WORKING PAPER

19-15 Measuring the Rise of Economic

Nationalism
Monica de Bolle and Jeromin Zettelmeyer
August 2019

Abstract
Since the mid-2000s, the platforms of major political parties in both advanced and emerging-market economies have
increasingly emphasized policies that stress national sovereignty, reject multilateralism, and seek to advance national
interests through measures that come at the expense of foreign interests. This paper documents this shift by evalu-
ating the policy platforms of the largest political parties (about 55 in total) in the Group of Twenty (G-20) countries
with regard to trade policy, foreign direct investment (FDI), immigration, and multilateral organizations. Preference
shifts with respect to industrial policy, competition policy, and macroeconomic populism are also examined. In
advanced economies, the biggest shifts were toward restrictions on immigration and trade and toward macroeco-
nomic populism. In emerging-market economies, the largest preference shifts were toward industrial policies favoring
specific sectors, macroeconomic populism, and industrial concentration. Trade protectionism and skepticism toward
multilateral organizations and agreements have increased in both advanced and emerging-market economies. As of
2018, economic policy preferences in emerging-market economies were more nationalist and less liberal than in
advanced countries, but the gap has narrowed. Right-wing parties tend to be more nationalist than left-wing parties
in the areas of immigration restrictions, FDI restrictions, and antimultilateralism, but there is no significant differ-
ence with respect to trade protectionism.

JEL Codes: F5, F1, F2

Keywords: nationalism, populism, capitalism, trade policy, industrial policy, protectionism

Monica de Bolle is a senior fellow at the Peterson Institute for International Economics and the Riordan Roett Chair
in Latin American Studies at the Johns Hopkins School for Advanced International Studies. Jeromin Zettelmeyer is
the Dennis Weatherstone Senior Fellow at the Peterson Institute for International Economics.

Authors’ note: We thank Alvaro Leandro and Madi Sarsenbayev for outstanding research assistance; Madona
Devasahayam, Cameron Fletcher, and Egor Gornostay for careful proofreading; Olivier Blanchard, Cullen Hendrix,
Gary Hufbauer, Marc Noland, Adam Posen, Dennis Quinn, Ted Truman, Ángel Ubide, Nicolas Véron, and seminar
participants at the Peterson Institute for International Economics and the International Monetary Fund for comments;
and the Smith Richardson Foundation for financial support.

1750 Massachusetts Avenue, NW | Washington, DC 20036-1903 USA | +1.202.328.9000 | www.piie.com

I. INTRODUCTION
Economic policies that seek to advance national interests at the expense of international integration and coop-
eration appear to have enjoyed a resurgence in recent years. Examples include the imposition of tariffs (United
States), withdrawal from international fora or treaties (United States, United Kingdom), promotion of national
industrial champions (China, France, Germany), more restrictive immigration policies (United States, many
other advanced countries), and restrictions on inward foreign direct investment (United States, European
Union). In keeping with a literature going back to the early 20th century, we refer to policies of this type as
“economic nationalism.” Economic nationalism is correlated with political nationalism, populism, and what
has been called “national populism” (Judis 2018, Eichengreen 2018, Eatwell and Goodwin 2018). President
Donald Trump, for example, represents all of these. But the match is far from perfect: Some political national-
ists, such as the government of Jair Bolsonaro in Brazil, have so far embraced traditional free-market liberalism.
Conversely, some economic nationalist ideas—for example, encouraging mergers to create large national cham-
pions—are being promoted by EU politicians that cannot be classified as either politically nationalist or populist
(Zettelmeyer 2019b).
This paper takes a stab at measuring the resurgence in economic nationalism—or more precisely, the policy
preferences underlying these nationalist policies, as articulated in the programs of political parties. How wide-
spread is the shift in preferences? Is it taking place only in advanced countries or also in emerging-market
countries? Which, if any, countries buck the trend? What policy areas are affected, and to what extent?
Unlike recent empirical literature on the sources of support for populist and extremist political parties (for
example, Funke, Schularick, and Trebesch 2015; Guiso et al. 2018; and Mounk 2018), this paper does not
attempt to investigate the causes of economic nationalism (except for confirming that economic nationalism
and populist politics are correlated). And while we hope to lay the basis for research on the effects of nationalist
economic policies, this is outside the scope of the present paper. As such, we take no position on whether prefer-
ence shifts in the direction of economic nationalism are good or bad. While most economists would come down
in favor of international integration and cooperation—that is, against economic nationalism—there is also a
long and influential intellectual tradition arguing that policies such as tariffs can be good and even necessary
at some stages of economic development (e.g., List 1841). Indeed, there is evidence suggesting that countries
acquire capabilities in a wide range of industrial activity before specializing (Imbs and Wacziarg 2003). To the
extent that some forms of economic nationalism help countries acquire such capabilities, they might conceivably
be globally welfare improving in the long run. The same type of argument could be (and has been) made to
justify “defensive” economic nationalism in advanced countries that see themselves at risk of losing industrial
capabilities (Altmaier 2019). This is an important debate, to which this paper does not contribute; it focuses
only on measurement.
To measure changes in policy preferences, we examine the manifestos (when available) of all major political
parties in G-20 countries at two points in time: the mid-2000s and recent general elections (between 2015
and 2018). We focus on policy preferences rather than policies themselves because the former is a structural

2
determinant of the latter and arguably a leading indicator. There may be a groundswell of support for economic
nationalism in a country, but whether and on what timetable this translates into actual policies is to some extent
random—depending, for example, on the political costs of withdrawing from an international arrangement, or
on whether the nationalist party only narrowly won a majority. The second reason for focusing on preferences is
that they can be inferred from readily available sources, namely party manifestos or platforms.1 The representa-
tiveness of the views expressed in a platform is measured by the success of the party in the subsequent election.
Our yardsticks for measuring policy preferences are eight scales, each focusing on one area of economic
policy. In each case, policy preferences are assessed against a benchmark defined by the liberal economic consensus
that had become almost universal after the collapse of the Soviet empire in the early 1990s. This consensus
emphasized free markets in a strong competition framework, “horizontal” rather than “vertical” industrial poli-
cies, stability-oriented macroeconomic policy, free but rules-based trade, liberal immigration policies, openness
to inward foreign direct investment (FDI), and international economic integration overseen by multilateral rules
and institutions (see, for example, EBRD 1999, 2009; Besley, Dewatripont and Guriev 2010).
Four of our scales measure departures from the liberal benchmark for (increasingly) nationalist reasons:
protectionist trade policies, restrictions on FDI, immigration restrictions, and the rejection of constraints
imposed by multilateral arrangements (antimultilateralism). These quantify economic nationalism stricto sensu.
The other four scales measure departures from the liberal consensus for reasons that might be motivated by
economic nationalism but could also have other motives, such as left-wing politics (for example, distributional
objectives) or capture by private interests. They involve competition policy, macroeconomic policy (macroeco-
nomic populism), and (vertical) industrial policy, as well as a scale that measures antitrade sentiment for broader
reasons than protectionism (trade restrictions).
The main result is that preferences have shifted in a nationalist and/or illiberal direction in most of the large
advanced and emerging-market economies. In advanced economies, the biggest shifts were toward trade and
immigration restrictions, as well as toward more populist macroeconomic policies. These occurred both within
established parties and as a result of the entry of new parties, which on average tended to be more nationalist.
In emerging-market economies, the largest preference shifts were toward industrial policies favoring specific
sectors, macroeconomic populism, and greater industrial concentration. Both advanced and emerging-market
economies experienced a shift toward more protectionism and greater skepticism toward multilateral organiza-
tions. As of 2018, policy preferences in emerging-market economies remain more nationalist and less liberal than
in advanced countries in several dimensions: trade protectionism, FDI protectionism, industrial policies targeted
at specific sectors, and macroeconomic populism. However, the gap has generally narrowed.
With respect to political ideology, right-wing parties tend to be more nationalist/illiberal than left-wing
parties in the areas of immigration restrictions, FDI restrictions, and antimultilateralism, while left-wing parties

1. Data on policy actions are much harder to obtain. Notable exceptions include the Global Trade Alert database,
which monitors a broad set of policies affecting trade (www.globaltradealert.org/), including immigration restric-
tions, but is available only since 2008. FDI restriction indices compiled by the IMF (www.elibrary-areaer.imf.org/
Pages/ChapterQuery.aspx) and the OECD (www.oecd.org/investment/fdiindex.htm) are available for a longer
period.

3
are more attached to vertical industrial policy and macroeconomic populism. But again, most of these gaps
have narrowed. Particularly noticeable in the data is a rise in protectionism among left-wing parties; as a result,
left-wing and right-wing party attitudes with respect to trade protectionism are no longer significantly different
on average.
We begin by defining economic nationalism and putting that definition in the context of the literature. The
subsequent sections explain the eight policy scales used in this paper, followed by our main results.

II. DEFINITION
The term “economic nationalism” goes back at least to the beginning of the 20th century. Figure 1 shows the
frequency with which the term appears in books digitized by Google published in English, French, Spanish, and
Italian since 1900; a related term, “national self-reliance” was used for Chinese books. In the Western languages,
the term begins to appear around World War I. This is followed by a large peak in usage in the 1930s and 1940s,
a long plateau in the postwar period, and a gradual decline since 1990. Recently, however, the online use of the
term in English has soared (figure 2), perhaps related to the election of President Trump and the use of the term
by his campaign manager and initial political advisor, Steve Bannon (see, for example, Drezner 2017).
There is a large literature on economic nationalism to which both economists and political scientists have
contributed. The economics literature goes back more than a century: The first article that we could identify is
by American economist Alvin Johnson (1917), one of the cofounders of the New School for Social Research.2
This literature tends to be written in reaction to economic nationalist policies prevailing at the time and is often
critical of these policies.3 The political science literature is more recent and discusses economic nationalism as
a facet of nationalism (see Crane 1998, Abdelal 2001, Helleiner 2002, Pickel 2003, and Nakano 2004), in the
context of international relations (Gilpin 1987), or both (Mayall 1990).4
These literatures are preceded by authors, going back to the late 18th century, who gave economic nation-
alism its intellectual foundation without calling it by that name. Seminal contributions include those of Alexander
Hamilton (1791) and Friedrich List (1841). Both argued that there could be no national prosperity without
industrialization and that industrial development justified—indeed, required—protection of domestic industry
against foreign competition. Economic nationalism of a more radical kind is associated with German philosopher
Johann Fichte (1800), whose Closed Commercial State provided a “blueprint for a complete collectivist society in
which production, distribution and exchange would be planned as national activities,” and international trade
would be eliminated altogether as the nation expanded to its “natural frontiers” (Mayall 1990, 80).
Economists and political scientists have not always agreed on the definition of economic nationalism.
Economists have generally identified economic nationalism with specific policies—either those focused on

2. Other contributors in this tradition include Gregory (1931), Rappard (1937), Heilperin (1960), Johnson (1967), and
Hieronymi (1980).
3. For example, Johnson (1917, 225) writes that “Economic nationalism may indeed be credited with important
gains in industrial efficiency. These are no adequate compensation for the evils produced by its aggressions. It is
national selfishness erected into a religion that menaces the peace and liberty of the world.”
4. Other contributions include those of Levi-Faur (1997), Shulman (2000), Abdelal (2001), Helleiner (2005),
Helleiner and Pickel (2005), Pickel (2005), and Szlajfer (2012). For a survey, see Pryke (2012).

4
 Figure 1
Mentions of the term “economic nationalism” in books in five languages, 1900–2008

percent percent
0.0000012 0.00009
English
French 0.00008
0.0000010 Spanish
Italian 0.00007
Chinese (right axis)
0.0000008 0.00006
Start of WWI
(1914–1918) 0.00005
0.0000006
0.00004

0.0000004 0.00003

0.00002
0.0000002
0.00001

0 0
1900 1909 1918 1927 1936 1945 1954 1963 1972 1981 1990 1999 2008

WWI = World War I

Note: For books in French, this chart refers to the term “nationalisme économique.” For books in Spanish,
“nacionalismo económico.” For books in Italian, “nazionalismo economico.” For books in Chinese, “自力更生”,
which means “self reliance” in English. The units represent the proportion of each phrase (“economic nationalism”
in English) in all two-word phrases contained in Google’s sample of books written in the same language.
Source: Google Books Ngram Viewer, books.google.com/ngrams.

limiting international integration and foreign influence (Rappard 1937, Heilperin 1960) or those designed
to benefit domestic firms in their competition with foreign firms, not only through trade policies but also via
subsidies, foreign investment restrictions designed to either accelerate industrial catchup or prevent competing
nations from catching up, and measures allowing or fostering cartels (Gregory 1931, Johnson 1965, Kahan
1967, Hieronymi 1980).
In the late 1990s and early 2000s, this definition came under fire from political scientists, who argued that
it lacked rigor as well as historical understanding, allowing liberally minded economists to brand essentially any
policy they did not like as nationalist (Crane 1998; Abdelal 2001, 2005; Helleiner 2002, 2005; Pickel 2003,
2005). In their view, economic nationalism should instead be defined in terms of motives—or as Helleiner
(2005, 220) puts it, “nationalist content”: Antiliberal policies are not economic nationalism unless motivated
by nationalist thought. This definition does not necessarily result in a specific set of policy prescriptions of the
sort that economists have typically identified with economic nationalism, because in some settings even liberal
economic policies may serve a nationalist purpose.
With this paper we hope to lay the ground for research that will establish whether policies of the kind that
have been advocated by new nationalists such as President Trump—including protectionism, restrictions on
inward foreign investment and immigration, and rejection of multilateral constraints—can make economic
sense for the countries that pursue them, at least in some historical settings. As such, this paper cannot go all
the way to adopting a definition of economic nationalism that is entirely context-specific and may thus include
dramatically different policy choices—sometimes liberal, sometimes illiberal. At the same time, we agree with

5
 Figure 2
Online use of the term “economic nationalism” in English

a. Google searches, 2004–18

interest over time
100
United States
90 Worldwide
80
70
60
50
40
30
20 President Trump
assumes office
10 (January 20, 2017)

0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Note: Search interest relative to the maximum. A value of 100 denotes peak popularity. A value
of 50 means that the term was half as popular.
Source: Google Trends.

b. Mentions in online news, 2010–18

mentions
500
450
400
350
300
250
200
150
100
50
0
2010 2011 2012 2013 2014 2015 2016 2017 2018

Note: Number of nonduplicate articles in which the phrase “economic nationalism” is mentioned.
Source: NOW Corpus (News on the Web).

the political science literature that motives matter, in part because they help distinguish nationalist from statist
or socialist policies. For example, for Harry Johnson (1965) “a preference for public enterprise over private
enterprise” is part of the economic nationalism observed in the new postcolonial states. But this preference could
be motivated by socialism, and as such is not necessarily nationalist. Even rejection of free trade could have
nonnationalist motives. In the 2017 German parliamentary election, the Greens were opposed to multilateral
free trade agreements because they believed that these agreements would benefit multinational corporations at
the expense of workers and consumers, not because they wanted to protect domestic workers at the expense of
foreign workers.

6
 The approach taken in this paper is hence to adopt a motives-based definition of economic nationalism, but
one that is sufficiently narrow to exclude liberal policies. Specifically, we define economic nationalism as policies
designed to further domestic economic interests (the interests of domestic producers, consumers, and/or workers) at
the expense of foreign economic interests, at least in the short run. The qualification “at least in the short run” allows
us to include policies, such as the infant industry tariffs advocated by Hamilton and List, that do not attempt to
give domestic interests a permanent advantage but are intended to level the playing field during a phase of catch-
up. In the same vein, some US or EU advocates of FDI restrictions against Chinese companies would argue that
these restrictions are needed only so long as China does not “play by market rules.”
This definition leads to the following policy classifications.
First, we define four policy categories that clearly meet our definition of economic nationalism: trade protec-
tionism, FDI protectionism, immigration restrictions, and rejection of multilateral constraints. The first three
capture policies that seek to restrict—or influence the terms of—international integration to protect domestic
interests at the expense of foreign interests. The last is concerned with protecting national autonomy from supra-
national or foreign interference, even at the expense of cooperation that may be collectively beneficial. Hence,
these four policies reflect economic nationalism stricto sensu.
Second, we define four further policy categories in which a specific policy choice may or may not reflect
economic nationalism, depending on how it is motivated: industrial policies, competition policies, macroeco-
nomic populism, and restrictions on trade for any reason (not just protectionism). For example, industrial
policies could be motivated by the desire to strengthen “strategic” industries against foreign competition—a
nationalist motive—but they could also reflect a desire to address perceived market failures. The latter might
make sense even in a world with just one country and no foreign competition, and as a result is not nationalist.
Similarly, allowing or promoting anticompetitive mergers and cartels could reflect a desire to build national
champions but could also reflect capture by special interests, or a version of Spencerian social Darwinism—the
view that size reflects superiority and strength. Only the first of these motives is nationalist according to our
definition. With respect to macroeconomic populism, we follow the definition of Rudiger Dornbusch and
Sebastian Edwards (1991), namely, macroeconomic risk taking in pursuit of a specific social or political objective
(with risks including high inflation, currency collapse, or financial crisis).5 This objective may be nationalist (for
example, when the central bank or the treasury directs credit to strategic sectors), but it may also be to reduce
inequality or benefit a particular group of producers. Finally, trade restrictions for any reason may combine
protectionism with other motives to restrict trade, such as maintaining environmental or labor standards.
In the remainder of the paper we attempt to establish how preferences for these policies differ across G-20
countries and between two points in time: the mid-2000s and the time of a country’s most recent general
election.

5. “For us, ‘economic populism’ is an approach to economics that emphasizes growth and income redistribution
and deemphasizes the risks of inflation and deficit finance, external constraints, and the reaction of economic
agents to aggressive nonmarket policies” (Dornbusch and Edwards 1991, 9).

