an environment of
more balanced trade (indeed, most trade models assume trade is balanced),
the ability to have large trade imbalances could enhance trade growth
because countries can import more when demand is strong. The growth in
cross-border capital flows magnifies the effects of the existing explanations
because budget constraints are no longer binding. In terms of demand as
an explanation, the ease of borrowing from abroad means that demand
can exceed supply for longer periods, without price increases. Similarly, in
the presence of greater capital mobility, the buildup of supply chains was
likely faster than it otherwise would have been.
The trade growth puzzle may therefore be the next of kin to one of the
major paradoxes in macroeconomics from the 1980s, the high correlation
between savings and investment across countries. Theory predicts that
savings should flow to the best investment opportunities. Because the top
investment prospects may not be in the domestic market, the correlation
between savings and investment across countries should be low. Instead
economists found it to be very high in the 1960s and 1970s. This paradox,
known as the Feldstein-Horioka puzzle (Feldstein and Horioka 1980) for the

400 SUSTAINING ECONOMIC GROWTH IN ASIA

economists who uncovered it, has unraveled in recent decades, as the cross-
country correlation between savings and investment declined. The widening
and then narrowing of global imbalances that resulted from greater capital
mobility contributed to more volatile trade growth. If instead the gap
between savings and investment had remained limited, trade growth would
have very likely been more balanced over time as shown in figure 16.7.

Implications for Trade and Growth

The unprecedented trade growth that followed the rise of cross-border
capital flows is linked to widening trade imbalances. Similarly, the recent
period of slow trade growth is associated with a narrowing of global imbal-
ances. From this perspective, the dramatic trade slowdown stems not just
from weak global growth but also from weaker growth combined with a
return to more balanced capital flows.
Going forward, even as global growth picks up, the new weaker relation-
ship between trade and income may remain in place if global imbalances
remain constrained. Overall, the dismantling of the Feldstein-Horioka
puzzle may have ushered in a period when the relationship between global
trade growth and global income growth is more volatile. In periods when
demand is strong in large countries, their widening imbalances fuel trade
growth to a greater extent than if they were constrained by their exports.
Similarly, the eventual pressure to narrow trade deficits could eventually
result in slower trade growth.
An alternative way of viewing the results is that the export-led growth
policies in Asia, especially China, fueled strong trade growth in the late
1990s and early 2000s and widening imbalances. Those policies effectively
shifted export growth from the future to that period, resulting in slower
trade growth in recent years as pressure for more balanced trade increased.
An important implication for the East Asian surplus countries is that they
will now need to rely more on domestic reform and less on export-led
growth. Over time, a new wave of global income and trade growth could
result if the roles reversed and surplus East Asia absorbed a higher share of
global capital flows.

References
Borio, Claudio, Harold James, and Hyun Song Shin. 2014. The International Monetary and Fi-
nancial System: A Capital Account Historical Perspective. Working Paper 204. Dallas: Federal
Reserve Bank of Dallas, Globalization and Monetary Policy Institute.
Bussière, Matthieu, Giovanni Callegari, Fabio Ghironi, Giulia Sestieri, and Norihiko Yama-
no. 2013. Estimating Trade Elasticities: Demand Composition and the Trade Collapse of
2008-2009. American Economic Journal: Macroeconomics 5, no. 3: 118–51.

GLOBAL IMBALANCES AND THE TRADE SLOWDOWN 401

Constantinescu, Cristina, Aaditya Mattoo, and Michele Ruta. 2015. The Global Trade Slow-
down: Cyclical or Structural? IMF Working Paper 15/6. Washington: International Mon-
etary Fund.
Feldstein, Martin, and Charles Horioka. 1980. Domestic Saving and International Capital
Flows. Economic Journal 90, no. 358: 314–29.
Gaulier, Guillaume, Gianluca Santoni, Daria Taglioni, and Soledad Zignago. 2015. The Pow-
er of the Few in Determining Trade Accelerations and Slowdowns. In The Global Trade
Slowdown: A New Normal?, ed. Bernard Hoekman. VoxEU.org eBook. London: Centre for
Economic Policy Research. Available at http://voxeu.org/sites/default/files/file/Glob-
al%20Trade%20Slowdown_nocover.pdf.
IMF (International Monetary Fund). 2014. Are Global Imbalances at a Turning Point? In
World Economic Outlook. Washington.
IMF (International Monetary Fund). 2016. Global Trade: What’s Behind the Slowdown? In
World Economic Outlook. Washington.
Shin, Hyun Song. 2011. Global Banking Glut and Loan Risk Premium. Mundell Fleming Lec-
ture presented at the 12th Jacques Polak Annual Research Conference, November 10–11,
Washington.

402 SUSTAINING ECONOMIC GROWTH IN ASIA

Comment
Davin Chor

Caroline Freund presents an original and thought-provoking hypothesis

on a mechanism that could have contributed to the current global trade
slowdown. Specifically, she links the steady unwinding of global imbalances
over the past five to ten years to the decline in the growth rate of interna-
tional trade flows. The logic behind this relationship is an intuitive one: For
several decades, the ability to run persistent current account deficits has
enabled countries such as the United States to sustain (and even increase)
their imports from the rest of the world, well above levels that would have
been possible if these imports had to be balanced in each period against
the value of the countries’ exports. But as these current account imbalances
have shrunk in recent years, growth in trade relative to output has inevi-
tably declined, which we are now seeing in the data.
The chapter marshals several key pieces of evidence in support of this
hypothesis. First, it shows very clearly that since 1995, global imbalances
and global trade have indeed become more tightly correlated; this is true
regardless of whether one traces the evolution of these variables in levels
(figure 16.1) or in their growth rates (figure 16.2). This correlation stands
up to closer scrutiny in a series of panel regressions: Movements in current
account imbalances tend to be positively associated with corresponding
shifts in the trade-to-GDP ratio, particularly among large economies in
the post-1995 period (table 16.1). Second, the chapter documents how the
trade position of individual countries has shifted since the global financial
crisis. Among countries that were running large current account deficits
prior to the crisis—such as the United States—import growth has since
moderated in tandem with narrowing global imbalances. On the other
hand, in countries that were initially building up large current account
surpluses—such as China—export growth has tapered off. These observed
patterns (reported in figures 16.4 to 16.6) are entirely in line with how one
would expect an unwinding of global imbalances to affect these countries’
export and import positions.
Freund’s argument thus carries some natural appeal. The economic
mechanisms articulated here and the evidence that has been presented
would strike a chord with observers familiar with the ongoing policy debates
over the United States’ large current account deficit and how best to address
it. While economists have long been aware of the relationship between
global imbalances and trade flows, credit is nevertheless due to the author

Davin Chor is professor in the Department of Economics at the National University of Singapore.

GLOBAL IMBALANCES AND THE TRADE SLOWDOWN 403

for making explicit in this chapter how this relationship could account for
(at least some part of) the global trade slowdown. At the same time, it bears
mentioning that the association between imbalances and trade flows is not
simply a mechanical accounting identity, since trade in goods is only one
component of a country’s current account.
That said, the ideas that this chapter has seeded do raise questions for
further research. As compelling as the descriptive evidence presented here
is, it remains an open (and intensely interesting) question whether these
correlations actually reflect a causal relationship running from narrowing
global imbalances to a slowdown in global trade growth. To make the case
that this effect is indeed causal, one would in principle like to be able to
exploit instances of plausibly exogenous policy changes that were intro-
duced with the express objective of reducing current account imbalances,
yet at the same time did not seek to directly target exports and imports to
meet this policy objective. This exercise, however, raises clear challenges
from a research design perspective.
Consider, for example, a country running a current account deficit. To
reduce such deficits in practice, expenditure-switching policies have often
been adopted—via say an exchange rate depreciation or an import tariff—
to encourage consumers to purchase domestic goods instead of imported
ones. (Such thinking, in particular the threat of unilateral tariffs, has in
fact pervaded the recent rhetoric of the Trump administration to address
the large bilateral US deficits with key trading partners like Mexico and
China.) If successful, such policies would shrink the country’s external
imbalance position, but they would also be doing so by directly reducing
the demand for imports. One would then be hard-pressed to disentangle
the effects of the policy to address imbalances from the effects of weak
import demand. Put otherwise, the finding in recent studies that weak
global demand has been the dominant factor in the global trade slowdown
could indirectly also be capturing the effect of current account reductions
by large deficit countries.
An alternative and possibly more promising research strategy might
be to focus on surplus countries. To take a prominent example, China
has widely been seen in recent years to be seeking to lower its burgeoning
current account surplus and instead pursue a growth strategy that is driven
more by domestic consumption rather than by a reliance on exports. There
may well be interesting policy measures—such as restrictions on lending of
surplus funds to foreign entities or a reduction in China’s holdings of US
treasuries—that could allow researchers to parse out the effect of a decrease
in China’s surplus on its and other countries’ trade performance. Research
along these lines would be extremely useful to more definitively resolve the
direction of causality. Such an exercise would help to provide more reliable

404 SUSTAINING ECONOMIC GROWTH IN ASIA

estimates from which to trace the effects of shifts in global imbalances on
countries’ subsequent trade positions.
In sum, this chapter has presented an intriguing hypothesis that clearly
deserves to be investigated more extensively. I hope it will encourage further
study of the links between developments in international macroeconomic
policy and international trade flows, which all too often have been treated
as disjoint subjects by academic researchers. The findings documented here
also bear a key lesson for policymakers, namely that measures to reduce
global imbalances can and should be expected to have effects that show
up in a country’s trade statistics, and these repercussions should not be
ignored.

GLOBAL IMBALANCES AND THE TRADE SLOWDOWN 405

 17
Toward a “New” Asian Model
SUBIR GOKARN

Two themes shaped the discussions during this conference. The first is that
the “old” Asian model is obsolete and that countries in the region have to
reorient their strategies for sustaining growth. The second is that an impor-
tant element of any such reorientation must take into account the oppor-
tunities the region itself provides. This chapter explores both themes and
draws some implications for a “new” Asian model.

The “Old” Asian Model

All of the countries referred to during the conference grew because there
was demand for their products in the advanced economies, mainly the
United States and Europe. The contribution of specific export-led strategies
to growth performance varied across countries. Overall, the model delivered
very successfully from the 1950s onward.
Many observers view the flying geese formation as an apt metaphor for
the Asian model. It implies strategic replication over succeeding genera-
tions of countries. There are two aspects to this replication, both of which
are important when considering the future of the model.
The first is what might be referred to as leverage. The growth process
begins at the lowest end of the supply chain: garments, footwear, and
other goods that directly satisfy consumer needs. Competitiveness in these

Subir Gokarn is executive director for Bangladesh, Bhutan, India, and Sri Lanka at the International Monetary
Fund. The views expressed in this chapter are those of the author and do not necessarily represent the views of the
IMF, its Executive Board, or its management.

407
sectors required combining low-cost labor, organized in efficient structures,
with the lowest-cost intermediate and capital goods. Initially, the latter were
sourced from abroad (for countries entering the formation later, possibly
from predecessors who had developed these capabilities). The key to lever-
age was that even as countries began to integrate backward into upstream
segments of the supply chain, preserving the competitiveness of the down-
stream remained paramount.
Leverage contributed to efficient backward integration, the basis of the
second aspect of replication, which might be referred to as succession. Even as
downstream competitiveness in earlier generations of countries eroded as
labor costs rose, upstream competitiveness allowed the export-led growth
process to sustain. Of course, as standards of living rose, domestic demand
became increasingly significant in the growth process. However, the virtuous
circle of efficiency and competitiveness moving upstream over the supply
chain was a key contributor to the sustainability of exports. Downstream
products shifted location and the export baskets of individual countries
changed over time, but the importance of exports in the growth of the region
remained.

Opportunities for a “New” Model

It is against this backdrop that the issue of obsolescence has to be viewed.
Changing demand patterns in the advanced economies, perhaps induced
by slower growth and demographic dynamics, suggest that these markets
will become less and less important sources of demand for Asian products.
This change was inevitable. But even as the role of the advanced economy
markets in Asian growth declines, the key element of the Asian model—the
promotion of efficiency and competitiveness up the supply chain—provides
the basis for the “new” model.
The second theme of the conference was the significance of being in
a particular neighborhood at a particular time. In his inaugural presenta-
tion, Adam Posen displayed the shares of different countries in the global
economy over time, as laid out in Subramanian (2011). In the late 19th
century, the two growth powerhouses, Great Britain and Germany, were
neighbors. There was significant economic benefit to the entire region from
having two large engines of growth located in it.
Asia is in a similar situation today, with China and India—whose shares
of global GDP will rise to 18 and 6 percent, respectively, over the next
decade—potentially providing significant growth momentum to the region.
Of course, that potential will be realized only if the countries in the region
implement appropriate policies. But with so many countries having de-
veloped competitive capabilities in a variety of downstream and upstream

408 SUSTAINING ECONOMIC GROWTH IN ASIA

sectors, they should be able to take advantage of opportunities in the neigh-
borhood.
The power of clusters and agglomeration is well understood. At the
sectoral level—Silicon Valley, Seattle, Cambridge, and Bengaluru—or the
country level, being in a dynamic and rapidly growing neighborhood pro-
vides all members of the cluster an opportunity to leverage their competen-
cies. The tailwinds that can be generated by having two very large players in
a neighborhood are significant and represent an opportunity that should
not be missed by the smaller countries in the region.
Under the “new” Asian model, countries in the region move away from
supplying the advanced-economy markets to meeting the needs of the
growing markets in the region itself. There is a great deal of diversity and
heterogeneity in the region, including a range of per capita income levels.
But many of its markets are growing rapidly and, under a reasonable set
of global and domestic conditions, will continue to do so for a long time.
The region provides opportunities for producers in several countries to
collaborate to exploit complementary competencies in a variety of sectors.
Regional supply chains are already well established in several countries. The
full potential across sectors and geographies is yet to be fully tapped.

Challenges of a “New” Model

Several challenges also need to be addressed. Aging and its implications
for labor supply and public spending in the form of health and social secu-
rity are going to be important considerations for all countries—for many,
sooner rather than later. Most Asian countries will cross dependency
thresholds at far lower levels of per capita income than the advanced econ-
omies of North America and Europe did.
Technology is another challenge. It will affect different countries in
the region in different ways. In countries with abundant labor resources,
the implications of automation are a cause of grave concern. In many Asian
countries, low-skilled work is likely to be the predominant mode of employ-
ment for years to come. Transiting to a work situation in which relatively
skilled people work collaboratively with robots is a distant prospect at
the macro level. The fact that the labor-intensive manufacturing/services
pathway to affluence is no longer open is very difficult to tell constituents.
How do countries in this situation deal with a threat that may be immi-
nent and against which they have no obvious buffer?
A third set of challenges relates to the nature of trade between coun-
tries that are similarly endowed. Classical trade theory posits that trade is
most likely to take place and to generate the most benefits between coun-
tries with different endowments. The old Asian model was firmly rooted in

TOWARD A “NEW” ASIAN MODEL 409

this theoretical framework, with its emphasis on exporting labor-intensive
products to relatively labor-scarce advanced economies. The new Asian
model involves a pivot to the neighborhood, which means establishing
significant trading relations with similarly endowed countries. How does
it square with trade theory?
Fortunately, there are enough grounds to believe that the pivot is achiev-
able. Gravity models indicate that geographical proximity is always condu-
cive to trade. Evidence based on new trade theories emphasizes features such
as intraindustry trade and the importance of cross-border supply chains
in achieving efficiency and increasing trade volumes. Intra-Asian trade is
already an established phenomenon, diluting concerns about similar en-
dowments. Multiple trade agreements are in place within the region. And
there are important historical precedents. Trade between advanced econo-
mies is significant. The European Union is a vivid example of the power of
geographic contiguity in enlarging trade opportunities, even between coun-
tries that are similarly endowed. These trends bode well for the new Asian
model.
The conditions necessary for the new model to work will not fall into
place automatically. Various kinds of barriers will have to be addressed by
countries, with a high level of coordination. Trade-facilitating infrastruc-
ture will have to be expanded or built to accommodate potential increases in
volumes. To the extent that some of the increase in trade will be in services,
rules, standards, and procedures will have to be framed and harmonized.
Encouraging the development and expansion of supply chains will
require clarity and predictability in both domestic and cross-border proce-
dures. Training and skills systems will have to be geared up in anticipation
of the new opportunities that will arise as a result of the pivot. All of these
issues will pose significant challenges to governments in the region, in that
they reflect a significant departure from the old model, which was rela-
tively simple in its focus on labor-intensive manufacturing.

Summary
Several key messages emerged from the conference. First, the changing land-
scape in advanced-economy markets is combining with structural changes
in Asia to render the traditional export channels less and less significant for
the region as a whole.
Second, the obsolescence of the old model should be a cause for concern
only if there is nothing to succeed it. But the two elements of the old
model—leverage and replicability—have created centers of efficiency across
the region along the entire supply chain in many sectors. And decades of
sustained growth in Asia have created a large market and set of opportuni-

410 SUSTAINING ECONOMIC GROWTH IN ASIA

ties for all countries in the region. The new model requires the pivoting of
Asian economies away from advanced-economy markets toward the region.
The fact that two large and fast-growing economies, China and India, are in
the neighborhood will facilitate this process.
Third, although the opportunities for a successful pivot are well
grounded in the historical experiences of other regions, several challenges
exist. Automation requires a strategic response. Countries seeking to do
more business with their neighbors will have to coordinate on many fronts,
from building infrastructure to harmonizing rules and procedures. To a
large extent, they are already doing so. But the issue gains some urgency
when seen in the context of a paradigm shift in regional strategy.
Whether the new Asian model is a move toward mediocrity or renewed
excellence depends on how effectively the countries in the region recognize
and act on the opportunities and respond to the challenges.

