Vietnamese - German University

Faculty of Economics and Management

Lecture Summary and Problem Set Manual

Prepared for 2017 students

Basic Macroeconomics - BMAK

A qualification course for FA and BA students at the Vietnamese - German University

By
Dr. Le Van Ha

Vietnamese - German University

Binh Duong, Vietnam

April 2020
2

ˆ This is a preliminary version which will be updated continuously. Strictly meant to be used within
FA/BA2017 BMAK Class.

ˆ Issues thoroughly discussed in the lecture slides will not be repeated in this note.
Contents

1 Macroeconomic Measurement 5
1.1 GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.1 The three approaches to measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.2 Real vs. Nominal GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Other important macroeconomic gauges: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2.1 GNP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2.2 PPP-based GDP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2 The Long Run Macroeconomy 9

2.1 The Solow Growth Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.1 The capital intensive form of the SGM . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1.2 The dynamics of labor and technology . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1.3 The dynamics of the capital stock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1.4 Some steady state results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.5 The golden rule of saving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.6 Problemset exercises on the SGM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3 The Short Run Macroeconomy 29

3.1 A Simple Theory of Household Consumption - C(Y-T,r) . . . . . . . . . . . . . . . . . . . . . 29
3.2 A Simple Theory of Firm’s Investment - I(r) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.3 Government Spending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.4 The Trade Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.4.1 The Interest Rate Parity Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.4.2 The Balance of Trade Function Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.6 The IS Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.6.1 Exercises on the IS curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.7 IS-TR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.7.1 Monetary Policy Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.7.2 The lending rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.7.3 The TR curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.7.4 The IS-TR model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.7.5 Quantitative analysis of the IS-TR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.8 Exercises on the IS-TR model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.8.1 Problem 3.14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.8.2 Problem 3.15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.8.3 Problem 3.16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.8.4 Problem 3.17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.9 The Extended IS-TR Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.9.1 The Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.9.2 Problemset Exercises on the Extended IS-TR Model . . . . . . . . . . . . . . . . . . . 53

4 The Macroeconomy in the Medium Run 59

4.1 The Supply Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.1.1 The Firm’s Pricing Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.1.2 Wage bargaining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.1.3 The Supply Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2 The Demand Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.3 The AD-AS model in the normal time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3
4 CONTENTS

4.3.1 Problem 4.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

4.4 The AD-AS Model During Deep Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.1 Problem 4.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.2 The Non-conditioned π ZLB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.4.3 Problem 4.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.4.4 Problem 4.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Chapter 1

Macroeconomic Measurement

The measurement of economic performance is the prelude to sensible economic analysis which provides us
with the understanding of how the economy works. Understanding how the economy works helps business
leaders (I mean you!) making the right decisions given the changes in major macroeconomic variables such
as unemployment, inflation, economic growth, stock market indexes, and the exchange rate of the domestic
currency against major trade partner currencies.

Good measurement of the development of main economic indicators is also of vital importance for the gov-
ernment to make good policy responses to the ailments of the economy, thus avoiding painful and prolonged
periods of economic downturns. For politicians, having a good team of economic advisers is also the key to
keep the economy’s performance at an acceptable level, thus raising their chance of being in power.

1.1 GDP
1.1.1 The three approaches to measurement
Gross domestic product, or GDP, is the market value of all final goods and services produced within a
country in a given period of time (recall your OVWL discussions). Although GDP is not a perfect measure
of economic well being1 , higher GDP is strongly correlated with higher living standards, longer longevity,
better health care, more quality education as well as a larger number of other services that people enjoy.
Three approaches to measuring GDP are: Product approach (measuring the contribution of each stage of
the production process, answering the question What is being produced); Income approach (how the profit
of the production process is divided among owners of the factors of production in the economy, answering
the question (Who gets the income generated by the production mentioned in the product approach); and
Expenditure approach answering the question Who buys the goods/services created in the production process
mentioned in the product approach. Regardless of which approached being employed, the same value of GDP
should be obtained. In practice, however, there often exist the differences between these approaches due to
the problem of statistical discrepancies.

Talking of GDP, two salient features arise: 1) In most if not all countries, GDP increases over a long
period of time (increasing trend); however, 2) the development of GDP is far from smooth, meaning that
actual GDP fluctuates around this trend. These observations give rise to the study of the economy in the
long run, and the short run, as we will proceed in this courses.

1.1.2 Real vs. Nominal GDP

In the year 1985, a lunch portion similar to that at the VGU canteen cost about 4 VND. Now you pay about
20,000 VND. In the same year 1985, the lunch would cost 0.4 USD. Now you pay roughly 0.78 USD.
X
(Nominal GDP)t = p · qit ,
| it{z } | {z }
t price of quantity of
good/service good/service
i in year t i in year t

When comparing values of nominal GDP across different years, this generally involves both changes in
quantities and prices, and thus does not allow to infer about changes in the standard of living. To allow for
such inference, real GDP adjusts for price changes and reflects the sum of the value added produced by all
1 Suffers from such limitations as excluding home production, disregarding environmental degradation, not taking income

distribution and other indicators of happiness into account.

5
6 CHAPTER 1. MACROECONOMIC MEASUREMENT

firms in the country in terms of only one set of prices, namely those of the so-called base year.

ˆ The Laspeyres approach to computing real GDP involves using initial year prices as the base-year
prices:
N
X
Laspeyres Laspeyres
(Real GDP)t = Yt = pi0 × qit , t = 0, 1, . . . , T,
i=1

implying the growth rate of real GDP:

N
P
pi0 × qit
i=1
1+ gYLaspeyres
t
= N
, t = 0, 1, . . . , T,
P
pi0 × qi,t−1
i=1

The main problem with the Laspeyres approach is that as one moves from the initial year (t = 0) to the final
year (t = T), Laspeyres-based real GDP over-states more and more the contribution to GDP of goods and
services that are becoming less and less expensive.

ˆ The Paasche approach to computing real GDP involves using final year prices as the base-year prices:
N
X
(Real GDP)Laspeyres
t = YtLaspeyres = piT × qit , t = 0, 1, . . . , T,
i=1

implying the growth rate of real GDP:

N
P
piT × qit
i=1
1 + gYLaspeyres
t
= N
, t = 0, 1, . . . , T,
P
piT × qi,t−1
i=1

The main problem with the Paasche approach is that as, starting in the final year (t = T), one moves back in
time towards the initial year (t = 0), Paasche-based real GDP over-states more and more the contribution to
GDP of goods and services for which the price is continuously rising from the initial year to the final year.

ˆ The Annual Chain Weight approach aims to overcome these (opposing) biases of the Laspeyres and
Paasche approaches and computes real GDP using the geometric average of results for the Laspeyres
and Paasche approaches, with annually updated base years:
v
u N N
u P P
u
u i=1 p i,t−1 × qit pi,t × qit
i=1
1 + gYAnnual Chain Weight
= u × ,
t
u N N
uP P
u pi,t−1 × qi,t−1 pi,t × qi,t−1
u i=1 i=1
u| {z } | {z }
t Annual Annual Paasche

Laspeyres

t = 0, 1, . . . , T.
N
X
Annual Chain Weight
(Real GDP)0 = Y0Annual Chain Weight = pi0 × qi0
i=1
and,
YtAnnual Chain Weight = (1 + gYAnnual
t
Chain Weight
) × Yt−1
Annual Chain Weight

1.2 Other important macroeconomic gauges:

1.2.1 GNP
GNP = GDP + Net Factor Income From Abroad (NFIA)

ˆ NFIA = Values of (exports - imports) of factor services

1.2. OTHER IMPORTANT MACROECONOMIC GAUGES: 7

ˆ Export of factor services:

– A Vietnamese person working in Japan on a labor export contract for 3 years who earns an annual
income of 10,000 USD.
– Viettel Group invests in Myanmar and made a profit of 10 million USD in 2017.

ˆ Import of factor services:

– Dr. Michel Toulouse, a Canadian, working at VGU as the coordinator of the Computer Science
program. His income is an example of the import of labor (a factor of production of education
service).
– Samsung Vietnam posted 5 billion USD in profit from their smart phone business in Vietnam in
2017.

1.2.2 PPP-based GDP

How to compare the levels of GDP per capita across countries, which are measured in different currencies?
This task is complicated by the fact that prices of traded goods relative to those of non-traded goods tend
to be sizably higher in "poor" countries than in "rich" countries, and that pricing of market exchange rates
is more closely related to the pricing of traded than of non-traded goods. Purchasing power parity (PPP)
exchange rates use a standardized basket of traded and non-traded goods and services across the countries
in question, estimating the market value of the basket in the rich countries currency, Vrich , using the rich
country’s currency; and the market value of the same basket in the poor country using the poor country’s
V
currency, Vpoor . The PPP-based exchange rate is computed as: EXppp = Vpoor rich
.