7
III. METHODOLOGY AND DATA
Several databases attempt to quantify the policy preferences of political parties. The Chapel Hill expert survey
(www.chesdata.eu) measures the general ideological orientation of European parties and their position on
European integration as well as several policy issues for about 30 European countries, based on expert panels for
each country. The Comparative Manifesto Project Database (https://manifesto-project.wzb.eu) codes party elec-
toral programs around the world (over 50 countries). Electoral programs are first “unitized” by dividing them
into “statements”. In a second step, each statement is assigned a single code based on whether it refers positively
or negatively to a specific policy goal or issue. The numbers of positive and negative mentions are subsequently
counted and expressed as a percentage of the number of statements in the dataset.6
Unfortunately, neither of these databases is sufficiently rich to allow users to extract changes in economic
nationalism in a broad cross-section of countries. The Chapel Hill survey is too limited in its coverage of both
countries and policy issues, which include only one of the eight policy categories that we are interested in
(immigration). The Manifesto Project is much broader in terms of country coverage, but limited in terms of
policy categories, covering only three that relate to the definition of economic nationalism used in this paper:
protectionism, internationalism (a negative proxy for the rejection of multilateral constraints), and economic
orthodoxy (a negative proxy for macroeconomic populism). Furthermore, because the Manifesto Project’s
coding criteria are defined quite narrowly—for example, obtaining a score on “negative protectionism” requires
a statement opposing tariffs rather than just stating that free trade or trade agreements are desirable—many
parties covered are assigned zero values in the categories that we are interested in. In addition, the Manifesto
Project’s approach cannot help us decide whether opposition to trade liberalization reflects nationalist or green/
progressive motives—a distinction that is important for the purposes of this paper.
As a result, the empirical work presented in this paper is based on our own coding of policy preferences for
economic nationalism. Like the Manifesto Project, the raw material consists of party platforms and/or election
programs. Unlike that project, however, we go beyond a binary coding (positive versus negative statements)
approach. Instead, we define cardinal scales, from 1 to 5, for each of our eight policy categories. Party manifestos
are coded using these scales, which are described in detail in appendix A. This approach provides for a much
richer dataset in terms of both the policy categories covered and the granularity of classifications. The disadvan-
tage is that it injects a degree of subjectivity in the policy ratings. We address this in three ways.
n Each policy manifesto was separately coded by two researchers—one of the coauthors and one of the research
assistants—followed by a reconciliation to deliver a final set of scores. Initial disagreements were recorded
and quantified. The Cohen kappa coefficient for interrater agreement, measured on a scale of −1 to +1,
varied from 0.52 to 0.66, indicating moderate to strong agreement.7 The pairwise correlation coefficients

6. For example, for the 2006 electoral platform of the Conservative Party of Canada, variable 406, “protectionism:
positive,” takes the value 0.523, while variable 407, “protectionism: negative,” takes the value 4.007. This means
that 0.523 percent of the statements in the party platform contained favorable references to protectionist mea-
sures (tariffs, quotas, export subsidies) while 4 percent of the statements contained negative references.
7. A kappa of −1 corresponds to perfect disagreement, 0 is the agreement one would expect if the ratings had
been produced randomly, and +1 represents perfect agreement. See McHugh (2012).

8
 between the policy ratings of the two raters, in categories for which both raters found sufficient information
to rate, was between 0.77 and 0.93 (see appendix B for details).
n All sources, as well as working files documenting and backing each individual rating decision, including
extracts from party platforms on which decisions were based, are available in full on the PIIE website,
making it easy for other researchers to reproduce (or modify) our ratings.
n As a robustness check, we also examined preference shifts according to the Manifesto Project data in three
categories: “protectionism,” comparable to our trade protectionism category; “internationalism,” which
should be negatively correlated with our “antimultilateralism” category; and “economic orthodoxy,” which
should be negatively correlated with macroeconomic populism.

The remainder of this section summarizes how our policy scales were constructed, which party manifestos
were selected for coding purposes, and in which policy dimensions they contained sufficient information to
allow coding.

Policy Scales
The construction of the policy scales was based on the following principles:
n Rating 1 describes what we view as a modern liberal consensus. By “consensus” we do not mean to imply
that this view is necessarily consensual in societies, only that it is consensual among today’s economic
liberals. This consensus does not necessarily describe free-market or laissez-faire views, but rather the views
of mainstream economists in academia or international organizations such as the International Monetary
Fund (IMF) or the Organization for Economic Cooperation and Development (OECD). Rating 1 can
assign a significant role to the state, so long as this is justified by market failures or competition enforce-
ment. Rating 1 is always nonnationalist in the sense defined above. While governments should of course
worry about national prosperity, consensus liberalism believes that national prosperity generally benefits
from international cooperation and does not require policies that come at the expense of other nations.
n Rating 5 is the farthest departure from this modern liberal consensus (in an illiberal direction) that can
arguably be observed in history.
n The progression from 1 to 5 is based on some combination of ideology and policy proposals expressed by
the party manifesto, explained below.

In what follows, we briefly state how these extremes, and the progression between them, are defined for each
of the eight policy scales (for details, see appendix A).

9
Economic Nationalism
Trade Protectionism
Rating 1. Absence of protectionism. This could be either because the party embraces free trade subject to the rules
of the multilateral trading system or because trade restrictions are justified with arguments that have nothing to
do with protectionism (for example, that unfettered trade hurts the environment, that it benefits multinationals
to the detriment of workers, or that it does not reflect the interests of developing-country trading partners).

Rating 5. Economy is closed to imports, except with respect to products that are deemed essential and cannot
be produced at home.

Progression: defined both by the extent to which the party advocates using policy instruments that are intended
to further domestic interests at the expense of foreign interests (e.g. tariffs, quotas, discriminatory procurement
policies) and the extent to which international trade, and multilateral and bilateral agreements that facilitate it,
are viewed with skepticism for nationalist reasons.

FDI Protectionism
Rating 1. No restrictions or regulatory impediments to FDI except for narrowly defined national security reasons
as determined by a politically independent body.

Rating 5. FDI is not welcome except in specific areas of national interest (“positive list”).

Progression: defined by the extent and way in which FDI restrictions are employed. The motivation for these
restrictions could include protecting a technological lead, creating know-how or jobs at home (via local content
requirements), or simply maintaining national control of companies.

Immigration Restrictions
Rating 1. Country is open to immigration except for restrictions that are meant to protect health and public
safety, avoid congestion of public services, and facilitate the integration of immigrants.

Rating 5. Country is closed to immigration except for immigrants that share the ethno-national background of
the domestic population or have ties with the ruling elites.

Progression: defined by increasingly nationalist motivations for restricting immigration and the resulting extent
of restrictions.

Rejection of Multilateral Constraints

Rating 1. Country is or wishes to be a member of virtually all international organizations whose mission it agrees
with. In working within this organization, it puts the purpose of the organization ahead of narrowly defined
national interest.

Rating 5. Country is a member of an international organization only if it effectively runs it or for propaganda
purposes and only if membership poses no constraints on national policies.

10
Progression: focuses on how a government resolves the potential tension between national interest in specific
instances and its support for the mission of the international organization.

Deviations from the Liberal Consensus for Any Reason, Nationalist or Not
Trade Restrictions
Rating 1. Free trade subject to the rules of the multilateral system. Trade is viewed as good, not just exports.
Tariff barriers to trade are minimal (historical relics). Nontariff barriers to trade are a consequence of regulation
that is justified by public policy aims extraneous to trade. Some state intervention in trade (e.g., export promo-
tion) is consistent with this rating so long as it is justified by overcoming information failures or reducing fixed
costs of trade.

Rating 5. Same as the trade protectionism scale (i.e., economy essentially closed to imports).

Progression: same as in the case of trade protectionism, except that deviations from the liberal baseline policy
prescription could be motivated both by protectionism and by trade skepticism for reasons unrelated to
protectionism.

Competition Policy (Tolerance of Market Power)

Rating 1. Competition law makes cartels illegal, aims to prevent abuse of economic power, and is effectively
enforced by an independent competition authority.

Rating 5. No effective competition regime. Cartels or monopolies exist in many sectors of the economy, typi-
cally run or encouraged by the state (whether in the form of state enterprises or via state supervision of privately
owned enterprises).

Progression: defined by effectiveness of competition enforcement and the degree to which the competition
regime allows the suspension or restriction of competition to pursue other aims (for example, to create national
champions).

Vertical Industrial Policy

Rating 1. Industrial policy is “horizontal” in the sense that the state provides public goods or services that are
meant to encourage economic activity in general rather than in specific sectors, and state financial support
for firms is provided competitively and based on the presumed externalities of the activities that benefit from
support (e.g., R&D subsidies).

Rating 5. The state seeks to develop specific sectors with means that include nationalization, price controls, and/
or direct state instructions.

Progression: defined by the extent to which the state seeks to promote specific sectors and the means it uses to do
so (including the extent to which it seeks to exercise direct control).

11
Macroeconomic Populism
Rating 1. Monetary and fiscal policies are focused solely on economic stability: maintaining low and stable infla-
tion, smoothing recessions and booms, and preventing financial or fiscal crises. Monetary policy is executed by
an independent central bank.

Rating 5. Monetary and fiscal policies are subordinated to objectives other than sustainability or stabilization.
The government will not compromise with respect to its preferred objectives, no matter what happens to macro
stability.

Progression: defined by the willingness of the government to take macroeconomic risks in pursuit of other objec-
tives such as (short-term) growth or redistribution.

PARTY SELECTION AND COVERAGE

This paper focuses on 18 of the G-20 countries (Saudi Arabia is excluded). Advanced and emerging-market
economies are evenly split in this group, which contains the G7 countries, Australia, Korea, and nine emerging-
market countries (Argentina, Brazil, China, India, Indonesia, Mexico, Russia, South Africa, and Turkey). These
18 countries account for about three quarters of the world economy in terms of GDP.8
To get a sense of whether political preferences for economic nationalism have changed, we compared party
platforms before a general (parliamentary or presidential) election in the mid-2000s and before the most recent
general election (as of December 2018). This time interval includes the global financial crisis (GFC), which
many analysts believe contributed to the rise of nationalism and populism in advanced countries (Eichengreen
2018, Eatwell and Goodwin 2018).
To keep the analysis manageable, we focused on political parties that secured at least 10 percent of votes
cast in at least one of these elections.9 In the case of China, the 11th Five-Year Plan (2006–10) and 13th Five-
Year Plan (2016–20) were used for pre- and post-GFC analysis, respectively. In the case of Italy, elections were
contested by broad center-right and center-left coalitions (with the Five Star Movement running separately
in 2018); hence, the electoral platforms of the coalitions were analyzed rather than those of individual parties
(except for the Five Star Movement). In a few cases, listed in the note to table 1, written party platforms could
not be obtained or were insufficiently informative to be rated. The parties or coalitions included in the analysis
nonetheless captured over half of the votes cast, and normally much more, with two main exceptions: Indonesia,
reflecting missing information, and Brazil, reflecting the extreme fragmentation of the Brazilian party system.10
Table 1 shows the parties (or electoral coalitions) analyzed. For the advanced countries, the sample contains
26 pre-GFC and 32 post-GFC parties/coalitions; for the emerging-market countries, 20 and 24 pre- and
post-GFC parties, respectively. The post-GFC party count reflects six new entrants in advanced countries: La

8. In purchasing power terms, about 72 percent; at market exchange rates, about 77 percent.
9. We also analyzed the platforms of two parties that missed the 10 percent threshold in these elections, the
German Greens and the UK Independence Party (UKIP). The Greens received nearly 10 percent in both elections,
while UKIP was of interest because of its outsized influence on the Brexit vote. In the case of elections for bicam-
eral legislative bodies, the selection was based on the election results for the lower chamber.
10. In Brazil’s 2018 election, the largest five parties together captured less than 40 percent of the vote.

12
 Table 1
Party platforms analyzed
Pre-GFC Post-GFC
Country Election Parties Election Parties
Frente para la Victoria
Chamber of Deputies, Frente para la Victoria Chamber of Deputies,
Argentina Cambiemos
2007 Coalición Cívica 2015
United for a New Alternative (UNA)*
Australian Greens Australian Greens
House of Representatives, House of Representatives,
Australia Australian Labor Party Australian Labor Party
2007 2016
Liberal Party of Australia Liberal Party of Australia
Workers’ Party (PT)
Chamber of Deputies, Workers’ Party (PT) Chamber of Deputies,
Brazil Social Democracy Party (PSDB)
2006 Social Democracy Party (PSDB) 2018
Social Liberal Party (PSL)
Conservative Party Conservative Party
House of Commons, House of Commons,
Canada Liberal Party Liberal Party
2006 2015
New Democratic Party New Democratic Party
China 11th Five-Year Plan Communist Party of China 13th Five-Year Plan Communist Party of China
Front National
Front National Socialist Party (PS)

12
13
13th National Assembly, 15th National Assembly,
France Socialist Party (PS) Les Républicains
2007 2017
Union for a Popular Movement (UMP) La France Insoumise
La République En Marche!
Alliance 90/The Greens
Alliance 90/The Greens CDU/CSU
Christian Democratic Union (CDU)/
16th Bundestag, 19th Bundestag, Free Democratic Party (FDP)
Germany Christian Social Union (CSU)
2005 2017 Social Democratic Party (SPD)
Free Democratic Party (FDP)
Alliance for Germany (AfD)
Social Democratic Party (SPD)
14th Lok Sabha, Bharatiya Janata Party 16th Lok Sabha, Bharatiya Janata Party
India
2004 Indian National Congress 2014 Indian National Congress
Democratic Party (PD)
People’s Representative Democratic Party (PD) People’s Representative
Indonesia Party of the Functional Groups (Golkar)
Council, 2004 Party of the Functional Groups (Golkar) Council, 2014
Great Indonesia Movement Party (Gerindra)
Center Left Coalition
Chamber of Deputies, Center Left Coalition Chamber of Deputies,
Italy Center Right Coalition
2006 Center Right Coalition 2018
Movimento Cinque Stelle
Constitutional Democratic Party
House of Representatives, Democratic Party House of Representatives,
Japan Liberal Democratic Party
2005 Liberal Democratic Party 2017
Kibō no Tō (Party of Hope)
Table continues
 Table 1 (continued)
Party platforms analyzed
Pre-GFC Post-GFC
Country Election Parties Election Parties
Institutional Revolutionary Party (PRI)
Alliance for Mexico (PRI-led)
Chamber of Deputies, Chamber of Deputies, Por México al Frente (PAN-led)
Mexico National Action Party (PAN)
2006 2018 Movimiento Regeneración
Democratic Revolutionary Party (PRD)
Nacional (MORENA)
Communist Party Communist Party
5th State Duma, 7th State Duma,
Russia Liberal Democratic Party of Russia Liberal Democratic Party of Russia
2007 2016
United Russia United Russia
South National Assembly, African National Congress National Assembly, African National Congress
Africa 2004 Democratic Alliance 2014 Democratic Alliance
Saenuri Party
South 18th National Assembly, Grand National Party (Hannara) 20th National Assembly,
Democratic Party
Korea 2008 United Democratic Party 2016
People’s Party
AK Parti
AK Parti
Grand National Assembly, Grand National Assembly, Nationalist Movement
Turkey Nationalist Movement
2007 2018 Peoples’ Democratic Party
Republican People’s Party

13
14
Republican People’s Party
Conservative Party Conservative Party
United House of Commons, Labour Party House of Commons, Labour Party
Kingdom 2005 Liberal Democrats 2017 Liberal Democrats
UK Independence Party UK Independence Party
United Presidential election, Democratic Party Presidential election, Democratic Party
States 2004 Republican Party 2016 Republican Party
GFC = global financial crisis
*Parties listed in semi-bold are new entrants.
Note: The table includes all parties that received at least 10 percent of the votes cast in one of the two elections, plus the German Greens and the UK Independence
Party. Brazil’s Democratic Movement (MDB) could not be rated for lack of a party platform. Canada’s Bloc Québécois was excluded because its policies express
preferences at the level of a region, making aggregation with other Canadian parties impossible. In the case of Australia’s Liberal Party, pre- and post-GFC ratings
were based on party or coalition materials covering several years (2002, 2004, and 2007; and 2013, 2016, and 2018, respectively) as party or coalition statements
preceding the 2007 and 2016 elections were insufficiently detailed. For Indonesia, we were able to obtain and analyze the platform of only the Golkar party in the
pre-GFC period, could not obtain the platform of the National Awakening Party (PKB) in the post-GFC period, and were not able to evaluate the Struggle party in
both periods because of lack of detailed information. For Japan, the Komeito party platform was not separately analyzed as Komeito was part of a coalition with
the Liberal Democratic Party in both periods. For South Korea, the platform of the Pro-Park Coalition (Future Hope Alliance) was not analyzed because it joined
the Grand National Party, later renamed Saenuri Party, in 2012. Parties or party platforms that are considered predecessor/successor organizations occupy the
same line in the table even if they have different names. Specifically, France’s Les Républicains is treated as the successor of UMP rather than as a new entrant.
Similarly, Korea’s Democratic Party is treated as the successor of the United Democratic Party. Mexico’s MORENA is treated as the successor of the pre-GFC PRD
on the grounds that it fielded the same presidential candidate. For Argentina, the 2015 electoral coalition Cambiemos is treated as successor of Coalición Cívica on
the grounds that the latter was one of its members and both fielded the main challenger to Frente de la Victoria in the 2007 and 2015 elections.
Source: Authors’ selection based on criteria described in the note.
 Table 2
Share of party platforms for which sufficient information was available to
allow coding (percent)
Pre-GFC Post-GFC
Emerging- Emerging-
Advanced market Advanced market
Policy category economies economies economies economies
Trade protectionism 85 85 88 75
FDI protectionism 19 45 27 71
Immigration restrictions 78 20 82 17
Antimultilateralism 93 90 97 79
Trade restrictions 85 85 88 75
Industrial policies toward specific sectors 89 95 94 96
Tolerance of concentration 67 75 67 58
Macroeconomic populism 89 80 94 96
GFC = global financial crisis; FDI = foreign direct investment
Source: Authors’ analysis based on platforms of parties identified in table 1.