Reference
Subramanian, Arvind. 2011. Eclipse: Living in the Shadow of Chinese Economic Dominance. Wash-
ington: Peterson Institute for International Economics.

TOWARD A “NEW” ASIAN MODEL 411

Index

Abe, Shinzo, 130 age profile, in Korea, 166, 168f

Abenomics, 130, 131, 139, 141 age-related spending, 344, 353
academic and industry sectors, connections, age-wage profile, in Japan, 144
76b aggregate capital accumulation, declining
academic salaries, in Japan, 78 in Korea, 167–168
adult education, contributing to human aggregate demand, exceeding aggregate
capital accumulation, 367 supply, 20
advanced economies aging
behavior in Asia, 124 creating a demographic “tax,” 40
demand patterns in, 5, 408 as a disinflationary factor, 175n
external demand growth from, 6 impact on investment, 49–50
lowered interest rates after the crisis, impact on labor force participation
116 rates (LFPRs), 46–49
policies following the financial crisis, impact on total factor productivity
237 (TFP), 43–46, 48f
productivity slowdown in, 347 implications for labor supply and
public debt as percent of GDP, 345f public spending, 409
quality of infrastructure, 355, 356f of Korea’s population, 169
secular stagnation transmitted to in the late- or post-dividend stages, 64
domestic output dynamics, 253 lowering growth, 6
slowdown perceived as durable, 237 low per capita income levels, 37, 67–68
syndromes perpetuating stagnation, migration softening the impact of
135 rapid, 43
at the technological frontier, 107, 257 mitigating the economic effects of,
weakness in demand leading to lower 20, 134
exchange rates abroad, 243 not leading to declining aggregate
advanced economy group, imbalances demand, 20
within, 281 pushing for deflation and
aerodynamics, evolution of, 17n protectionism, 136
Africa, 19, 38 reducing aggregate demand and
age of Enlightenment, 26 aggregate supply, 175n
slowing potential growth, 344

413
aging speed macroeconomic developments in,
associated with higher current account 287–289, 287f
balances, 52 projected effects of fiscal measures,
change between 2020 and 2030, 58f 368f
creating a demographic tax on growth, Asia-15 economies, inflation rates in,
3 288–289, 288f
defined, 58f Asian crisis (1997–98), 269
effects on interest rates, 58, 60 Asian economies
rapid in Asia, 37, 66–67, 344 danger from secular stagnation, 1–2
resulting in lower investment, 51 dominated by China and India, 116
setting to create a demographic tax on largest, 33n
growth, 67 public debt, 346f
airplane design, trial and error used in, 17n public spending on pensions and
“All 100 Million Playing an Active Role” health care, 346f
(Ichi-Oku Sou-Katsuyaku), 141 Asian model
Android platform, as Google’s innovation, efficiency and competitiveness up the
91 supply chain, 408
annual growth rates, in Korea and the as obsolete, 5
global economy, 166, 167f asset churning, 275
Annual Report on Exchange Arrangements and “asset market meltdown” hypothesis, 62
Exchange Restrictions (AREAER), 110, asset prices, 194, 304
125 asset returns, demographics and, 62–64,
Anti-Monopoly Law, in Japan, 82 62t, 63t
APDMOD semi-structural model of the atmospheric pressure, Torricelli’s discovery
global economy (IMF), 365, 365n, of, 17
365n, 366n Australia
arbitrage, between foreign and domestic alternative policy strategies, 326f, 327f
bonds, 114 change in 10-year government yield,
artificial intelligence (AI), replacing 55f
workers, 28 demographic characteristics, 35t
ASEAN (Association of Southeast Asian demographic impact on current
Nations), real GDP growth relative to, account, 53f
2000–16, 241f higher old-age dependency by 2020, 52
ASEAN-4 (Malaysia, Thailand, and the immigration, 37
Philippines), 244 impact of aging on TFP, 48f
ASEAN-9 (Cambodia, Indonesia, Laos, impact of continued immigration on
Malaysia, Myanmar, the Philippines, the workforce, 43
Singapore, Thailand, and Vietnam), impact of demographics on 10-year
249 real interest rates, 61f
Asia impact of demographic trends, 42f
adopting full-fledged inflation-forecast labor force participation rates, 48f, 49f
targeting, 323 as a large commodity-exporting
China and India providing growth economy, 324
momentum, 408 as a late-dividend economy, 35, 37
contribution to real trade growth, 392f lower speed of aging, 52
countries crossing dependency maintained higher real interest rates,
thresholds, 409 116
demographic impact on 10-year real managing downside risks and avoiding
interest rates, 61f a deflation trap, 324–328
demographic impact on current old-age dependency ratio, 36t, 38f, 57f
account, 53f open-economy New Keynesian model,
demographic trends, 34–40 337
growth leadership shifting from Japan per capita income level, 39f
to China, 270 real neutral interest rates, 56f

414 SUSTAINING ECONOMIC GROWTH IN ASIA

 share of older workers, 44f scope for quantitative easing,
share of workforce with productivity December 2016, 303t
rising or falling, 45f tax revenue sources, 364f
youth dependency ratio, 57f Bank for International Settlements (BIS)
authoritarian regimes, rise of, 22 alternative path for the federal funds
automatic adjustment mechanisms, in rate, 306
pension systems, 375 concerns about overextended financial
automation, 409, 411 markets, 305
credit gap measures produced by, 119
baby boomers, retiring and drawing down credit-to-GDP gap statistics, 125
savings, 62 banking cartel, in Japan, 81
backward integrating, leverage contributed banking crisis, in Japan in 2003, 135
to, 408 banking glut, in the United States, 400
backward-looking expectations (lags), 337 banking problems, depressing Japanese
Bacon, Francis, 18 productivity growth, 107
bad equilibrium, 316, 320 bank lending, to zombies, 170
bad governance, 23 Bank of Canada, 306, 319
balanced trade countries, 390, 391f Bank of England, 306
balance-of-payments equation, 112 Bank of Japan (BoJ), 131–132, 290, 290f,
“balance sheeting working group,” 190n 291, 303
balance sheets Bank of Korea, 2, 295
banks making greater use of, 297 banks, Park nationalized Korea’s, 85
central banks expanding, 286 Basic Survey on Wage Structure, 152
crises in the twenty-first century, 278 below-target inflation rates, persisting, 335
overstretched limiting the effects of best practices, adopting in Japan, 164
monetary policy, 334 Betzig, Eric, 24
policies of central banks, 286 Betzig-Hell microscope, 24
potential of policies, 4 birth rates, not rising with income, 10–11
vulnerabilities, 282–283 BIS. See Bank for International Settlements
Baltic nations, scores on governance (BIS)
indicators, 22 bond markets
Bangladesh domestic influencing financial
CPI inflation rates, 288f conditions, 4
current account balances, 310f measure of integration, 125
direct taxes, 366 more important in monetary policy
fertility rates, 19 transmission, 286
fiscal packages, 380t, 383t risks from global, 297–298
foreign exchange reserves and net worries about destabilizing, 304
official assets, 311f bonds, 307, 335
in Group 3, 347b booms, emergence of unsustainable, 124
infrastructure, index of quality of, 356f borrowing
months inflation rate below inflation from abroad, 400
target, 371f by local governments in China, 188
personal income tax (PIT) yields, 363f risks of highly leveraged, 307
PPP framework needing further Bragg, William Henry, 24
development, 366n Bragg, William Lawrence, 24
public debt, actual and projected, 346f Brent crude oil price, 1991–2017, 275f
public spending efficiency frontier, budgetary aggregates, numerical limits on,
378t 373
public spending on education, 359f budget constraints, as no longer binding,
public spending on health care, 358f 400
real per capita GDP, growth rate of, Bulgaria, 22, 117, 117n
287f “burden of knowledge,” increasing, 26
real short-term interest rate, 293f Burke, Edward, 23

INDEX 415
business cycle synchronization, 239n in Korea and selected countries,
business enterprise research and 1970–2015, 168f
development (BERD), in Japan, 151f carbon prices, 183
business environment, in Indonesia, 242 Cartagena Protocol on Biosafety, 23
business models, development in China of cartels, in Japan, 82
new, 185 Ceauşescu, Nicolae, 22, 23
business R&D and government support, in CEIC database, 224
Japan, 151f Central Asian republics, governance
business regulation and taxation, indicators, 22
simplifying in India, 222 central bank(s)
active use of balance sheets, 301–302,
Cambodia 302t
fiscal packages, 380t, 383t avoiding the traps of fiscal and
in Group 3, 347b financial dominance, 304
infrastructure, index of quality of, 356f balance sheets
personal income tax (PIT) yields, 363f advancing objectives, 286
public debt, actual and projected, 346f keeping markets for local financial
public spending efficiency frontier, assets operating, 298
distance from, 378t policies, 4
public spending on education, 359f clear signals, providing, 321
public spending on health care, 358f conducting QE through purchases of
tax revenue sources, 364f government bonds, 302
trade growth in, 399t deflation risks, fighting, 323
capacity utilization gap, in the US, easing monetary policy but tightening
299–300, 299t macroprudential policies, 309
capital inflation, control of low, 298–304
cost of, 124 interest rate, publishing, 323
taxation of, 353 interest rates, 125
capital accumulation, in Indonesia, limited influence of in China, 187, 188
256–257 monetary autonomy of, 312
capital exporters and importers, 51, 53 monetary policy
capital flows in China, 189
across borders, 275, 276f conduct of, 318
demographic factors strengthening, nominal interest rate, determining,
52–53 114
management of, 118 official inflation targets of, 316
opportunities for, 67 policy interest rate, setting the path
spurred by unconventional monetary of, 321
policy, 237 policy rates, reducing below zero, 300
capital formation, in Japan, 143 scope when inflation expectations
capital gains, to financial firms holding settle at low levels, 286
bonds, 307 tolerating longer periods of inflation
capital inflows, 114 below target, 307
capital mobility, 112, 387 world real interest rate, ready for
capital openness index (CO), 58 external shocks to, 297
capital stock, in Indonesia, 262, 262f central bankers, managing inflation, 269
capital surcharges, enhancing resilience to central bank transparency index, 323, 324f
shocks, 334 Centre for Monitoring the Indian Economy
capital-to-effective-labor ratio, remaining (CMIE), 225
constant, 41 CGE model, showing China’s annual GDP
capital-to-labor ratio, 50 growth, 183
capital-to-output ratio chaebols (Korea’s industrial
approaching ratios in advanced conglomerates), 86, 86n, 88
countries, 166, 168f channels of spillover, for low productivity
growth, 104

416 SUSTAINING ECONOMIC GROWTH IN ASIA

channels of transmission, from advanced financial sector risks, prevention of,
economies to Indonesia, 242–244 181
chemicals, market share of exports over fiscal packages, 380t, 382t
time, 212f foreign capital, volatility of the supply
Chen Shui-bian administration, in Taiwan, of, 2000–15, 251f
95 foreign exchange reserves and net
Chiang Kai-shek, authoritarian regime of, official assets, 311f
91–92 GDP growth targets, abandoning, 188
child care, provision of, 370 global exports share versus India, 198f
childcare spending, permanently increasing, global real trade growth, 393f
367 governance indicators, improvements
child rearing, support for in Japan, 141 in, 22
children, having higher-quality, 11 in Group 2, 347b
China growth
absolute decline in, 38 deceleration, 182
account balances and reserve higher since the global financial
purchases, 271f crisis, 194
age-related spending, 344, 346f a more balanced pattern of, 315
aging slowdown as inevitable, 184–185
experiencing rapid until 2030, 52 strategy driven more by domestic
fiscal pressures from, 353 consumption, 404
impact on growth, 41 imbalances, role of, 394–396
impact on TFP, 48f import share of exports, 398
aging speed change between 2020 and Indonesian exports to, 246f, 247n
2030, 58f Indonesian imports from, 246f, 247n
capital control regime, 280n infrastructure, index of quality of, 356f
capital inflows, 122 labor force aging, 344
central government setting growth labor force participation, 48f, 49f
targets, 186 loans to domestic banks offsetting the
CPI inflation rates, 288f decline in foreign-currency assets,
current account balances, 274f, 310f 311
debt ratio change under Scenario 4b, long-term public debt projection, 349t
354f macroeconomic policy framework for,
deceleration pressures, 185–186 181–191
demographic characteristics, 35t macroeconomic stability screening
demographic impact on current mechanism, 189
account, 53f median age, projected rise in, 20n
demographics, impact on 10-year real monetary policy independence,
interest rates, 61f enhancing, 188
demographic trends, impact of, 42f most-favored nation (MFN) applied
development zones, 187, 189 tariff rates, 230f
direct taxes, 366 natural rates remaining high, 54
drag on growth, 41n net investment, 2002–17, 279f
engineering, 18 nonfinancial credit-to-GDP ratio, 181
environmental costs, 182–184 old-age dependency ratio, 36t, 38f, 57f,
expansion of credit, 119, 120 344
export growth tapered off, 403 older workers, share of, 44f
export-led growth policies, 401 overinvestment, aftereffects of, 165
export market, catching up with Korea patents, average quality of, 186
in, 171 pensions, 375
exports, fast-growing, 391 per capita income level, 39f
exposure to, affecting trade-related personal income tax (PIT) yields, 363f
spillovers, 243 phone makers. market shares of
fertility rates, 19 increasing, 91n

INDEX 417
 polity not democratic and open, 22 cloud technology, just getting started, 26
as a post-dividend economy, 35 clusters and agglomeration, power of, 409
primary commodity exports from Cobb-Douglas production function, 339
Indonesia, 243n Cœur, Jacques, 12n
productivity, slowdown in, 347 commodities, 339
public debt, actual and projected, 351f competition, in Japan, 82–83
public spending competitiveness indicator, 339
on education, 359f Complete Dictionary of Arts and Sciences
efficiency frontier, distance from, (Croker), 26n
378t composition effects, accounting for drag on
on health care, 358f wage growth in the US, 134
on pensions and health care, 346f Comprehensive and Progressive Agreement for
quasi-fiscal debt of local governments, Trans-Pacific Partnership (CPTPP), 130
190 computer memory, decreasing cost of, 27n
real GDP growth correlated to computers, improvements in, 24–25
Indonesia’s, 253t conditional forecast path, publication of,
real GDP volatility, 2000–16, 241f 319
real neutral interest rates, 56f conditionality, communicating, 323
real per capita GDP, growth rate of, Condorcet, 1795 essay on human progress,
287f 16n
real short-term interest rate, 293f constant true prices, reflected by an
rebalancing from investment toward inflation rate of 1 percent, 292
consumption, 243 consumer and industrial goods, US market
research and development spending, for, 76b
362f consumer preference, shifting from goods
reserves fell over the past three years, to services in China, 184
310–311 consumer price and wage inflation,
restructuring in, 316 modeled by reduced-form Phillips’
Scenario 3 and, 350 curves, 339
scope for quantitative easing, consumer price inflation, domestic, 251
December 2016, 303t consumption, 267, 338
services and manufacturing sectors as consumption tax, in Japan, 131, 132, 289n
percent of GDP, 185f contagion, channels of, 3, 108
shift in consumer preference from contingent liabilities, emanating from
goods to services, 184 SOEs, 376
slowdown in economic growth, 3–4, conventional forward guidance, 322
191 convergence argument, 169n
SOEs in, 376 core CPI inflation
spillovers to the rest of the world, 325 for Japan, 333f
State Council setting monetary growth in Korea, 296f
targets, 187 in Thailand, 296f
tariff rates, 230 core inflation, keeping at or above 2
tax revenue sources, 364f percent, 285
10-year government yield, 55f Corporate Enterprise Annual Statistics, from
trade growth in, 399t the Ministry of Finance in Japan, 150
trade slowdown, role in, 400 corporate income tax, effects of cutting, 369
trade surpluses, expanded, 388 correlations, reflecting a causal
trade-to-GDP ratio driven by exports- relationship, 404
to-GDP ratio, 396 corruption indices, 14
WTO, accession to, 270 corrupt nations, interacting with nations
youth dependency ratio, 57f with better institutions, 22
Chinn-Ito index, 109–110, 118, 125 corrupt regimes, being deposed, 24
chonsei, rental system of Korea, 174n countercyclical fiscal rules, 373
city-states, set up city governments, 13