The PPP-based GDP is calculated using the EXppp instead of the market exchange rate. As illustrated
in the class, using market exchange rate the GDP per capita in Vietnam is about 2,200 USD. However,
using the PPP-based exchange rate, it’s about 7,000 USD. The idea of the purchasing power parity is to
ask: to live like a typical Vietnamese, how much would an American has to pay? Normally, the prices of
non-tradable goods and services in Vietnam are lower than in the US, thus PPP-based GDP per capita in
Vietnam is higher than the GDP per capita in current GDP.

The story will be reversed if we compare a country like Qatar and the USA. In Qatar, most services are
more expensive than in the USA, thus PPP-based GDP per capita for Qatar is lower than the GDP per
capita in current USD.
8 CHAPTER 1. MACROECONOMIC MEASUREMENT
Chapter 2

The Long Run Macroeconomy

Figure 2.1: Long-run Economic Growth

ˆ Figure (1.1) presents the graph of GDP per capita (measured in purchasing power parity dollar) from the
year 0001 to 2010 for some countries, including Vietnam1 . Of course with such critical VGU minds, you all
may simultaneously utter some objection: how do they measure income from such a far distance in history?
I can’t tell you. All I know is that some prestigious professors have spent all of their life devising methods to
estimate the mankind’s living standards over the last two millennia. Still, they may have made some gross
mistakes and all the data created are nothing more than the guesswork. True! It is hard to accept but most
of what we know about our own history is guesswork (and the rest is pure prejudice). So if we all for this
moment contend with what we have at hand, Figure (1.1) tells some interesting story.

Figure 2.2: Vietnam’s GDP Per Capita Since 1820

We, mankind as one, have lived most our life in the last two thousand years around the subsistence level,
i.e. we just had barely enough to eat and hardly enough to cover our body. In developed countries, the
living standard only experienced a sustainable pick up sometime at the middle of the 18th century. Yeah,
if you happen to like the history subject during the high-school time, you may also recall that mid-18th
1 The data can be downloaded from here: http://www.ggdc.net/maddison/oriindex.htm

9
10 CHAPTER 2. THE LONG RUN MACROECONOMY

century also marked the introduction of new technologies that had never existed before, namely the steam
engines, and together with its, came the steamships, steam spinning, steam trains and many other ensuing
inventions. These newly invented technologies lifted people’s lives above the subsistence level. People in
some Western countries suddenly found themselves unable of eating everything they had made. So they
saved those redundant products for future use. The future life, thus, became better thanks to the fact that
people would be able to consume both the saving now with the production of the future. As the story goes,
humankind has become richer and richer over the last two centuries.

ˆ Figure (1.2) in combination with Figure (1.1) tells the story of Vietnam. Like the rest of the world,
our ancestors spent all of their lives living at the subsistence level. Things only started to look a little bit
brighter in the 1990s. Before that, we suffered decades of wars, war against the US, war against China, and
war against ourselves. Each war was followed by a sharp drop in our parents’ and grandparents’ hard-earned
standard of living.

ˆ It’s beyond the scope of this manual, to say anything less superficial than that when it comes to the
history of economic development. However, this manual as part of the great BMAK course, will equip you
with the models so that once mastering them, you are able to tell a much deeper story in your own way.
Before we start, let’s send our special thank to Nobel Laureate Robert Solow, who put up such a powerful
analytical framework we know as the Solow Growth Model.

2.1 The Solow Growth Model

The Solow model is powerful in explaining why humankind has, over the last two centuries, and Vietnam
has, over the last three decades, experienced unprecedented improvement in our living standards. In our
world today, it seems that production and distribution of goods and services involves a lot of things: human
minds, human hands, machines, tools, equipments, factories, roads, ports, know-hows, recipes, and so on.
Solow suggested that we bundle them into three categories only, namely labor, capital and technology. These
three ingredients combine available inputs in certain way to produce all the goods and services we enjoy in
our daily life. A simple, but yet, powerfully way some summarizing this process is to define a production
function as:
Yt = F (Kt , At , Lt ) = Ktα (At Lt )1−α

where Kt , Lt , and At are capital, labor, and technological progress, respectively. At any point in time,
we have Kt units of capital, Lt of labor, and At units of know-hows. ∂K ∂Y
t
= αKt1−α (At Lt)1−α is defined
as the marginal product of capital. This is also the share of goods and services that we pay to capital
owner per unit of capital they own. Capital owners own Kt units of capital at time t, and their income is:
∂Y
Kt · ∂K t
= Kt · αKt1−α (At Lt)1−α or α · Ktα (At Lt)1−α = α · Yt .

ˆ You may wonder why we need to bundle labor and technology into one and call this composite factor as
effective units of labor. For the sake of exposition, imagine that each worker is a capsule of two things, their
labor (L), and their skills (A) that are embodied within them. So an ALRonaldo = 1000·ALCongP huong at the
market price as of 2013. As individual human beings, there is no meaning, and we have no right to compare
Cristiano Ronaldo with Nguyen Cong Phuong. However, as effective workers in the football industry the
market valued Ronaldo 1000 times more than Cong Phuong.

ˆ This is the very point I reiterated during the lectures, students are workers working so hard to increase
the A factor within themselves. If you do not read the text by Burda and Wyplosz, and other accompanying
documents, not to mention dozing over during the lectures, blame yourself for the fact that Vietnam is poorer
than Germany even though the two countries are comparable in terms of the number of human beings. Let
us dream of a day when one Vietnamese worker is valued as 15 Chinese, and we will forever escape the
agonizing feeling of living next to a giant and cruel neighbor.

ˆ Similar to the case of capital, the salary paid to an effective unit of labor is its marginal product,
∂Y
∂AL = (1−α)Ktα (At Lt)−α , and so the total wages and salaries paid to workers are At Lt·(1−α)Ktα (At Lt)−α =
(1 − α) · Ktα (At Lt)1−α = (1 − α) · Yt . In other words, if α = 0.3 and Vietnam’s GDP is USD 240 billion,
USD 80 billion will be paid to capital owners in the form of rent and depreciation, while USD 160 billion
will be paid to the workers (and their effectiveness factors) in the form of wages and salaries.
2.1. THE SOLOW GROWTH MODEL 11

ˆ One last point, the production function Y = K α · (AL)β is specified in the manner that α + β = 1. This
is interpreted as the production function exhibits the constant returns to scale. That is,

F (λ · Kt , λ · (At , Lt )) = (λ · Kt )α (λ · At Lt)1−α = λ · Ktα (At Lt)1−α = λ · Yt

This constant returns to scale implicitly assumes two underlying assumptions. What are they?

2.1.1 The capital intensive form of the SGM

For the ease of analysis (remember the simple Solow model?), we transform the SGM into the capital intensive
form as:
α
Ktα (At Lt)1−α

Yt Kt
f (κt ) = = = = κα
t
At , Lt At , Lt At , Lt
Kt
where κt = At ,Lt .

2.1.2 The dynamics of labor and technology

Labor is assumed to grow at a constant rate of n, that is Lt+1 = (1 + n) · Lt . Similarly, technology also
grows at a constant rate g, At+1 = (1 + g) · At . We find ourselves smarter each period (i.g. each year) by a
factor of (1+g). Do you believe that your g is positive? For the society as a whole, we assume that n and g
are positive, but rather small in magnitude. Empirical evidence shows that g ≈ 0.015 in the US in the last
150 years. For Vietnam, the current n ≈ 0.01.

2.1.3 The dynamics of the capital stock

Vietnamese people are known for our thriftiness. For the given amount of income Yt , i.e. everything we
produce, we consume only part of it, and save s · Yt = s · Ktα (At Lt)1−α . This saving is reinvested to cover
the wears and tears of the existing capital, known as depreciation (δ · Kt ) and also to increase the capital
stock, i.e. the number of structures, machines, and tools, of the economy in the next period, Kt+1 . The
dynamics of the capital stock can be described as:

Kt+1 = Kt − δ · Kt + s · Ktα (At Lt)1−α

Kt+1 = (1 − δ) · Kt + s · Ktα (At Lt)1−α

Kt+1 Kt Ktα (At Lt)1−α

At Lt = (1 − δ) · At Lt +s· At Lt

Kt+1 At+1 Lt+1 Kt Ktα (At Lt)1−α

At+1 Lt+1 At Lt = (1 − δ) · At Lt +s· At Lt

κt+1 · (1 + n)(1 + g) = (1 − δ) · κt + s · κα
t

κt+1 · (1 + n + g + ng) = (1 − δ) · κt + s · κα
t

Note that the ng term drops out as it is very small.

κt+1 · (1 + n + g) = (1 − δ) · κt + s · κα
t

1−δ s α
κt+1 = 1+n+g κt + 1+n+g κt

The net change in κ is

1−δ s α
∆κt+1 = ( 1+n+g − 1)κt + 1+n+g κt

δ+n+g s α
∆κt+1 = − 1+n+g κt + 1+n+g κt
12 CHAPTER 2. THE LONG RUN MACROECONOMY

Figure 2.3: When κt < κ∗

Figure 2.4: When κt > κ∗

At the steady state, κt+1 = κt = κ∗ , and therefore, ∆κt+1 = 0, or

δ+n+g ∗ s ∗α
1+n+g κ = 1+n+g κ

(δ + n + g)κ∗ = sκ∗α

(δ + n + g)κ∗ = sκ∗α
κ∗ s
κ∗α = δ+n+g

κ∗1−α = s
δ+n+g

 1
 1−α
s
κ∗ =
δ+n+g
Please note this steady state value and the dynamics of κ from the lecture, as demonstrated by Figures (1.3)
and (1.4).