République En Marche!, La France Insoumise, the Alliance for Germany (AfD), the Five Star Movement, the
Party of Hope (Japan), and the People’s Party (Korea, merged into Bareunmirae Party in 2018).11 In emerging-
market economies, new entrants were Argentina’s United for a New Alternative (UNA), Turkey’s People’s
Democratic Party, and Indonesia’s Gerindra party. Jair Bolsonaro’s Social Liberal Party (PSL) was also regarded
as an entrant: Although it existed in the 2000s, it received no congressional seats in the 2006 election but became
the largest party in the Brazilian congress in the 2018 election after becoming associated with the far right.12
Table 2 shows the extent to which the policy issues that we are interested in are discussed in the electoral
platforms of these parties or coalitions. Four points stand out.
First, trade policy, immigration restrictions, industrial policy, antimultilateralism, and macroeconomic
policy all gained salience in advanced country platforms over the period analyzed, in the sense that the coverage
of these topics rose.
Second, among parties in emerging-market economies, there was a large rise in the salience of FDI policies
and macroeconomic policy. In contrast, trade policy, antimultilateralism, and competition policy (tolerance of
concentration) became less salient.
Third, immigration was and remains a side issue in the party platforms of emerging-market economies
whereas it is a major issue in the platforms of advanced economies.
Fourth, the reverse is true for FDI policies. These are mentioned more frequently in the platforms of
emerging-market parties—particularly in the post-GFC period—than those of advanced countries. Although
the share of advanced country party platforms that refer to FDI has risen, it remains low—just 27 percent.13

11. Cases in which parties changed names—e.g., from Union for a Popular Movement (UMP) to Les Républicains
(France), from United Democratic Party to Democratic Party (Korea)—were not treated as exits or new entries.
12. Mexico’s MORENA and Argentina’s Cambiemos were not considered new entrants; see note to table 1.
13. This reflects the following parties (in the post-GFC period): Conservative Party of Canada, La République En
Marche!, the Alliance for Germany (AfD), the Five Star Movement, the Labour Party, the UK Conservative Party,
and the Trump administration (which was rated separately from the 2016 Republican platform; see below).

14
15
 We next turn from the question of what was discussed in party platforms to how it was discussed.

IV. RESULTS
The main results of the paper can be summarized under four headings, which describe the levels and changes
in policy preferences in the categories listed in table 2 from various perspectives. We consider preference shifts
among existing parties, differences between new and old parties, party-level correlations between economic
nationalism, populism and political orientation, and country-level changes in economic nationalism. As a
robustness check, we subsequently compare our findings with results based on the Manifesto Project Database.

Preference Shifts of Existing Parties or Coalitions

Figure 3 shows changes in policy preferences of parties that existed (and could be rated) in both the pre- and
post-GFC periods (see appendix C for the party-level data underlying the figure). The top two rows show
results for advanced-economy parties, the bottom two for emerging-market parties. For each group, the upper
row shows ratings in policy categories regarding economic nationalism stricto sensu—trade protectionism, FDI
protectionism, immigration restrictions, and the rejection of constraints imposed by multilateral organiza-
tions—while the lower row captures deviations from the liberal consensus in trade policy, industrial policy,
competition policy, and macroeconomic policy, for reasons that may reflect both nationalist and other motives.
For each policy category, the size of the circles represents the number of party platforms with a specific
rating. For example, in the trade protectionism category, nine advanced-economy parties were rated 1 in the
pre-GFC period, ten were rated 2, two were rated 3, one was rated 4 (France’s Front National), and none was
rated 5. The post-GFC distribution, in the same order, was six, twelve, three, one, and zero. That is, most parties
continued to be rated 2, but whereas in the pre-GFC period nine parties were rated 1 and only two parties
received a 3 rating, in the post-GFC period just six parties were rated 1 and three parties received a 3 rating.
The arrows between circles give a sense of the flow between ratings. The width of each arrow is proportional
to the number of parties that switched from one to another rating. Gray arrows denote flows between the
same ratings; dark red arrows denote flows from a lower (less protectionist) to a higher (more protectionist)
rating, green arrows the reverse. For example, in the trade protectionism category, the gray arrow from rating
1 in the pre-GFC period to rating 1 post-GFC represents six parties that remained firmly antiprotectionist
throughout the period: the Australian Liberals, the UK Conservatives and Liberal Democrats, and the German
Free Democrats, Social Democrats, and Greens (the latter opposed trade agreements such as the EU-Canada
Comprehensive Economic and Trade Agreement, but for nonprotectionist reasons). The two downward-sloping
red arrows represent parties classified as moving toward protectionism: the Australian Greens, UKIP and the UK
Labour Party (from 1 to 2), and the US Democratic and Republican Parties (from 2 to 3). The green upward-
sloping arrow represents the only advanced-country party that was classified as less protectionist in the post-GFC
period than pre-GFC: France’s UMP/Les Républicains (from 3 to 2).
The main result is that, depending on the policy category and the country group, the distribution of policy
preferences either moved away from the liberal consensus or was roughly unchanged. For advanced economies,

16
Figure 3
Distribution of party policy ratings before and after the global financial crisis
a. Advanced economies
Trade protectionism FDI protectionism Immigration restrictions Antimultilateralism
22 parties 22 parties 22 parties 25 parties

1 1 1 1 1 1 1

2 2 2 2 2 2 2 2

3 3 3 3 3 3 3

4 4 4 4 4

5 5

Trade restrictions Industrial policy Tolerance of concentration Macroeconomic populism

22 parties 24 parties 14 parties 23 parties

1 1 1 1 1 1 1 1

2 2 2 2 2 2

3 3 3 3 3 3

4 4 4 4 4 4

b. Emerging-market economies
Trade protectionism FDI protectionism Immigration restrictions Antimultilateralism
13 parties 7 parties 2 parties 15 parties

1 1 1 1 1 1

2 2 2 2 2 2 2

3 3 3 3 3 3 3

4 4 4 4

Trade restrictions Industrial policy Tolerance of concentration Macroeconomic populism

13 parties 19 parties 11 parties 16 parties

1 1 1 1 1 1 1

2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3

4 4 4 4 4 4

5 5 5 5 5

FDI = foreign direct investment

Note: Ratings in circles: 1 = liberal/nonnationalist; 5 = illiberal/nationalist. Size of the bubbles to the left and right represents
the share of parties with that rating before and after the global financial crisis (GFC), respectively. Size of arrows represents
the volume of flows from one rating before the GFC to another after the GFC.
Source: Authors’ calculations based on data shown in appendix C.

17
 Table 3
Average changes in party ratings for existing parties
(p values shown in parentheses)
Advanced Emerging-market
Policy category Number of parties economies economies
0.18 0.23
Trade protectionism 35
(0.100) (0.476)
−0.67 0.14
FDI protectionism 10
(0.305) (0.741)
0.29 0.00
Immigration restrictions 23
(0.033) (1.000)
0.08 0.27
Antimultilateralism 40
(0.425) (0.319)
0.27 0.23
Trade restrictions 35
(0.027) (0.476)
0.13 0.26
Industrial policy 43
(0.522) (0.359)
0.00 0.18
Tolerance of concentration 25
(0.164) (0.427)
0.04 0.56
Macroeconomic populism 39
(0.814) (0.081)
GFC = global financial crisis; FDI = foreign direct investment
Note: The table is based on eight regressions—one for each policy category—of the type r(i,t) =
b0 + b1*postGFC*AE + b2*postGFC*EME + d(i), where r(i,t) is the rating assigned to party i in the
category at time t (before or after the GFC), postGFC is a dummy variable taking on the value
one in the post-GFC period and zero otherwise, AE is a dummy variable for advanced economy
parties, EME is a dummy variable for emerging-market parties, and d(i) is a full set of dummy vari-
ables for specific parties. The number of observations for each regression is twice the number of
parties shown in the second column. The coefficient estimates b1 and b2 are equal to the average
change in policy ratings for advanced and emerging-market countries, respectively. The third
and fourth columns of the table show the coefficient estimates for b1 and b2, respectively, with p
values in parentheses. Robust standard errors were used.
Source: Authors’ calculations based on data presented in appendix C.

it moved toward more nationalist preferences and away from the liberal consensus in five of the eight policy
categories: trade protectionism, immigration, trade restrictions, industrial policy, and macroeconomic populism.
With respect to FDI protectionism, competition policy, and attitudes toward multilateralism there was little
or no change on balance.14 In emerging-market economies, there was a shift toward nationalism/illiberalism in
seven of the eight categories, the only exception being immigration restrictions.
Table 3 shows the average magnitude of these changes. In advanced countries, the largest average shift away
from the liberal consensus was in the areas of immigration restrictions (0.29) and trade restrictions (0.27). The
shift with respect to trade restrictions was larger than that with respect to trade protectionism (0.18) because the
former, but not the latter, captures the antitrade shift among some left-of-center parties, such as the Australian
and German Greens and the German Social Democrats. This is why the red, downward-sloping arrows in figure
3 are slightly wider, for the advanced countries, in the trade restrictions chart than in the trade protectionism
chart.

14. The policies of President Trump are not reflected in this chart, which in the case of the United States compares
Democratic and Republican platforms in the 2016 and 2004 elections. This is discussed separately below.

18
15
 For the emerging-market parties, the largest changes were in the areas of macroeconomic populism (0.56),
antimultilateralism (0.27), vertical industrial policy (0.26), and trade protectionism/trade restrictions (0.23).
Note that trade protectionism and trade restriction ratings are identical among emerging-market parties: The
phenomenon of trade skepticism for nonnationalist reasons does not (yet) appear to exist in these countries, at
least not in the major parties.
Because of the small sample numbers involved, most of the average changes shown in table 3 are not statisti-
cally significant. The exceptions are increases in support for immigration restrictions, trade protectionism and
trade restrictions in advanced economies (p = 0.033, p = 0.1, and p = 0.027, respectively) and macroeconomic
populism in emerging-market economies (p = 0.081).

New versus Old Parties

As described in the previous section, the sample of post-GFC election platforms includes ten parties that did
not exist before the GFC. In addition, we rated three sets of government policies or policy documents that
turned out to be quite different from the preferences expressed by the preceding party platforms: Made in
China 2025 (State Council 2015),15 the policies of the Trump administration, and the policies advocated in
Germany’s National Industrial Strategy 2030 (Altmaier 2019).16 These policy preferences were treated as if they
corresponded to those of new political parties, bringing the total of major new entrants to 13.
Figure 4 compares the distribution of ratings among these new entrants—eight in the advanced countries,
five in the emerging markets—with those of the existing parties, both in the post-GFC period. In each policy
category, the sum of shares of the old parties and new entrants sum to one, so that the charts give a sense of
the number of new entrants relative to existing parties in addition to their policy preferences. Existing parties
are represented in blue, new entrants in orange. In each policy category the bar of the group with the higher
average rating is shown with a darker shading. For example, in the “trade protectionism” category, the bars corre-
sponding to the new entrants are represented with a darker shade (dark orange) than those of the existing parties
(light blue), reflecting the fact that the new entrants were, on average, more nationalist than the existing parties.
The main results are as follows:
n For the advanced countries, new entrants were, on average, more nationalist than existing parties. This is
true for all policy categories (the mode and most of the probability mass of the distribution of ratings of the
existing parties is to the left of that of the new entrants).
n For the emerging markets, the picture is more mixed. In three categories, existing parties were more nation-
alist than new entrants; in two the reverse was true, and in two (industrial policy and FDI restrictions) there
is a virtual tie. That said, the number of new entrants among the emerging-market parties evaluated is so
small that this comparison is not very meaningful.

15. See McBride and Chatzky (2018) for a description.

16. See Zettelmeyer (2019a) for an analysis.

19
Figure 4
Party distribution of policy ratings of existing and new parties, post-GFC

.56
a. Advanced economies
.39 .38
.3 .3
.27
.23

.1 .1 .1 .1 .1 .12 .12
.06 .06 .08 .09 .09
.03 .04 .04 .04 .04 .03 .03

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Trade protectionism FDI protectionism Immigration restrictions Antimultilateralism

.74

.53

.42
.39

.16 .16 .18 .16

.15
.1 .09 .09 .09 .09
.06 .06 .06 .06 .04 .04 .06
.03 .03 .03 .03

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Trade restrictions “Vertical” industrial policy Tolerance Macroeconomic
of concentration populism

b. Emerging-market economies
.5 .47
.39
.32
.26 .25 .25
.22

.11 .11 .11 .11 .11 .11 .11 .11

.05 .05 .06 .06 .06 .05 .05

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Trade protectionism FDI protectionism Immigration restrictions Antimultilateralism

.38
.32 .33 .32
.26
.20 .18
.17 .14 .14
.11 .11 .11 .14 .13 .13 .13
.08 .08 .07 .07 .07 .09 .09
.05 .05 .04 .05

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Trade restrictions “Vertical” Tolerance Macroeconomic
industrial policy of concentration populism

Existing parties
New parties

GFC = global financial crisis; FDI = foreign direct investment

Note: Ratings: 1 = liberal/nonnationalist; 5 = illiberal/nationalist. Number above bars shows the proportion of existing and
new parties. Within each category, the height of all bars sums to one and the bars of the group (existing or new parties)
with the higher average rating are highlighted through a darker shading.
Source: Authors’ calculations based on data shown in appendix C.

20
 Although new entrants in the advanced countries had more nationalist policy preferences, on average, than
existing parties, there was considerable variation. Three of the eight advanced country entrants—La République
En Marche!, the Party of Hope (Japan), and the People’s Party (Korea)—were rated less nationalist, on average,
than the incumbents. Among the remainder, two—the Alliance for Germany and the Five Star Movement—
represent interesting mixes. The Alliance for Germany espouses a mix of ordoliberal and nationalist policies, the
latter with respect to immigration (rated 4), antimultilateralism (rated 4 because of a “Germany first” attitude
toward international fora), macroeconomic populism (rated 3, on the grounds of the party’s attacks on the
European Central Bank in defense of the German saver), and FDI (rated 3 for emphasizing the need to protect
German technology from acquisition by foreign state-owned companies). The Five Star Movement was rated 3
in trade protectionism and restrictions for its ambivalence about trade, and 4 in the antimultilateralism category,
for making “sovereignty” vis-à-vis the European Union a central theme of its platform (including vowing to
ignore EU fiscal rules), but 1 on FDI protectionism (by expressing very friendly views on FDI) and 2 on immi-
gration. It was also rated 5 on macroeconomic populism (in part for arguing that the presence of a national bank
could make fiscal constraints disappear).
The three documents or policy sets that were rated separately from party platforms in China, the United
States, and Germany also received relatively high ratings. Made in China 2025 was rated in trade protectionism,
FDI protectionism, trade restrictions, tolerance of concentration, and industrial policy toward specific sectors,
and received 4, 3, 3 and 4, respectively. The policies of the Trump administration were rated 4 for trade protec-
tionism, 2 for FDI restrictions, 4 for immigration, 4 for antimultilateralism, 3 for industrial policy toward specific
sectors, and 3 for macroeconomic populism. German economy minister Peter Altmaier’s National Industrial
Strategy 2030 was rated 3 for FDI protectionism, 4 for industrial policy, and 3 for tolerance of concentration,
but just 1 for antimultilateralism and 2 for trade protectionism.
The figure also allows a comparison between the policy preferences of existing parties (blue bars) in advanced
and emerging-market economies as well as between new entrants (orange bars) in advanced and emerging-
market economies. As far as existing parties go, policy preferences in the advanced economies tend to be more
liberal/less nationalist than those in emerging-market economies, except with respect to immigration (where
advanced-economy parties tend to be more restrictive) and multilateralism (where the distributions are very
similar). The differences are particularly stark in the categories of trade restrictions/protectionism and vertical
industrial policies, where the mode of the preference distribution of existing advanced-economy parties is 2, and
4 for emerging-market parties. In contrast, the policy preferences of new entrants in advanced countries tend to
be at least as nationalist as those of new entrants in emerging-market economies, and not very different from the
traditional policy preferences of emerging-market parties.