418 SUSTAINING ECONOMIC GROWTH IN ASIA

countries current account(s)
with bigger current account deficits, balances
278 in Asia, 1985–2016, 310f
changing political structure, 22 correlated with credit, 120
groups of in regard to trade, 390, 391f correlated with initial net foreign
illustrating risk-avoidance strategy for assets, 278
monetary policy, 324–333 measuring net capital outflows, 121
trade positions shifting, 403 1991–2017, 274f
unique parameterizations in the in 2015, 54f
G20MOD, 340 behavior leading net international
country groupings, in Asia, 347b investment positions, 281f
coverage, for fiscal rules, 373 deficits, 403, 404
CPI inflation rates, in Asia, 288f demographic impact on, 51t, 53f, 54f
credibility, earning, 338 driving divergence in net foreign
credit assets, 280, 282f
buildup of excessive, 119 imbalances, associated with shifts in
changes in reflecting demand the trade-to-GDP ratio, 388f, 389f,
rebalancing, 122 403
surge in before the financial crisis, 400 surpluses, 309, 310f
credit and asset prices, excessive growth in
domestic, 114 Daewoo, 86n
credit availability, as a driver of resource dark corner, for India, 338
misallocation in India, 232 Data Envelopment Analysis (DEA), 357f,
credit booms 378
countries with seeing increases in net data mining, in artificial neural networks,
capital inflows, 121 25
precipitating major growth slowdowns data science, new era of, 25
or financial crises, 123 debt financing, 188, 371, 372f
risk of unsustainable, 124 debt liabilities, declined in Indonesia, 249
credit expansions, 119, 400 debtors, going farther into debt, 278, 281f
credit gap, 119–122, 120f, 122f debt ratio, change in under Scenario 4b,
credit growth, in India, 231 354f
creditor countries, deciding to spend more, debt ratios, under Scenario 1, 350
278 debt ratio under Scenario 4b, 353, 354f
creditors, seeking to ensure resources for debt rules, 373
the future, 278 debt-to-income ratios, imposing, 308
credit-to-GDP gap, 119, 125 deceleration of demand, lowering import
cross-border capital flows, growth in, 400 demand, 242
cross-border financial stability risks, deceleration pressures, measures taken to
addressing, 283 offset in China, 185–186
cross-border risk sharing, opportunities de facto financial integration, as the ratio of
for, 67 foreign assets plus liabilities to GDP,
cross-border supply chains, 410 118
cross-country differences, in interest rate de facto integration, having no impact on
transmission, 118 productivity growth, 110
cross-country regressions, of the change in deficit countries, 390, 391f
interest rate associated with the global countering macroeconomic forces with
financial crisis, 117–118 import barriers, 280
CTFP, measuring the productivity level demographic trends exerting
across counties, 125 downward pressure, 52
Cullen, William, 17 facing the risk that lenders will
currency depreciation, seeking to prevent withdraw, 282
unwanted, 310 importing more by borrowing, 387
currency mismatches, 280, 309 reduction in demand, 396

INDEX 419
deficits, as a systemic risk, 396 scenario for public debt projection,
deficits (and surpluses), in slower-growing 348
advanced debtors (and creditors), 280 taking as destiny, 134
defined-benefit pension system, 375 demographic “tax,” on growth, 40
deflation demographic transitions, 6
Japan’s policies to exit, 330–333 demographic trends
response to in the early 1990s, 291 affecting savings, investment, and
threat of faded in most countries, 285 current account balance, 50
deflationary spiral, fear of, 291 in Asia, 34–40
demand for Asian growth, 67
boosting aggregate in the short term, baseline growth impact of, 42f
267 global impact of, 42f
changes in domestic correlated with implications of, 40–64
changes in domestic credit, 122 policy implications of, 64–66
China’s shift to domestic, 400 for post-dividend countries, 41
cyclical changes in, 108 reducing growth in post-dividend
domestic in Indonesia, 239–240 countries, 50
exceeding supply without price demographic variables, 52n, 60n
increases, 400 demography, Korea’s following Japan’s, 170
trade slowdown and, 400, 404 deregulating, targeted sectors, 141
demand elasticity, decline in, 229 Desaguliers, John T., 18
demand rebalancing, planting seeds of developing Asia, public debt as percent of
secular stagnation, 123 GDP, 345f
demand shocks, central banks dealing with, development zones, in China, 187, 189, 191
328 difference-in-difference methodology, 108n
demand-side issues, in Indonesia, 267 diffusion of technologies, determining
demand slump, countering, 237 productivity growth, 107
demand stimulus, tradeoff with financial diminishing returns to research, in
stability, 7 individual sectors, 27
demand weakness, assertive policy actions Dincer-Eichengreen index of transparency,
following, 316 for four Asian countries, 323, 324f
demographic changes direct taxes, 366
baseline impact of, 41 discount, high rate of, 14
effects of, 176, 177f discretionary monetary policy, 317
demographic classification, of Asia, 35t distribution system, in Japan, 82
demographic dividend DNA molecule, 24
for early- and late-dividend countries, “doctrine of equivalents,” 79
41 domestic and global factors, correlation
ending for many Asian economies, 33 to domestic real GDP growth in
harnessing for early-dividend Indonesia, 253, 253t
countries, 64 domestic assets, expanding holdings of, 312
in Indonesia, 261 domestic factors, explaining deviation of
reversing in many countries, 6 Indonesia’s growth, 253, 254f
demographic model, on the effect of the domestic financial assets, buying, 302
two-child policy in China, 182 domestic financial markets, 286. See also
demographic negative feedback, 10 financial markets
demographic projections, long-term, 43n domestic interest rate, 112, 113, 114
demographics domestic supply linkages, strengthening in
in Asia, 316 India, 230
in Korea and Japan, 171f dominant elites, extracting rents, 12
labor productivity and, 47t downside scenario, in China, 325
rapid decline in the labor force in downward nominal wage rigidity, 292
China, 182 DRAM chips, Korean production of, 89,
89n

420 SUSTAINING ECONOMIC GROWTH IN ASIA

drugs and pharmaceuticals, market share of moving to universities and vocational
exports over time, 212f schools, 163
“dual economy,” in Japan, 150 policies to encourage in Japan, 164
dual labor market structure, in Korea, 169 public spending as a percent of GDP,
dynamics of foreign assets equation, 112 359f
public spending on, 355, 357, 359f
early-dividend demographic characteristic, South Korean compared to Japan’s, 88
35t education-to-employment pipeline, in
early-dividend economies, 37 Japan, 164
ease of doing business, Indonesia ranking effective lower bound, constrained the Bank
poorly on, 259 of Japan, 330
East Asia efficiency frontier
with balanced trade or trade deficits, calculated using Data Envelopment
398 Analysis techniques, 355n
contribution to trade growth, 391, of public investment, 357f
393f of public spending on education,
current account balances, 274f 360f–361f
economies maintaining currencies at of public spending on health care, 358f
undervalued levels, 280 Electronic Research and Service
imports correlating with US imports, Organization (ERSO), 93
398 electronics-based industrial park, built by
more market-based flexible exchange Taiwanese industrial planners, 94–95
rate systems, 101 electronics industry, in Taiwan, 93
regressing real import and export EMDEs (emerging-market and developing
growth, 398, 399t economies)
surplus countries needing to rely more accumulation of foreign reserves by,
on domestic reform, 401 271f, 273
transmission of US shock to, 398, 399t higher net saving by, 272
as the world’s fastest-aging region, 34 pervasive foreign currency
East Asia (other), net investment, 2002–17, denomination of foreign liabilities,
279f 280
EBA (external balance assessment) model, Emerging Asia, defined, 294f
50, 50n, 52, 56n emerging economies, 3, 237
economic embargoes, 14 Emerging Market Bond Index (EMBI)
economic equilibrium, 292 spread, 251
economic growth, low, reorganized as the in Indonesia, 252f, 253t
“new normal,” 193 emerging-market capital growth, declining
economic/high-tech development zones, in over the medium term, 262
China, 186 emerging-market economies
economic integration, pursuing credit booms in, 124
international, 8 growth decelerated, 315
economic nationalism, threat of, 15 having higher nominal interest rates,
economic structure, of Korea similar to that 117
of Japan, 171 reduced policy interest rates, 116
economic underperformance, not turning emerging-market funds, 297
into public outcry in Japan, 136 emerging markets
economies deceleration in trend growth, 237
before the Industrial Revolution, 12 fraction of the global economy, 236
less advanced catching up to the macroeconomic dynamics, 236
frontier, 107 TFP accounting for postcrisis decline
education in potential growth rate, 257
efficiency frontier of public spending emigration, of working-age people, 37
on, 360f–361f empirical literature, on the link between
exporting and productivity growth,
108

INDEX 421
empirical models, data for, 224–228 impact larger with greater foreign
employees by firm size, all industries, Japan currency exposure, 244
and United States, 2001, 2006, 159t Japan’s real and nominal, 80
employment movements not affecting exports in
gap, in the US, 299–300, 299t India, 202
as the most important macroeconomic policies, 101, 309–311
policy objective in China, 186, 191 of the yen and the won in the mid-
scenario analysis for Indonesia, 262f 1980s, 89
security, in Japan, 81 excise taxes, raising, 376
system, in Japan, 146n exogenous path, for the policy rate, 320n
endogenous credibility-building process, expected stock returns, 62n
328 expenditure on wages and salaries,
endogenous interest rate, producing, 320 strengthening the rupee, 205,
endogenous monetary policy, 320 218t–219t
endogenous policy credibility, model for expenditure rules, 373
India, 337–338 expenditures, types of, 339
endogenous policy rate, 322 expenditure-switching policies, 404
energy export(s)
clean more expensive than dirty, 183 changing composition of India’s,
pricing reforms, 376 198–199
subsidies, 376 constraining current account deficits,
engineering goods, exporting from India, 396
198 as a critical driver of India’s slowdown,
Enlightenment Europe, artisans and 197, 197f
scientists, 18 decline of Indonesia’s, 244
enterprise creation and growth, Japan elasticity of India’s, 200–201
behind the United States, 163 elasticity of with respect to the
entry barriers, in Korea, 170 exchange rate, 204, 206t–207t
environmental costs, in China, 182–184 examining the drivers of India’s, 229
environmental taxes, introducing, 376 growing twice as fast as GDP, 389
epistemic base, in technology, 18 growth rates on for Indonesia, 244n
“equality before the law,” 12 increasing in sophistication and value
“equal work, equal pay” rule, in Japan, 141 addition from India, 198
equilibrium rates of interest, decline in, India increasing its share of global,
292, 294 197, 198f
equities, official purchases of, 302 India’s, 194f
equity and real estate price bubbles, in market share of by sectors in India,
Japan, 289 205, 211f
equity prices, run-up in US, 272 measuring the elasticity of India’s, 200
equity purchases, as the most untapped observed and predicted, 397, 397f
channel for QE, 303 reduction translating to an import
escape clauses, for fiscal rules, 374 reduction in surplus countries, 397
EU-28, workforce with productivity rising by region for Indonesia, 245, 246f, 247
or falling, 45f response to exchange rate movements,
EU-Japan trade agreements, 130 222
Euler, Leonhard, 17n role in India’s growth slowdown, 196f,
Europe, old-age dependency ratio, 38f, 57f 197f
exchange rates sensitivity of India’s, 200–202
changes not affecting exports, 229 share of India’s in GDP, 195f
diminished passthrough not specific supporting economic growth in
to Japan, 135 Indonesia, 266
effects of weaker, 243 from surplus countries, 390
fixed keeping Japan’s cost of export basket, composition of Indonesia’s,
production relatively low, 79–80 241n

422 SUSTAINING ECONOMIC GROWTH IN ASIA

export diversification, reducing export Federal Reserve, purchased long-term
volatility, 266 bonds, 303
export elasticities, decline in, 229 feedback loops, negative, 6
export goods, Japan’s expanding range of, Feldstein-Horioka puzzle, 400–401
79 female employment, potential of, 46
export industries, Japan’s relocating female labor force participation rates
production abroad, 149 (LFPRs), 46n
exporting firms, reacting to exchange rate female workers
changes in India, 202, 204–221 better utilizing, 134
export-led growth policies in Asia, fueled in Japan, 133, 141, 143, 143f, 146
strong trade growth, 401 fertility rates
export-led strategies, contribution to in Asia, 34f
growth performance, 407 in Bangladesh, 19
export-oriented manufacturing industries, in China, 182
in Korea and Japan, 171, 172f control of by inexpensive means, 11
“export-oriented policy,” effects of, 101 decline in encouraging more women in
exports elasticity, post-2005 decline in, 232 the labor force, 46
exports-exchange rate equations, estimation declining since the late 1960s, 34
of, 202 effect of after 10 years, 176, 177f
export side, measuring India’s exposure on, in Japan, 148
194–195, 195f in Korea, 166n
exports-to-GDP ratio, for the United States, in Korea and select countries, 1980–
394 2015, 168f
export volatility, 266 lower with richer nations or persons,
external assets, reflecting an increase in 11
outward FDI and other investment, in the Philippines, 37
249n slowdown worldwide, 19
external balance assessment (EBA) model, in the United States, 168f
50, 50n, 52, 56n fertilizers, from emergence of organic
external balances, implications for, 50–53 chemistry, 17
external balance sheet, foreign currency financial capital, provided to “insider” firms
exposure of Indonesia’s, 250f in Japan, 81
external demand financial conditions, easy promoting
in India, 229 leverage, 278
in Indonesia, 241, 242 financial crisis, in Japan, 289n
external factors, impacting Indonesia’s financial cycle-augmented Taylor rule,
growth, 251, 253, 253t, 255 federal funds rate implied by, 306
external financing, cost of for Indonesia, financial deregulation, in Japan proceeded
250, 252f slowly, 82
external liabilities, in Indonesia, 249, 250f financial distortions, powering the US
external sector current account deficit, 275
in Indonesia, 244, 245f financial exposure, stock and flow
supporting growth, 309 indicators of, 248–251
External Sector Reports (IMF), 278, 280, 282 financial frictions, spilling over across
External Wealth of Nations dataset, 125 countries, 107
financial globalization, 312
factors of trade, inefficient allocation of, financial imbalances, buildup of, 334
258 financial integration, 66n, 110, 117
“fair” trade, 15 financial intermediaries, interest rate risks
family circumstances, making regular on the balance sheets of, 307
employment compatible with, 148 financial linkages, 243, 244
famines, man-made in the 20th century, 11 financial markets
FDI Regulatory Restrictiveness Index allowing monetary policy to affect a
(OECD), 248–249 wide array of market prices, 334

INDEX 423
 demographic trends for, 67 response to demand shocks in Korea,
efficiency of in Korea, 170 169
expecting a permanent slowdown in firm-size differences, in labor quality and
advanced economies, 244 TFP, 158
implications of demographic trends firm-size wage differences, 149, 156
for, 53–64 “the first 21st century crisis,” 269
improving access to in India, 222 fiscal and quasi-fiscal expenditures,
in Japan, 81–82 planning of, 186, 191
financial openness fiscal institutions, need for supporting, 371,
Chinn-Ito index of, 109–110, 118, 125 373–376
of Indonesia, 245, 245f fiscal measures, effects on real GDP in
financial policies, in Indonesia, 242 selected Asian economies, 368f–369f
financial risks, sector-specific, 308 fiscal packages
financial sector, supervisory oversight in composition and cost of, 379–383,
Indonesia, 242 380t–381t, 382t–383t
financial side effects, risks during recovery covering public infrastructure, tax
phases, 7 rebalancing, and labor markets, 363
financial stability fiscal policy
from accommodative monetary policy, addressing long-term implications of
334 the new mediocre, 377
considerations regarding, 334–335 addressing the risk of a new mediocre,
mitigating risks, 64 343
risks to, 286, 304–309 assisting monetary policy, 325
worries not justifying keeping the complementing monetary policy, 328
policy rate higher than warranted, governed by long-run fiscal policy
312 targets, 339
financial vulnerabilities, in various parts of political desire for wasteful procyclical,
the world, 315 132–133
financing, of labor sector reforms, 367 supporting sustained and inclusive
finite resource, planet earth as, 21 growth, 353, 355
firm-level data, results often difficult to fiscal reforms
translate to the aggregate level, 108 effects across Asia, 367–370
firms. See also large firms financing through debt, 377
deferring compensation in Japan, 144 illustrative packages, 365–367
early emergence of leading, 75b payoff in a low-inflation environment,
increasing part-time workers in Japan, 370–371
146 potential to improve growth, 343
incumbent resisting rendering existing quantifying the payoff of, 363–371
business models obsolete, 76b fiscal room, determining available, 132
investment determined by a Tobin’s fiscal rules, 373–374, 374f
Q combined with the Bernanke- fiscal space
Gertler-Gilchrist model, 338 in Indonesia, 242
in job training contributing more to using, 7
firm-size wage differences, 155 fiscal stabilization, in Japan, 130, 133
as net borrowers, 338 fiscal stimulus
not adjusting investments in response in Abenomics, 139
to exchange rate fluctuations, 205, early improving outcomes in Australia,
220t–221t 325
paying a premium to part-time fiscal sustainability, 5
workers, 145 flexible inflation targeting, 318–319, 324
productivity and wage gaps between floating exchange rate era, 269
large and small in Japan, 149–159 flow deficit, 278
reluctant to employ more regular flow indicators, 248
employees in Japan, 162 flying geese formation, 407