2.1.4 Some steady state results

ˆ The growth rate of GDP, Yt :

Yt Ktα (At Lt)1−α

Yt = At Lt · At Lt = At Lt · At Lt = At Lt · κ∗α

log(Yt ) = log(At ) + log(Lt ) + log(κ∗α )

Take the total differentiation of both sides of the previous equation gives:

∆Yt ∆At ∆Lt ακ∗(α−1)

Yt = At + Lt + ∆κ∗ · κ∗α
2.1. THE SOLOW GROWTH MODEL 13

At the steady state, κ is constant, that means the growth rate of Yt is equal to (n+g), which is the sum
of the growth rate of At , g, plus the growth rate of Lt , n. In words, the total economy becomes bigger
if we have more hands to work (yet, also more mouths to feed!), and if each pair of hands become more
and more skillful (bravo!).
 1
 1−α ∗
ˆ What happens if s is becomes larger? Please take notice that κ∗ = δ+n+g s
. Therefore, ∂κ
∂s =
 α
 1−α
1 s
1−α · δ+n+g > 0. An increase in the saving rate will lead to an increase in investment, bringing
the economy to a new level of steady state capital per effective worker κ∗ . Higher κ∗ means higher
Yt ∗ Yt Yt Yt
At Lt = κ . Higher At Lt means higher At · At Lt = Lt , i.e. higher output per worker. However, this
increase if only temporary as the additional growth due to higher κt stops once the economy enters
the new steady state.

ˆ The growth rate of GDP per worker, Yt

Lt :

Yt Yt Ktα (At Lt)1−α

Lt = At · At Lt = At · At Lt = At · κ∗α

Yt
log( L t
) = log(At ) + α · log(κ∗ )

Take the total differentiation of both sides of the previous equation gives:

Y
∆ Lt ∆At ∆κ∗
Yt
t
= At +α· κ∗
Lt

The growth rate of GDP per capita at the steady state is equal to the growth rate of technology. In
other words, in the long run we can only have a better life if we train ourselves to have more skillful
pairs of hands and smarter heads.

2.1.5 The golden rule of saving

The result in the above equation states that the economy will settle down at a steady state level of capital
per effective worker, which is determined by the exogenous parameters δ, n, and g; and the rate of saving,
s. Given the parameters, each choice of s will give us one steady state, with just a temporary change in
Yt
the Yt /Lt growth. In the end, we don’t care much about L t
, but rather the amount left for consumption
Yt
(1 − s) · Lt . One natural question is: How much to save so that we can maximize the amount of consumption.

ˆ At each steady state corresponding to a saving level s, the consumption per effective worker, also means
per worker, is:
 ∗  ∗
∗ Ct Yt
ct = = (1 − s) · = (1 − s) · κ∗α = κ∗α − s · κ∗α
At Lt At Lt
and take notice from Figures (1.3) and (1.4) that at the steady state, s · κ∗α = (δ + n + g) · κ∗ . Using this
result, the equation becomes:
c∗t = κ∗α − (δ + n + g) · κ∗
Choosing the rate of saving is at the same time choosing the capital stock per effective worker. It is easier
to maximize c∗t with respect to κ∗ . Therefore, taking the derivative of c∗t w.r.t. κ∗ yields:
∂c∗
t
∂κ∗ = −(δ + n + g) + ακα−1
∂c∗
Setting the derivative t
∂κ∗ = 0 yields:
∗(α−1)
−(δ + n + g) + ακopt =0

∗(α−1) δ+n+g
κopt = α

 1
 α−1
δ+n+g
κ∗opt = α .
14 CHAPTER 2. THE LONG RUN MACROECONOMY

Noting that a = ( a1 )−1 , we get:

 1
 1−α
α
κ∗opt =
δ+n+g
 1
 1−α
combining the steady state capital per effective worker result, κ∗ = δ+n+g s
and the above result, it is
apparent that at the optimum, α = s. That is the golden rule to save is s = α.
 1
 1−α
Each level of saving, s, leads to a steady state with a κ∗ = δ+n+g s
. When s = α, the level of saving is
optimal because it maximizes the consumption per effective worker (and therefore, consumption per worker)
at the steady state.
It is best to save the income share of capital to invest back to capital. When α is high, the marginal product
of capital is also high, so it is worth investing a higher share of income to build a larger stock of capital.
Another interpretation of this theoretical result is that capital per effective worker is accumulated to the
point when the marginal cost of holding capital per effective worker is equal to the marginal product of
capital per effective worker.

2.1.6 Problemset exercises on the SGM

Problem 1.2
Consider the following Solow-type growth model:
Yt = Ktα , 0 < α < 1,
Kt+1 = (1 − δ) · Kt + It , 0 < δ < 1,
It+1 = St − CAt ,
St = s · Yt , 0 < s < 1, and
CAt = m · Yt , 0 < m < 1.
2.1. THE SOLOW GROWTH MODEL 15

ˆ a) Derive the steady-state level of income, and the steady-state growth rate of income.

ˆ Answer:

ˆ This simple model only features capital as the only input in the production process. Therefore, there is
no need to derive the intensive form. Let’s go straight to the dynamics of capital:

Kt+1 = (1 − δ) · Kt + St − CAt

Kt+1 = (1 − δ) · Kt + sKtα − mKtα

Kt+1 = (1 − δ) · Kt + (s − m) · Ktα

When Kt is relatively small, δ · Kt < (s − m) · Ktα , the capital stock is increasing because net investment is
positive. Kt keeps increasing until δ · Kt ≈ (s − m) · Ktα , the level of capital stock will stay constant, say at
K ∗ and the economy enters the steady state.
To solve for K ∗ , let

∆Kt+1 = Kt+1 − Kt = −δ · Kt + (s − m) · Ktα

Setting ∆Kt+1 = 0 and denoting the value of Kt satisfying this condition as K ∗ yield:

δ · K ∗ = (s − m) · K ∗α
K∗ s−m
K ∗α = δ

K ∗1−α = s−m
δ

1
K∗ = s−m

1−α
δ

The steady state leve of income is:

α
s−m

1−α
Yt = δ

b) Derive how the steady-state level of income depends on the coefficient m, reflecting the current account
surplus,0 < m < 1. Provide the economic reasoning for your results.

α
ˆ Answer: First notice that ∂K ∗
∂m
1
== − δ(1−α) s−m 1−α
δ , which indicates that the level of capital stock
depends negatively on the level of current account. If the country is a net lender, the current account is
positive, the domestic capital stock at the steady state is lower than the level of capital that would otherwise
accumulated in a close economy. Reversely, if the current account CA < 0, the country is a net borrower,
the capital stock at the steady state is higher than the closed economy case because domestic investment is
higher than domestic saving. With a lower level of capital stock, the output Y ∗ = K ∗α is also lower.

Figure 2.5: Natural disaster destroys half of the economy’s capital stock
16 CHAPTER 2. THE LONG RUN MACROECONOMY

c) Suppose that some years after the economy had reached steady state, a natural disaster destroys half
of the economy’s capital stock. Describe both graphically and verbally what will happen in this economy
after the natural disaster has struck.
Answer: When half of the capital stock is destroyed, the economy find itself at the level of capital that is
lower than the capital stock at the steady state, gross investment, (s − m) · Ytα is larger than depreciation,
δKt . The level of capital stock will keep increasing until it reaches the steady state level. This process is
demonstrated in Figure (1.5).

Problem 1.3
A country’s production function is given by Yt = At ·Kt0.65 ·L0.35
t . In the year 2016 we observed Kt = 10, 000;
Lt = 100; and Yt = 10, 000. Suppose that from 2016 to 2017, income (Yt ) grew by 2.6%, the capital stock
(Kt ) by 1.2%, and employment (Lt by 0.5%. What is the rate of technological progress between 2016 and
2017? What is the level of technology in 2017?

ˆ Answer: Plugging the giving information to the production function 10000 = At · 100000.65 · 1000.35 yields
A2016 = 5.011872336. Next taking the natural logarithm of the production function yields:

log(Yt ) = log(At ) + 0.65 · log(Kt ) + 0.35 · log(Lt ).