Correlations with Political Orientation and Populism

Although this paper does not investigate the causes of economic nationalism, it is interesting to see to what
extent party preferences for economic nationalism are correlated both with political orientation (left-wing
versus right-wing) and with political populism as classified by other researchers. For right-/left-wing orienta-

21
tion, two measures were used: the Manifesto Project’s “right-left index” (RILE, a continuous variable defined
between −100 and 100, calculated by subtracting the proportion of party manifestos designated as “left” from
the proportion of those designated as “right”; see Mölder 2016); and a left-/right-wing dummy variable based
on the Database of Political Institutions 2017 (Cruz, Keefer, and Scartascini 2018) as primary source, and the
Comparative Political Data Set 1960–2016 (Armingeon et al. 2018) as secondary source. The classification of
parties as populist or nonpopulist was based on the PopuList (Rooduijn et al. 2019, primary source) and Kyle
and Gultchin (2018) (secondary source). Appendix table D1 shows the party classifications.
Consider first the correlations between RILE and current (post-GFC) economic nationalism (figure 5). The
figure shows scatterplots with RILE on the horizontal axis and post-GFC ratings on the vertical axis. A positively
sloped regression line means that more right-wing parties tend to have higher (more nationalist/illiberal) ratings,
while a negatively sloped line indicates a correlation between left-wing parties and higher ratings. The figure
indicates that FDI protectionism, immigration restrictions, and antimultilateralism tend to be correlated with
right-wing orientation (the first two are statistically significant). Left-wing orientation is correlated with a prefer-
ence for industrial policies toward specific sectors and (more weakly) for macroeconomic populism and tolerance
of concentration. Interestingly, there is no significant correlation between RILE and attitudes toward trade.
Figure 6 examines the correlation between post-GFC ratings and political orientation using a different set
of party classifications, by running regressions of ratings on left-/right-wing dummy variables (based on Cruz,
Keefer, and Scartascini 2018 and Armingeon et al. 2018) and plotting the difference in the coefficients (light
green bars). The results are broadly consistent with those of figure 5: immigration restrictions, antimultilater-
alism, and FDI protectionism tend to be correlated with right-wing orientation and industrial policy with left-
wing orientation. For example, with respect to immigration restrictions, the right-wing parties in our sample are
1.4 points more nationalist, on average, than the left-wing parties; for antimultilateralism, the average difference
is 0.7 points. Figure 6 also suggests a correlation between left-wing orientation and trade protectionism/trade
restrictions, but this is small and not statistically significant (see table D2 for details).
Figure 6 also shows the correlations of various policy preferences with populism, based on a similar regres-
sion (dark blue bars). Nationalist/illiberal policy preferences turn out to be positively correlated with populism
in all policy categories, with statistically significant coefficients in most categories (see table D2). The average
difference between populist and nonpopulist parties with respect to trade protectionism is almost 1 point, with
respect to immigration restrictions 1.2 points, with respect to antimultilateralism 1.5 points, and with respect to
macroeconomic populism 1.4 points.
Figures 7 and 8 focus on the correlations between political ideology and populism with changes in (rather
than levels of) policy preferences, using a similar regression-based approach (see notes to the figures and table
D2). Figure 7 shows that parties that were classified as populist in 2018 (dark blue bars) shifted their prefer-
ences more in a nationalist/illiberal direction than parties classified as nonpopulist (light blue bars). Among the
populist parties, a large shift toward trade protectionism stands out (by 0.6 point, as opposed to just 0.1 point
on average among the nonpopulist parties).

22
Figure 5
Post-GFC party ratings and RILE
a. Trade protectionism b. FDI protectionism
post-GFC rating post-GFC rating
5 5

4 4

3 3

2 2
y = 0.0023x + 2.1608
(0.776) y = 0.0348x + 2.1177
1 1 (0.010)

0 0
–60 –50 –40 –30 –20 –10 0 10 20 30 40 –60 –50 –40 –30 –20 –10 0 10 20 30 40
RILE RILE

c. Immigration restrictions d. Antimultilateralism

5 5

4 4
y = 0.0181x + 2.0273
3 3 (0.165)

2 2
y = 0.0227x + 2.5249
1 (0.012) 1

0 0
–60 –40 –20 0 20 40 –60 –50 –40 –30 –20 –10 0 10 20 30 40

e. Trade restrictions f. Industrial policies toward specific sectors

5 y = –4E–05x + 2.2854 5
y = –0.0169x + 2.2856
(0.996) (0.045)
4 4

3 3

2 2

1 1

0 0
–60 –50 –40 –30 –20 –10 0 10 20 30 40 –60 –40 –20 0 20 40

g. Tolerance of concentration h. Macroeconomic populism

5 5
y = –0.0089x + 1.8166
y = –0.0059x + 1.5584 (0.479)
4 (0.532) 4

3 3

2 2

1 1

0 0
–60 –50 –40 –30 –20 –10 0 10 20 30 40 –60 –50 –40 –30 –20 –10 0 10 20 30 40

GFC = global financial crisis; FDI = foreign direct investment

Note: Ratings on vertical axis: 1 = liberal/nonnationalist; 5 = illiberal/nationalist. RILE stands for the Manifesto Project’s
“Right-Left index” measuring political orientation on a scale of –100 to 100, where larger positive values mean more
right-wing. P values of slope coefficients shown in parentheses (bold indicates p≤0.05).
Sources: Manifesto Project Dataset (Volkens et al. 2018b); authors’ classification based on data shown in appendix C.

23
 Figure 6
Average differences in post-GFC ratings, right-wing versus left-wing and populist
versus nonpopulist parties

1.53
1.4 1.44

1.2

.94 .94

.66 .71

.43 .47
.39

–.15
–.24
–.42
–.56
–.77

lis ic
m

ns e

lic ial

io e
ni ade

sm I

ns n
ni FD

tio ad

at c
tio io

pu om
is

m
y

n
sm

po str

tr an
ric at

al

ric Tr
tio Tr

en ler

po on
du
st gr

er
tio

at
re mi

nc To

ec
In
til
ec

ec

st
Im

ro
ul

re

ac
ot

ot

im

co
pr

pr

M
nt

of
A

Average difference between right-wing and left-wing parties

Average difference between populist and nonpopulist parties

GFC = global financial crisis; FDI = foreign direct investment

Note: The figure shows the average differences in rating, in each category, between right-wing and left-wing parties
(left bar, in light green) and populist and nonpopulist categories (right bar, in dark blue). It is based on coefficients
from the regressions r(i) = b0 + b1*L(i) + b2*R(i) + d(i) and r(i) = c0 + c1*P(i) + d(i) (see table D2 for details), where
r(i) is the rating assigned to party i and R(i), L(i), and P(i) are dummy variables taking on the values one for right-wing,
left-wing, and populist parties, respectively, and zero otherwise; and d(i) is a full set of dummy variables for specific
parties. The left bars show the difference between the coefficient b2 and b1 (i.e., b2 – b1), the right bars the coefficient
c1. In the first regression, the residual category is center parties and unclassified parties, while in the second regression
the residual category is nonpopulist parties.
Source: See note to table D2 in appendix D.

Figure 8 compares preference shifts between left- and right-wing parties. Whereas figure 6 indicated that
right-wing parties are more nationalist/illiberal than left-wing parties in categories such as immigration restric-
tions, FDI restrictions, and antimultilateralism, figure 8 shows that the shift in the direction of nationalism/
illiberalism since the GFC has in fact been larger among left-wing parties in almost all policy categories (FDI
restrictions are the only exception). The shift was particularly pronounced with respect to macroeconomic popu-
lism (by 0.7 point) and in the trade policy areas (by 0.5–0.6 point), where differences with right-wing parties are
no longer statistically significant in the post-GFC period (table D2).

Country-Level Changes
Combining shifts in preferences among existing parties with the preferences of new entrants, how did average
preferences at the country level change in the eight policy categories studied in this paper? This question is
answered in figure 9. The horizontal axis of each chart shows pre-GFC average country-level ratings, the vertical
axis average country-level ratings in the most recent post-GFC elections. Averages were computed by weighting
the party-level ratings with the share of the vote that the parties obtained in the elections listed in table 1.
Advanced-economy averages are denoted by blue dots, emerging-market averages by dark red dots. The three

24
 Figure 7
Average rating changes from pre- to post-GFC, populist versus nonpopulist
parties

.6 .6
.5 .5
.43
.33 .33 .33
.25 .09 .24
.2 .04
.13 .14

–.29

lis ic
ns e

lic ial
ni de

ns n

io e
sm I
ni D

tio ad
tio io

at c

pu om
is

m
sm

n
po str
io F

tr an
io ra

ric at

al

ric Tr
T

en ler
du

po on
st gr

er
at
re mi

nc To

ec
In
t

til
ec

st
ec

Im

ro
ul

re
ot

ac
ot

im

co
pr

pr

M
nt

of
A

Populist
Nonpopulist

GFC = global financial crisis; FDI = foreign direct investment

Note: The figure shows the average change in ratings, in each category, from the pre-GFC to the post-GFC
period for parties classified as populist or nonpopulist. The left bars (dark blue) and right bars (light blue)
plot the coefficients b1 and b2, respectively, from the regression r(i,t) = b0 + b1*postGFC*P(i) + b2*postGFC
*NP(i) + d(i), where r(i,t) is the rating assigned to party i in each category at time t (before or after the GFC),
postGFC is a dummy variable taking on the value one in the post-GFC period and zero otherwise and P(i)
and NP(i) are dummy variables taking on the values one for populist and nonpopulist parties, respectively,
and zero otherwise; and d(i) is a full set of dummy variables for specific parties. See table D2 in appendix D
for details.
Source: See note to table D2 in appendix D.

Figure 8
Average rating changes from pre- to post-GFC, left-wing versus right-
wing parties

.69
.62
.46
.33 .33
.27 .24 .21 .22
.13

–.07 –.07 –.07

–.25
io e

lis ic
ns e

lic ial
ni ade

sm I

ns n

m
ni FD

at c
tio d
tio io

pu om
is

m
y
sm

po str

tr an
ic Tra
ric at

al
tio Tr

en ler
du

po on
st gr

er
tio

at
re mi

nc To

ec
In
r
til
ec

ec

st
Im

ro
ul

re
ot

ot

ac
im

co
pr

pr

M
nt

of
A

Left-wing
Right-wing

GFC = global financial crisis; FDI = foreign direct investment

Note: The figure shows the average change in ratings, in each category, from the pre-GFC to the post-
GFC period for parties classified as either left-wing or right-wing. The left bars (dark green) and right
bars (light green) plot the coefficients b1 and b2, respectively, from the regression r(i,t) = b0 + b1*postGFC*
L(i) + b2*postGFC*R(i) + d(i), where r(i,t) is the rating assigned to party i in each category at time t (before
or after the GFC), postGFC is a dummy variable taking on the value one in the post-GFC period and zero
otherwise and L(i) and R(i) are dummy variables taking on the values one for left-wing and right-wing
parties, respectively, and zero otherwise; and d(i) is a full set of dummy variables for specific parties. See
table D2 in appendix D for details.
Source: See note to table D2 in appendix D.

25
Figure 9
Country averages (including new parties), pre-GFC versus post-GFC

a. Trade protectionism b. FDI protectionism

post-GFC post-GFC
5 5
RUS IND
4 4 IDN
USA BRA CHN
3 3 DEU IND
ITA FRA FRA
TUR ARG CHN
2 BRA
GBR MEX
2 CAN
AUS
DEU JPN CAN ARG TUR
1 1
ITA

1 2 3 4 5 NA 1 2 3 4 5
pre-GFC pre-GFC

c. Immigration restrictions d. Antimultilateralism

5 5
RUS
4 4
IDN
ZAF
3 ITA 3 ITA
AUS FRA TUR USA
DEU GBR GBR FRA
2 2 ARG CHN
USA DEU
CAN BRA
1 ZAF
1 KOR IND
MEX CANJPN MEX
AUS

1 2 3 4 5 1 2 3 4 5

e. Trade restrictions f. Industrial policies toward specific sectors

5 5
RUS IND RUS
4 4 IDN CHN
ZAF BRA
USA BRA CHN MEX ITA
3 TUR ARG 3 TUR IND
ITA FRA ARG FRA
DEU GBR CAN
2 MEX 2 JPN KOR
CAN AUS
GBR AUS JPN DEU
1 1 USA

1 2 3 4 5 1 2 3 4 5

g. Tolerance of concentration h. Macroeconomic populism

5 5

4 4 ITA
IND
RUS
ZAF ARG
3 3 CHN
BRA
ITA CHN MEX
2 IDN MEX 2 GBR
TUR FRA
KOR TUR DEU
IDN USA
1 GBR DEU 1 CAN KOR
JPN CANUSA AUS JPN
AUS BRA

1 2 3 4 5 1 2 3 4 5
USA-Trump Advanced economies
Germany-NIS2030 Emerging-market economies
Made in China 2025 Advanced and emerging-market economies sharing same data points

GFC = global financial crisis; FDI = foreign direct investment

Note: Country averages computed by weighting party-level classifications by share of votes obtained. The three
ratings for China, the United States, and Germany that refer to recent government pronouncements or policies rather
than party platforms were not included in country averages and are shown separately, using squares (bright red for
Trump, green for Germany’s National Industrial Strategy 2030, and black for Made in China 2025). In graph b, countries
for which there was insufficient information to assign pre-GFC scores but sufficient information to assign post-GFC scores
in a specific category were classified as “n.a.” on the horizontal axis in order to be able to show a vertical axis score.
26
Source: Authors’ classification based on data shown in appendix C.
ratings for China, the United States, and Germany that refer to recent government pronouncements or policies
rather than party platforms were not included in country averages and are shown separately, using squares (bright
red for Trump, green for Germany’s National Industrial Strategy 2030, and black for Made in China 2025).17
There are three main takeaways from the figure.
First, in all eight categories, most points are above the 45-degree line, indicating a rise in economic nation-
alism (figures 9a–9d) and/or a move away from the liberal consensus paradigm (figures 9e–9h). Among the
advanced economies, the largest shifts occurred in Italy and the United States, particularly with respect to
antimultilateralism, trade protectionism, immigration, and macroeconomic populism. Among the emerging-
market economies, the largest shifts toward economic nationalism were in Russia, South Africa, and, on some
policy dimensions, India.
Some countries bucked the trend, however. Among the advanced economies, these include Canada, Japan,
and France; among the emerging markets, Mexico and China (measured based on the comparison between
the 13th and 11th five-year plans). For the most part, these are countries that started out with fairly high (i.e.,
nationalist or illiberal) ratings in the pre-GFC period. France’s recorded decline in economic nationalism reflects
two facts: the nationalism of several party platforms in the pre-GFC period (more so than other advanced-
economy parties at that time) and the entry and electoral success of the liberal La République En Marche! in
the 2017 election. Canada, on the other hand, is a genuine outlier in the sense that its economic nationalism
did not rise in spite of the fact that it was low to begin with. Between the 2006 and the 2015 general elections,
neither of the country’s two large parties, the Canadian Liberals and the Conservatives, shifted positions on the
policy dimensions evaluated in this paper. Japan looks like Canada in the sense that none of the major parties
seem to have changed their position in the direction of more economic nationalism, but, as in France, some of
this may have to do with the fact that on some dimensions—immigration in particular18—they started out fairly
far from the liberal consensus. For Mexico, the results are driven by the fact that the positions of the country’s
two large center-right parties, the PAN and the PRI, were more aligned with the liberal consensus in 2018 than
in the mid-2000s—possibly, to more sharply differentiate themselves from the positions of current president
Andrés Manuel López Obrador’s Movimiento Regeneración Nacional (MORENA), which were much more
nationalist.
Second, as one would expect based on the last section, emerging-market economies remain further from
the liberal consensus than advanced economies in most policy categories (except on immigration and multilater-
alism). Most advanced economies continue to be less protectionist than emerging-market economies, although
less so than pre-GFC. The United States is an exception in this respect: The 2016 platforms of the Democratic

17. To plot these on a two-dimensional plane, pre-GFC country average party preferences corresponding to China,
the United States, and Germany were used as horizontal axis values.
18. A complication, in the case of Japan, is that there is not enough information to rate the immigration stance
of parties in the pre-GFC period, perhaps because there was broad agreement with the country’s comparatively
restrictive immigration regime. There is a bit more information in the post-GFC period, which appears to go in
the direction of liberalizing immigration and naturalization for highly skilled immigrants, with the 2017 Liberal
Democratic Party manifesto stating that “If excellent foreign talents who live in Japan wish to naturalize to Japan,
we will promote efforts to quickly make decisions about their permission.”

27
and Republican parties were both rated 3, implying ambivalence about international trade, while the Trump
administration was rated 4. Broadening the trade policy scale to incorporate nonnationalist antitrade views
(figure 9e) does not change the overall picture: Its only effect is to raise the average ratings for Australia and
(more noticeably) Germany, driven by (nonnationalist) skepticism about trade agreements among progressive
parties in these countries.
With respect to inward FDI, emerging-market economies continue to be more nationalist than advanced
economies, reflecting the embrace of local content requirements and policies intended to encourage technology
transfer. Preferences about multilateral organizations vary widely in both groups. Italy and the United States are
nationalist outliers among advanced economies, and Russia in the emerging-market group.
Finally, the policy preferences of the Trump administration, Made in China 2025, and the German National
Industrial Strategy 2030 are much more nationalist—in the policy categories in which they provide informa-
tion—than the most recent party platforms in the same countries.
Figure 10 gives a sense of the overall magnitude of the shifts in policy preferences for advanced and emerging-
market countries when the ratings of “new entrants” are included in the calculation of the post-GFC averages.19
Because most of these new entrants are more nationalist than existing parties (particularly in the advanced
countries), the changes are somewhat larger than those shown in table 3.
n For the advanced countries, the largest shifts occurred with respect to preferences for immigration restric-
tions (0.44), macroeconomic populism (0.36), and trade restrictions (0.34). Most of the rise in trade skepti-
cism is attributable to protectionism (0.27), the remainder to progressive parties opposing trade agreements
on the grounds that they threaten workers or the environment. Attitudes toward FDI moved in the reverse
direction on average, but this result should be discounted as it reflects the small sample (only three parties
expressed preferences in this policy dimension ahead of the crisis) as well as an outlier.20
n In the emerging-market countries, preferences have shifted the most with respect to industrial policies
focused on specific sectors (0.63) and macroeconomic populism (0.61).
n Trade protectionism and antimultilateralism have risen in both emerging-market and advanced economies,
by roughly similar amounts on average (by 0.32 and 0.25, respectively, in emerging-market economies, and
by 0.27 and 0.29, respectively, in advanced economies).