424 SUSTAINING ECONOMIC GROWTH IN ASIA

forecast, deviating from a previous, 323 data in current prices from the CEIC,
foreign assets, 112, 309 224–225
foreign capital, volatility of, 251f gap shrinking in Japan, 141
foreign currency growth
debts in the corporate sector, 309–310 abandoning targets in China, 188
exposure of liabilities, 250n higher real, 332
liabilities, rapid decline in, 249 impact of demographic trends, 42f
foreign direct investment (FDI) in India, 197
attracting, 267 overemphasis on performance, 188
gross external positions representing, growth rate
275, 278 in China, 181
regulatory environment impeding, 248 in Korea and Japan, 173f
foreign exchange reserves imbalances to, 392
accumulation of after the Asian per capita
financial crisis, 301–302 correlation with government
continued accumulation not justified, revenue as percent of GDP, 2015,
310 362f
in many Asian economies, 309–311, impacted by demographic trends,
311f 42f
foreign exchange reserves and net official ratio of trade to, 388, 388f
assets in Asia, 1985–2016, 311f GEMPACK software, 183n
foreign interest rates, 113, 114 gems and jewelry, as exports, 212f
foreign investors, 297 gender inequality index, 49n
foreign secular stagnation. See also secular generalized method of moments (GMM)
stagnation estimator, 204n
developments representing, 111 Germany, 20n, 39f, 55f, 136
impact on the real exchange rate, 113 “gig economy,” 20n
leading initially to a boom in capital global business cycles, export basket in
inflows, 114 India sensitive to, 199
responses to, 112 global current account balances and reserve
foreign workers, 65, 65n purchases, 271f
forward guidance, on the policy rate, 301, global current account imbalances, 270,
322 271f
forward-looking, model-consistent global deflationary spiral, likelihood of, 237
expectations (leads), 337 global economy
forward-looking models, anchoring forward-looking, multieconomy model
inflation expectations in the short run, of, 338
320n Indonesia’s financial integration with,
“frankenfoods,” concerns about, 23 248f
frontier firms, TFP differences with profound changes, 239
nonfrontier firms, 149n stagnation of, 166
fuel subsidies, lowering in Indonesia, 242 underperforming for several years, 315
full disclosure, 323 global financial crisis
full employment, mandated for in background to, 273–278
Australia, 324 channels of contagion during, 105
Funding for Lending scheme, of the Bank effects on different countries, 104
of England, 302 as an important step in diffusion of
secular stagnation, 104
G-7 economies, growth slowdown Indonesia not suffering from the
besetting, 1 recessionary impact of, 238
G20MOD country model, 338–340 interest rates before and after by
gaslighting, supported by pneumatic country, 2000–16, 117f
chemistry, 17 interest rates before and after by
GDP. See also real GDP region, 115f
as a “natural experiment,” 123

INDEX 425
 natural rate of interest declining after, India’s growing participation in, 229
54 global trade
of 2008–09, 270 slowdown symptomatic of secular
global financial cycle, 121 stagnation, 242
global financial integration, across slowdown unprecedented in recent
emerging markets, 249f history, 387
global financing shocks, 250 world trade growth under more
global gross financial flows, 276f balanced global trade, 396–401
global growth, 194, 195, 199 global value chains, India’s limited role in,
global imbalances. See also imbalances 230, 231f
calculated, 388, 388f Gourinchas-Rey effect, 280
correlated with global trade, 403 government(s)
effects of shifts in on countries’ fiscal stimulus targeting the lower
subsequent trade positions, 405 middle class in Indonesia, 267
in the emerging markets, 270 policies encouraging entrepreneurship
excess, 278, 282 and global competitiveness in Japan,
and global trade since 1982, 366f 164
linking to the decline in the growth publishing balance sheets in China,
rate of international trade flows, 403 190
mono-causal explanations for purchases of bonds, 302
excessive, 4 reducing exposure to SOE-related
owing little to foreign exchange fiscal risks, 376
intervention by EMDEs, 281 resisting higher bond yields, 304
policy implications, 282–283 revenue as percent of GDP, 362f
relationship with trade, 5 revenue ratios compared with other
relationship with trade flows, 403–404 economies, 359, 362f
relative to global trade remaining spending or taxation mattering more
constant, 389 than debt levels, 132
rise of, 270–273 supported the low-inflation objective,
risks from excess, 278–282 318
role in fueling trade growth, 400 gravity models, indicating geographical
twenty-five years of, 269–283 proximity conducive to trade, 410
widening and narrowing of, 387, 397f, “gray market” for capital, tolerated in
401 Taiwan, 92
global imbalances/GDP, versus global Great Depression, described, 305
imbalances/trade, 390f Great Enrichment, 9, 18
global implications, of Asia’s demographic Great Recession, 105, 111
evolution, 38 Greece, governance indicators, 22
global integration, India’s, 194, 223 gross capital flows, 282–283
global investment liabilities stock, 277f gross external positions, 275, 277f
global liquidity trap, 105 Group 1 (Japan), 347b, 368f, 370
globally integrated economy, channels of Group 2 countries
international transmission of secular effect of fiscal reforms, 370
stagnation, 103 including China, Hong Kong, Korea,
global neutral real rate, settling at around Singapore, and Thailand, 347b
1 percent in the medium to long run, increasing spending pressures from
293 demographic trends, 359
global real prices, determining, 339 projected effects of fiscal measures,
global real trade growth, contributions to, 369f
390, 392f, 393f Group 3 countries
global risks, India vulnerable to, 4 gaining from reforms and positive
global supply chains. See also supply chains spillovers, 370
development of explaining the recent
trade slowdown, 398

426 SUSTAINING ECONOMIC GROWTH IN ASIA

 including Bangladesh, Cambodia, Haber-Bosch nitrogen-fixing process, 12n
India, Indonesia, Laos, Malaysia, “handbooks,” compiled by groups of
Maldives, Mongolia, Myanmar, the experts, 26
Philippines, Sri Lanka, Vietnam, and Hansen, Alvin, 28
the Pacific Island countries, 347b Hartz IV labor market reform, in Germany,
needing to step up public spending, 133
359 healthcare
projected effects of fiscal measures, efficiency frontier of public spending
369f on, 358f
growth making available for older workers, 65
in absolute balance relative to trade, public spending on, 355, 358f
392, 394 reform of, 374–375
accounting heterogeneity, in external exposure, 245
components of decelerated, 166, high-tech startup incubators, in China, 185
167f historical primary balance scenario, for
of Korea, 167t public debt projection, 348
results of, 166–169 Hong Kong
deceleration, in China, 182–185 aging
demographic trends, implications of, change between 2020 and 2030, 58f
40–50 impact on growth, 41
in imbalances, 390 impact on TFP, 48f
impact of demographic trends and rapid until 2030, 52
higher labor force participation, 49f contribution to real trade growth, 393f
as natural-resource-using or resource- CPI inflation rates, 288f
saving, 21 credit gaps, increases in, 119
normal in the 20th and 21st centuries, current account balances, 1985–2016,
10 310f
outlines of a model for sustaining, 8 demographic characteristics, 35t
payoff of fiscal reforms, 363–371 demographic trends, impact of, 42f
promoting slower but higher-quality fiscal packages, 380t, 382t
in China, 186 foreign exchange reserves and net
rate(s) official assets, 311f
in Korea decreasing, 169 in Group 2, 347b
ratio of trade to GDP and the ratio immigration, 37, 43
of global imbalances to GDP, 388, infrastructure, index of quality of, 356f
389f interest rates
of real per capita GDP, 287–288, lowered after the crisis, 116
287f nominal fell close to zero, 124
in ratios of trade/GDP and real short-term, 293f
imbalances/GDP, 389f labor force participation, 48f, 49f
role of demographic and structural old-age dependency ratio, 36t, 38f, 52,
factors in, 133–135 57, 57f
role of exports in India’s, 195, 196f, older workers, share of, 44f
197 personal income tax (PIT) yields, 363f
slowdown in China, 184–185 as a post-dividend economy, 35
“Smithian” in nature in the pre- public spending
Industrial Revolution era, 14 on education, 359f
strategy on health care, 355, 358f
structural reforms in Abenomics, R&D spending, 357
139 real per capita GDP, growth rate of,
targeting manufacturing in 287f
Indonesia, 267 research and development spending,
guest worker programs, targeting specific 362f
skills, 65

INDEX 427
 scope for quantitative easing, prolonging the demographic dividend,
December 2016, 303t 65
tax revenues, direct taxes as a small softening the impact of aging, 67
share of, 366 impact theory, 17n
tax revenue sources, 364f import constraint, in deficit countries, 396
trade growth in, 399t imported intermediate inputs, in sectors
workforce, share of, with productivity with low domestic value, 205, 210t
rising or falling, 45f imported intermediates, response to
youth dependency ratio, 57f an appreciation of the rupee, 204,
Hong Kong Monetary Authority, 298 208t–209t
household deposits, lack of pass-through of imports
negative rates, 300 by deficit countries, 390
household property mortgages, rules on, growth rates declining for Indonesia,
308 244n
households, types of in APDMOD, 365 by region to Indonesia, 245, 246f, 247
housing bubble, 306 temporary trade restrictions on
Hsinchu Science Park, 94–95 Indonesian, 258, 260f
Huawei, 91, 91n imports-to-GDP ratio, for China, 396
“human bridges,” 76b income growth, relationship with trade
human capital growth, 387
concentrated in large firms in Japan, income per capita, on a global scale as self-
81 limiting, 21
East Asian investment in, 78 incomes policy, aimed directly at sluggish
growth rate of slowing, 258 wage-price dynamics, 332
in Japan, 77–78 India
raising in India, 222 aging, impact on TFP, 48f
upgrading, 65 aging speed change between 2020 and
human capital accumulation 2030, 58f
affecting TFP, 258 banks’ nonperforming and
boosting, 353 restructured assets, 232f
economic loss by reducing, 162 central bank, attempting to reduce
growth rate in Indonesia, 258 long-term expectations of inflation,
high-skill in Indonesia, 258, 259f 317
supporting, 367 central bank, transparency, 323, 324f
hydrostatics, needed for ship design, 17n Consumer Price Index (CPI) inflation
hyperglobalization, 270, 400 declining, 324
Hyundai, 86n CPI inflation rates, 1985–2016, 288f
credibility, building, 328–330
ideas, 27, 28 current account balances, 1985–2016,
imbalances. See also global imbalances 310f
enhancing transmission from slow US dark corner for, 338
growth to East Asian exports and debt paths, stabilizing, 353
imports, 398 debt ratio change under Scenario 4b,
expanding openness in China, 396 354f
moving together with trade within demographic characteristics, 35t
countries, 394 demographic impact on current
relationship with trade, 388, 389–390 account, 53f
and trade growth at the country level, demographics, impact on 10-year real
391–392, 394–396 interest rates, 61f
IMF. See International Monetary Fund demographic trends, impact of, 42f
(IMF) as an early-dividend economy, 37
immigration endogenous policy credibility, model
effects on the workforce, 43, 43n of, 337–338
exchange rate appreciation, 231

428 SUSTAINING ECONOMIC GROWTH IN ASIA

exporting firms reacting to exchange merchandise and service exports,
rate changes, 202, 204–221 200–201, 201t, 203t
exports monetary policy, establishing
changes in, 193 expectations on inflation, 329
changing composition of, 198–199 months inflation rate below inflation
global share versus China, 198f target, 371f
of goods and services in current US most-favored nation (MFN) applied
dollars, 196f tariff rates, 230f
and imports, 194f old-age dependency ratio, 38f, 57f
pillars for increasing, 222 older workers, share of, 44f
reflecting the new normal, 193–223 openness of the economy, 193–198,
revelation, 4 232–233
sensitivity to changes, 200–202 per capita income level, 39f
as a share of GDP, 195, 195f, 196f personal income tax (PIT) yields, 363f
weak relationship with exchange policy responses, delay of, 331f
rates, 202 primary commodity exports, demand
fertility rate, 37 for from Indonesia, 243n
financial integration with the rest of public debt, actual and projected, 346f,
the world, 194 351f
firm outcomes and exchange rates public spending
by domestic value added, 210t on education as percent of GDP,
investment, 220t–221t 359f
profits, 217t efficiency frontier, 378t
salaries and wage expenses, on health care as percent of GDP,
218t–219t 358f
total expenses, 215t–216t on pensions and health care, 346f
total sales, 213t–214t quarterly projection model for, 328
value of exports, 206t–207t real per capita GDP, growth rate of,
value of imported raw materials, 1985–2016, 287f
208t–209t real short-term interest rate, 292
fiscal packages, 380t, 383t real short-term interest rate, 1985–
foreign capital, volatility of the supply 2016, 293f
of, 251f research and development spending,
foreign exchange reserves and net 362f
official assets, 1985–2016, 311f resources, misallocation of, 232
GDP growth, 197 Scenario 3 and, 350, 352
global value chains, limited role in, scope for quantitative easing,
230, 231f December 2016, 303t
in Group 3, 347b scope limited to reallocate spending,
growth, continued to enjoy higher, 194 366n
growth slowdown, role of experts in, services in export basket (including
196f, 197f oil), 199f
infrastructure, index of quality of, 356f SOEs in, 376
insulation from global growth and/or structural rigidities, needing to
protectionist shocks, 194 address, 233
internal trade barriers, 231 supply shocks, 328–330, 330f
labor force participation, 48f, 49f tariff barriers, relatively high, 230, 230f
long-term public debt projection, 349t tax revenue sources, 364f
manufacturing and services exports: VAT increases, already factoring in,
check for nonlinearity, 227t–228t 366
manufacturing exports, sectoral working age populations growing
change in, 2003–15, 199f substantially, 344, 346f
market share of exports over time, youth dependency ratio, 57f
211f, 212f indirect taxes, 366, 368–369, 376

INDEX 429
Indonesia in Group 3, 347b
aging, impact on TFP, 48f growth
aging speed change between 2020 and domestic and external contributions
2030, 58f to, 251, 253–255
average openness (exports + imports), higher since the global financial
2000–16, 245, 245f crisis, 194
budget deficit legally limited, 266 rate of real per capita GDP, 287f
CPI inflation rates, 288f rate of the working-age population,
credit gaps, increases in, 119 261f
current account balances, 310f sources of lower potential, 256
debt ratio change under Scenario 4b, health care, gap between actual and
354f expected public spending on, 355n
demographic characteristics, 35t imports by region, 245, 246f, 247
demographic impact on current infrastructure, index of quality of, 356f
account, 53f labor, future paths for, 261
demographics, impact on 10-year real labor force participation, impact of
interest rates, 61f higher, 49f
demographic trends, impact of, 42f labor force participation rate, 48f,
domestic and external factors role in 261–262, 261f
the deceleration of TFP growth in, long-term public debt projection, 349t
258 medium-term potential growth
domestic demand, 239, 241 increasing, 262
domestic real GDP growth, correlation months inflation rate below inflation
with domestic and global factors, target, 371f
253, 253t more exposed than other large
as an early-dividend economy, 37 economies, 245
Emerging Market Bond Index (EMBI) most-favored nation (MFN) applied
spread, 252f tariff rates, 230f
energy and commodities, producer of, as a net debtor, 243n
266 net emigration from, impact of, 43
export demand, insulated from weaker, old-age dependency ratio, 36t, 38f, 57f
247 older workers, share of, 44f
exports by region, 245, 246f, 247 output dynamics, 238–242, 255–257
external balance sheet foreign currency per capita income level, 39f
exposure, 250f personal income tax (PIT) yields, 363f
external demand from advanced population with tertiary education,
economies, 247, 247f 259f
external financing, cost of, 250, 252f potential growth rate, 256–257, 257f
external sector, 244, 245f potential output growth rate, 255–256,
fertility rates, 19, 37 256f
financially integrated economies, one protectionism, increase in, 258
of the least, 249, 249f public debt, actual and projected, 346f,
fiscal buffers, 353 351f
fiscal packages, 381t, 383t public spending
foreign exchange reserves and net on education as percent of GDP,
official assets, 311f 359f
foreign participation, bans on, 249 efficiency frontier, distance from,
global developments, insulated from, 378t
263 on health care as percent of GDP,
global economy, financial integration 358f
with, 248, 248f on pensions and health care, 346f
global economy, reduced financial R&D spending, 357
integration with, 248f real GDP growth, 239f, 240f, 241f, 247f
global financial integration, 249f real GDP volatility, 241f

430 SUSTAINING ECONOMIC GROWTH IN ASIA

 real neutral interest rates, 56f under a central bank’s target, 67
real output growth, low volatility of, correlation to Indonesia’s real GDP
239, 241f growth, 253t
real short-term interest rate, 293f dangers of ultra-low, 289–298
real trade growth, 393f deviation from targeted, 321
regulatory environment impeding expectations, 321–322, 337
foreign direct investment (FDI), 248 failing to increase as no longer Japan-
regulatory quality, 1996–2015, 260f specific, 130
research and development spending, keeping at or above 2 percent, 4
362f keeping within policy mandate, 312
Scenario 3 and, 350, 352 lower expected accompanied by low
scenario analysis, 262f nominal rates of interest, 292
scope for quantitative easing, overshooting target, 323
December 2016, 303t persistence, higher levels of, 328
secular stagnation secular downward pressure on, 294
the age of, 235–263 inflationary pressures, containing for India,
countering, 255–262 338
exposure to in advanced economies, inflation forecast-based reaction function,
244–251 ignoring inflation shocks, 320
SOEs in, 376 inflation forecasts, 318
spillovers of secular stagnation, inflation-forecast targeting, 319, 320, 324
238–255 inflation rate(s)
students with higher education, 259f in the Asia-15 economies, 288–289,
tax revenue sources, 364f 288f
temporary trade restrictions on failure to raise in Japan, 130
imports, 258 of at least 2 percent needed, 291–292
temporary trade restrictions on low coupled with a low equilibrium
imports, 1996–2015, 260f real interest rate, 294
10-year government yield, change in, measured by CPI in Japan, 290f
55f number of months below inflation
trade and financial engagement, target in selected Asian economies,
ongoing decline in, 238 371f
trade-based exposure to secular inflation spiral, 320
stagnation, 244–248 inflation surprises, 295, 300, 301f
trade growth in, 399t inflation targeting, 285, 317–319
turning inward, 4 informality, contributing to low tax yields,
volatility of the supply of foreign 359
capital, 251f infrastructure
working age populations growing, 344, building and improving in India, 222
346f increasing spending, 355
youth dependency ratio, 57f index of quality of, 356f
inductive method, in science, 25 investment, 353, 367
Industrial Bank of Korea, rate for 3-year in China, 187
bonds issued by, 175f impact on growth, 368
industrial concentration, in Korea, 86 in physical and human in India, 222
industrial research productivity, decline in innate ability, gap across workers in Japan,
Japanese, 83–85, 84f 155
Industrial Revolution, 15, 16 innovation
Industrial Technology Research Institute bad institutions supporting resistance
(ITRI), in Taiwan, 93 to, 23
industry creation, science-based and closely determining productivity growth at
connected to university research, 76b the frontier, 107
inflation incremental, 90, 100
in Asia, 2