Taking the total differentiation of both sides of the previous equation yields:
∆Yt ∆Kt ∆Lt ∆At
Yt = 0.65 · Kt + 0.35 · Lt + At

∆At ∆Yt ∆Kt ∆Lt

At = Yt − 0.65 · Kt − 0.35 · Lt

Plugging the information on the growth rate yields ∆A

At = 0.01645, thus the growth rate of At is 1.65% from
t

2016 to 2017. The level of technology progress in 2017 is (1 + 0.01645) · 5.011872336 = 5.094317636

Problem 1.4
Let us suppose that income per worker in Mexico in 2017 is equal to $ 16,000. Also suppose that the
economy-wide rate of saving s = 0.15 (with a balanced government budget and net exports, net factor
payments from abroad as well as net unilateral transfers equal to $ 0), the rate of worker growth n = 0.01,
the rate of depreciation of capital δ = 0.03, the rate of technological progress g = 0.02, and the Cobb-Douglas
production function
Yt = ·Ktα · (At Lt )1−α ,
with α = 0.4. Assume that capital per effective worker in Mexico in 2017 is in steady state. a) If no further
information is given, what do you predict regarding the level of income per worker in Mexico in 2020?

ˆ Answer: At the steady state, the income per worker growth rate is the same as the growth rate of
technological progress. As a result, the income per worker in Mexico in 2020 is given by:
Yt Yt Yt Yt
(L t
)2020 = ( L t
)2019 · (1 + g) = ( L t
)2018 · (1 + g) · (1 + g) = ( Lt
)2017 · (1 + g) · (1 + g) · (1 + g)
Yt Yt
(L t
)2020 = ( L t
)2017 · (1 + 0.02)3 = 16979.328

b) Now take into account the following additional information: from 2017 to 2018 onward, the rate of
technological progress was to decrease to g = 0.01, What is your new prediction regarding the level of
income per worker in Mexico in 2020? Provide detailed economic rationale for the change in your prediction.

ˆ Answer: At the 2017 steady state, the level of capital per effective worker is given by: κ∗2017 =
 1
 1−α
s
δ+n+g2017 From 2017 to 2018 onward, the new technology progress growth rate is g2018 = 0.01. The
change in the technological progress growth rate pushes the Mexican economy into a new path of growth
 1
 1−α
with a different steady state level of capital per effective worker as: κ∗new = δ+n+g
s
2018
To make the prediction of the level of income per worker by 2020, we compute the level of
capital per effective worker in 2017 and the following years, as well as the level of technology
in 2017 and the following years. This information will allows the prediction of income per
capita in 2020.
2.1. THE SOLOW GROWTH MODEL 17

ˆ The level of technology in 2017 is computed as:

Y2017 Y2017 α
L2017 = A2017 L2017 A2017 = A2017 · κ2017

Y2017
L2017
A2017 = κα
2017

16000
A2017 = 4.6050393730.4 = 8686.136373

ˆ We need first to derive the formula of κ∗ at the steady state. For the sake of completeness, I also
include the derivation here. Please do this in similar exam questions

Kt+1 = Kt − δ · Kt + s · Ktα (At Lt)1−α

Kt+1 = (1 − δ) · Kt + s · Ktα (At Lt)1−α

Kt+1 Kt Ktα (At Lt)1−α

At Lt = (1 − δ) · At Lt +s· At Lt

Kt+1 At+1 Lt+1 Kt Ktα (At Lt)1−α

At+1 Lt+1 At Lt = (1 − δ) · At Lt +s· At Lt

κt+1 · (1 + n)(1 + g) = (1 − δ) · κt + s · κα
t

κt+1 · (1 + n + g + ng) = (1 − δ) · κt + s · κα
t

κt+1 · (1 + n + g) = (1 − δ) · κt + s · κα
t

1−δ s α
κt+1 = 1+n+g κt + 1+n+g κt

The net change in κ is

1−δ s α
∆κt+1 = ( 1+n+g − 1)κt + 1+n+g κt

δ+n+g s α
∆κt+1 = − 1+n+g κt + 1+n+g κt

At the steady state, κt+1 = κt = κ∗ , and therefore, ∆κt+1 = 0, or

δ+n+g ∗ s ∗α
1+n+g κ = 1+n+g κ

(δ + n + g)κ∗ = sκ∗α

(δ + n + g)κ∗ = sκ∗α

κ∗ s
κ∗α = δ+n+g

κ∗1−α = s
δ+n+g

 1
 1−α
κ∗ = s
δ+n+g

ˆ Now the level of capital per intensive worker for 2017 can be computed as:
 1
 1−α  1
 1−0.4
κ∗2017 = s
δ+n+g2017 = 0.15
0.03+0.01+0.02 = 4.605039373

ˆ Next, we compute the value of At and κt recursively as:

1−δ s α
κ2018 = 1+n+g2018 κ2017 + 1+n+g2018 κ2017 = 4.650186818
18 CHAPTER 2. THE LONG RUN MACROECONOMY

1−δ s α
κ2019 = 1+n+g2019 κ2018 + 1+n+g2019 κ2018 = 4.694180338

1−δ s α
κ2020 = 1+n+g2020 κ2019 + 1+n+g2020 κ2019 = 4.737043506

A2020 = A2017 · (1 + g2018 ) · (1 + g2019 ) · (1 + g2020 ) = A2017 · (1 + 0.01)3 = 8949.334991

ˆ The predicted income per worker in 2020 for Mexico is:

Y2020 0.4
L2020 = A2020 · K2020 = 16672.23082

ˆ Notice that the new level of income per capita is lower because of the lower technology progress growth.
c) For the year 2017, derive the golden rule level of consumption per effective worker. If technological
progress for 2017 and all following years was equal to g = 0.02, what is the time path of consumption per
worker implied by this golden rule level of consumption per effective worker in the 2018-2020?

ˆ Answer (for the sake of completeness, I include the full answer here, which means a repe-
tition of what has been said in the summary):
ˆ At each steady state corresponding to a saving level s, the consumption per effective worker is:
 ∗  ∗
Ct Yt
c∗t = At Kt = (1 − s) · At Kt = (1 − s) · κ∗α = κ∗α − s · κ∗α

and take notice that at the steady state, s · κ∗α = (δ + n + g) · κ. Using this result, the above equation
becomes:

c∗t = κ∗α − (δ + n + g) · κ

ˆ Choosing the rate of saving is at the same time choosing the capital stock per effective worker. It is
easier to maximize c∗t with respect to κ. Therefore, taking the derivative of c∗t w.r.t. κ in (1.8) yields:

∂c∗
∂κ
t
= −(δ + n + g) + ακ∗(α−1)

∂c∗
Setting the derivative ∂κ
t
= 0 yields:

−(δ + n + g) + ακ∗(α−1) = 0

δ+n+g
κ∗(α−1) = α

 1
 α−1
δ+n+g
κ∗ = α .

Noting that a = ( a1 )−1 , we get:

 1
 1−α
κ∗ = α
δ+n+g

 1
 1−α
combining the general steady state level of capital per effective worker, which is κ∗ = δ+n+g s
,
with this golden rule steady state saving rate, it is apparent that at the optimum, α = s. That is the
golden rule to save is s = α. It is best to save the income share of capital to invest back to capital.
ˆ Similar to question b), a change in the saving rate put the economy on a new growth path toward a
new steady state. The new level of capital per worker is determined as:
 1
 1−α
s0.4
κ∗ = δ+n+g

ˆ From the previous question κ2017 = 4.605039373. A2017 = 8686.136373; A2018 = A2017 · (1 + 0.02) =
8859.859101; A2019 = A2018 · (1 + 0.02) = 9037.056283; A2020 = A2019 · (1 + 0.02) = 9217.797408;; and
κt can be computed recursively as:
2.1. THE SOLOW GROWTH MODEL 19

1−δ s0.4 α
κ2018 = 1+n+g2018 κ2017 + 1+n+g2018 κ2017 = 5.052130575

1−δ s0.4 α
κ2019 = 1+n+g2019 κ2018 + 1+n+g2019 κ2018 = 5.500188339

1−δ s0.4 α
κ2020 = 1+n+g2020 κ2019 + 1+n+g2020 κ2019 = 5.947811249

ˆ Following the golden rule of saving, the time path of consumption per worker, C pw , between 2017 and
2020 for Mexico is:
pw Y2017
C2017 = (1 − s2017 ) · L2017 = 0.6 · 16000 = 9600

pw Y2018
C2018 = (1 − s2018 ) · L2018 = 0.6 · A2018 · κ0.4
2018 = 10161.73564

pw Y2019
C2019 = (1 − s2019 ) · L2019 = 0.6 · A2019 · κ0.4
2019 = 10723.31989

pw Y2020
C2020 = (1 − s2020 ) · L2020 = 0.6 · A2020 · κ0.4
2020 = 11285.51235

The consumption per worker time path is:

Right after the change in the saving rate, their is a drop in consumption (as Yt = Ct +St ). The new sav-
ing rate pushes the economy out of balance, GDP per worker (and thus consumption per worker)grows
at the rate

GR( C Yt Yt α
Lt ) = GR( Lt ) = GR( At Lt At ) = GR(κt At ) = α × GR(κt ) + GR(At ) = α × GR(κt ) + g.
t

Remember that before the change in the saving rate, the economy was in the old steady state with
 1
 1−α
s=0.15
κ2015 = κ2016 = κ2017 = κ∗before 2018 = = 4.605039373
δ+n+g

The growth rate of κt before 2018 was 0. From 2017 to 2018, the saving rate changed to 0.4, shifting the
 1
 1−α
s=0.4
saving line up, the economy of Mexico has a new steady state, with κ∗new = δ+n+g = 6.240251469.
∗
κ2018 < κnew , the economy is on the left of the steady state level of capital per effective worker. The
growth rate of κ2018 on ward > 0. Therefore, GR( CLt )2018 on ward > g.
t

ˆ Which Factor is More Relevant to economic growth?