19. Advanced-country and emerging-market averages were calculated as unweighted means of the country-level
averages shown in figure 9, except for the United States and China, for which country averages were recomputed
by substituting post-GFC party-based ratings with the policy/document-based ratings of Trump and Made in
China 2025, where available (i.e., post-GFC ratings of the Republican Party and the Communist Party of China
were replaced with ratings of Trump and Made in China 2025, respectively). These substitutions were motivated
by the observation that, since 2017, the Republican Party has been supportive of President Trump’s policies and
by an implicit notion that Made in China 2025 reflects the views of China’s political leadership (i.e., the Communist
Party of China). It is too early to judge whether the German National Industrial Strategy 2030 has received broad
support in the CDU/CSU, which is why it was omitted from calculations.
20. Canada’s New Democratic Party took an anti-FDI stance in its 2006 platform (“Across the country, people
are seeing Canada lose control of its own economic future as foreign ownership of Canadian businesses and
resources reaches historic highs” (NDP 2006, 28) while arguing in its 2015 platform that Canada needed to attract
more FDI.

28
 Figure 10
Average change in policy preferences, including new entrants

Trade protectionism

FDI protectionism

Immigration restrictions

Antimultilateralism

Trade restrictions

Industrial policies

Competition policy

Macroeconomic populism

0 1 2 3 4

FDI = foreign direct investment; GFC = global financial crisis

Note: Dark blue represents advanced economies and red represents emerging-
market economies. Dots represent pre-GFC averages, squares post-GFC averages.
Footnote 20 describes how these averages were constructed. Arrows show the
direction of movement. A movement from left to right means that party policy
preferences have become more nationalist and/or less liberal.
Source: Authors’ calculations based on data shown in appendix C.

Comparison with Results Based on the Manifesto Project Database

As a robustness check, in this section we examine whether the basic result reported in this paper—a rise in
economic nationalism around the world since the mid-2000s—holds if an entirely different data source is used,
the Manifesto Project Database. As briefly discussed above, this database does not allow analysis of all policy
categories that are of interest in this paper, but it contains relevant information in three categories: protectionism
(comparable to our trade protectionism category), internationalism (which should be negatively correlated with
our antimultilateralism category), and economic orthodoxy (which is a proxy for the absence of macroeconomic
populism). As discussed previously, the Manifesto Project Database creates measures for party preferences in
these categories by counting the frequency of positive and negative mentions of protectionism and interna-
tionalism (for economic orthodoxy, only positive mentions are counted). From these, a “net” measure can be
constructed by subtracting the frequency of negative mentions from the frequency of positive mentions.
In interpreting the Manifesto Project data, one difficulty is how to deal with instances in which there were
neither positive nor negative mentions in party platforms. One approach is to treat them as missing values:

29
 Figure 11
Country averages based on the Manifesto Project Dataset
a. Protectionism
Zero entries interpreted as missing values Zero entries interpreted as zero values
post-GFC post-GFC
France 2
2 France
Russia United Kingdom Russia
1 Germany Australia
1 South Korea
Japan Italy Japan South Africa
0 0 United States
United States Mexico Italy

–1 –1 Turkey
Canada
Canada
–2 –2
–3 –2 –1 0 1 –3 –2 –1 0 1
pre-GFC pre-GFC

b. Anti-internationalism
1 1
Russia Russia
0 0
France
–1 France –1 Italy
Italy South Africa South Japan South
Japan Africa Korea
–2 –2
United
United States Canada Kingdom Mexico
Turkey Canada Australia
–3 –3
United Kingdom Australia United Turkey
Germany States Germany
–4 –4
–5 –4 –3 –2 –1 0 –5 –4 –3 –2 –1 0

c. Economic orthodoxy
4 4
United Kingdom Turkey
Australia Japan
Turkey United Kingdom
3 3

France Japan
2 2
France Canada
Canada Germany Italy
United Germany Italy
1 1 States
Mexico
Russia South Korea
0 0 South Africa
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Advanced economies
Emerging-market economies

GFC = global financial crisis

Note: Country averages computed by weighting party-level balance of positive and negative mentions by share
of votes obtained.
Source: Manifesto Project Dataset (Volkens et al. 2018a).

This assumes that the party had no views in the corresponding category. Another is to treat them as zeros,
i.e., as neutral views. Figure 11 shows country-level results for the three economic nationalism–related policy
categories covered by the Manifesto Project, based on both approaches. As in figure 9, country-level average
party ratings were computed using the proportion of votes obtained by the parties for which data were available
both in the closest election before the GFC (x-axis) and in the most recent election (y-axis). To give the results

30
for internationalism the same interpretation as the antimultilateralism ratings shown in figure 9, the sign of
net internationalism is reversed: Bigger numbers mean greater opposition to (or a less favorable view of) inter-
national cooperation, as in the antimultilateralism policy category. In contrast, the policy category economic
orthodoxy, which shows only positive mentions, is plotted without change. The interpretation of this figure is
thus different from the other two: In figures 11a and 11b, points above the 45-degree line imply an increase in
economic nationalism; in contrast, in figure 11c, points below the 45-degree line imply a decline in economic
orthodoxy, which we interpret as a proxy for an increase in macroeconomic populism.
The main result is that, consistent with figures 9a and 9d, most points in figures 11a and 11b lie above
the 45-degree line, indicating an increase in economic nationalism. This is true regardless of how parties with
zero ratings are treated. The two datasets also agree that Australia is one of the few countries to dodge that
trend (there is no agreement on the remaining countries, which according to figures 11a and 11b are Italy
and Turkey for protectionism and Germany and Turkey for anti-internationalism, while figures 9a and 9d
indicate small increases in economic nationalism for these countries). Figure 11c is consistent with the view that
macroeconomic populism has increased in most countries since the pre-GFC period, with France being one of
the exceptions, in line with the results of figure 9h (again, there is no agreement on the remaining exceptions).
Although the magnitude of these changes is impossible to compare across the two datasets because of the
different methods for generating policy ratings, the messages are broadly consistent. Most G-20 countries for
which data are available in both datasets have experienced a rise in economic nationalism since the mid-2000s.

V. CONCLUSION
Economic nationalism, defined as a preference for economic policies that seek to promote national economic
interests at the expense of foreign interests, has been on the rise since the mid-2000s. This rise is broad-based,
encompassing advanced and emerging-market economies, right-wing and left-wing parties, and both existing
parties and new entrants. While parties that are classified as populist by political scientists tend to have much
more nationalist and illiberal economic policy preferences than parties classified as nonpopulist, preference shifts
toward economic nationalism are visible in both groups.
At the same time, these shifts are not universal. Exceptions include some countries that were already rela-
tively nationalist in their economic policy orientation prior to the GFC, such as France and several emerging-
market economies, as well as Canada, where the main parties have not shifted their positions away from the
liberal and multilateralist consensus.
This paper focused on the measurement of political preferences. It did not address two important questions.
First, what are the causes for the increasing popularity of economic nationalism, which appears to be on the
rise not just among populist parties and politicians but also in mainstream parties? How do shifts to economic
nationalism relate to economic and social change? Second, to what extent are these changes in preferences
translating into actual policies? In some countries, economic nationalism may be a paper tiger. But based on the
experience in the United States, where the policies of the administration have been more nationalist than could
have been inferred from the 2016 party platforms—in turn triggering economic nationalist responses in other
countries—this cannot be assumed.

31
REFERENCES
Abdelal, Rawi. 2001. National Purpose in the World Economy: Post-Soviet States in Comparative Perspective.
Ithaca and London: Cornell University Press.
Abdelal, Rawi. 2005. Nationalism and International Political Economy in Eurasia. In Economic Nationalism in a
Globalizing World, ed. Eric Helleiner and Andreas Pickel. Ithaca and London: Cornell University Press.
Altmaier, Peter. 2019. National Industrial Strategy 2030: Strategic Guidelines for a German and European Industrial
Policy. Berlin: Federal Ministry for Economic Affairs and Energy.
Armingeon, Klaus, Virginia Wenger, Fiona Wiedemeier, Christian Isler, Laura Knöpfel, David Weisstanner, and Sarah
Engler. 2018. Comparative Political Data Set 1960–2016. Bern: Institute of Political Science, University of Bern.
Besley, Timothy, Mathias Dewatripont, and Sergei Guriev. 2010. Transition and Transition Impact: A Review of the
Concept and Implications for the EBRD. Report for the EBRD’s Office of the Chief Economist. London: European
Bank for Reconstruction and Development.
Crane, George T. 1998. Economic Nationalism: Bringing the Nation Back In. Millennium: Journal of International
Studies 27, no. 1: 55–75.
Cruz, Cesi, Philip Keefer, and Carlos Scartascini. 2018. Database of Political Institutions 2017. Washington: Inter-
American Development Bank Research Department.
Dornbusch, Rudiger, and Sebastian Edwards. 1991. The Macroeconomics of Populism. In The Macroeconomics of
Populism in Latin America, ed. Rudiger Dornbusch and Sebastian Edwards. Chicago and London: University of
Chicago Press.
Drezner, Daniel W. 2017. Who benefits from Bannon’s economic nationalism? Washington Post, February 7.
Available at www.washingtonpost.com/posteverything/wp/2017/02/07/who-benefits-from-bannons-economic-
nationalism/?utm_term=.ddacd7128221 (accessed on August 4, 2019).
Eatwell, Roger, and Matthew J. Goodwin. 2018. National Populism: The Revolt Against Liberal Democracy. London:
Pelican Books.
Eichengreen, Barry. 2018. The Populist Temptation: Economic Grievance and Political Reaction in the Modern Era.
Oxford: Oxford University Press.
EBRD (European Bank for Reconstruction and Development). 1999. Transition Report 1999: Ten Years of Transition.
Economic Transition in Central and Eastern Europe, the Baltic States and the CIS. London.
EBRD (European Bank for Reconstruction and Development). 2009. Transition Report 2009: Transition in Crisis?
London.
Fichte, Johann Gottlieb. 1800. Der Geschlossene Handelsstaat. Tübingen (Cotta). Republished by Michael Holzinger,
2015, Berlin.
Funke, Manuel, Moritz Schularick, and Christoph Trebesch. 2015. Politics in the Slump: Polarization and Extremism
after Financial Crises, 1870–2014. Available at https://ec.europa.eu/economy_finance/events/2015/20151001_
post_crisis_slump/documents/c._trebesch.pdf (accessed on July 26, 2019).
Gilpin, Robert (with the assistance of Jean M. Gilpin). 1987. The Political Economy of International Relations.
Princeton University Press.
Gregory, Theodore E. 1931. Economic Nationalism. Address given at Chatham House, March 12. International Affairs
(Royal Institute for International Affairs 1931–39) 10, no. 3: 289–306.
Guiso, Luigi, Helios Herrera, Massimo Morelli, and Tommaso Sonno. 2018. Populism: Demand and Supply. Available
at www.heliosherrera.com/populism.pdf (accessed on July 26, 2019).
Hamilton, Alexander. [1791] 1966. Report on Manufacturers. In The Papers of Alexander Hamilton, December 1791–
January 1792, vol. 10, ed. Harold C. Syrett, 230–340. New York: Columbia University Press.
Heilperin, Michael A. 1960. Studies in Economic Nationalism. Geneva: L’Institut Universitaire de Hautes Études
Internationales.
Helleiner, Eric. 2002. Economic Nationalism as a Challenge to Economic Liberalism? Lessons from the 19th
Century. International Studies Quarterly 46, no. 3: 307–29.
Helleiner, Eric. 2005. Conclusion: The Meaning and Contemporary Significance of Economic Nationalism. In
Economic Nationalism in a Globalizing World, ed. Eric Helleiner and Andreas Pickel. Ithaca and London: Cornell
University Press.
Helleiner, Eric, and Andreas Pickel, eds. 2005. Economic Nationalism in a Globalizing World. Ithaca and London:
Cornell University Press.
Hieronymi, Otto. 1980. The New Economic Nationalism. Geneva: Battelle Geneva Research Center.

32
Imbs, Jean, and Romain Wacziarg. 2003. Stages of Diversification. American Economic Review 93, no. 1: 63–86.
Johnson, Alvin. 1917. The Passing of Economic Nationalism. Harper’s Magazine 135, no. 806: 221–25.
Johnson, Harry G. 1965. A Theoretical Model of Economic Nationalism in New and Developing States. Political
Science Quarterly 80, no. 2: 169–85.
Johnson, Harry G., ed. 1967. Economic Nationalism in Old and New States. Chicago: University of Chicago Press.
Judis, John B. 2018. The Nationalist Revival: Trade, Immigration, and the Revolt Against Globalization. New York:
Columbia Global Reports.
Kahan, Arcadius. 1967. Nineteenth-Century European Experience with Policies of Economic Nationalism. In
Economic Nationalism in Old and New States, ed. Harry G. Johnson. Chicago: University of Chicago Press.
Kyle, Jordan, and Limor Gultchin. 2018. Populists in Power Around the World. London: Tony Blair Institute for
Global Change.
Levi-Faur, David. 1997. Economic Nationalism: From Friedrich List to Robert Reich. Review of International Studies
23, no. 3: 359–70.
List, Friedrich. 1841. Das Nationale System der Politischen Ökonomie. Stuttgart: J.G. Cotta Verlag.
Mayall, James. 1990. Nationalism and International Society. Cambridge: Cambridge University Press.
McBride, James, and Andrew Chatzky. 2018. Is ‘Made in China 2025’ a Threat to Global Trade? Council on Foreign
Relations Backgrounder, August 2. Available at www.cfr.org/backgrounder/made-china-2025-threat-global-trade.
McHugh, Mary L. 2012. Interrater Reliability: The Kappa Statistic. Biochemia Medica 22, no. 3: 276–82.
Mölder, Martin. 2016. The Validity of the RILE Left–Right Index as a Measure of Party Policy. Party Politics 22, no.
1: 37–48.
Mounk, Yascha. 2018. The People vs. Democracy: Why Our Freedom Is in Danger and How to Save It. Cambridge,
MA: Harvard University Press.
Nakano, Takeshi. 2004. Theorising Economic Nationalism. Nations and Nationalism 10, no. 3: 211–29.
NDP (New Democratic Party of Canada). 2006. Jack Layton. Getting Results for People. Platform 2006. Available
at www.poltext.org/sites/poltext.org/files/plateformes/can2006ndp_plt_en._14112008_165642.pdf (accessed on
August 16, 2019).
Pickel, Andreas. 2003. Explaining, and Explaining with, Economic Nationalism. Nations and Nationalism 9, no. 1:
105–27.
Pickel, Andreas. 2005. Introduction: False Oppositions: Reconceptualizing Economic Nationalism in a Globalizing
World. In Economic Nationalism in a Globalizing World, ed. Eric Helleiner and Andreas Pickel. Ithaca and London:
Cornell University Press.
Pryke, Sam. 2012. Economic Nationalism: Theory, History and Prospects. Global Policy 3, no. 3: 281–91.
Rappard, William E. 1937. Economic Nationalism. In Authority and the Individual: Harvard Tercentenary Conference
of Arts and Science, 74–112. Cambridge, MA: Harvard University Press.
Rooduijn, M., Van Kessel, S., Froio, C., Pirro, A., De Lange, S., Halikiopoulou, D., Lewis, P., Mudde, C. & Taggart,
P. 2019. The PopuList: An Overview of Populist, Far Right, Far Left and Eurosceptic Parties in Europe. Available
at www.popu-list.org (accessed on July 25, 2019).
Shulman, Stephen. 2000. Nationalist Sources of International Economic Integration. International Studies Quarterly
44, no. 3: 365–90.
State Council. 2015. Made in China 2025 (May 8). The Chinese language version is available at www.gov.cn/
zhengce/content/2015-05/19/content_9784.htm (accessed on March 1, 2019), and the English language version
is available at www.cittadellascienza.it/cina/wp-content/uploads/2017/02/IoT-ONE-Made-in-China-2025.pdf
(accessed on March 1, 2019).
Szlajfer, Henryk. 2012. Economic Nationalism and Globalization: Lessons from Latin America and Central Europe.
Leiden: Brill.
Volkens, Andrea, Pola Lehmann, Theres Matthieß, Nicolas Merz, Sven Regel, and Bernhard Weßels. 2018a. The
Manifesto Data Collection. Manifesto Project (MRG/CMP/MARPOR). Version 2018a. Berlin: Wissenschaftszentrum
Berlin für Sozialforschung (WZB). Available at https://doi.org/10.25522/manifesto.mpds.2018a (accessed on May
15, 2019).

33
 Volkens, Andrea, Werner Krause, Pola Lehmann, Theres Matthieß, Nicolas Merz, Sven Regel, and Bernhard
Weßels. 2018b. The Manifesto Data Collection. Manifesto Project (MRG/CMP/MARPOR). Version 2018b. Berlin:
Wissenschaftszentrum Berlin für Sozialforschung (WZB). Available at https://doi.org/10.25522/manifesto.
mpds.2018b (accessed on July 25, 2019).
Zettelmeyer, Jeromin. 2019a. The Return of Economic Nationalism in Germany. PIIE Policy Brief 19-4. Washington:
Peterson Institute for International Economics.
Zettelmeyer, Jeromin. 2019b. The Troubling Rise of Economic Nationalism in the European Union. Realtime
Economic Issues Watch, March 29. Washington: Peterson Institute for International Economics.

© Peterson Institute for International Economics. All rights reserved.

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 neces-
sarily 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, intellectu-
ally 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 in its latest fiscal year 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.