INDEX 431
 Japanese companies underinvested in, exogenous having no nominal anchor,
163 320
Japanese style focused on relatively exposures rising to risks, 308
incremental invention, 79n financial stability risks, keeping too
leadership in East Asia, 74 raising, 334
maintaining a high level of, 29 foreign secular stagnation, insulating
pro-incremental and pro-incumbent from, 114
biases in East Asia, 96 high enough to counter a potential
rates of, 6 financial threat as crippling, 305
reform of Japan’s system for, 84 levels before the global financial crisis
shifting from radical experimentation and after, 115
to incremental improvements of low
existing designs, 75b creating larger maturity mismatch
systems, 3, 75b risks, 308
innovative activity, incremental, 75b encouraging debtors to continue
innovative firms, motivating employees in borrowing, 278
Japan, 85 having problematic side effects, 103
innovators, output sustaining growth leading central banks, 335
without increasing effort, 27 transmitted across countries, 104
institutional changes, reorienting from in the macroeconomic spillover data,
external to domestic demand, 237 115
institutions near their effective lower bound, 316
extending credit to households and not responding to much higher public
firms, 334 debt levels, 132
quality of, key to adopting superior policy in Taiwan, 92
technologies, 259 world real, 55f
insufficient final demand (excess saving interest rate growth differential. See IRGD
problem), in Japan, 139 (interest rate growth differential)
insularity, in Indonesia, 244n interest rates, low, reducing debt service
insurers and service providers, competition burdens of borrowers, 304
among, 375 interfirm alliances, in Japan, 82–83
intellectual property protection, in Japan, international financial cooperation, need
78–79 for, 283
intellectual property rights, 6, 76b international financial institutions,
intellectual property system, strengthened structural impediments in Indonesia,
in Japan, 84 259–260, 260f
interest parity equation, 112 international financial integration,
interest rate(s) measures of, 125
adjusting to the target, 320 International Monetary Fund (IMF)
adverse effects of low, 124 Asia and Pacific Department of, 2
before and after the global financial on export performance of India, 229
crisis, 115f, 117f monitoring external excess imbalances
aversion to large changes in, 321 annually, 282
decline in (real) attributable to changes projecting a bigger US current account
in real sector fundamentals, 176 deficit in coming years, 281
decline in long-term, 54, 55f projections of growth prospects for
declining in Asia, 2 advanced economies, 280
decreasing everywhere since the early international monetary regime, period of
1990s, 123 adjustment to the new, 269
demographic trends, impact of, 54–62 international spillovers, exposing lower-
demographic variables, impact of, 56t income economies to secular
domestic demographic factors, impact stagnation, 2
of, 59 intra-Asian trade, as an established
phenomenon, 410

432 SUSTAINING ECONOMIC GROWTH IN ASIA

inventions debt ratio change under Scenario 4b,
epistemic base of, 15 354f
as incremental, 75b deflation, policies to exit, 330–333
land-augmenting, 12n deflation spirals, never fell into, 130
recombining existing devices, 26 deflation trap, struggling to escape
temporary impact in the Middle Ages, from, 165
15 demographic characteristics, 35t
inventors, 26 demographic impact on current
investment account, 53f
adjusting over time to the labor force, demographics in, 171f
41n demographic trends, impact of, 42f
impact of aging on, 49–50 early postwar era, novelty in, 77
impact of demographic variables, 51 economic development of, 77–85
tending to drive trade movements, 400 economic integration, pursuing greater
weak, accounted for three-quarters of with the world, 130
the global trade slowdown, 229 economy still in a dark corner in early
zero population growth changing the 2017, 331
composition of, 20 educational excellence in the postwar
inward-looking stance, of Indonesia, 238, era, 77–78
263 employees by firm size, all industries,
inward migration, prolonging the 159t
demographic dividend, 43 exchange rate advantages and effects,
iPhone, by Apple, 90 79–80
Iran, fertility rates, 19 fertility rates in, 168f
Ireland, taking advantage of inventions financial markets, 81–82
made elsewhere, 28 firms
IRGD (interest rate growth differential), filing a large number of patent
348, 350, 353, 353n applications, 79
Italy, per capita income level, 39f losing ground in the 1990s, 84
outengineering Western rivals,
James Webb Space Telescope, 24 73–74
Japan fiscal packages, 380t, 382t
aging foreign exchange reserves and net
impact on growth, 41 official assets, 311f
impact on TFP, 48f foreign technology purchases,
rapid until 2030, 52 government policy of limiting,
of the working-age population, 169n 93n–94n
aging speed change between 2020 and GDP, spent on pensions and health
2030, 58f care, 344, 346f
banking system, subsidized lending Government Pension Investment Fund
to, 302 reforms in, 66
business R&D and government in Group 1, 347b
support for business R&D, by firm growth
size, 151f drag on stable for, 41n
capital-to-output ratios, 168f driven by domestic sources of
central bank transparency, 323, 324f demand, 82
colonial rule markedly less harsh in policies to achieve higher
Taiwan than in Korea, 91 sustainable, 324
competition in, 82–83 human capital development, 77–78
consumer price index (CPI), 174f import share of exports, 398
CPI inflation rates, 288f industrial research productivity,
current account balances, 274f, 310f decline in, 83–85, 84f
debt path, stability of, 353 inflation, 286, 289–291
debt ratio as an outlier in Asia, 344 inflation rate measured by CPI, 290f

INDEX 433
 infrastructure, index of quality of, 356f as a post-dividend economy, 35
innovation system, 78, 83n productivity growth, 73
interest rates lowered after the crisis, public debt, actual and projected, 351f
116 public spending
Korea caught up with in the export on education as percent of GDP,
market, 171 359f
Korea’s monetary policy compared efficiency frontier, distance from,
with, 174n 378t
labor, decline in the quality of, 163 on health care as percent of GDP,
labor force participation, impact of 358f
higher, 49f on pensions and health care, 346f
labor force participation rate by age QE, currently engaged in, 302
and employment status, 2013, 147f quality of infrastructure, 355, 356f
labor force participation rates, 48f R&D productivity, trends in, 84f
labor market flexibility enhancing with real 10-year bond yield, 272f
human capital accumulation, 148 real estate market in a bubble in the
labor market institutions, 80–81 1990s, 173–174
labor productivity, 140f, 150f real estate prices and consumer price
labor quality gap, 157f index, 174f
labor supply, transformation of, 134 real neutral interest rates, 56f
LFPRs increased the most in, 49 real per capita GDP, growth rate of,
literacy proficiency, 142f, 143f 287f
long-term public debt projection, 349t real short-term interest rate, 293f
lost decades, 3, 129 real trade growth, 393f
macroeconomic developments, 290f research and development spending,
monetary policy in the 1990s, 174n 362f
months inflation rate below inflation revealed comparative advantages of,
target, 371f 172f
natural interest rate around zero, 178n Scenario 3 and, 350
natural interest rate below zero, 178 Scenario 4a and, 352
net investment, 279f Scenario 4b and, 352
nominal GDP, growth rates, 290f scope for quantitative easing,
nominal GDP growth rates, 173f December 2016, 303t
nonmanufacturing sector wage and secular stagnation, 2, 139
productivity differences by firm size, students, government subsidized study
156f abroad for, 77
nonregular employees, share of in total students, regularly outperform
workers, by sector, 144f Americans, 87
nonregular employment problem, tax revenue sources, 364f
143–149 10-year government yield, change in,
observed phenomena today occurring 55f
worldwide, 135 Three Arrows Plus policy package, 333f
off-the-job training expenses by firm total factor productivity, 96, 141f
size, 158f trade growth, slowdown in, 194
old-age dependency ratio, 35, 36t, 38f, trade growth in, 399t
52, 57f trade imbalances, widening, 391
older workers, share of, 44f trade surpluses, expanded, 388
output gap, 1954Q1–2017Q1, 290f unemployment rate, 1954Q1–2017Q1,
patent system of Bismarck-era 290f
Germany, 78–79 universities underperforming in
pension system with automatic graduate education, 78
adjustment mechanisms, 375 VAT increases, factoring in, 366
per capita income level, 39f wage and productivity differences,
personal income tax (PIT) yields, 363f 153f, 154f, 155f

434 SUSTAINING ECONOMIC GROWTH IN ASIA

 wage level, 145f demographics, impact on 10-year real
workforce, share of, 45f interest rates, 61f
working-age population, decline in, 38 demographic trends, impact of, 42f
youth dependency ratio, 57f economic development of, 85–91
Japan Industrial Productivity (JIP) economic performance of, 165
Database, 144–145, 163 educational excellence in the postwar
Japanization, Korea at risk of, 170–175 era, 77–78
job card system, in Japan, 148, 163 educational system expanded in reach
Jorgenson-Griliches-type labor quality and quality, 87
indices, 152 fertility rates, 168f, 177f
firms rarely the source of major
keiretsu networks, in Japan, 82 product innovations, 90
Kircher, Athanasius, 26n fiscal packages, 380t, 382t
KLEMS-type data of the market economy, foreign exchange reserves and net
in Japan, 150 official assets, 1985–2016, 311f
kleptocracies, 14 fundamentals similar to those in Japan
knowledge 20 years ago, 170
exponential growth of, 27 in Group 2, 347b
making accessible, 26 growth, declining potential, 165–178
propositional and prescriptive, 18 growth accounting of, 167t
of underlying natural rules, 16 growth rate of real per capita GDP,
knowledge-based jobs, age making no 287f
difference in performance, 20n growth trend, 166–170
knowledge-based technological change, 16n health care, gap between actual and
knowledge transmission, improved expected public spending on, 355n
methods of, 27 housing prices rising with general
Korea (South Korea) prices, 174, 174f
age-related spending projected to rise, human capital policy choices, 78
344, 346f import share of exports, 398
aging with inflation below 1 percent in 2016,
fiscal pressures from, 353 289
impact on growth, 41 infrastructure, index of quality of, 356f
impact on TFP, 48f innovative strengths, 91
rapid until 2030, 52 interest rates, lowered after the crisis,
aging speed change between 2020 and 116
2030, 58f Japanese patent system, inherited by,
capital-to-output ratios, 168f 88
central bank transparency, 323, 324f “Japanization,” at risk of, 170–175
commodity semiconductor products, in Japan’s footsteps by the end of the
producers of, 89 1980s, 74
consumer price index (CPI), 174f labor force aging, impact of, 344
core CPI inflation, 296f labor force participation, impact of
core inflation, 295, 296f higher, 49f
CPI inflation rates, 288f labor force participation rates, 48f
current account balances, 310f labor productivity, 2014, 140f
debt ratio change under Scenario 4b, level of productivity remained the
354f same over the past 20 years, 96
deflation, not replicating Japan’s, 173 literacy proficiency among 16–65-year-
demographic characteristics, 35t olds, 142f
demographic impact on current long-term public debt projection, 349t
account, 53f median age, projected rise in, 20n
demographic projections for, 166n monetary policy compared with
demographics, 171f Japan’s, 174n

INDEX 435
 monetary policy since the Asian real trade growth, 393f
financial crisis, 296f recession, heading for a prolonged, 3
months inflation rate below inflation research and development spending,
target, 371f 362f
mortality rate, 177f resource allocation, efficiency of,
natural interest rate, declining, and 169–170
monetary policy, 175–178, 175f, revealed comparative advantages of,
176t, 177f 172f
natural interest rate, estimating, 178n Scenario 3 and, 350
negative shocks, preparing policy scope for quantitative easing,
measures for large, 178 December 2016, 303t
nominal and real interest rates, 175f secular stagnation, exhibiting features
nominal GDP growth rates, trends in, of, 2
173f short-term interest rates, 295, 296f
nominal interest rates, 124, 175 students regularly outperform
old-age dependency ratio, 36t, 57f, 344 Japanese students, 87
expected to rise relatively quickly, 57 tax revenue sources, 364f
higher by 2030, 52 10-year government yield, change in,
increasing from 15 to 20 percent, 55f
38f total factor productivity, 177f
older workers, share of, 44f trade growth, 194, 399t
pensions in as partially defined-benefit trade imbalances, widening, 391
systems, 375 trade surpluses, expanded, 388
per capita income level, 39f trend growth in nominal GDP
per capita income trailing Japan’s by deviating from that of Japan, 173
20 years, 172, 173f 20-year lag, mimicking Japan, 3
personal income tax (PIT) yields, 363f underpriced Japanese rivals in the
PISA scores, on the frontier for, 357, 1980s, 89
361f universities, limited research
policies and pro-incremental bias in, capabilities of, 88
87–91 workforce, share of with productivity
as a postdividend economy, 35 rising or falling, 45f
production sites squeezed by Chinese working-age population, absolute
manufacturing capabilities, 90 decline in, 38
public debt, actual and projected, 346f, youth dependency ratio, 57f
351f Korean Ministry of Strategy and Finance, 2
public health care spending, 355 Kuwait, invasion of in August 1990, 21
public spending
on education as percent of GDP, labor
359f future paths in Indonesia, 261
efficiency frontier, distance from, increases in the supply of, 133–134
378t increasing supply through provision of
on health care as percent of GDP, social benefits, 353
358f reforms having a long-term impact,
on pensions and health care, 346f 368f–369f, 370
R&D expenditure, high, 357 reforms package, 366–367
R&D spending among, 88f labor force
real estate prices and consumer price declining in China, 184–185
index, 174f impact of aging on, 344
real natural interest rates, effects of impact on economic growth, 40–43
changes in fundamental variables participation
on, 176t in China, 183f
real neutral interest rates, 56f, 177f increasing, 65
real short-term interest rate, 293f participation rate

436 SUSTAINING ECONOMIC GROWTH IN ASIA

 by age and employment status, in in Japan and United States, 1947–
Japan, 147f 2012, 140f
ages 15–64 in, 48f in major OECD countries,2014, 140f
impact of aging on, 46–49 raising, 367
impact on growth for small changes labor quality
in, 49n as critical to Japan’s worsening
in Indonesia, 261–262 productivity performance, 162
Indonesia vs. EU-28 and OECD, differences declining between large
261f and small firms, 152
male, 46n differences in based on Jorgenson-
projections and assumptions about, Griliches approach, 153, 157f
261 effects of improvement in in Japan,
remaining unchanged, 41 145n
shrinkage, postponement of in Japan, gap, 157f
134 implications for policy and companies,
significant decline in since 2013 in 163
China, 182, 183f labor unions, prioritizing job security in
labor income, 62n, 353 Japan, 148
labor input, in Japan, 150 land and natural resources, diminishing
labor-intensive manufacturing/services returns from, 11
pathway, 409 land contamination, costs likely to rise, 183
labor market Laos
bias against workers over 50, 20n actual and projected public debt, 346f
changes in Japan, 133, 162 distance from public spending
decrease in the number of workers efficiency frontier, 378t
entering in Korea, 166 in Group 3, 347b
enhancing flexibility, 134 personal income tax (PIT) yields, 363f
institutions in Japan, 80–81 public spending on education as
issues key to Japan’s revitalization, 142 percent of GDP, 359f
policies increasing government public spending on health care as
spending, 367 percent of GDP, 358f
policies promoting active, 65 tax revenue sources, 364f
pro-incumbent bias of institutions in trade growth in, 399t
Japan, 81 large balance sheet effects, absence of not
reforms in Japan, 130 specific to Japan, 135
reforms offsetting adverse growth large firms. See also firms
effects of aging, 65 decline in technology spillovers from
reforms relieving Japan’s two in Japan, 149
structural problems, 143 labor quality based on the Jorgenson-
rigidity of Japan’s, 148 Griliches approach, 153–154, 157f
tightening in Japan, 141 providing more job training to workers
labor mobility in Japan, 155
across firms limited in Japan, 78, 155 lasers, 24–25
importance of, 75b, 76b late-dividend demographic characteristic,
in response to demand shocks in 35t
Korea, 169 late-dividend economies, 35, 37
labor productivity Laue, Max von, 24
differences between small and large learning-by-exporting hypothesis, 107
firms by country, 150f Lee Jae-yong, 87
gap between Japan and the United Lee Kuan Yew, 22
States widening, 162 Lee Kun-hee, 87
growth declining reducing annual Lee Myung-bak, 86
economic growth, 184 legal systems, biased, 12
in Japan, 149, 150f, 152, 162 legislative support, for fiscal rules, 373