To this point, we have discuss the golden rule, and the apparent indication is that Mex-
ico should save more, increasing its saving from 15% to 40%. This policy prescription is
painful and impossible in practice because people are shortsighted and asking them to give
up 17.64% of their current consumption for the optimal consumption in the future sounds
20 CHAPTER 2. THE LONG RUN MACROECONOMY

nothing more than a crazy idea.

Figure 2.6: Golden Rule Vs. Techonological Progress

An alternative policy prescription could be that Mexico should go on with the current consumption,
however, there are measures to boost the skills of its people so that the country can increase its
technological progress by 1 more %, so g increases from 0.02 to 0.03 (saving stays the same). Figure
(2.6) illustrates these alternative policy prescriptions by graphing the development of consumption per
capita for the base scenario discussed in question a), the golden rule scenario, and the 3% technological
progress.

1. CPC stands for consumption per capita

2. CPC lines are measured on the left vertical axis
3. The red long-dash line is the difference between CP Cs=0.15,g=0.03 and CP Cs=0.4,g=0.02 , measured
in the right vertical axis.

ˆ It can be seen that the red line became positive in 2017 due to the sudden increase in saving to
meet the golden rule while the 3% growth in technology scenario does not require any adjustment in
consumption. The difference narrows gradually and reaches its lowest point in some 30 years but still
positive. After that, the difference starts to go to infinity.

ˆ The discussion in this section emphasizes the overwhelmingly important role of technological
progress in general, and human capital investment in particular. Only by investing in human can
we sustain the improvement in our standard of living. A nation is doomed to fail if its young
generations are not provided with the proper education to increase their wealth of human
capital.

Problem 1.5
Consider an economy, in which labor input grows by 1% per year and technology improves by 3% per year.
Every year, 3% of the capital stock wears out. The households save a constant fraction of 35% of their
income. The elasticity of output with respect to capital in the country’s standard Cobb-Douglas production
function is 0.5.

a) Determine the steady state value of capital per effective worker.

ˆ Answer: Given the standard Cobb-Douglas production function in the context of the Solow growth
model, the production function is:

Yt = Ktα (At Lt )1−α

2.1. THE SOLOW GROWTH MODEL 21

The elasticity of output with respect to capital is defined as:

YKtt = ∂K
∂Yt
t
·Kt
Yt
= α · Ktα−1 (At Lt )1−α · K α (AK t
t Lt )
1−α
t
=α
The production function for this country is therefore:
Yt = Kt0.5 (At Lt )0.5
In the context of the Solow growth model, the capital dynamics equation is:

Kt+1 = Kt − δ · Kt + s · Ktα (At Lt)1−α

Kt+1 = (1 − δ) · Kt + s · Ktα (At Lt)1−α

Kt+1 Kt Ktα (At Lt)1−α

At Lt = (1 − δ) · At Lt +s· At Lt

Kt+1 At+1 Lt+1 Kt Ktα (At Lt)1−α

At+1 Lt+1 At Lt = (1 − δ) · At Lt +s· At Lt

κt+1 · (1 + n)(1 + g) = (1 − δ) · κt + s · κα
t

κt+1 · (1 + n + g + ng) = (1 − δ) · κt + s · κα
t

κt+1 · (1 + n + g) = (1 − δ) · κt + s · κα
t

1−δ s α
κt+1 = 1+n+g κt + 1+n+g κt

The net change in κ is

1−δ s α
∆κt+1 = ( 1+n+g − 1)κt + 1+n+g κt

δ+n+g s α
∆κt+1 = − 1+n+g κt + 1+n+g κt

At the steady state, κt+1 = κt = κ∗ , and therefore, ∆κt+1 = 0, or

δ+n+g ∗ s ∗α
1+n+g κ = 1+n+g κ

(δ + n + g)κ∗ = sκ∗α

(δ + n + g)κ∗ = sκ∗α
κ∗ s
κ∗α = δ+n+g

κ∗1−α = s
δ+n+g

 1
 1−α
κ∗ = s
δ+n+g

Plugging the model parameters into this formula: n = 0.01; g = 0.03; δ = 0.03, and s=0.35 yields:
 2
κ∗ = 0.01+0.03+0.03
0.35
= 25

b) Suppose the government could provide incentives for households to either increase or reduce the saving
rate. If the government’s aim was to maximize consumption per effective worker, should the households save
more or less of their output?

ˆ Answer: Firstly, we need to find the optimal saving rate for this country.
At each steady state corresponding to a saving level s, the consumption per effective worker is:
 ∗  ∗
c∗t = ACt K
t
t
= (1 − s) · Yt
At Kt = (1 − s) · κ∗α = κ∗α − s · κ∗α

and take notice that at the steady state, s · κ∗α = (δ + n + g) · κ. Using this result, the above equation
becomes:

c∗t = κ∗α − (δ + n + g) · κ
22 CHAPTER 2. THE LONG RUN MACROECONOMY

Choosing the rate of saving is at the same time choosing the capital stock per effective worker. It is easier
to maximize c∗t with respect to κ. Therefore, taking the derivative of c∗t w.r.t. κ in yields:
∂c∗
∂κ
t
= −(δ + n + g) + ακα−1
∂c∗
Setting the derivative ∂κ
t
= 0 yields:

−(δ + n + g) + ακ∗(α−1) = 0
δ+n+g
κ∗(α−1) = α
 1
 α−1
δ+n+g
κ∗ = α .

Noting that a = ( a1 )−1 , we get:

 1
 1−α
κ∗ = α
δ+n+g

 1
 1−α
combining the general steady state level of capital per effective worker, which is κ∗ = δ+n+gs
, with
this golden rule steady state saving rate, it is apparent that at the optimum, s = α=0.5. The government
should encourage households to save more.

c) Given the result you found in (b) what are the consequences for output, investment, capital, consump-
tion per effective worker following a change of your policy recommendation. Provide a detailed economic
rational on how the outcomes arise.

ˆ Answer:

ˆ Capital per effective worker:

Economic Reasoning: Due to the change in the saving rate from s=0.35 to s1 = 0.50, gross investment
increases from It = 0.35 · Yt to It = 0.50 · Yt , the net investment per capita

δ+n+g s1 α
∆κt+1 = − 1+n+g κt + 1+n+g κt

becomes positive, the level of capital per effective workers start to increase from the former steady
state level of 25 to the new steady state level of 51.02040816.

ˆ Output per effective worker:

Economic Reasoning: Initially, the economy is at the steady state and output per effective worker
was constant and equal to κ∗α
1 . With the new saving rate, the level of capital per effective workers
2.1. THE SOLOW GROWTH MODEL 23

starts growing and the economy converges to the new steady state k2∗ . Output per effective worker
also converges to the new level κ∗α
2 .

ˆ Investment per effective worker:

Economic Reasoning: Investment was constant at 35% of the constant output per effective worker
initially. The jump in the saving rate to 50% leads to a jump in the investment per effective worker.
As the economy converges to the new steady state level, investment per capita also converges to the
new constant steady state level of 0.5 · κ∗α
2 .

ˆ Consumption per effective worker:

Economic Reasoning: Initially, consumption was constant at 0.65 · κ∗α

1 . With the increase in saving
implied by the golden rule, consumption dropped 1 to 1 with the increase in saving. Over time,
the economy grows due to more capital per effective worker, leading to the growth in consumption
24 CHAPTER 2. THE LONG RUN MACROECONOMY

per effective worker. The increase in consumption is more than enough to compensate for the initial
drop because the new saving rate is the optimal choice. However, it takes almost 40 years for the
consumption to reach the initial level following the increase in saving, which is painfully slow.

Problem 1.6
Suppose an economy’s output per worker was $50,000 and its labor force had a of size 1 million people in
2015. The saving rate was 20%, labor force grew at a rate of 0.5%, capital depreciated at a rate of 3% while
technology improved by 1%. Further the capital share of production in the economy was roughly 0.35.

a) Suppose that the economy could be modeled using the standard Cobb-Douglas production function
and the market is perfectly competitive, compute the level of technology for this economy in 2015 given the
economy is in steady state.