34
APPENDIX A
MEASURING ECONOMIC NATIONALISM: DEFINITION OF SCALES
The scales below define a modern liberal, “social market economy” consensus as well as increasing deviations
from this consensus. A rating of 1 denotes adherence to the consensus, 5 the maximum deviation. Eight policy
categories are covered.
n The first four—trade protectionism, FDI protectionism, immigration restriction, and rejection of multilat-
eral constraints—describe deviations from the liberal consensus that occur for economic nationalist reasons
(based on the definition proposed in the main text, that is, policy measures that seek to promote national
interests at the expense of foreign interests, at least in the short run).
n Four further categories—trade restrictions, competition policy (tolerance of concentration), vertical indus-
trial policy, and macroeconomic populism—describe deviations from the liberal consensus that might
reflect a variety of motives, both nationalist and nonnationalist. The latter could include a desire to protect
the interests of workers against big business, or the interests of business at the expense of consumers, or the
poor against the rich.

One area where the distinction between nationalist and nonnationalist motives is important is with respect
to trade policies. Because the liberal consensus has been criticized from two sides that otherwise have nothing
in common—nationalists caring mainly about domestic business, and progressives caring about global envi-
ronmental social objectives and the welfare of poorer countries—two trade-related scales are defined. “Trade
protectionism” rates deviations that are specifically motivated to protect domestic interests; “trade restrictions”
adds deviations from the consensus for any other motive, focused on the willingness to tolerate trade barriers or
discourage trade agreements.21

Economic Nationalism Stricto Sensu

Trade Protectionism
This scale captures deviations from liberal trade preferences only to the extent that they are motivated by economic
nationalism (in this context that means protectionism). Trade restrictions or rejection of trade agreements motivated
by strengthening the interests of consumers vis-à-vis firms, labor vis-à-vis firms, small firms vis-à-vis multinationals, or
noneconomic (e.g., environmental) versus economic objectives are disregarded in this scale and reflected in the “trade
restrictions” scale below.
Because this scale is concerned with deviations from the liberal benchmark only to the extent that they are driven
by protectionism, a rating of 1 can indicate not only the liberal consensus benchmark but also deviations from that

21. Note that the distinction between nationalist/protectionist motives and progressive motives does not unam-
biguously map to the right/left distinction.
n To the extent that the political left is opposed to free trade/trade agreements, it is usually for progressive rea-
sons, but sometimes also for protectionist reasons (example: Democrats in the United States).
n When the political right does not embrace free trade, it is generally for protectionist reasons. But even here
there are exceptions. In 2017, the AfD opposed bilateral trade agreements with essentially the same arguments
as the German Greens: that they would reduce democracy/national sovereignty with respect to regulation.
And like the Greens, the AfD argued that the developing countries require some tariffs to develop, while the
European Union should open its markets further.

35
consensus motivated by nonnationalist objectives (e.g., protecting the interests of consumers and workers relative to the
interests of business, small firms relative to multinationals) or noneconomic (e.g., environmental) relative to economic
(e.g., growth) objectives.

1. EITHER:
Free trade subject to the rules of the multinational trading system:
n There is no stated preference of exports over imports (trade is viewed as good, not just exports).
n There are no tariffs, quotas, or export subsidies (except for purely defensive safeguards or countervailing
duties that are justified within multinational trading rules).
n Any export promotion takes the form of lowering information costs and helping firms navigate foreign
entry requirements, rather than financial incentives.
n Government procurement does not discriminate against foreign firms.
n Nontariff trade barriers are the consequence of regulation that is justified by public policy aims extraneous
to trade (such as environmental or health standards) rather than the desire to keep out foreign firms. It is
applied in the same way to domestic and foreign firms.
n Trade negotiations and agreements (whether multilateral, plurilateral, or bilateral) aim at reducing regula-
tory barriers to trade, including by aligning or coordinating regulation. The motivation for the latter,
however, is solely the desire to lower trade barriers as much as possible while maintaining domestic regula-
tory preferences, reflected in consumer protection, labor standards, or environmental standards, not to
“level the playing field” (i.e., to make other countries lose a competitive edge afforded by cheaper labor and
lower regulatory cost).

OR:
n Some trade barriers are maintained, and/or plurilateral/bilateral trade agreements are opposed, on the
grounds that some trade barriers are in the interests of developing countries, consumer protection, the
environment, labor, or small and medium-sized enterprises (SMEs); that trade arrangements would add
to the power of multinational corporations; and that trade arrangements would weaken domestic regula-
tion—particularly regarding environmental, health, and labor standards.
Note: To qualify for a 1 rating in this scale, any defense of trade barriers and/or rejection of trade deals or trade
liberalization must not be motivated by a desire to put the interests of domestic entities (consumers, workers, firms)
over those of the corresponding foreign entities. Resisting liberalization to protect domestic consumers from foreign
firms or (the lack or) foreign regulation is consistent with a rating of 1; protecting domestic firms from foreign
firms is not. Arguments for trade barriers or against trade agreements consistent with 1 either make no distinction
according to nationality or residency or, if they make a distinction, emphasize the interests of foreign countries.
For example, from the perspective of advanced countries, such arguments might state that developing countries

36
 should have a right to impose tariffs to protect their industries but that advanced countries should lower tariffs and
subsidies.

2. A view that emphasizes the value of free trade within a multilateral rules-based system, but at the same time
argues for or defends one or several of the following:
n a preference for exports over imports;
n government procurement rules favoring domestic firms (except in cases where a free trade arrangement
prohibits such discrimination);
n minimum labor and environmental standards as a way to create a “level playing field”—that is, ensuring that
weaker regulatory standards do not give foreign firms a competitive edge over domestic firms. (Attempts to
create minimum labor or environmental standards purely to limit the bargaining power of corporations—
whether foreign or domestic—vis-à-vis workers and to protect workers’ rights or the environment are not
regarded as nationalist in this scale; see 1.)

3. Ambivalence about free trade for nationalist reasons:

n Trade and international integration are viewed as good only if they create jobs and increase incomes at
home. Imports are viewed with skepticism and are unambiguously supported only in the case of important
intermediate inputs.
n New trade liberalization and bilateral/plurilateral trade agreements are opposed when viewed as unlikely to
create net job or broad-based income gains for domestic workers and/or improve the trade balance, or if
they excessively restrict national sovereignty.
n Persistent trade deficits and/or domestic job loss as a result of import competition are viewed as indicative
of “unfair” trade practices that could justify trade policy actions.
n There is a view that negative domestic outcomes associated with free trade can and should be addressed
through the vigorous use policy instruments such as tariffs, nontariff regulatory requirements, or the nego-
tiation, or renegotiation, of free trade agreements (FTAs), but still within the rules of the multilateral
system.

In spite of this ambivalence, the multilateral trading system continues to be viewed as having more benefits
than costs because it prevents destructive trade wars. There is no intention to return to unilateralism or a purely
“transactional” approach to trade policy.

4. Aggressive use of trade policies to increase exports and/or reduce imports, without, however, aiming at autarky
in the long run. The motivation for this can differ but typically includes (a) mercantilist ideas and (b) the
creation or defense of domestic productive capacity, in some combination.
a. The objective of trade policy is to achieve a high market share in international export markets, both because
this creates jobs at home and because it is viewed as conducive to long-run growth and developments

37
 (learning by doing, leading to high domestic productivity and higher living standards). To achieve this,
any combination of trade tools could be legitimate—tariffs, quotas or administrative/regulatory measures
to protect domestic markets, subsidized credit, direct subsidies, export subsidies, foreign exchange inter-
vention, and trade deals to open foreign markets and level the competitive playing field. The use of these
policies is constrained only by the threat of protectionism or related tools on the other side that might defeat
the objective of promoting exports.
b. The primary objective of trade policy is to create or defend domestic productive capacity and protect
domestic jobs, particularly in industrial sectors. Exports are viewed as desirable but increased market share
abroad is not an intrinsic aim. Neither is autarky. Imports are fine so long as they do not threaten domestic
jobs or preferred industries and do not sap valuable foreign exchange needed to import inputs essential for
the preferred industries. As long as domestic objectives are met and preferred sectors are in external surplus
or at least balance, the level of trade is a secondary concern (neither good nor bad). Protectionism is a
legitimate tool but, unlike in 5, it is limited to preferred industries and generally sold as either temporary
(the infant industry argument) or defensive (it offsets the unfair trade practices of others and is needed as
long as these unfair practices are in effect). Protectionism can be in the form of tariffs, quotas or administra-
tive/regulatory measures to protect domestic markets, subsidized credit, direct subsidies, export subsidies,
foreign exchange intervention, or restrictions on outward FDI.
Examples: Japan from the 1950s to the 1970s, South Korea in the 1960s and 1970s, Argentina under the
Kirchners, Brazil under Dilma Rousseff

5. The economy is closed to imports, using high tariffs, tight quotas, or outright prohibition, except for prod-
ucts that are deemed necessary either for domestic consumption or as intermediate inputs and that cannot be
produced at home. This is viewed as a desirable—or at least necessary—long-run state of affairs, not just a
temporary evil. The motivation for this is some combination of
n a more permanent version of the ideas underlying 4b: to build domestic industry in an environment where
(1) domestic industry is not competitive abroad nor expected to become competitive anytime soon or (2)
outside trade practices are viewed as unfair without much hope for improvement, and/or
n an effort to achieve autarky to make the country immune to trade disruptions associated with wars or trade
embargos imposed by hostile foreign powers.
Examples: Nazi Germany, Fascist Italy, Francoist Spain during the 1940s and 1950s, Brazil from the
Vargas administrations until the 1980s

FDI Protectionism
This scale describes restrictions with respect to FDI. As with the trade policy scale, each point on the scale describes the
extent and manner in which restrictions are used. The specific motivation for these restrictions may differ, however.
We focus on three: protecting a technological lead (for example, with respect to FDI motivated by the desire to acquire
domestic technology), creating know-how or jobs in the domestic country, or simply maintaining national control.

38
1. There are no restrictions or regulatory impediments to FDI except for narrowly defined national security
reasons (applying only to access to military technology, the protection of military secrets, and the protection
of critical infrastructure). Any such restrictions are imposed by an independent agency. Restrictions apply
only to majority control. There is no formal screening or mandatory notification process.

2. There is either formal screening or a mandatory notification process for FDI, and either:

n an attempt to protect a technological lead: The national security test described in 1 above is widened with
the aim of preventing majority control of firms operating in “strategic sectors” whose technologies or
products are potentially relevant for defense, even if they have much broader applications (e.g., artificial
intelligence). Minority shares continue to be fine in these sectors; or

n an attempt to catch up technologically: To promote domestic development, or acquire technology or
know-how from foreign firms, the country imposes restrictions on the employment of foreign personnel
and/or a minimum local content requirement with respect to inputs (e.g., foreign firms must purchase
at least 10 percent of inputs in the country). Employment restrictions are either time bound or subject
to an economic needs test.

3. One of the following:

n Attempt to protect a technological lead: There is no majority ownership in sectors in which the country
wishes to protect its technological know-how for commercial purposes (i.e., even when there are no links to
defense); OR
n Attempt to catch up technologically: As in 2, but with unconditional or more stringent requirements (e.g.,
minimum local employment that is not time bound or needs tested, local content requirement exceeds 50
percent); OR
n Desire to maintain national control of the economy: Restrictions on majority ownership of either large compa-
nies or companies in specific sectors, even without links to defense.

4. EITHER:
n Attempt to protect technological lead: No majority or minority ownership in sectors in which the country
wishes to protect its technological know-how, for any purpose; OR
n Desire to maintain national control of the economy: Broad restrictions on majority ownership (i.e., even of
small companies, and/or in many sectors); and/or restrictions on minority ownership of large companies or
companies in specific sectors.

5. Foreign direct investment is not welcome except in specific areas of specific national interest (“positive list”).
Existing foreign firms outside these sectors are nationalized.
Example: Venezuela under Hugo Chávez

39
Immigration Restrictions
This scale is organized in terms of both (increasingly nationalist) motivations for restricting immigration and the
restrictions that typically follow from these motivations. The motivations include minimal protections of health and
public safety, avoidance of congestion of public services, selection based on expected contribution to the national
economy, selection based on expected ability to assimilate, protection of domestic jobs, and the desire to maintain society
as ethnically and culturally homogeneous as possible.
1. Country is generally open to immigration. Some immigration restrictions may apply, but they are justified
solely by concerns about public health (e.g., no immigrants with contagious diseases), safety and security
(no criminals), or the desire to avoid congestion and maintain the ability to integrate immigrants, which
may justify limits or quotas. There is no attempt to select immigrants on the basis of economic value or
cultural fit, and limits are never applied to those seeking refugee or asylum status. Attempts to enforce
immigration restrictions are (obviously) consistent with 1. To the extent that there is still illegal immigra-
tion, there is a path for illegal immigrants to apply for legal status.
2. In addition to criteria/mechanisms to facilitate integration and avoid congestion, immigration is regulated
by criteria that aim to select immigrants expected to make a positive economic contribution to society and/
or assimilate easily. Such criteria could include a points-based assessment, requiring an applicant to have
a job offer, or requiring family ties. This category is consistent with concerns about and promises to crack
down on illegal immigration, but the means of doing so consists only in the better or faster enforcement of
existing laws.
Examples: Canada, Australia, some EU countries, the United States after 1917 but prior to Trump

3. In additon to concerns about the ability to assimilate and positively contribute to society, immigration
criteria reflect a concern that immigrants will take away jobs from residents or threaten public safety. As a
result,
n there is a cap on the total number of immigrants (for all categories, including refugees) or the total
number of asylum seekers/refugees that can be accepted in a given year (note: a cap on “economic
migrants” may not yield a 3 rating if it can be justified by the desire to avoid congestion; see 1); or
n high standards are imposed to qualify for immigrant/refugee/work-visa status, including an economic
needs test: employers must prove that they cannot find a native for the job; or

In addition to better enforcement, there are calls to rethink the laws themselves and/or think about deter-
ring illegal immigrants (e.g., through limited access to public services) or expanding enforcement measures,
including better securing of the borders with physical means (fences, walls, large increases in border patrols).
4. In addition to reflecting the concerns described above, immigration criteria place substantial value on main-
taining the nation’s ethnic and/or cultural homogeneity. As a result, immigration is significantly restricted,
but tradeoffs between ethnic/cultural homogeneity and expected economic benefits are allowed. Individuals
who are viewed as ethnically or culturally fitting face few restrictions; all others are granted work visas only

40
 temporarily (seasonal workers) or when their skills are in short supply. In addition to the stance on illegal
immigration as in 3 above, there are calls for deporting undocumented migrants.
5. With few exceptions, immigration is prohibited unless immigrants/refugees share the ethnic background of
or have family ties with the ruling elites. Same stance on illegal immigration as in 3 above.

Rejection of Multilateral Constraints

International organizations and fora such as the United Nations, WTO, IMF, Financial Stability Board, or Basel
Committee on Banking Supervision aim to solve international collective action problems and/or set international
standards by exposing a country to international influence and sometimes by imposing hard constraints. Efforts may
involve attempts at influencing, and sometime constraining, national policy. This scale describes how a government
resolves tensions between narrowly defined national interests and potential interest in supporting an organization that
can solve collective problems.
1. Country is or wishes to be a member of all international organizations whose mission it supports. In its work
within an organization, it generally puts the purpose of the organization ahead of narrowly defined national
interest. When the organization adopts a policy or view that is viewed as conflicting with the national
interest, the country nonetheless implements the policy. When a country wishes to leave or not be part of
a specific multilateral organization, it is not because the organization impinges on the country’s sovereignty
but because the country fundamentally disagrees with the mission of the organization, while it still sees the
need for multilateralism in other policy dimensions.
2. Country is or wishes to be a member of virtually all international organizations except those whose gover-
nance and mission are viewed as undesirable (rather than because the country does not wish to lose sover-
eignty). In its work within this organization, the country generally puts narrowly defined national interest
ahead of the purpose of the organization. When the organization adopts a policy or view that is viewed as
conflicting with the national interest, the country may drag its feet and put up resistance within the rules of
the organization but not go so far as to break those rules.
3. Country is or wishes to be a member of all international organizations whose mission it supports. However,
when the organization adopts a policy or view that is viewed as conflicting with the national interest, the
country does not comply with this policy, even if this involves breaking the rules of membership, while
attempting to minimize the fallout for the credibility of the organization.
4. Country is a member of some international organizations but refuses to join (or wishes to withdraw from)
others, even when it is not ideologically opposed to them, on the grounds that they would constrain national
policymaking or national courts. With respect to the organizations in which it is a member, the country
refuses to comply with a decision or policy adopted by the organization that is viewed as conflicting with the
national interest, even if doing so might threaten the organization’s credibility and ultimately its existence.
5. Country is not a member of any international organization that could meaningfully constrain national
policies. It joins or remains in an international organization only if it effectively runs it or for propaganda
purposes, so long as membership poses no constraints on national policies.