INDEX 437
Lehman Brothers, collapse of, 294f long-term and policy interest rate, for
leverage, relating to strategic replication, Japan, 333f
407–408 long-term bond yields: Emerging Asia and
LG, 86n selected advanced economies, 294f
liberal democracy, in Italy and Germany long-term inflation expectations
after 1945, 23 for India, 338
life-cycle savings, positive effect of aging in Japan, 291
speed, 58 long-term interest rates
life expectancy declined since 2000, 292, 294f
in Asia, 34f demographics and, 60t
increases in, 46 of emerging Asian economies, 292
linking changes in the retirement age in Korea, 175, 175f
(or benefits) to, 65–66 lowered, 307
rising, 34 long-term output, estimating, 41
lifetime employment system, 80–81, 148 long-term public debt projections, 349t
life-work balance, in Japan, 141, 148 long-time employment relationships
light radar (Lidar), 25 declining in importance in Japan, 147f
“limited regular employment,” in Japan, in Japan, 146
148 loss absorbency requirements, reducing
linear inflation forecast-based reaction spillover risks, 334
function, for Australia, 325, 326f loss function, procedure minimizing, 321
linear reaction functions, performing loss-minimization approach, to policy
poorly in abnormal times, 321 formulation, 320–322
liquidity-constrained households, 365 loss-minimizing policy response, simulated,
liquidity constrained households, 365n 329, 330f
liquidity traps lost decade, in Japan, 129
avoiding the danger of, 295 lower-for-longer interest rate, 325
emerging-market and developing lower-income countries, tending to grow
economies at a safe distance from, faster, 169n
124 low-inflation environment, payoff to fiscal
falling into, 114 reforms in, 370–371
global, 105 low-inflation trap, 316, 316n, 320
international propagation of, 105n
pushing economies into, 124 M2 growth, in China, 181–182, 187
literacy proficiency, 142f, 143f macroeconomic developments
loan-to-value (LTV) ratios, compared to in Asia, 287–289, 287f
debt-to-income ratios, 308 in Japan, 290f
local bond market liquidity, swings in macroeconomic forecast, publication of,
affecting local financial conditions, 322
297 macroeconomic frameworks, undertaking
local governments (China) stabilizing stimulus, 8
avoiding further expansion of debt, macroeconomic leverage, 181, 184, 187
189 macroeconomic management framework,
central government issuing mandates in China, 4, 181, 187, 191
for, 186 macroeconomic policies, 7, 64, 68
contingent and implicit liabilities of, macroeconomic policy frameworks, 7,
190 186–190
debt surging, 187 macroeconomic slowdown, in Japan, 84
promotion of officials, 186 macroeconomic spillovers, in the
topping up the national growth target, transmission of secular stagnation,
188 111
longevity, impact on current account macroeconomic stability screening, setting
balances, 51–52 a limit on total government borrowing,
189

438 SUSTAINING ECONOMIC GROWTH IN ASIA

macroeconomic theory, on spillovers in efficiency frontier, distance from,
productivity growth, 104 378t
macro forecast, publication of, 319 on health care as percent of GDP,
macroprudential instruments, design of, 358f
308 on pensions and health care, 346f
macroprudential policies real GDP growth relative to ASEAN-4,
addressing vulnerabilities from 2000–16, 241f
accommodative monetary policy, real GDP volatility, 2000–16, 241f
334 real neutral interest rates, 56f
on high loan-to-value lending for real per capita GDP, growth rate of,
mortgages, 309 1985–2016, 287f
intensity of, 120, 121f real short-term interest rate, 1985–
new tools, 287 2016, 293f
using, 114 real trade growth, 1998–2008,
Malaysia 2012–2015, 393f
aging, impact on TFP, 48f regulatory quality, 1996–2015, 260f
aging speed change between 2020 and research and development spending,
2030, 58f 362f
CPI inflation rates, 288f Scenario 3 and, 350
credit gaps, increases in, 119 scope for quantitative easing,
current account balances, 310f December 2016, 303t
debt paths, stabilizing, 353 SOEs in, 376
debt ratio change under Scenario 4b, tax revenue sources, 364f
354f trade growth in, 399t
demographic characteristics, 35t workforce, share of with productivity
demographic impact on current rising or falling, 45f
account, 53f youth dependency ratio, 57f
demographics impact of on 10-year Maldives, in Group 3, 347b
real interest rates, 61f Malthus, Thomas, 10
demographic trends, impact of, 42f Malthusian checks, evidence for the
fiscal packages, 381t, 383t existence of, 11n
foreign capital, volatility of the supply Malthusian curse, 10–12, 19
of, 251f manufacturing center, China’s rise as, 400
foreign exchange reserves and net manufacturing exports, 198, 199f
official assets, 1985–2016, 311f manufacturing firms, connections with
government yield, change in 10-year, commercial banks, 81
55f manufacturing sector
in Group 3, 347b differences in labor productivity and
infrastructure, index of quality of, 356f wages increased in Japan, 153, 154f,
labor force participation, 48f, 49f 155f
as a late-dividend economy, 35, 37 variation in partner income elasticities
long-term public debt projection, 349t across, 202, 203t
most-favored nation (MFN) applied marginal productivity of capital,
tariff rates, 230f determining the (real) interest rate,
old-age dependency ratio, 36t, 38f, 57f 176n
older workers, share of, 44f market leaders, attracting talented pioneers,
per capita income level, 39f 75b
personal income tax (PIT) yields, 363f market liquidity, in bond markets, 297
public debt, actual and projected, 346f, markets, buffering role against shocks, 323
352f median age, in the United States increasing,
public spending 19n–20n
on education as percent of GDP, median public debt-to-GDP ratio, 344, 345f
359f medical care, crushing expenses of, 20

INDEX 439
medieval economy, as dynamic in many adjusting to the outlook for financial
areas, 15 stability, 307
Meiji Restoration of 1868, 77 aggressive easing in Abenomics, 139
mercantilist protectionist policies, 14 allowing the interest rate to vary, 323
merchandise and service exports, avoiding prolonged underemployment
determinants of India’s, 200–201, 201t of resources and to sustain
merchandise exports, from India, 4, 222, investment, 288
231f buffer to counter negative shocks in
microeconomic foundations, of Korea and Japan, 178
consumption and investment, 338 building credibility amid frequent
microscopes, improvements to, 24 supply shocks, 328–329
middle-income trap, idea of, 134 countering weakness in aggregate
migration, 37, 41, 43 demand, 307
minimum pension guarantees, providing, credibility of, 337
66 easy, encouraging risky behavior in
ministries, setting targets for sectoral financial markets, 286
growth, 186 effects on the real economy, 334
Ministry of Finance, pilot programs in enhancing the independence of in
China, 190 China, 186, 188, 191
model(s) expansionary improving financial
deep learning, engaging in data stability, 309
mining, 25 following an inflation-targeting regime
of endogenous policy credibility for in APDOMD, 365
India, 337–338 governed by an interest rate reaction
of a small open economy responding function, 339
to secular stagnation abroad, impact on inflation, 286
111–114 implications of slower growth rates,
summary of, 337–340 285
for sustaining growth for Asian improving communications, 317
economies, 8 in Indonesia, 253t
used by central banks to conduct interest rate by region 2000–16, 115f
policy, 320 in Korea since the Asian financial
model-based simulations, illustrating crisis, 296f
the growth payoff of fiscal reform measures of the BOJ, 291
packages, 377 minimizing a quadratic loss function,
Moerner, William, 24 338
monetary accommodation, with debt in the new mediocre, 315–340
financing leading to higher real GDP, not providing a firm enough anchor in
371, 372f India, 329
monetary and exchange rate policies, for reacting promptly to negative shocks
sustained growth in Asia, 285–312 to growth, 285
monetary and fiscal frameworks, regular reports, 317
reinforcing one another, 332 responses to negative output shocks,
monetary and fiscal policies, room for 316
maneuver, 335 results from low credibility, 328
monetary ease, reduced the risks of banking risk avoidance strategy for, 317–324
system loan portfolios, 334–335 risks to financial stability constraining,
monetary expansion, natural bias toward 304–309
excess in China, 188 scope for unconventional, 300–304
monetary growth targets, setting in China, slowdown in growth for, 4
187 in Thailand, 296f
monetary policy unable to offset the impact of fiscal
accommodative in Thailand, 372f shocks in Japan, 131

440 SUSTAINING ECONOMIC GROWTH IN ASIA

 unlikely to jolt an economy out of a Netherlands, 14, 28. See also low countries
trap, 332 net international investment position
monetary transmission, changing with (NIIP), 243n, 278, 279f
aging, 64 “new” Asian model, 408–410, 411
money laundering, addressing, 283 new mediocre
Mongolia avoiding in Asia with fiscal policy,
distance from public spending 343–383
efficiency frontier, 378t implications for medium- to long-term
fiscal packages, 381t, 383t fiscal sustainability, 343
in Group 3, 347b as a problem in Asia, 316
infrastructure, index of quality of, 356f public debt dynamics under, 344–353
months inflation rate below inflation Newton, Isaac, theory on hydrostatics, 17n
target, 371f New Zealand
public spending on health care as aging, lower speed of, 52
percent of GDP, 358f aging impact on TFP, 48f
research and development spending, demographic characteristics, 35t
362f demographic dividend, 41, 43
tax revenue sources, 364f demographic impact on current
mortality rates, 11, 11n account, 53f
most-favored nation (MFN) applied tariff demographic trends, impact of, 42f
rates, 230f higher old-age dependency by 2020, 52
Mubarak, overthrow of, 24 immigration, impact of on the
multilateral dispute resolution, 283 workforce, 43
Myanmar immigration in, 37
actual and projected public debt, 346f labor force participation, 48f, 49f
in Group 3, 347b as a late-dividend economy, 35, 37
old-age dependency ratio, 36t, 38f, 57f
national integration, promoting in India, older workers, share of, 44f
230–231 per capita income level, 39f
natural phenomena and patterns, observing real interest rates, maintained higher,
and analyzing, 24 116
natural philosophers, 16 youth dependency ratio, 57f
natural rate of interest Nigeria, theft and corruption by ruling
in advanced economies, 1990–2016, elites, 14
273f 90 percent debt-to-GDP ratio, as a
change in, 176t tipping point for government debt
decline in, 54, 56f, 175, 178 sustainability, 132
domestic falling by the same amount nitrogen-fixing process, 12n
as foreign, 112 nominal and real interest rates, in Korea,
drivers of the secular decline in, 175f
55n–56n nominal GDP
estimating for Korea, 178n growth rate in Japan, 290f
keeping close to the natural level, 111 growth rates in Korea and Japan, 173f
variables of the decline in Korea’s, 176 nominal interest rate
natural resources, exported from Indonesia, close to zero in the future, 294
243n constrained by the effective lower
net capital inflow with an increase in bound, 323
domestic demand, 112 equation for, 114
net debtor, lower NIIP, 243n hitting the zero lower bound, 123
net-debt-to-GDP ratio, downward for source for, 125
Japan, 332 zero bound on, 114
net export improvement, limited following non-Asian advanced economies, comparing
a depreciation, 135 with emerging-market and developing
net government debt, for Japan, 333f economies, 115

INDEX 441
non-Asian emerging-market economies, in Asia, 36t
inflation was significantly higher, 116 change between 2020 and 2030, 57f
nonfinancial corporate debt, low in India, defined, 36t, 38f, 57f, 58f
232 in East Asia, 34
nonfrontier firms, TFP differences with increasing from 15 to 20 percent in
frontier firms, 149n China, Korea, and Thailand, 344
nonmanufacturing sector, in Japan, in Korea following that of Japan, 170,
143–144, 144f, 153, 156f 171f
nonmarket economy, in Japan, 143n rising globally, 19
nonmonetary factors, having a major years to increase from 15 to 20 percent,
influence on inflation, 133 38f
nonperforming loans, 232 “old” Asian model, 407–410
nonregular employees “old corruption,” declined in Britain after
defined, 162 1750, 23
investing in the human capital of, 163 older people, 65, 375
in Japan, 143n, 154, 157f older workers
in total workers by sector, in Japan, associated with reduction in labor
144f productivity growth, 46, 47t
training for, 144 male, as nonregular employees in
nonregular employment, in Japan, 143–149, Japan, 146
162 positive effects on labor force
nonregular workers participation, 375
accumulating less human capital, 162 reducing annual growth, 46
employment adjustment in Korea, 169 in working-age population in Asia, 44f
increase in rising, 146 one-child policy, 182
providing benefits to, 163 “one county, one zone,” as inappropriate in
wages rising in Japan, 148 China, 189
nontraditional employment, in the one-year US Treasury bond rate, 251, 253n,
nonmanufacturing market economy, 253t
162 open capital markets, 76b
notional defined contribution pension open-economy New Keynesian model, for
scheme, 375 Australia, 337
nuclear blackmail, by medium-sized but openness and integration, in Asia, 136
poor nations, 21 Oppo, 91, 91n
numerical target, clarifying monetary Organization for Economic Cooperation
policy, 317 and Development (OECD), Japan the
first nonwestern nation to join, 77
obsolescence, issue of, 408 “original sin,” 250n
off-the-job training expenses, by firm size, output, 321, 329
1975–2010, in Japan, 158f output dynamics
oil exporters in Indonesia, 238–242, 255–257
current account, 275f Indonesia’s output dynamics, 238
with exchange rate pegs, 280n output gap
oil imports, countries understanding the alternative measures of, 299–300, 299t
cost of, 273n BOJ’s estimate of, 290, 290f
Okun’s law, on an excess unemployment defined, 321
rate, 290–291 for Japan, 333f
old-age dependency negative persisting, 335
capital-to-labor ratio and, 51 output growth, in Indonesia, 238
higher found to reduce interest rates, output growth rates, paths for potential,
59 260–262
impact on interest rates, 57, 57f “output of ideas,” 28
reducing current account balance, 51 outward-looking strategy, in Indonesia, 266
old-age dependency ratio

442 SUSTAINING ECONOMIC GROWTH IN ASIA

overlapping generations model, of Kwon debt ratio change under Scenario 4b,
for Korea, 176, 176t 354f
overnight interbank call rate, in Korea, 178 demographic characteristics, 35t
overpopulation, risk of thoroughly demographic dividend, 41, 43
discredited, 19 demographic impact on current
overshooting, of the inflation rate, 336 account, 53f
demographics impact on 10-year real
Pacific Island countries interest rates, 61f
fiscal packages, 381t, 383t demographic trends, impact of, 42f
in Group 3, 347b as an early-dividend economy, 37
Parent, Antoine, 18 fertility rate, 37
Park Chung-hee, 85–87 fiscal buffers, 353
Park Gyeun-hye, 87 fiscal packages, 381t, 383t
partial equilibrium analysis, 53 foreign capital, volatility of the supply
part-time workers, reliance of firms on, 146 of, 2000–15, 251f
passivity, reinforced in Japan, 136 foreign exchange reserves and net
“patent production function,” regression official assets, 1985–2016, 311f
of, 83 global financial integration across
patents emerging markets, 249f
awarded to Taiwanese inventors, 94f in Group 3, 347b
in exporting firms in Korea, 89 infrastructure, index of quality of, 356f
holders of bargaining with contract labor force participation, 48f, 49f
manufacturers and other business long-term public debt projection, 349t
partners, 76b months inflation rate below inflation
infringement in Japan, 79 target, 371f
as a measure of output of R&D, 28 most-favored nation (MFN) applied
one-stop services in China, 185 tariff rates, 230f
Paulson, Henry, 270 net emigration from, impact of, 43
pensions, 20, 66, 375 old-age dependency ratio, 36t, 57f
pension systems old-age dependency ratio increasing
generosity of, 53n from 15 to 20 percent, 38f
reform of, 65–66, 374–375 older workers, share of, 44f
per capita GDP growth, in Japan, 129 per capita income level, 39f
per capita income, relative to the United personal income tax (PIT) yields, 363f
States, 37–38, 39f public debt, actual and projected, 346f,
performance-based wage system, 65 352f
personal and corporate income tax public spending
relying more on, 363 on education as percent of GDP,
relying more on than on general 359f
consumption taxes, 364f efficiency frontier, distance from,
personal income tax (PIT) 378t
cut in as progressive, 366 on health care as percent of GDP,
cutting increasing the labor supply 358f
and lowering wage costs, 369 on pensions and health care, 346f
measures of yields, 359, 363f R&D spending, 357
yields from in selected economies, 363f real GDP growth relative to ASEAN-4,
Philippines 241f
aging impact on TFP, 48f real GDP volatility, 241f
aging speed change between 2020 and real per capita GDP, growth rate of,
2030, 58f 287f
CPI inflation rates, 1985–2016, 288f real short-term interest rate, 293f
current account balances, 1985–2016, real trade growth, 393f
310f regulatory quality, 260f

INDEX 443
 research and development spending, increasing rapidly, 34
362f as inflationary, 175n
Scenario 3 and, 350, 352 and lower growth in Asia, 2
scope for quantitative easing, macroeconomic impact on saving and
December 2016, 303t investment, 51
tax revenue sources, 364f negative effects of on TFP, 169n
trade growth in, 399t posing a challenge for public debt
working-age people, emigration of, 37 sustainability, 344
working age populations growing threatening long-term fiscal
substantially, 344, 346f sustainability, 377
youth dependency ratio, 57f population dynamics, holding back growth,
Phillips curve 10
containing a nonlinear output gap, population growth. See also youth
337 dependency
declining estimates of the slope of, 286 in Asia, 34f
as dormant, 298–300 capital-to-labor ratio and, 51
estimate of an expectations- changes expected to be relatively small
augmented, 298–299, 299t between 2020–2030, 52n
regressions on US GDP deflator, 299t lower lowering growth, 6
physical capital accumulation, 353 rate for Asia, 34
Pitt, William, 23 triggering technological responses, 11
Poland, governance indicators, 22n “positive checks,” Malthusian, 11
policy positive supply shock, deflationary effect on
adapting to older population, 37 aggregate wages, 134
delay of responses in India, 331f post-65 cohort, extended number of years
emerging markets making changes of productivity, 20
to, 237 post-dividend demographic characteristic,
implications for global imbalances, 35t
282–283 post-dividend economies, 35
implications of demographic trends, potential growth, considering the evolution
64–66 of, 260–262
loss-minimization approach to potential growth rate
formulation, 320–322 declined in emerging markets, 255
need for a forward-looking, 5–8 decomposition into the actual
preserving space as the wrong strategy capital growth rate and potential
for an economy at risk of deflation, employment growth rate, 256n
295 decomposition of Indonesia’s,
strategies for Australia, 325–328, 326f, 2001–14, 256–257
327f in Indonesia, 257f, 262
tools as imperfect, 308 potential output
policy interest rates declined in Indonesia as a result of
choosing, 321 lower TFP growth, 263
as conditional, 322 as endogenous, 339
keeping higher than warranted by scenario analysis for Indonesia, 262f
macroeconomic conditions, 304 potential output growth rate in Indonesia
against the same variable after the actual and forecast, 256f
crisis, 116–117, 117f actual and forecast, 2000–22, 255–256
population potential shocks, sources of, 335
of Korea projected to decrease, 170 practical knowledge, sense of contempt
Malthusian responses to, 11 for, 18
projections affecting output through Prandtl, Ludwig, theory of aerodynamics,
aggregate labor and capital, 41 17n
population aging “precautionary principle,” 23
effect on inflation, 175n