ˆ Answer:
The economy is represented by the production function:

Yt = Ktα (At Lt )1−α

Given that the market is perfectly competitive, firms are free to maximize their profit by choosing the
optimal level of labor and capital without influencing the market price, interest rates, and wages (non com-
patitive market structure such as monopoly,oligopoly, and monopsony have to take the market price/wages
into account). The profit maximization problem is:

maximize Πt = Pt · Yt − Wt · Lt − Rt · Kt = Pt · Ktα (At Lt )1−α − Wt · Lt − Rt · Kt

W,L

The F.O.Cs are:

∂Πt
= Pt · αKtα−1 (At Lt )1−α − Rt = 0
∂Kt
∂Πt
= Pt · (1 − α) · Ktα · (At )1−α · (Lt )−α − Wt = 0
∂Lt
Rearranging terms yields:

∂F (Kt , AT , Lt ) Rt
= αKtα−1 (At Lt )1−α = ≡ rt
∂Kt Pt

∂F (Kt , AT , Lt ) Wt
= (1 − α) · Ktα · (At )1−α · (Lt )−α = ≡ wt
∂Lt Pt
where rt and wt are real interest rate and real wages. The maximization problem yields the results that in
the perfectly competitive market, each worker/unit of capital is paid their marginal products. Therefore,
2.1. THE SOLOW GROWTH MODEL 25

the income share of capital is:

∂F (Kt ,AT ,Lt )
∂Kt · Kt α · Ktα−1 · (At Lt )1−α · Kt
= =α
F (Kt , AT , Lt ) Ktα (At Lt )1−α
Using the given information that the capital share of production in the economy was 0.35 yields α = 0.35.
The production function for this economy is, therefore:

Yt = Kt0.35 (At Lt )0.65

The capital dynamics equation is:

Kt+1 = Kt − δ · Kt + s · Ktα (At Lt)1−α

Kt+1 = (1 − δ) · Kt + s · Ktα (At Lt)1−α

Kt+1 Kt Ktα (At Lt)1−α

At Lt = (1 − δ) · At Lt +s· At Lt

Kt+1 At+1 Lt+1 Kt Ktα (At Lt)1−α

At+1 Lt+1 At Lt = (1 − δ) · At Lt +s· At Lt

κt+1 · (1 + n)(1 + g) = (1 − δ) · κt + s · κα
t

κt+1 · (1 + n + g + ng) = (1 − δ) · κt + s · κα
t

κt+1 · (1 + n + g) = (1 − δ) · κt + s · κα
t

1−δ s α
κt+1 = 1+n+g κt + 1+n+g κt

The net change in κ is

1−δ s α
∆κt+1 = ( 1+n+g − 1)κt + 1+n+g κt

δ+n+g s α
∆κt+1 = − 1+n+g κt + 1+n+g κt

At the steady state, κt+1 = κt = κ∗ , and therefore, ∆κt+1 = 0, or

δ+n+g ∗ s ∗α
1+n+g κ = 1+n+g κ

(δ + n + g)κ∗ = sκ∗α

(δ + n + g)κ∗ = sκ∗α
κ∗ s
κ∗α = δ+n+g

κ∗1−α = s
δ+n+g

 1
 1−α
κ∗ = s
δ+n+g

Plugging the model parameters into this formula: n = 0.005; g = 0.01; δ = 0.03, and s=0.2 yields:
 1
 0.65
κ∗ = 0.01+0.03+0.005
0.2
= 9.922983743
At the steady state,

Y2015 Y2015
L2015 = A2015 L2015 · A2015 = κ∗ · A2015
Y2015
L2015 50,000
A2015 = κ∗α = 9.9229837430.35 = 22, 394.6978031362.

b) Assume the economy stays on its balanced growth path. What is your prediction for output per
effective worker, output per worker and output in 2030 for this economy. Also comment on the growth rate.
26 CHAPTER 2. THE LONG RUN MACROECONOMY

ˆ Answer:

Output per effective worker: At the steady state, output per effective worker is κ∗α , which is constant
because κ∗ is constant. Its growth rate is, therefore, zero.

Output per worker: At the steady state, output per worker

Yt Yt
yt ≡ Lt = At Lt · At = κ∗ · At

Taking the logarithm of both sides yields:

log(yt ) = log(κ∗ ) + log(At )

Taking the total differential of both sides with respect to time yields:
∆yt ∆κ∗ ∆At
yt = κ∗ + At

At the steady state, the level of capital per effective worker is constant, and therefore, income per worker
grows at the rate of technological progress yt+1 = yt (1 + g), yt+2 = yt (1 + g)2 and yT = yt+1 (1 + g)T −t , for
T ≥ t. Plugging t = 2015, and T = 2030 to get y2030 = y2015 · (1 + 0.01)2030−2015 = 58048.44777.

Yt
Output: At the steady state, Yt = At Lt · At · Lt = κ∗ · At · Lt . Taking the logarithm of both sides yields:

log(Yt ) = log(κ∗ ) + log(At ) + log(Lt )

Taking the total differential of both sides with respect to time yields:

∆Yt ∆κ∗ ∆At ∆Lt

Yt = κ∗ + At + Lt

At the steady state, the level of capital per effective worker is constant, and therefore, output grows
at the rate of technological progress plus the rate of growth in the labor force. Yt+1 = Yt (1 + n + g),
Yt+2 = Yt (1 + n + g)2 and YT = Yt+1 (1 + n + g)T −t , for T ≥ t. Plugging t = 2015, T = 2030, and
Y2015 = 50, 000 · 1, 000, 000 to get Y2030 = Y2015 · (1 + 0.01 + 0.005)2030−2015 = 62, 511, 603, 332.72.

c) Now suppose the growth rate of the labor force was to decrease to -1%. What is your prediction for
consumption per effective worker, consumption per worker and consumption in 2030?

ˆ Answer:

Using the results from question a), the level of capital per effective worker at the steady state is:
 1
 1−α
κ∗ = δ+n+g s
When the rate of labor force grows changed from n=0.5% to n1 = −1%, the steady
state level of capital per effective worker changed to:
 1
 1−α
κ∗1 = δ+ns1 +g
Plugging the new values of the parameters into this formula yields:
 1
 0.65
κ∗1 = 0.01+0.03+(-0.01)
0.2
= 18.51620196

From question a) the 2015 value of capital per effective worker was: 9.922983743 and the level of technol-
ogy progress was 22,394.6978031362. Due to the change in the population growth rate, capital per effective
worker started growing until it reaches the new steady state level of 18.51620196. Using the results from
question a), we can compute the value of capital per effective worker recursively as:
1−δ s
κt+1 = 1+n+g κt + 1+n+g κα
t

Therefore,

1−0.03 0.2 0.35

κ2016 = 1−0.01+0.03 κ2015 + 1−0.01+0.03 κ2015 = 10.0718285
2.1. THE SOLOW GROWTH MODEL 27

t 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030
κt 10.071828 10.218541 10.36313 10.505605 10.645977 10.78426 10.920466 11.05461 11.186707 11.316774 11.444826 11.570881 11.694958 11.817075 11.937251

Consumption in 2030 under this scenario is:

(1−0.2)·κα α
2030 ·A2030 ·L2030 = κ2030 ·A2015 ·(1+0.01)
2030−2015
·L2015 ·(1−0.01)2030−2015
= 42, 608, 890, 731.18

Consumption per worker in 2030 under this scenario is:

C2030 48,047,472,679.54
L2030 = 860,058.35 = 49, 541.86

Consumption per effective worker in 2030 under this scenario is:

C2030 42,608,890,731.18
A2030 ·L2030 = 860,058.35·25,999.55 = 1.905489219

Problem 1.7

Assume labor-augmenting technological progress in the production function with a share of capital on pro-
duction of 0.3. In the richest country in the world, output per person is 50 times as high as in the poorest
country. The poorest country has a saving rate of 5%, the richest has one of 30.125%. Also there are no
differences in countries’ depreciation, population growth and technological progress rates. Further the initial
capital stock of the economies’ is their respective steady state. What would you conclude concerning the
level of technology A in the richest country compared to the poorest one?

ˆ Answer:

The capital per effective worker in the rich country at its steady state is:
1
srich 1−α
κ∗rich = ( δ+n+g )

The capital per effective worker in the poor country at its steady state is:
s 1
κ∗poor = ( δ+n+g
poor
) 1−α

The ratio of capital per effective worker between the two countries in the steady state is:
s 1
rich ) 1−α
( δ+n+g 1 1
κrich
κpoor = spoor 1 = ( sspoor )
rich 1−α
= ( 30.125
5 )
1−0.3 = 13.00841426
( δ+n+g ) 1−α

Furthermore, at any point in the steady state:

Yrich Yrich
yrich = Lrich = Arich Lrich · Arich = κα
rich · Arich

and
Ypoor Ypoor
ypoor = Lpoor = Apoor Lpoor · Apoor = κα
poor · Apoor

Therefore,
yrich
ypoor = { κκpoor
rich
}· Arich
Apoor
yrich
Arich ypoor 50
Apoor = κrich α
{ κpoor }
= 13.008414260.3 = 23.15808782877508

The technological progress level is about 23.1581 times higher in the rich country in comparison with
that of the poor country.