41
DEVIATIONS FROM CONSENSUS CAPITALISM BASED ON VARIOUS MOTIVES—
INCLUDING NATIONALISM
Trade Restrictions
This scale describes a spectrum of policies between free trade subject to the rules of the multilateral system and autarky.
The motives for why countries or parties prefer points above 1 on this scale may include nationalism (strengthening
domestic jobs/firms/economic development at the expense of foreign jobs/firms) and other motives (protecting the interests
of consumers and workers relative to the interests of business, small firms relative to multinationals), or noneconomic
(e.g., environmental) relative to economic (e.g., growth) objectives. The former maps onto what is commonly known as
protectionism, whereas the latter does not.
Because deviations from the liberal baseline policy prescription could be due to both protectionist and nonprotec-
tionist motives, a party receiving a 1 or 2 rating under the trade protectionism scale might still receive a higher rating
(2 or 3) under the trade restrictions scale (this would typically be the case for a “progressive” party that does not like
free trade but not because it wants to protect national producers). However, protectionist motives tend to be the main
driver of large deviations (points 4 and 5 on the scale).
1. Free trade subject to the rules of the multilateral trading system:
n There is no stated preference of exports over imports: trade is viewed as good, not just exports.
n There are no tariffs, quotas, or export subsidies (except for purely defensive safeguards or countervailing
duties that are justified within multinational trading rules).
n Export promotion is fine (only) as long as it takes the form of lowering information costs and helping firms
navigate foreign entry requirements, rather than financial incentives.
n Government procurement does not discriminate against foreign firms.
n Any nontariff trade barriers are the consequence of regulation that is justified by public policy aims extra-
neous to trade (such as environmental or health standards) rather than the desire to keep out foreign firms.
It must be applied in the same way to domestic and foreign firms.
n Trade negotiations and agreements (whether multilateral, plurilateral, or bilateral) aim to reduce regulatory
barriers to trade, including by aligning or coordinating regulation. The motivation for the latter is to lower
trade barriers as much as possible while maintaining domestic regulatory preferences, reflected in consumer
protection, labor standards, or environmental standards. Importantly, the motivation for these standards is
neither to help workers and the environment in other countries nor to “level the playing field” (i.e., to make
other countries lose the competitive edge afforded by cheaper labor and lower regulatory cost); any of these
motives would warrant a 2 rating (see 2c).

2. Like 1, except that one or several of the following apply:

a. A preference for exports over imports;

42
b. Government procurement rules favor domestic firms (except in cases where a free trade arrangement
prohibits such discrimination);
c. Bilateral and multilateral trade arrangements carry provisions that embody labor and environmental stan-
dards beyond what can be justified by merely attempting to reduce nontariff trade barriers, which may reflect
either nationalist motives (protecting domestic producers and workers) or progressive motives (protecting
the environment, protecting workers everywhere, or limiting the power of multinational corporations).
Examples: The European Union versus “third countries,” the United States during most of the postwar period

3. Ambivalence about free trade

n While “fair” trade is viewed as good, there are concerns that trade could lead to domestic job loss (nationalist
perspective) or favor either business at the expense of workers or large corporations at the expense of small
ones (progressive perspective).
n New trade liberalization and bi-/plurilateral trade agreements are opposed when these are viewed as:

n unlikely to create net gains in jobs or income for domestic workers (nationalist perspective),
n conducive to a regulatory “race to the bottom” (progressive perspective), or
n implying excessive restrictions on national sovereignty (in particular, with respect to setting domestic
regulatory standards and undercutting or bypassing domestic courts) (this could have both nationalist
of progressive motives).
n There is a view that negative outcomes associated with free trade can and should be addressed through policy
instruments such as safeguards, countervailing duties, nontariff regulatory requirements, or the negotiation,
or renegotiation, of FTAs that are viewed as inadequate from either a nationalist or progressive perspective,
but still within the rules of the multilateral system.
In spite of this ambivalence, the view remains that the multilateral trading system has more benefits than
costs because it prevents destructive trade wars. There is no intention to return to unilateralism or a purely
“transactional” approach to trade policy.

4. Trade policies are aggressively used to increase exports and/or reduce imports, but without aiming at autarky
in the long run. The motivation for this stance can differ but typically includes (a) mercantilist ideas and (b) the
creation or defense of domestic productive capacity, in some combination.
a. The objective of trade policy is to achieve a high market share in international export markets, both because
this creates jobs at home and because it is viewed as conducive to long-run growth and developments
(learning by doing, leading to high domestic productivity and high living standards). To achieve this, any
combination of trade tools could be legitimate—tariffs, quotas or administrative/regulatory measures to
protect domestic markets, subsidized credit, direct subsidies, export subsidies, foreign exchange interven-
tion, and trade deals to open foreign markets and level the competitive playing field. The use of these

43
 policies is constrained only by the threat of protectionism or related tools on the other side that might defeat
the objective of promoting exports.
b. The primary objective of trade policy is to create or defend domestic productive capacity and protect
domestic jobs, particularly in industrial sectors. Exports are viewed as desirable but increased market share
abroad is not an intrinsic aim. Neither is autarky. Imports are fine so long as they do not threaten domestic
jobs or preferred industries nor sap valuable foreign exchange needed to import inputs essential for the
preferred industries. As long as domestic objectives are met, and preferred sectors are in external surplus
or at least balance, the level of trade is a secondary concern (neither good nor bad). Protectionism is a
legitimate tool, but unlike in 5, it is limited to preferred industries and generally sold as either temporary
(the infant industry argument) or defensive (to offset the unfair trade practices of others and needed so long
as these unfair practices are in effect). Protectionism can be in the form of tariffs, quotas or administra-
tive/regulatory measures to protect domestic markets, subsidized credit, direct subsidies, export subsidies,
foreign exchange intervention, or restrictions on outward FDI.
Examples: Japan from the 1950s to the 1970s, South Korea in the 1960s and 1970s, Argentina under the
Kirchners, Brazil under Dilma Rousseff

5. The economy is closed to imports, using high tariffs, tight quotas, or outright prohibition, except for products
that are deemed necessary for domestic consumption or as intermediate inputs and that cannot be produced at
home. This is viewed as a desirable—or at least necessary—long-run state of affairs, not just a temporary evil.
The motivation for this is some combination of
n a more permanent version of the ideas underlying 4b: to build domestic industry in an environment where
it is not competitive abroad and not expected to become competitive anytime soon, or outside trade prac-
tices are viewed as unfair without much hope for improvement;
n efforts to achieve autarky to make the country immune to trade disruptions associated with wars or trade
embargos imposed by hostile foreign powers.
Examples: Nazi Germany, Fascist Italy, Francoist Spain during the 1940s and 1950s, Brazil from the
Vargas administration until the 1980s

Competition Policy (Tolerance of Market Power)

This scale describes the continuum between the view that competition between firms is all-important, regardless of
nationality, and the view that national policy aims—such as support for preferential sectors, building national cham-
pions that can compete globally, or state control over strategic industries—can justify limiting or even completely
abandoning competition between firms.
1. Competition law makes cartels illegal and explicitly aims to prevent abuse of economic power (market
dominance) by requiring approval for mergers and monitoring other potentially anticompetitive behavior.
An independent competition authority effectively enforces this law.

44
 Examples: best practices in industrial countries—the United States after passage of the Sherman Act (1890)
and dismantling of trusts in the 1920s; Germany after World War II; EU competition law and institutions

2. Like 1 but with weak or mixed enforcement, either because competition authority is weak or lacks indepen-
dence or because the authority (or the party program or the government, if that’s what is being rated) takes
a permissive view on market dominance for nationalist reasons—for example, based on the argument that
national champions are good for competing in international markets.
3. Either (1) there is no effective competition law and/or no effective competition authority, but the state
stops short of encouraging cartels or national champions, or (2) competition law/enforcement exists and is
enforced for some areas of the economy but is suspended in others to allow cartels or national champions.
4. There is no effective competition law and/or competition authority. In addition, the state encourages cartels
or oligopolies in some sectors.
Examples: Japan in the 1950s and 1960s (keiretsu), Korea in the 1970s (chaebols), Nazi Germany (1933–36)

5. There is no effective competition law and/or competition authority. In addition, cartels or monopolies
exist in many sectors of the economy, either run or encouraged by the state (whether in the form of state
enterprises or via state supervision of privately owned enterprises).
Examples: Nazi Germany (1937–45), Francoist Spain in the 1940s and 1950s22

Vertical Industrial Policy

This scale describes whether the state wishes to support specific sectors of the economy and what means it is willing to use.
1. There is no preference for specific sectors of the economy. Any industrial policy is “horizontal”: the state
provides public goods or services that are meant to encourage economic activity in general, including infra-
structure, education, health care, and access to finance. SME credit takes the form of revolving and self-
sustained credit lines. R&D subsidies or tax breaks, and grants or subsidized credit for startups, are also
consistent with this category, provided they are sector-neutral and subsidies are provided competitively. So
are advisory or “extension” services and public-private consultation and coordination efforts (for example,
with respect to infrastructure, education, and public services).
2. Some policies promoting specific sectors or companies exist in the form of subsidies, deregulation, tax
breaks, or subsidized credit. Typically, this category includes limited subsidies to agriculture and/or specific
industrial sectors (such as the airline industry) that are viewed as worthy of support for national prestige or
technological development purposes.
Example: Some EU countries

22. It is unclear whether Italy during Fascism also belongs here or whether it is in category 4—the Italians invented
the ideas but may have taken them less seriously than the Germans or Spaniards.

45
3. Use of vertical policies such as sector-, firm-, or goods-specific subsidies (including export subsidies), tax
breaks, preferential access to credit, selective price controls, and trade protection to benefit specific sectors or
firms—without, however, trying to set targets for the development of that sector, and without nationaliza-
tion or a policy of partial state ownership of firms considered “strategic.” The state tilts the playing field,
but leaves it to the sector to develop.
Example: Argentina under Néstor Kirchner

4. EITHER:
a. Like 3 (i.e., a stated preference for and promotion of specific sectors and/or firms) but with the aim of
achieving predefined developmental targets, such as growing a sector or firm to a certain size or achieving
a specific investment volume in a particular area. The targets are typically laid out in a four- or five-year
development plan.
Examples: Japan in the 1950s–60s, Korea in the 1960s and 1970s

OR
b. Like 3 (and unlike 4a without the planning element), but enhanced by additional instruments that go
beyond taxes, subsidies, and preferential access to credit, to include nationalizations, a policy of direct state
participation in firms that are considered strategic, and/or direct instructions to industry captains.
Examples: Argentina under Perón, Brazil under Vargas

5. BOTH 4a and 4b. The execution of plans involving targets for the development of specific sectors through
means that include nationalization, price controls, and direct state instructions. Full state planning is a specific
case, but this categorization is broader in that it does not require state ownership and can include indirect means
of achieving production or investment targets, such as preferential access to credit and/or special tax or subsidy
regimes, so long as these indirect means allow the state to steer most of the economy.
Examples: Nazi Germany after 1936, Francoist Spain in the 1940s and 1950s

Macroeconomic Populism
This scale is defined in terms of the aims of macroeconomic policy—specifically, the extent to which stability objectives
are subordinated to other objectives that may conflict with stability; willingness to take macroeconomic risks.
1. Monetary and fiscal policies are constrained by the desire to maintain economic stability: smoothing reces-
sions and booms, maintaining low and stable inflation, and preventing financial or fiscal crises. Monetary
policy is executed by an independent central bank. The government is committed to fiscal sustainability
(this may, but does not have to, include the use of fiscal rules). This does not preclude increases in public
debt, e.g., for fiscal stimulus or public investment purposes, so long as the debt remains clearly sustainable.
Fiscally relevant forecasts are either undertaken or monitored by independent institutions.
Examples: the United States during most of its postwar history, several EU countries today

46
2. Without openly questioning the premise that monetary and fiscal policy should be stability-oriented, or
that the central bank should be independent, the government occasionally uses fiscal policy, or encourages
the central bank to use monetary policy, to meet other aims even when this creates stability risks (e.g., by
pushing inflation beyond the central bank’s policy objective or by creating fiscal risks).
3. Monetary and fiscal policies aim to balance stability objectives with other objectives that may conflict with
macroeconomic and financial stability, such as supplying credit to preferred sectors, achieving redistribu-
tion through deficit-financed social spending, or reducing the tax burden (without equivalent spending
cuts).
4. Monetary and fiscal policies are primarily devoted to objectives that may conflict with macroeconomic and
financial stability. However, the government still cares about macro stability (just a lot less than it cares for
the other objectives) and acknowledges that there could be a tradeoff: for example, catastrophic macroeco-
nomic outcomes (such as a hyperinflation) may lead governments to refocus on stability to the detriment of
objectives that would normally be preferred. Budget constraints are ignored until the economy is close to or
in a crisis.
5. Monetary and fiscal policies are subordinated to objectives other than sustainability or stabilization. Unlike
4, there is no tradeoff: The government will not compromise with respect to its preferred objectives, no
matter what happens to macro stability, because it either (1) cares about macro stability only in a lexico-
graphic sense (or not at all), or (2) is convinced that there is never a tradeoff. Budget constraints are always
ignored so long as the primary objectives are not met.

47
APPENDIX B
MEASURING DISAGREEMENT ACROSS RATERS
The coding of party manifestos was performed by the two authors and two research analysts (RAs). Each party
manifesto was separately rated by one of the authors and one of the RAs. Differences in coding were subse-
quently resolved and final grades assigned. To gauge the degree of initial agreement among raters (interrater
reliability), Cohen’s kappa coefficients were computed based on the original scores for each pair of raters. Results
are reported in table B1.
Based on the standard interpretation of kappa coefficients (McHugh 2012), the reported coefficients indi-
cate moderate agreement among raters for pairs Author 1–RA 1, Author 2–RA 1, and Author 2–RA 2 and
substantial agreement for pair Author 1–RA 2.
An alternative measure for describing agreement between raters is to compute correlation coefficients. Unlike
the kappa statistic, this requires dropping observations in which one or both raters viewed the information in the
party platforms as insufficient to come up with a rating (“not enough information”). Both standard (Pearson)
and Spearman rank correlation coefficients were between 0.77 and 0.93 (table B2). The pattern of correlation
across the four pairs of raters is more or less the same as that of Cohen’s kappa coefficients.

Table B1
Cohen’s kappa coefficients
RA 1 RA 2
Author 1 0.544 0.658
Author 2 0.520 0.532
Note: The formula used to compute Cohen’s kappa coefficients
is κ = (po − pe)/(1 − pe), where po is the relative observed agree-
ment and pe is the hypothetical probability of chance agree-
ment between raters. All original scores, including “not enough
Information,” were included in the calculations.
Source: Authors’ calculations based on raters’ original grades.

  Table B2
Correlation coefficients
Table B2. Correlation coefficients 
Pearson correlation  Spearman correlation 
      Pearson correlation coefficients
coefficients  coefficients  Spearman correlation coefficients
      RA 1  RA 1 RA 2  RARA 1 
2 RA 2  RA 1 RA 2
Author 1 
Author 1    0.840  0.840 0.926  0.854 
0.926 0.923  0.854 0.923
Author 2     0.772  0.775  0.791  0.791 
Author
  2 0.772 0.775 0.791 0.791
 
Note:
  The Pearson correlation coefficient is the standard linear correlation computed using the formula:
𝑟𝑟𝑟𝑟�� � � ∑�����𝑥𝑥𝑥𝑥� � 𝑥𝑥𝑥𝑥̅ ��𝑦𝑦𝑦𝑦� � 𝑦𝑦𝑦𝑦�����∑�����𝑥𝑥𝑥𝑥� � 𝑥𝑥𝑥𝑥̅ �� �∑�����𝑦𝑦𝑦𝑦� � 𝑦𝑦𝑦𝑦��� �.  
. The Spearman correlation is a correlation coefficient computed
 
using variables’ ranks.
  Source: Authors’ calculations based on raters’ original grades.

48
 Table C1
Ratings by country, category, and period
Immigration
Trade protectionism FDI protectionism restrictions Antimultilateralism Trade restrictions
Country Party Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC
APPENDIX C

Advanced economies
Australian Labor Party 2 2 — — 2 3 1 1 2 2
Australia Liberal Party of Australia 1 1 — — 2 3 1 1 1 1
Australian Greens 1 2 — — 2 2 1 1 2 3
Liberal Party 2 2 — — 2 2 1 1 2 2
Canada Conservative Party 2 2 2 2 2 2 1 1 2 2
New Democratic Party 2 2 3 1 1 1 1 1 2 2
Socialist Party (PS) 3 3 — — 2 2 1 1 3 3
Les Républicains (UMP) 3 2 — — 3 3 2 2 3 2
France Front National 4 4 3 3 5 5 4 4 4 4
La France Insoumise — 3 — — — 1 — 2 — 3
La République En Marche! — 2 — 3 — 2 — 1 — 2
Christian Democratic Union/
2 2 — — 3 2 2 1 2 2
Christian Social Union (CDU/CSU)
Social Democratic Party (SPD) 1 1 — — 1 2 1 1 1 2

49
Germany
Alliance 90/The Greens 1 1 — — 1 1 1 1 2 3
Free Democratic Party (FDP) 1 1 — — 1 2 1 1 1 1
Alliance for Germany (AfD) — 2 — 3 — 4 — 4 — 3
Center Right Coalition 3 — — — 3 4 1 3 3 —
Italy Center Left Coalition 2 2 1 — 1 2 1 1 2 2
Movimento Cinque Stelle — 3 — 1 — 2 — 4 — 3
Liberal Democratic Party 2 2 — — — 3 1 1 2 2
Japan Constitutional Democratic Party 2 2 — — — — 1 1 2 2
Kibō no Tō (Party of Hope) — 1 — — — — — 1 — 1
Democratic Party — — 4 — — — 1 1 — —
South Korea Saenuri Party (Hannara) — — — — — — 1 1 — —
People’s Party — 1 — — — — — 1 — 1
Labour Party 1 2 — 1 2 2 1 2 1 2
Conservative Party 1 1 — 1 3 3 2 2 1 1
United Kingdom
Liberal Democrats 1 1 — — 2 2 1 1 1 1
UK Independence Party 1 2 — — 3 3 4 4 1 2
Democratic Party 2 3 — — 1 1 1 1 2 3
United States
Republican Party 2 3 — — 2 3 4 4 2 3
Table continues
 Table C1 (continued)
Ratings by country, category, and period
Tolerance of Macroeconomic Party vote share
Industrial policy concentration populism (percent)
Country Party Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC
Advanced economies
Australian Labor Party 3 3 1 1 1 1 43.4 34.7
Australia Liberal Party of Australia 1 1 1 1 1 1 36.6 28.7
Australian Greens 1 2 1 1 1 — 7.8 10.2
Liberal Party 2 2 1 1 1 1 30.2 39.5
Canada Conservative Party 2 2 1 1 1 1 36.3 31.9
New Democratic Party 2 3 1 1 1 1 17.5 19.7
Socialist Party (PS) 3 2 — — 3 3 24.7 7.4
Les Républicains (UMP) 3 2 — 3 2 1 39.5 15.8
France Front National 2 3 3 — 3 3 4.3 13.2
La France Insoumise — 4 — — — 4 — 11.0
La République En Marche! — 2 — — — 1 — 28.2
Christian Democratic Union/
1 2 — 2 1 1 35.2 32.9
Christian Social Union (CDU/CSU)
Social Democratic Party (SPD) 1 2 — 1 1 1 34.2 20.5