444 SUSTAINING ECONOMIC GROWTH IN ASIA

predatory foreigners, attracted to riches of product market efficiency, declining in
economic success, 14 Korea, 170
predatory wars, likelihood of, 21 Product Market Regulation Index of the
“premature deindustrialization,” OECD, Korea’s ranking in 2013, 170n
phenomenon of, 6 product markets, Korea’s among the most
prescriptive knowledge, connection with heavily regulated, 170
propositional knowledge, 18 professional education system, reforming
price stability policy, time-consistent, 317 Japan’s, 148
primary and preventive care, supporting, professions, impact of aging differing
375 across, 44
prime-age population, choice for examining proficiency, greater utilization of needed in
savings-investment relationships, 50n Japan, 143
Principle of Price Stabilization, 305 profits, no significant relationship with
Principle of Productive Credit, 305 exchange rate movements, 205, 217t
private consumption, in Indonesia, 267 Programme for International Student
privateers, use of, 14 Assessment (PISA), scores, 361f, 378
private investment, stimulating in targeted “pro-incremental” bias, in innovation
sectors in Japan, 141 systems, 100
private retirement saving schemes, 375 pro-incumbent bias, 87, 92, 100
private sector R&D, changes in Japanese, projected purchasing power parity, per
83–84 capita income growth rate, 39f
procyclical fiscal policy, reducing, 373 prompt corrective action, as a risk-avoiding
production function in China, labor share strategy, 335
in, 182n prompt effective action, following supply
productivity shocks, 329–330, 331f
determined by the efficiency of factor propositional knowledge, connection with
allocation, 108 prescriptive knowledge, 18
gap between part-time and regular protectionism, 258, 283, 398. See also “fair”
employees in Japan, 145 trade
in the generation of ideas, 27 protectionist sentiment, on the rise, 316
impact of aging on, 44 provider payment systems, improving, 375
improving Japan’s performance, 162 prowess database, collected by the CMIE,
Indonesia focusing on, 267 225–226
spillovers, 106–111 prudential policies, adapting to curb risks,
productivity and interest rates scenario, for 308
public debt projection, 348, 350 Prussia, rent-seeking in, 23
productivity-enhancing reforms, discussed public capital, impact on potential output,
in the IMF’s Regional Economic Outlook 339
(2017), 68 public debt
productivity-enhancing spillovers, limited actual and projected in selected Asian
diffusion to Indonesia, 263 economies, 346f, 351f–352f
productivity growth dynamics under the new mediocre,
deceleration of in China, 184 344–353
decreasing everywhere since the early long-term projections for reflecting
1990s, 123 the impact of the new mediocre,
literature on the determinants of, 347–350
105–106 lower in Asia, 344
lower transmitted across countries as percent of GDP, by region, 345f
through financial channels, 107 ratio to GDP, 347–348
measuring in Japan and other East public debt-to-GDP ratios, 344, 346f
Asian economies, 100 public debt-to GDP ratios, 350, 352–353
policies prioritizing to tackle country- public healthcare spending, 53n, 355, 358f
specific factors impeding, 101 public infrastructure push package,
365–366

INDEX 445
public investment efficiency frontier, 355, public spending on, 357–359, 362f
357f spending on in selected economies,
public pension funds, reforming the 2015, 362f
management of, 66 strategy, in Japan, 84
public policy choices in East Asia, during radical innovation, effects of, 100
years of rapid growth, 74 rate of real exchange rate appreciation
public-private partnerships (PPPs), for versus the US dollar, 251
infrastructure spending, 355 real 10-year bond yield, in advanced
public spending economies, 1990–2018, 272f
challenges of the new mediocre and, real 10-year Treasury rates, after the Asian
353–363 crisis, 272, 272f
on education as percent of GDP, 359f real depreciation, shifting domestic and
efficiency frontiers, 378, 378t foreign spending, 112
on health care as percent of GDP, 358f real effective exchange rate
level, composition, and efficiency of, correlation to Indonesia’s real GDP
355–359 growth, 253t
on pensions and health care, 346f for Japan, 333f
purchasing power, in Indonesia, 267 real equilibrium level, of the interest rate as
“pure imitator,” transitioning to close to zero, 335
“incremental innovator,” 90 real estate
markets in Japan and Korea, 173–174
quadratic loss function, 320–321 prices and CPI in Korea and Japan,
quantitative easing (QE), scope for in the 174f
Asia-15 economies, 302–304, 303t relationship with demographic
quasi-equilibrium (a “dark corner”), of low- variables, 63–64
inflation or deflation trap, 316 real exchange rate
quasi-fiscal borrowing, 190, 191 constant in the long run, 113
quasi-fiscal expenditures, 189 defined, 112
depreciating less in the short run, 113
depreciating over time, 114
R&D (research and development) index for, 225
East Asian approach having both source of, 125
strengths and weaknesses, 74 real GDP. See also GDP
increased government expenditure on effects of fiscal reforms, 368, 368f
in China, 185 growth rate of in Indonesia and in the
investment, appropriating the returns United States, 238–239, 240f
to in Japan, 79 volatility for selected countries, 241f
labs, top corporate effectively led by real GDP growth
PhD recipients in the United States, contribution of domestic and global
78 factors to Indonesia’s, 254f
productivity contributions to Indonesia’s, 247f
declined in the US, 100 for Japan, 333f
declining in East Asia, 96 rates in Indonesia, 239f
slowdown in growth, 3 rates in Indonesia and the United
trends in Japanese, 84f States, 240f
sector, increasing the proportion of the relative to ASEAN-4, 241f
labor force engaged in, 28 trends in Korea and the world, 167f
spending real growth rate, in China, 251
concentration of among South real interest rate(s)
Korean firms, 88f change in, 118
concentration of in a single chaebol- components of, 323
the Samsung Group, 86 computing, 175f
productivity of Japan’s corporate, decreased, 54, 116, 123, 272
73n defined, 293f

446 SUSTAINING ECONOMIC GROWTH IN ASIA

 depending on global growth, 280 remediation costs, in China, 182–183
higher, spelling trouble for advanced rental rate, not always moving together
debtors, 280 with housing prices, 64n
impact of demographic variables on rental system of Korea, called chonsei, 174n
the 10-year, 58–61 rent-seeking
insulating, 114 at historically low levels today, 14
interpreted as the “natural” rate of internal, 14, 21–22
interest, 111 pairwise correlation coefficients of
by region, 115f measures closest to, 15n
real long-term rate, declining for about 30 pervasive in the age of mercantilism,
years, 292 12
real natural interest rate, effects of changes unlikely to stifle long-term economic
in fundamental variables in Korea, growth worldwide, 23
177f repetitive work, productivity declining with
real neutral interest rates, in selected age, 20n
countries, 56f repressing savings, trap of, 136
real output growth rate, in Indonesia, 251 Republic of Samsung. See Samsung Group
real short-term interest rate, 292, 293f research, questions for further, 404
real trade growth, 391f research and development. See R&D
rebalancing (research and development)
economies from foreign to domestic reserve accumulation, excess exerting a
demand, 119 powerful negative externality, 310
toward indirect taxes package, 366 Reserve Bank of India, bolstering the
recessionary impact, of tax hikes of 1997 credibility of the inflation target, 329
and 2014 in Japan, 131 resource allocation, efficiency in Korea,
reforms, effects across Asia, 367–370 169–170
regional integration, preventing domestic resources, finiteness of nonreproducible, 12
stagnation traps, 135–136 restrictions on trade, in Indonesia, 258–259
regional supply chains, 409 retirement age
regular employees average of 57 in Asia, 366
defined, 162 increasing by two years, 366
increases in marginal productivity with raising, 370, 375
age in Japan, 146 retirement system reforms, 66
in Japan, 143n retrenchment, of Indonesia from the global
training for, 144 economy, 238, 263
regular workers revealed comparative advantage, 172f
excessive projections for in Korea, 169 revenue
firms’ decisions on wage rates collection and composition, 359–363
depending on expectations of future efficient mobilization of, 375–376
inflation, 149 revenue policies, challenges of the new
very limited wage increases in Japan, mediocre and, 353–363
148 revenue rules, 373
regulation, undoing the consequences of Ridley, Matt, 26
monetary expansion, 308–309 risk
regulatory capture, in Japan, 83 assertive action called for to avoid, 316
regulatory policies, preserving financial classes of, 307
stability, 286–287 risk-avoidance strategy
regulatory quality based on loss-minimization for
declining path of, 260 Australia, 326f, 327f, 328
in Indonesia, Malaysia, the Philippines, for monetary policy, 4–5, 317–324
and Thailand, 260f resulting in better outcomes in
relative prices, adjusting, 291 Australia, 325
religion, in the secular and humanistic risk management strategy, policy
West, 11 preventing long-term inflation, 329

INDEX 447
robotization, replacing workers, 28 IRGD decreasing gradually to 1
robots, skilled people working percentage point below the 2022
collaboratively with, 409 level by 2030, 350
Romania, improved governance indicators, IRGD rising gradually reaching 2
22 percentage points above the 2022
Rome, technological progress not level by 2030, 350
sustainable, 18 for public debt projection, 348, 350
rule of bad law, 12 scholars, contempt for practical knowledge,
rupee, appreciation of, 204–205 18
rural labor migration, projected for 2028, science, advancing with better tools, 24
185n Science and Technology Basic Law,
Russia, theft and corruption by ruling implemented in Japan, 84
elites, 14 scientific knowledge, access to the existing
pool of, 25–26
“safe assets,” increasing the availability of, scientific products, for commercial use in
66 China, 185
Samsung Electronics, R&D expenditure in scientific research, 27
2010, 88f scientists, 17, 18, 26
Samsung Group second 21st century crisis, 269
high fraction of Korean exports, 86 secondary school enrollment, spending on,
produced more DRAM chips, 89n 360f
prominence in the global smartphone sectors, increased market shares with the
industry, 90, 91 strengthening of the rupee, 205, 212f
share of global smartphone shipments, secular stagnation
91n in advanced economies as transmitted
value of, 86n to Indonesia, 242–244, 255
savers, maintaining faith in domestic assets, affected nominal or real interest rates,
133 118
saving behavior, impact of aging on, 50n appropriate frequency to study, 121
saving determinants, survey of, 50n competing explanations for, 236
“saving” households, holding debt as a contagion across countries, 103
source of wealth, 365 diffusing via international trade in
savings goods and services, 107
aging-related, 52 exposure of Indonesia to, 244–251
demographic trends affecting, 50 foreign, 111, 112, 113, 114
excess in some countries, 105, 387 history and reality, 9–29
and investment across countries, 400 Indonesia countering, 255–262
Scandinavia, rent-seeking in by 1850, 23 Japan’s steps to address, 130
Scenario 1: status quo likelihood of in the foreseeable future,
debt ratio decreasing or stabilizing in 19
all economies, 350 limited transmission of from advanced
for public debt projection, 348, 349t economies to Indonesia, 263
Scenario 2: historical primary balance literature on, 105–106
debt ratio increasing in Japan, and as normal, 9
Malaysia and remaining flat in Sri reducing foreign interest rate and
Lanka, 350 demand in the long run, 113
for public debt projection, 348 results from a shortage of safe assets,
Scenario 3: demographics 105
pensions and healthcare spending, 350 risk of getting trapped in, 7
for public debt projection, 348 as the rule rather than the exception
Scenario 4: productivity and interest rates until 1800, 3
gradual decline in the IRGD, 352 spillovers to output dynamics in
increase in the IRGD raising the debt- Indonesia, 238–255
to-GDP ratio, 352–353 structural causes of Japan’s, 139

448 SUSTAINING ECONOMIC GROWTH IN ASIA

 symptoms of, 2, 103, 123 credit gaps, increases in, 119
as a syndrome, 103 current account balances, 1985–2016,
threat to Asian economies, 1–2 310f
transmitted through links to China, demographic characteristics, 35t
243 demographic trends, impact of, 42f
transmitting from advanced to fiscal packages, 380t, 382t
emerging economies, 236 foreign exchange reserves and net
self-fulfilling crises, model of, 105 official assets, 1985–2016, 311f
senior leadership roles, lack of women in in Group 2, 347b
Japan, 163 immigration, 37
serfdom, abolition of, 23 immigration, impact on the workforce,
service economy, in Japan, 83 43
service exports, surge in from 2003 to 2007, inflation below 1 percent in 2016, 289
198 infrastructure, index of quality of, 356f
services interest rates lowered after the crisis,
per capita consumption of in China, 116
184 labor force participation, 48f, 49f
share in India’s export basket, 199f as a late-dividend economy, 37
services exports most-favored nation (MFN) applied
in Indonesia, 267 tariff rates, 230f
plateaued over the last decade, 198, nominal interest rates fell close to zero,
199f 124
services sector, rise in the share of in China, old-age dependency, expected to rise
184, 185f relatively quickly, 57
shadow federal funds rate, 303–304 old-age dependency, higher by 2030,
shocks 52
effect of large, 335 old-age dependency ratio, 36t, 38f, 57f
flexibility of fiscal rules in the face of older workers, share of, 44f
large, 374 personal income tax (PIT) yields, 363f
identification of, 253n projected to transition to post-
possibility of further large negative, dividend status by 2030, 35
315 public debt, actual and projected, 346f
prompt reactions to establishing public health care spending, 355
credibility, 336 public spending
unhinging inflation expectations, 317 on education as percent of GDP,
short-run fiscal policy, resulting in 359f
significant deviations, 339 efficiency frontier, distance from,
short-term interest rate 378t
in Indonesia, 251 on health care as percent of GDP,
in Korea, 296f 358f
in Thailand, 296f real per capita GDP, growth rate of,
short-term “natural” policy rates (r*), 287f
pattern for, 272, 273f real short-term interest rate, 293f
simulations, spawning entirely new real trade growth, 393f
“computational” fields of research, 25 research and development spending,
Singapore 362f
aging scope for quantitative easing,
experiencing very rapid until 2030, December 2016, 303t
52 tax revenue sources, 364f
impact on growth, 41 10-year government yield, change in,
impact on TFP, 48f 55f
aging speed change between 2020 and trade growth in, 399t
2030, 58f workforce, share of, with productivity
CPI inflation rates, 1985–2016, 288f rising or falling, 45f

INDEX 449
 youth dependency ratio, 57f internal factors of, 105
zero lower bound, currently at, 295 standard economic models, naïve view of
65–80 bracket, increased labor force the inflation process, 133
participation by, 20 Standard International Trade Classification
SK, 86n (SITC) codes, 172f
small and medium enterprises (SMEs), startups and new entrants, limited in Japan,
R&D intensity of in Japan, 149, 151f 81
smaller firms, easier access to capital in stasis, likelihood of reverting to a world
Taiwan than in South Korea, 92 of, 29
small firms, more prevalent in Japan than State Council in China, 188
in the United States, 158 state-owned enterprises (SOEs), reform of,
social knowledge, defining, 26 376
social security system, improving in Japan, status quo
141 average annual growth impact relative
“social skills” jobs, productivity rising with to, 41
age, 20n primary balance as of 2022, 353
SOEs (state-owned enterprises), bailouts of scenario for public debt projection,
troubled, 376 348, 349t
spillovers steam engine, theory of thermodynamics
from advanced economies, 106, 109, and, 17
109t steel, discovery of the chemical composition
complexity of to Indonesia, 255 of, 16
Sri Lanka Stein-Hardenberg reforms, in Prussia after
debt ratio change under Scenario 4b, 1806, 23
354f stimulus programs, temporary, 133
direct taxes as a small share of tax stock and bond markets, regulation of in
revenues, 366 Japan, 82
fiscal packages, 381t, 383t stock indicators, 248
in Group 3, 347b stock returns, demographic trends
infrastructure, index of quality of, 356f affecting, 62, 63
long-term public debt projection, 349t storage-and-search techniques, 26
personal income tax (PIT) yields, 363f strategic replication, 407
public debt, actual and projected, 346f, stress tests, of the balance sheets of
352f financial intermediaries, 287
public spending structural break, in India around 2005, 222
on education as percent of GDP, structural reforms
359f addressing challenges of aging, 68
efficiency frontier, distance from, in Japan, 141, 332
378t in Korea, 178
on health care as percent of GDP, as the main policy option, 288
358f need for supporting, 371, 373–376
on pensions and health care, 346f offsetting adverse growth effects of
R&D spending, 357 aging, 65–66
research and development spending, slowing the increase in healthcare
362f costs, 375
Scenario 3 and, 350, 352 undertaking, 8
stabilizing debt paths in, 353 structural supply-side policies, in Indonesia,
tax revenue sources, 364f 267
stabilization policy, against negative students
shocks, 133 with higher education in Indonesia,
“stagflation,” 329 259f
stagnation sending to the best universities in the
in emerging markets, 236–237 world from Indonesia, 267
exceptions to the rule of, 9n