Problem 1.8

In the context of the Solow Growth Model, please explain why defeated countries like Germany and Japan
experienced higher growth rate than countries like the UK and the USA after the World Ward II.
28 CHAPTER 2. THE LONG RUN MACROECONOMY

Problem 1.9
A significant part of Vietnam’s import is machinery and equipment. Imported machines and equipment em-
body the technological progress that could influence the technological progress of the Vietnamese workers.
Suppose that there are two types of technologies, the outdated technologies such as those being moved out
of middle income countries like China and Thailand because of environmental problems (type B); and the
advanced technologies that go with the modern machines and equipment from countries such as Germany,
Japan and the USA (type G). Suppose further that type B equipment and machines drag the technological
progress in Vietnam down by 0.5% per year, while type G equipment and machinery boost the technological
progress by 1%. In the context of the standard Solow Growth Model:

a) Please explain the consequences on GDP, GDP per worker in Vietnam in the next 15 years if this
country chooses to import type B technologies only. Please draw the Solow model diagram and the time
paths of output, capital, investment, and consumption per effective worker for this scenario.

b) What is your prediction on GDP, GDP per worker in Vietnam in the next 15 years if instead this
country imports the type G technology. Please draw the Solow model diagram and the time paths of output,
capital, investment, and consumption per effective worker for this scenario.
Chapter 3

The Short Run Macroeconomy

ˆ In the long run, price and wages fully adjust, the output in the economy is purely determined by the
supply side of the economy (i.e. how much capital, labor that we have; how fast is the growth of technology;
and how obstructing government and government regulations are). We discuss these issues in depth using
the insights from the Solow Growth Model, and the Solow-Romer Model.

ˆ In the short run, price and wages are sticky, firms have the opportunity to adjust their production without
raising costs (at least substantially) because wage contracts, input supply contracts and the level of capital
are fixed in the short run. At the same time, demand is volatile because the economy faces constant shocks:
consumer confidence, domestic government policies, foreign demands and exchange rate policies, etc.

ˆ As apparent in equation (3.1), the interest rate, the cost of current consumption or the return on saving,
is a determinant of three out of 4 demand components. Also, the exchange rate works, together with the
interest rate, to determine the last component, i.e. imports and exports. Luckily, the central bank has the
power to set the interest/exchange rates, and therefore, capable of influencing the demand in the short run,
even though to what extent such policy can improve economic outcome is still open to debate.

ˆ A general short-run macroeconomic model is of the form:

Y
|{z} = C0 + Cy · (Y − T ) − Cr · r + I0 − Ir · r + G0 + Gy · (Ȳ − Y )
| {z } | {z } | {z }
Output produced/Supply Household Consumption Firms’ Investment Government Spending
(3.1)
+ T B0 + T Bex · (Y − T ) − T Bim · (Y − T ) + T B [0 − r (r − r∗ )] .
∗ ∗
| {z }
Trade Balance

ˆ Thus, it is important for us to understand the mechanism through which firms and households make
their consumption/investment decisions, governments decide their spending levels, and how the trade balance
varies with the changes in the interest rates and the exchange rate. Such understanding will pave the way
for us to build suitable short run macroeconomic models capable of explaining the short-run fluctuations,
thus enabling us to devise suitable countering policies. In the followings, we will step by step discuss all of
the four components of demand.

3.1 A Simple Theory of Household Consumption - C(Y-T,r)

By the household, we mean a typical household in the economy, which is endowed with certain level of
resources. For simplicity, we assume that the household only lives in two periods: now and the future. There
are three sources of wealth the household may possess, namely accumulated wealth at the beginning of period
1 (F W1 ); labor income during period 1 (Y1 ); and labor income during period 2 (Y2 ). If the household does
not have the motive to save at the end of the future, they will consume everything they have in period 2.
Instead, if they want to give some bequest F W2 to their future generation, they will leave certain amount
of wealth at period 2. Suppose that the real interest rate is r, the present value of all resources that the
household possesses is:
Y2 F W2
P V R = (1 + r)F W1 + Y1 + −
1+r 1+r

29
30 CHAPTER 3. THE SHORT RUN MACROECONOMY

ˆ Given this resources, the household face the problem of allocating consumption between now and the
future so that its level of utility obtained is maximized. Let’s denote C1 and C2 as the level of consumption
at the current and future period respectively. The decision by the household is constrained by the level of
resources, which could be formally expressed as:

C2
C1 + = PV R
1+r

ˆ By consuming C1 and C2 , the utility the household obtaining can be modeled as:

U (C2 )
U (C1 , C2 ) = U (C1 ) + ,
1+ρ

ρ is a positive term between 0 and 1, included in the model to account for the fact that households value
consumption now more than the postponed. The more impatient the household is, the larger the value of ρ.
0
U (C1 ) is assumed to be a concave function, with U (C) > 0 and U ” (C) < 0. Marginal utility from one more
unit of consumption is positive, however, the magnitude of the marginal effect decreases as consumption
increases.

ˆ The allocation facing the household is:

U (C2 )
V ≡ max U (C1 , C2 ) = U (C1 ) +
C1 ,C2 1+ρ

C2
s.t. C1 = P V R − 1+r

If you are so diligent, you may start setting up the Lagrangian and take the F.O.C. However, the simpler
C2
trick works here. Plugging C1 = P V R − 1+r into the maximized function V , and take the differentiation
w.r.t C2 to see the miracle:

C2 U (C2 )
V ≡ max U (C1 , C2 ) = U (P V R − )+
C2 1+r 1+ρ

0
∂V 0 −1 U (C2 )
= U (C1 ) +
∂C2 1+r 1+ρ
∂V
setting ∂C2 = 0, we obtain the optimal condition as:

0
U (C1 ) 1+r
= (3.2)
U 0 (C2 ) 1+ρ

ˆ This equation is known as the Euler Equation. Some further discussion here is guaranteed:
0 0
ˆ U (C1 ) and U (C2 ) are the marginal utility obtained by consuming one more unit of C1 , and one more
unit of C2 , respectively; and (1 + r) and (1 + ρ) are the marginal cost of consuming one more unit of
C1 and C2 , respectively. Intuitively, equation (3.2) states that the price of utility should be equalize,
i.e. U1+ρ
0
(C2 )
= U1+r
0
(C1 )
. Let’s imagine that you buy two chickens, weighing 1.5 and 2 kg, respectively;
you pay 30,000 VND for the smaller one and 40,000 for the bigger. The price per kilogram should be
0 0
the equal, i.e. 30,000
1.5 =
40,000
2 . Think of U (C1 ) and U (C2 ) as two chickens.

ˆ The optimal condition does not depend on Y1 and Y2 , which means that households smooth out their
consumption between now and the future, given their present value of resources and the ability to
borrow freely. In other words, what is important is the PVR, not when the PVR is earned. The
allocation of consumption between now and the future depends on the subjective discouting rate ρ and
the interest rate r. Given the same level of PVR, some households choose to be borrowers if their
subjective discount rate is higher than the interest rate (ρ > r), others choose to be lenders, when their
subjective discount rate is lower than the interest rate (ρ < r).
3.2. A SIMPLE THEORY OF FIRM’S INVESTMENT - I(R) 31

Figure 3.1: Optimal Consumption: Borrower

Figure 3.2: Optimal Consumption: Lender

ˆ Some other points to remember:

– Households may face credit constraints, preventing them from maximizing utility by borrowing
against the future. This leads to the empirical finding that consumption is closely related to
current disposable income.
– When utility function is strictly concave, consumption smoothing implies that a temporary in-
crease in income will lead household to consume only part of the increase and save part of it. A
permanent increase will lead the household to increase their consumption proportionately.
The aggregate consumption function is, therefore, specified as:

C= C0 + Cy ·(Y − T ) − Cr ·r
| {z } | {z } | {z }
Intercept: Cr > 0
capturing
Cy ∈ (0, 1)
all factors
other
than Y,
T, and r

3.2 A Simple Theory of Firm’s Investment - I(r)

The purpose of this section is to highlight the key principle of firms’ investment, equalizing marginal cost
and benefit of investment. Consider a two-period model, the periods being the current and a future period,
in which a firm considers at the beginning of the current period to purchase one additional unit of capital
so as to:
- Increase the amount of output produced in the future period (we assume that capital purchased in
the current period can only be used to produce output in the future period (it takes time to install
capital)), and
32 CHAPTER 3. THE SHORT RUN MACROECONOMY

- Sell that portion of the additional capital that has not depreciated at the end of production in the
future period.

To keep notation as simple as possible, we also assume that the prices of:

- One unit of capital, and of

- One unit of output in the current and future periods are equal to one.