19
50
Germany
Alliance 90/The Greens 1 1 1 1 1 2 8.1 8.9
Free Democratic Party (FDP) 1 1 1 1 1 2 9.8 10.7
Alliance for Germany (AfD) — 1 — 1 — 3 — 12.6
Center Right Coalition — — 1 — 2 4 49.7 41.5
Italy Center Left Coalition 3 2 1 1 1 1 49.8 17.0
Movimento Cinque Stelle — 4 — 3 — 5 — 32.7
Liberal Democratic Party 3 3 1 1 1 1 38.2 33.3
Japan Constitutional Democratic Party 2 2 1 — 1 1 31.0 19.9
Kibō no Tō (Party of Hope) — 2 — — — 1 — 17.4
Democratic Party 4 1 1 1 — 1 25.2 25.5
South Korea Saenuri Party (Hannara) 2 3 — 1 1 1 37.5 33.5
People’s Party — 3 — 1 — 1 — 26.7
Labour Party 2 3 1 — 1 2 35.2 40.0
Conservative Party 1 2 1 1 1 2 32.4 42.4
United Kingdom
Liberal Democrats 1 2 1 1 1 1 22.0 7.4
UK Independence Party 2 2 — — 4 2 2.2 1.8
Democratic Party 2 2 1 1 1 1 46.5 47.3
United States
Republican Party 1 1 — 1 3 1 49.3 48.3
Table continues
 Table C1 (continued)
Ratings by country, category, and period
Immigration
Trade protectionism FDI protectionism restrictions Antimultilateralism Trade restrictions
Country Party Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC
Emerging-market economies
Frente para la Victoria 3 4 — 4 — — 3 4 3 4
Argentina Cambiemos (Coalición Cívica) 2 2 — 1 — — 1 1 2 2
United for a New Alternative (UNA) — 1 — 1 — — — 1 — 1
Workers’ Party (PT) 3 4 1 3 — 1 2 2 3 4
Brazil Social Liberal Party (PSL) — 2 — 1 — — — — — 2
Social Democracy Party (PSDB) 2 — — 1 — 2 1 1 2 —
China Communist Party of China 4 3 3 2 — — 3 2 4 3
Bharatiya Janata Party 4 4 3 4 — — 2 1 4 4
India
Indian National Congress 2 4 2 1 — — — 1 2 4
Party of the Functional Groups
— 3 — — — — 1 3 — 3
(Golkar)
Indonesia Great Indonesia Movement Party
— 4 — 4 — — — 4 — 4
(Gerindra)

51
20
Democratic Party (PD) — — — — — — — — — —
Institutional Revolutionary Party
(PRI) (Alliance for Mexico (PRI– 2 1 — 1 2 1 2 1 2 1
led))
Mexico MORENA (Democratic Revolutionary
3 3 — 4 1 — 1 — 3 3
Party (PRD))
Por México al Frente (PAN–led)
3 1 1 — — — 1 1 3 1
(National Action Party (PAN))
United Russia 2 4 — 3 — — 2 4 2 4
Communist Party 4 — — — — — 5 — 4 —
Russia
Liberal Democratic Party of
4 4 — 4 3 — 3 5 4 4
Russia
African National Congress — 3 — — — — 1 1 — 3
South Africa
Democratic Alliance 1 — 1 — 2 3 2 — 1 —
AK Parti 2 2 1 1 — — 3 3 2 2
Republican People’s Party 2 — 1 1 — — 1 1 2 —
Turkey
Nationalist Movement 2 3 1 1 — — 1 1 2 3
People’s Democratic Party — — — — — — — 4 — —
Table continues
 Table C1 (continued)
Ratings by country, category, and period
Tolerance of Macroeconomic Party vote share
Industrial policy concentration populism (percent)
Country Party Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC Pre-GFC Post-GFC
Emerging-market economies
Frente para la Victoria 3 4 1 — 4 5 31.1 37.4
Argentina Cambiemos (Coalición Cívica) 1 2 — — 1 2 14.9 34.5
United for a New Alternative (UNA) — 1 — — — 1 — 19.9
Workers’ Party (PT) 3 4 — — 1 5 15.0 10.3
Brazil Social Liberal Party (PSL) — 4 — 1 — 1 — 11.7
Social Democracy Party (PSDB) 3 1 1 1 1 1 13.6 6.0
China Communist Party of China 4 4 3 2 3 3 100.0 100.0
Bharatiya Janata Party 4 3 — — 4 4 22.2 31.0
India
Indian National Congress 4 4 1 — 1 3 26.5 19.3
Party of the Functional Groups
1 4 1 1 1 1 21.6 14.8
(Golkar)
Indonesia Great Indonesia Movement Party
— 4 — 3 — 2 — 11.8
(Gerindra)

21
52
Democratic Party (PD) — 2 — — — 2 7.5 10.2
Institutional Revolutionary Party
(PRI) (Alliance for Mexico (PRI- 2 1 1 1 2 1 28.9 16.5
led))
Mexico MORENA (Democratic Revolutionary
5 5 2 2 4 4 29.7 37.3
Party (PRD))
Por México al Frente (PAN-led)
2 2 1 — 2 1 34.3 17.9
(National Action Party (PAN))
United Russia 3 4 2 3 — 2 64.3 54.2
Communist Party 5 5 5 5 — — 11.6 13.3
Russia
Liberal Democratic Party of
5 3 3 3 — 5 8.1 13.1
Russia
African National Congress 2 4 — — 2 4 69.7 62.2
South Africa
Democratic Alliance 1 1 1 — 1 1 12.4 22.2
AK Parti 2 2 1 1 1 1 46.6 42.6
Republican People’s Party 3 4 1 3 1 1 20.9 22.6
Turkey
Nationalist Movement 3 4 1 1 1 2 12.9 8.2
People’s Democratic Party — — — 4 — 4 — 11.2
Table continues
 Table C1 (continued)
Ratings by country, category, and period
Trade Immigration
protectionism FDI protectionism restrictions Antimultilateralism Trade restrictions
Post-GFC Post-GFC Post-GFC Post-GFC Post-GFC
Policies/documents
President Donald Trump 4 2 4 4 4
Germany’s National Industrial
2 3 — 1 2
Strategy 2030
Made in China 2025 4 3 — — 4
Tolerance of Macroeconomic

22
53
Industrial policy concentration populism
Post-GFC Post-GFC Post-GFC
President Donald Trump 3 — 3
Germany’s National Industrial
4 3 —
Strategy 2030
Made in China 2025 4 3 —
Note: Table shows the ratings given to each party, for each policy category, before and after the global financial crisis
(GFC). Party vote shares represent the share of votes gained by parties in the general elections identified in table 1. In each
category, party-level weights used in calculations of weighted country-level averages were computed by dividing a vote
share of a party with an assigned rating by the sum of vote shares of parties with available ratings (parties with no rating
were excluded from weighting). “—” means not enough information.
Source: Authors’ coding of party platforms identified in table 1.
APPENDIX D
APPENDIX D

Table D1
RILE index and party classifications
Right/Left Populism
Country Party RILE index classification classification
Advanced economies      
Australian Labor Party –34.397 Left Nonpopulist
Australia Liberal Party of Australia 12.240 Right Nonpopulist
Australian Greens –23.218 — Nonpopulist
Liberal Party –20.683 Left Nonpopulist
Canada Conservative Party 29.016 Right Nonpopulist
New Democratic Party –9.674 Left Nonpopulist
Socialist Party (PS) –28.947 Left Nonpopulist
Les Républicains (UMP) 13.619 Right Nonpopulist
France Front National 5.814 Right Populist
La France Insoumise –30.019 Left Populist
La République En Marche! 0.000 — Nonpopulist
Christian Democratic Union/
2.757 Right Nonpopulist
Christian Social Union (CDU/CSU)
Social Democratic Party (SPD) –21.437 Left Nonpopulist
Germany
Alliance 90/The Greens –21.058 Left Nonpopulist
Free Democratic Party (FDP) 0.578 Right Nonpopulist
Alliance for Germany (AfD) 17.430 Right Populist
Center Right Coalition 4.670 Right Populist
Italy Center Left Coalition –7.505 Left Nonpopulist
Movimento Cinque Stelle –49.032 — Populist
Liberal Democratic Party –2.036 Right Nonpopulist
Japan Constitutional Democratic Party –28.205 Left Nonpopulist
Kibō no Tō (Party of Hope) — — Nonpopulist
Democratic Party –42.508 — Nonpopulist
South Korea Saenuri Party (Hannara) –20.069 — Nonpopulist
People’s Party –35.328 — Nonpopulist
Labour Party –27.560 Left Nonpopulist
Conservative Party –2.607 Right Nonpopulist
United Kingdom
Liberal Democrats –21.751 Right Nonpopulist
UK Independence Party 2.037 Right Populist
Democratic Party –20.578 Left Nonpopulist
United States
Republican Party 32.969 Right Nonpopulist
Table continues

23

54
Table D1 (continued)
RILE index and party classifications
Right/Left Populism
Country Party RILE index classification classification
Emerging-market economies      
Frente para la Victoria — Left Nonpopulist
Argentina Cambiemos (Coalición Cívica) — Right Nonpopulist
United for a New Alternative (UNA) — — Nonpopulist
Workers’ Party (PT) — Left Nonpopulist
Brazil Social Liberal Party (PSL) — — Nonpopulist
Social Democracy Party (PSDB) — Left Nonpopulist
China Communist Party of China — — Nonpopulist
Bharatiya Janata Party — Right Populist
India
Indian National Congress — Left Nonpopulist
Party of the Functional Groups
— — Nonpopulist
(Golkar)
Indonesia Great Indonesia Movement Party
— — Nonpopulist
(Gerindra)
Democratic Party (PD) — — Nonpopulist
Institutional Revolutionary Party
(PRI) (Alliance for Mexico (PRI- 0.155 Center Nonpopulist
led))
Mexico MORENA (Democratic Revolutionary
0.000 — Nonpopulist
Party (PRD))
Por México al Frente (PAN-led)
14.416 Right Nonpopulist
(National Action Party (PAN))
United Russia 2.793 Center Populist
Communist Party –18.297 Left Nonpopulist
Russia
Liberal Democratic Party of
13.693 — Nonpopulist
Russia
African National Congress –31.926 Left Populist
South Africa
Democratic Alliance –8.120 — Nonpopulist
AK Parti –7.381 — Populist
Republican People’s Party –20.851 Left Nonpopulist
Turkey
Nationalist Movement –2.496 — Nonpopulist
People’s Democratic Party –32.163 — Nonpopulist
Policies/documents      
  President Donald Trump — Right Populist
  Germany’s National Industrial
— Right Nonpopulist
Strategy 2030
  Made in China 2025 — — —
Note: Table shows the RILE (Right–Left) index, right/left dummy variables and populism classifications for each
party. When the RILE index was not available for the elections to which the manifesto analyses in this paper
referred, the most recent RILE index available in the Manifesto Project Dataset (Volkens et al. 2018b) was used.
For Italy’s Center Left and Center Right coalitions, an average RILE index was computed by weighting the
available party RILE indices of coalition members by share of votes obtained. All data presented in the table
pertains to the post–GFC period. “—” means not enough information.
Source: See sources of table D2.

24
55
Table D2
Average changes from pre- to post-GFC and differences post-GFC in party ratings
(p values shown in parentheses, followed by the number of observations)
Correlation with changes from pre- to post-GFC
Regression 1 Regression 2 Regression 3
Left Right RILE Populist Nonpopulist
Policy category 1 2 3 4 5
0.46** –0.07 –0.01 0.60 0.13
Trade protectionism (0.016) (0.715) (0.451) (0.113) (0.359)
68 68 58 70 70
–0.25 0.33 0.01 0.33 –0.29
FDI protectionism (0.774) (0.316) (0.474) (0.305) (0.576)
18 18 12 20 20
0.33* 0.27 0.00 0.33 0.25
Immigration restrictions (0.056) (0.176) (0.778) (0.255) (0.101)
46 46 46 46 46
0.13 0.00 0.00 0.43 0.09
Antimultilateralism (0.147) (1.000) (0.397) (0.299) (0.418)
78 78 66 80 80
0.62*** –0.07 –0.01 0.60 0.20
Trade restrictions (0.002) (0.715) (0.260) (0.113) (0.188)
68 68 58 70 70
0.24 0.21 0.00 0.50 0.14
Industrial policy (0.320) (0.252) (0.791) (0.219) (0.457)
84 84 70 86 86
0.22 0.00 0.00 0.50 0.04
Tolerance of concentration (0.321) (N/A) (0.668) (0.188) (0.671)
48 48 44 50 50
0.69** –0.07 –0.02** 0.33 0.24
Macroeconomic populism (0.020) (0.814) (0.030) (0.566) (0.177)
76 76 62 78 78
Table continues

25
56
Table D2 (continued)
Average changes from pre- to post-GFC and differences post-GFC in party ratings
(p values shown in parentheses, followed by the number of observations)
Correlation with post-GFC ratings
Regression 4 Regression 5 Regression 6
Left Right RILE Populist
Policy category 6 7 8 9
0.13 –0.28 0.00 0.94***
Trade protectionism (0.731) (0.469) (0.776) (0.004)
47 47 35 48
–0.47 0.19 0.03*** 0.43
FDI protectionism (0.468) (0.737) (0.010) (0.400)
26 26 15 27
–0.33 1.07*** 0.02** 1.20**
Immigration restrictions (0.353) (0.010) (0.012) (0.028)
31 31 28 31
–0.90** –0.19 0.02 1.53***
Antimultilateralism (0.041) (0.697) (0.165) (0.000)
51 51 39 52
0.27 –0.29 0.00 0.94***
Trade restrictions (0.458) (0.461) (0.996) (0.002)
47 47 35 48
0.15 –0.62* –0.02** 0.47
Industrial policy (0.710) (0.092) (0.045) (0.210)
54 54 40 55
–0.20 –0.35 –0.01 0.39
Tolerance of concentration (0.697) (0.363) (0.532) (0.476)
36 36 31 37
0.23 –0.01 –0.01 1.44***
Macroeconomic populism (0.644) (0.988) (0.479) (0.001)
53 53 40 54
GFC = global financial crisis; FDI = foreign direct investment; N/A = not available
Note: The table presents regression coefficients that relate the measures of right/left-wing orientation and
political populism to both changes in economic nationalism from before to after the global financial crisis (GFC)
(columns 1–5) and levels of economic nationalism ratings in the post-GFC period (columns 6–9), based on the
following specifications:
1. Columns (1) and (2): coefficients b1 and b2 from the specification: r(i,t) = b0 + b1*postGFC*R(i) +
b2*postGFC*L(i) + d(i)
2. Column (3): coefficient b3 from: r(i,t) = b0 + b1*postGFC + b2*RILE(i) + b3*postGFC*RILE(i) + d(i)
3. Columns (4) and (5): coefficients b1 and b2 from r(i,t) = b0 + b1*postGFC*P(i) + b2*postGFC*NP(i) + d(i)
4. Columns (6) and (7): coefficients b1 and b2 from r(i) = b0 + b1*L(i) + b2*R(i)
5. Column (8): coefficient b1 from r(i) = b0 + b1*RILE(i)
6. Column (9): coefficient b1 from r(i) = b0 + b1*P(i)
r(i,t) denotes the rating assigned to party i in each category at time t (before or after the GFC), postGFC is a
dummy variable taking on the value one in the post-GFC period and zero otherwise, R(i), L(i), P(i), and NP(i) are
dummy variables taking on the values one for right-wing (R), left-wing (L), populist (P), and nonpopulist (NP)
parties, respectively, and zero otherwise; and d(i) is a full set of dummy variables for specific parties.
China was excluded from regressions using L and R dummies. The RILE index is an estimate of right-left positions
of parties, calculated by subtracting the proportion of party manifestos designated as “left” from the proportion
of party manifestos designated as “right.” It can theoretically range from −100 to +100 (Mölder 2016). P-values
are presented in parentheses, followed by the number of observations.
* denotes p ≤ 0.1; ** p ≤ 0.05; *** p ≤ 0.01. Robust standard errors were used.
For regressions using both L and R or both P and NP, a test was performed to see if the corresponding coeffi-
cients are statistically different from each other. Coefficients that are statistically different from each other at the
10% level (in each category) are shown in bold.
Sources: For the RILE index: Manifesto Project Dataset (Volkens et al. 2018b); for classification of parties as right or
left: Database of Political Institutions 2017 (Cruz, Keefer, and Scartascini 2018, primary source), Comparative Political
Data Set 1960–2016 (Armingeon et al. 2018, secondary source), The PopuList (Rooduijn et al. 2019, tertiary source);
for classification of parties as populist or nonpopulist: The PopuList (Rooduijn et al. 2019, primary source), Kyle and
Gultchin (2018) (secondary source). President Trump and Germany’s National Industrial Strategy 2030 were both
classified as right, and as populist and nonpopulist, respectively, based on authors’ judgment. Italy’s Center Left and
Center Right coalitions were classified as left and right, respectively, based on their coalition names.
26
57