450 SUSTAINING ECONOMIC GROWTH IN ASIA

subdued inflation, in many countries, 370, growth rate of real per capita GDP,
371f 287f
sublicensing, practice of, 93, 93n human capital policy choices effects,
succession, as an aspect of replication, 408 78
supply, boosting in India as critical, 223 increasingly squeezed between Silicon
supply chains, 400, 410. See also global Valley and China, 96
supply chains in Japan’s footsteps by the end of the
supply shocks, 329, 330f 1980s, 74
supply-side policies and interest rate cuts, less diversified than Korea, 93
not necessarily producing demand, level of industrial concentration much
267 lower than in Korea, 92
supportive arrangements, creating for fiscal lowered interest rates after the crisis,
rules, 374 116
surplus countries, 390, 391f nominal interest rates fell close to zero,
demographic trends exerting upward 124
pressure on current balances, 52 pro-incremental bias in innovation
facing little that would force them to system, 96
adjust excess, 282 real short-term interest rate, 293f
feedback effects to, 396–397 scope for quantitative easing,
focusing research on, 404 December 2016, 303t
import growth and export growth small firms focused on implementing
highly correlated with US import and refining technologies developed
growth, 398 by others, 95
Survey of Adult Skills (PIAAC), ranking trade flows to and from China, 95
Japan at the top, 142, 142f trade growth in, 399t
sustained growth, experienced only in 3 at the zero lower bound, 295
percent of human history, 9 Taiwan Semiconductor Manufacturing
Switzerland, 23, 28 Corporation (TSMC), 93
syndrome, defined, 103n Takenaka, Heizo, 135
talent sources, unlocking in Japan, 164
tailwinds, of very large players in a taper tantrum, in July 2013, 294f, 297–298
neighborhood, 409 tax administration, weaknesses in, 359
Taiwan tax bases, narrow, 359
activity in sectors that no longer hold taxes
technological opportunities, 96 rebalancing toward indirect, 368–369,
committed to educational excellence in 376
the postwar era, 77–78 types of, 339
concentration of US patent rights in taxes and tax collection, efficiency of in
electronics hardware, 94, 94f Asia, 362
CPI inflation rates, 288f tax minimization strategies, addressing,
current account balances, 310f 283
economic development of, 91–96 tax revenue, sources of in selected
electronics firms attracted by China’s economies, 364f
low costs, 95 tax structure, improving to support growth,
entrepreneurial character of managers, 362
92–93 tax yields, 359, 363f
exports and industrial structure Taylor rule, 306
concentrated in electronics and “technical compendia,” compiled by groups
information technology (IT), 93 of experts, 26
foreign exchange reserves and net technical progress, negative feedback loops
official assets, 311f from demographics to, 6
government directing and subsidizing technological advances, limited diffusion to
substantial international technology Indonesia, 263
transfer, 93

INDEX 451
technological breakthroughs, fueled by rapid until 2030, 52
rapid entry of a large number of new aging speed change between 2020 and
firms, 75b 2030, 58f
technological followers, catching up by, 28 core CPI inflation, 296f
technological frontier, moving ahead to core inflation remaining well below 2
avoid secular stagnation, 28 percent, 286
technological innovation CPI inflation rates, 288f
accounting for low growth in advanced credit gaps, increases in, 119
economies, 257 current account balances, 1985–2016,
promotion of in China, 185 310f
technologically dynamic sectors, risks and debt ratio change under Scenario 4b,
rewards for talented young American 354f
managers and engineers, 76b demographic characteristics, 35t
technological paradigm, creating a new, 75b demographic impact on current
technological progress account, 53f
Chinese and pre-1700 Western as demographics impact on 10-year real
experience-based, 16n interest rates, 61f
impact on economic well-being in the demographic trends, impact of, 42f
Middle Ages, 15 fiscal packages, 380t, 382t
land- and resource-augmenting, 12 foreign capital, volatility of the supply
most critical for East Asian economies, of, 251f
101 foreign exchange reserves and net
resource-augmenting, 21 official assets, 311f
technological spillovers, found to raise TFP gap between actual and expected
and growth, 257 public spending on health care,
technology 355n
acceleration of development, 186 global financial integration across
adopting new in Japan, 164 emerging markets, 249f
allowing the elderly to cope with in Group 2, 347b
handicaps of old age, 20n inflation below 1 percent in 2016, 289
as another challenge, 409 infrastructure, 356f, 371, 372f
coevolving with science, 24 labor force aging, impact of, 344
developing and improving, 24 labor force participation, 48f, 49f
gap narrowing in Korea, 169 long-term public debt projection, 349t
gradual widening epistemic base of, 16 monetary ease, conditions supporting
partially offsetting downward pressure further, 311
on economic growth in China, 186 monetary policy, effects of
telescopes, improvements to, 24 accommodative, 372f
Temporary Trade Barriers database (World monetary policy since the Asian
Bank), 258 financial crisis, 296f
10-year government yield, change in, 55f months inflation rate below inflation
10-year real interest rates, 58–61, 61f target, 371f
terms-of-trade growth, correlation to old-age dependency ratio, 36t, 38f, 57f,
Indonesia’s real GDP growth, 253t 344
tertiary education, low levels of in older workers, share of, 44f
Indonesia, 258, 259f pensions in a partially defined-benefit
TFP. See total factor productivity (TFP) systems, 375
Thailand per capita income level, 39f
age-related spending projected to rise, personal income tax (PIT) yields, 363f
344, 346f as a post-dividend economy, 35
aging public debt, actual and projected, 346f,
fiscal pressures from, 353 351f
impact on growth, 41 public spending
impact on TFP, 48f

452 SUSTAINING ECONOMIC GROWTH IN ASIA

 on education as percent of GDP, improving Japan’s macro-level by
359f improving quality of workers in
efficiency frontier, distance from, SMEs, 158
378t increasing in Korea, 178
on health care as percent of GDP, of Japan’s macroeconomy, 141
358f large differences between large and
on pensions and health care, 346f small firms, 163
real GDP growth relative to ASEAN-4, for older workers, 47t
2000–16, 241f from the Penn World Tables version
real GDP volatility, 241f 9.0, 125
real per capita GDP, growth rate of, scenario analysis for Indonesia, 262f
287f sources of changes in, 257–260
real short-term interest rate, 293f total factor productivity (TFP) growth
real trade growth, 393f assumed to rise, 262
regulatory quality, 260f causes of slowing in Indonesia, 238
research and development spending, contributors to slowdown in
362f Indonesia, 263
Scenario 3 and, 350 declining as a result of convergence to
Scenario 4b and, 353 the technological frontier, 258
scope for quantitative easing, in the economy as a whole, 27–28
December 2016, 303t of emerging-market economies
short-term interest rate, 296f correlated with that of advanced
sustained deflation, risk of falling into, economies, 106
295, 296f in Japan, 139
tax revenue sources, 364f as the only source of growth for Korea,
10-year government yield, change in, 169
55f postcrisis relationship with measures
trade growth in, 399t of economic openness, 111
workforce, share of with productivity remaining unchanged, 40
rising or falling, 45f results of three cross-country
working-age population, absolute regressions, 109
decline in, 38 in the United States, 186
youth dependency ratio, 57f total sales and costs, increased following
thermodynamics, theory of, 17 an appreciation of the rupee, 205,
Thirty Years’ War, 13 213t–214t, 215t–216t
Three Arrows Plus package, 332, 333f tradable sector, growth of India’s, 194, 194f
“the 3Cs” approach, comprehensive, trade
consistent, and coordinated, 316 and absolute balance, panel regressions
time-consistency problem, solution to, 318 on, 395t
time dimension, prudential tools building balance, 112
up buffers, 334 barriers, internal in India, 231
total factor productivity (TFP) deficits, eventual pressure to narrow,
accounting for factor utilization, 258n 401
aging, impact on, 43–46, 48f and financial spillover, lower limiting
at constant national prices (RTFPNA), the diffusion of technology, 257
125 flows, facilitating international
as endogenous, 339 knowledge spillovers, 107
factors boosting, 169 growth
human capital component of declining associated with rapid export growth
during downturns, 258 in Asia and rapid import growth in
impact on growth in Indonesia, the Americas, 391, 392f
256–257 decomposing, 390–391
implications for, 401

INDEX 453
 moderating imbalances associated trade-to-GDP ratio, increased, 394
with slowing, 5 training
not sufficient to explain widening differences in on-the-job and off-the-
trade balances, 389 job, 155, 158f
period of rapid, 390, 391f gearing up, 410
for the periods 1995–2008 and moving to universities and vocational
2012–14, 397–398, 398t schools, 163
periods of high accompanied by transformative innovations, requiring
relatively faster import and export incremental inventions, 75b
growth, 394 transgenic crops, opposition to, 23
in surplus Asian countries Trans-Pacific Partnership (TPP), Japan
correlated with US import growth, pushing for, 141
399t Transparency International, corruption
imbalances, 400, 401 indices, 14
improvement in Indonesia, 266 transparent communications, conventional
integration forward guidance for, 322–324
having a negative impact on TFP trend total factor productivity (TFP)
growth, 110 growth
in Japan, 130 boosting potential, 353, 355
measured by the ratio of exports to impact of aging on, 344
GDP in 2007, 109 Trump administration, rhetoric on bilateral
liberalization, encouraging firms to US deficits, 404
reduce the number of products, turning Japanese, risk of in the late 1990s
146n sense, 129
linkages, as a direct source of
transmission of spillovers of secular ultra-low inflation, dangers of, 289–298
stagnation, 242 undisbursed loans, steadily rising in
modeled by reduced-forms equations, Indonesia, 267
339 unemployment, lowering long-term, 367
nature of between countries similarly unemployment rate
endowed, 409 in Japan, 290, 290f
relationship with imbalances, 389–390 lowering the long-term, 370
responsiveness to income and global United Kingdom
imbalances, 398t change in 10-year government yield,
restrictions 55f
doing little to alter global current per capita income level, 39f
imbalances, 283 United States
in Indonesia, 258–259 absolute balance relative to GDP and
slowdown, explanations for, 398, to trade, 394
400–401 account balances and reserve
and specialization, growth from purchases, 271f
vulnerable to political shocks, 14 aging speed change between 2020 and
tensions between the United States 2030, 58f
and trading partners in Asia, 6 business R&D and government
warfare, sparked global imbalances, support for business R&D, by firm
280 size, 151f
trade-facilitating infrastructure, expanding capital-to-output ratios, 168f
or building, 410 central bank transparency, 323, 324f
trade/GDP ratio, without growing current account balances, 1991–2017,
imbalances, 397f 274f
Trade in Value Added (TiVA) database deficit, growing after the mid-1990s,
(OECD), 397n 270
trade-related exposure, of Indonesia to demand, fueling export growth in Asia,
secular stagnation, 244–248 398

454 SUSTAINING ECONOMIC GROWTH IN ASIA

dollar, strengthening, 281 10-year government yield, change in,
external deficit, widening of after the 55f
mid-1990s, 272, 274f tightened regulations and eased
fertility rates, 168f monetary policy, 307
firm-size wage differences explained by trade deficits, expanded, 388
labor characteristics, 154–155 trade frictions with the rest of the
global real trade growth, 393f world, 281
growth in trade, faster driven by trade growth before and after the
widening imbalances, 394 financial crisis, 391, 393f
growth rate in, 238–239, 238n, 240f unexpected fall in net exports, 280
housing bubble, 274 youth dependency ratio, 57f
import growth moderated, 403 zero bound constraining conventional
imports, correlation with the total monetary policy, 294
exports and imports of East Asian universities
surplus countries, 398 excelling in global science, 78
inflation, 251 graduates of top-ranked in Japan, 155
correlation to Indonesia’s real GDP Indonesia sending students to the best
growth, 253t in the world, 267
and the output gap in, 298–300, limited research capabilities of Korean,
299t 88
shocks, affecting all variables other moving education and training to, 163
than US growth, 253n reformed in Japan, 85
surprises, 300, 301f underperforming in graduate
labor productivity, 140f, 150f education, 78
lessons of real estate bubbles and upstream competitiveness, allowed export-
banking crises, 289 led growth to sustain, 408
literacy proficiencies, 142f, 143f urbanization, slowing in China, 184
long-run equilibrium real rate of
around 1 percent for, 293n validation, becoming increasingly more
long-term bond yields, 294f costly, 27
mortgage-related assets, surge in valuation effects, moderating tendency for
European purchases of, 275 net international investment positions
as the most important trade deficit to diverge, 282f
country, 398 VAT (value-added tax)
natural rate of interest, 273f collection
net investment, 279f closing gaps, 376
number of employees by firm size, all increasing under rebalancing
industries, 159t package, 366
old-age dependency ratio, 38f, 57f as efficient, 363
patent grants, 86, 94, 94f increases
per capita income level, 39f direct impact of, 332n
Phillips curves for GDP inflation and gradual, 332
employment gap, 301f vector auto-regression (VAR) analysis, 251,
prevalence of small firms, 158 251n, 253, 253n
real 10-year bond yield, 1990–2018, Venezuela, 14
272f venture capital market
real GDP growth, 240f, 251, 253t in Japan, 85
real neutral interest rates, 56f in Taiwan, 92
reestablished financial stability after vertical integration, deceleration in, 400
the global crisis, 335 Vietnam
revealed comparative advantages of, aging, rapid until 2030, 52
172f aging impact on TFP, 48f
aging speed change between 2020 and
2030, 58f

INDEX 455
 CPI inflation rates, 1985–2016, 288f wage and labor productivity differences,
current account balances, 1985–2016, between large and small firms, 152,
310f 153f
debt prospects similar to China’s, 353 wage gap
debt ratio change under Scenario 4b, between large and small firms in Japan,
354f 162
demographic characteristics, 35t between nonregular and regular
demographic trends, impact of, 42f employees in Japan, 144, 145f, 162
emigration from, impact of net, 43 regulations reducing between regular
fiscal packages, 381t, 383t and nonregular employees in Japan,
foreign exchange reserves and net 148
official assets, 1985–2016, 311f between the two firm groups, 153f, 154
in Group 3, 347b wage profile, relationship with marginal
infrastructure, index of quality of, 356f productivity of employees in Japan,
labor force participation, 48f, 49f 145
long-term public debt projection, 349t wages and prices, forcing down, 291
old-age dependency ratio, 36t, 38f, 57f water pollution costs, likely to rise, 183
older workers, share of, 44f water power, improvement in, 16–17
pensions, 375 Watt, James, 17
per capita income level, 39f weak demand, explaining the recent trade
public debt, actual and projected, 346f, slowdown, 398
352f wealth, of urban areas before the
public spending Enlightenment, 13
efficiency frontier, distance from, Wells, H. G., 10
378t women
on health care as percent of GDP, increasing labor force participation, 65
358f retaining in the workforce, 133
on pensions and health care, 346f “Womenomics,” in Japan, 133
real per capita GDP, growth rate of, won
1985–2016, 287f depreciated sharply against the yen, 89
real short-term interest rate, 292, 293f strength of the Korean, 90
real trade growth, 1998–2008, workers
2012–2015, 393f accumulating human capital, 144
research and development spending, flow across firms and industries
362f limited in Japan, 81
Scenario 3 and, 350 LFPRs and, 46n
scope for quantitative easing, not changing jobs because of deferred
December 2016, 303t compensation in Japan, 149
tax revenue sources, 364f retraining, 65
10-year government yield, change in, working hours in Japan, 133
55f workforce aging
trade growth in, 399t associated with higher capital per
workforce, share of with productivity worker, 49
rising or falling, 45f estimated impact of projected on
youth dependency ratio, 57f growth, 46, 48f
virtual reality, as an educational tool, 27 estimating the effect on productivity,
Vivo, 91n 44–45
volatility share with productivity rising or
defined, 241f falling, 45f
of Indonesia’s net nonofficial inflows workforce composition, changes in, 3
compares favorably with regional working-age population
peers, 250, 251f absolute changes for different
of the supply of foreign capital, 251f demographic country groups in
voluntary saving, encouraging, 66 Asia, 38, 40f

456 SUSTAINING ECONOMIC GROWTH IN ASIA

 change in in Asia and the rest of the world trade growth, under more balanced
world, 40f global trade, 396–401
changes in LFPRs for, 46, 48f, 49 World Trade Organization (WTO),
declining in 2017, 166 founding of, 269
defined, 40f, 44f, 50n
growth rate of in Indonesia, 261f Xiaomi, 91, 91n
projected to decline, 34 x-ray crystallography, 24
shrinking of in Korea, 166
working life, longer, 367
World Bank, indices of, 14, 15n Yanukovych, overthrow of, 23
World Development Indicators (World Bank), yen, appreciated sharply against the US
225 dollar, 89
World Economic Outlook, evidence of the Young, Thomas, 26n
Gourinchas-Rey effect for advanced youth dependency. See also population
economies, 280 growth
world economy, rise in financial impact of, 51, 56, 56t, 59
integration, 248 ratio defined, 57f
“world” interest rate, 292
world long-term rate, depressed, 297 zero bound constraint, on nominal interest
world oil price and oil exporters’ current rate, 115–116
account, 275f zero inflation, twenty years of in Japan, 291
World Population Prospects: 2015 Revision zero or negative population growth, 19
(United Nations), 34n zombie firms
world real interest rates, decline in, 54, 55f in India, 232
world real long-term interest rate, hovering in Korea, 170
around zero, 292–293

INDEX 457