The firm will purchase the additional unit of capital if the marginal (extra) revenue (MR) from doing so
exceeds the marginal cost (MC):
 

   
1  
| M{zR } = | {z
0 }+ · | M{zPK −δ
} + | 1 {z
 
current
1+r  }

in }  additional
amount
portion of 
current period
| {z of output additional
period current capital that
value of ad- has not
values depreciated
ditional unit
of future | {z }
revenue future period
 
1
| M{zC } = | {z
1 }+ · | {z
0 }
current
1+r
in future
current period period
| {z }
period current
values value of
additional
unit of
future cost
The firm will purchase the additional unit of capital if:

M P K+1−δ
MR = 1+r ≥ 1 = MC

⇔ MR
MC = T obin0 sq = M P K+1−δ
1+r ≥1

This idea can be represented graphically as:

An increase in the interest rate r lower the present value of the marginal revenue of the additional in-
vestment unit, raising the opportunity cost of investment, leading firms to cut back on their purchases of
additional capital. This idea can be formalized by taking the derivative of the the Tobin’s q with respect to
the interest rate:

The aggregate investment function is, therefore, specified as:

3.3. GOVERNMENT SPENDING 33

I= I − I ·r
| {z0 } | {zr }
Intercept: Ir > 0
capturing
all factors
other
than Y,
T, and r

3.3 Government Spending

Government spending is the least complicated of the four components of aggregate demand:

G = G0 + Gy · (Ȳ − Y ). (3.3)

Equation (3.3) reads that government spending includes an autonomous factor (G0 ) representing the rigid
part of government spending that cannot adjust easily, at least in the short run such as minimum defense,
education, health care and employee compensation spending. The term Gy · (Ȳ − Y ) represents the counter-
cyclical nature of government spending in the sense that the government will cut its spending when the
economy shows too much demand, inflation is picking up to prevent the economy from getting overheated.
On the contrary, as the economy heads into recession, consumer confidence plummets, the government will
increase in spending to stimulate the economy, increase demand with the hope of bringing the economy back
to its long-run growth path. Of course, this theory mainly reflects the best practice. Government spending
in Vietnam in the last decades has always been increasing, regardless of the business cycle. Perhaps, we need
some VGU graduates to join the Ministry of Finance so change could be initiated.

ˆ Another important factor influencing government spending is government debt, especially government
external (in foreign currency) debt. We are all aware of the Greek public debt crisis, while the threat to
Vietnam is widely discussed on the mass media recently. However, to keep things simple, we will ignore
government debt for now and will come back in later sections.

3.4 The Trade Balance

As is apparent from the lecture, the trade balance is the difference between export and import. Export
represents foreign demand, which depends on how much income foreigners have to buy our products. Import,
on the contrary, depends on how much income domestic consumers have and how fancied of foreign (by this I
mean Japanese, American or German) made products. On top of that, both import and export are strongly
influenced by the level of the exchange rate. The income you have at this point of your life mainly comes from
your parents’ subsidy, mostly in VND of course. The current exchange rate is 22,800 VND/USD. If you want
to buy an Iphone priced at 599 USD, you have to pay 13,657,200 VND. The price will be 14,076,500 VND,
14,975,000, and 17,970,000 VND if the exchange rate increase to 23,500 VND/USD, 25,000 VND/USD, and
30,000 VND/USD respectively. Given your parents’ income are mostly in VND, and thus yours, the price
of the Iphone keeps increasing even though its price in USD stays constant. The Iphone is becoming more
expensive because the VND depreciates. The reverse is true to an American who loves Vietnamese Pho
noodle.

ˆ To capture such insights, the trade balanced in this course is modeled as:
T B = T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Y − T ) + T B [0 − r (r − r∗ )]

The parameters T Bex , T Bim , and T B respectively captures the impact of an increase in foreign disposable
income (on export (+)), domestic disposable income (on import(+)), and the real effective exchange rate
(on both import (-), and export (+)).

3.4.1 The Interest Rate Parity Theory

While the impact of (Y ∗ − T ∗ ) and (Y − T ) is obvious, it is important to understand the mechanism through
which we postulate the relationship between r and r∗ , and the real effective exchange rate. Let’s start with
a question constantly asked by financial investors: should I invest in my own country’s financial market and
earn the return of 1 + rt or should I invest in the international market to earn 1 + rt∗ , and then convert
the earnings back to my own country’s currency to earn the investment return of 1t · (1 + rt∗ )t+1 . At any
34 CHAPTER 3. THE SHORT RUN MACROECONOMY

given point in time, if 1 + rt > 1t · (1 + rt∗ )t+1 , investors will sell their assets in the international market
and buy the domestic assets, raising the demand for domestic currency. The domestic currency appreciates,
depressing the 1t term in 1t · (1 + rt∗ )t+1 . The reverse is true if 1 + rt < 1t · (1 + rt∗ )t+1 . The process of
adjustment will keep going until:
1
1 + rt = · (1 + rt∗ )t+1 . (3.4)
t
Rearranging terms in (3.4) to get:
1 + rt∗
t = t+1 . (3.5)
1 + rt
Equation (3.5) states that the current real effective exchange rate t depends on the domestic and interna-
tional interest rate, and the expectation of future exchange rate. If the central bank increase the interest
rate rt , t becomes smaller, the domestic currency appreciates. Keep replacing s by s+1 recursively to get:

1 + rt∗
t = t+1 .
1 + rt
∗
1 + rt∗ 1 + rt+1
t = t+2 .
1 + rt 1 + rt+1
∗ ∗
1 + rt∗ 1 + rt+1 1 + rt+2
t = t+3 .
1 + rt 1 + rt+1 1 + rt+2
∗ ∗ ∗
1 + rt∗ 1 + rt+1 1 + rt+2 1 + rt+3
t = t+4 .
1 + rt 1 + rt+1 1 + rt+2 1 + rt+3
s=3
Y ∗ 
1 + rt+s
t = t+s+1 .
s=0
1 + rt+s

finally,
s=∞
Y  ∗ 
1 + rt+s
t = ∞ .
s=0
1 + rt+s

In the long run, the nominal exchange rate fully adjust to align with the price level at home and abroad,
ensuring that ∞ = 1. Thus the interest rate parity condition:
s=∞  ∗ 
1 + rt∗ Y 1 + rt+s
t = , (3.6)
1 + rt s=1 1 + rt+s

Given no information on future domestic and foreign interest rates, equation (3.6) states that if the domestic
interest rate r increases, the domestic currency will appreciate. If international interest rate is higher than
the domestic interest rate, the domestic currency will depreciate. This insight is formalized as:

 = 0 − r · (r − r∗ ), 0 < r < 1

3.4.2 The Balance of Trade Function Form

The above discussion allows for the specification of a simple function form relating the trade balance with
its determinants as:
T B = T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Y − T ) + T B 
with
 = 0 − r (r − r∗ )

3.5 Exercises
Problem 3.1
Consider a household that makes consumption decisions for two periods, t = 1, 2. The household’s real
financial wealth at the beginning of period 1 amounts to 5,000 Euro; its real disposable income amounts to
25,000 Euro in period 1 and to 30,000 Euro in period 2. The household at the end of period 2 wishes to
3.5. EXERCISES 35

have real financial wealth in the amount of 4,000 Euro. The real interest rate between period 1 and period
2 is r = 0.01. The household maximizes utility from consumption according to the objective function
U (C2 ) (C1 )1−κ 1 (C2 )1−κ
U (C1 ) + = +
1+ρ 1−κ 1+ρ 1−κ
where ρ = 0.02 denotes the subjective rate of discounting, and κ = 0.5 denotes a coefficient in the instanta-
neous utility function u(C).

a) Obtain the ratio at which this household wishes to consume in period 1 relative to period 2. Also
obtain this household’s optimal levels of consumption in periods 1 and 2.

b) How does your answer to (a) change if the household in period 1 had real disposable income amounting
to 30,000 Euro? What therefore is this household’s marginal propensity to consume?

c) How would your answer to (b) change if the household’s subjective rate of discounting increased?

d) How does your answer to (a) change if the real interest rate increased to r = 0.015? What does this
answer suggest concerning the relative strengths of the substitution and income effects?

Problem 3.2
Consider the aggregate consumption function:
C = C0 + Cy · (Y − T ) − Cr · r
a) Which parameter refers to the marginal propensity to consume? Assuming a strictly concave utility
function, set up the household’s consumption maximization problem and explain why the marginal propen-
sity to consume is assumed to be positive and smaller than one.

b) Using the terms substitution effect and income effect to explain how a consumer who inter-temporally
maximizes consumption reacts to an increase in the interest rate. Does it matter whether the consumer is a
creditor or a borrower? What is the reaction when the interest rate decreases?

c) Based on your reasoning in part b), explain why the parameter Cr in the aggregate consumption
function is assumed to be positive.

Problem 3.3
Suppose that the utility function of a representative household could be represented by: u(C) = log(C).
Assuming that households maximize their utility by smoothing between consumption now (C1 ), and con-
sumption in the future (C2 ).

a) Derive the Euler equation for this household and show that current consumption, C1 could be ex-
pressed as C1 = α · P V R.

b) Suppose that the subjective discount factor ρ = 0.02, r = 0.01, Y1 = Y2 = 10, 000, T1 = T2 = 2000,
determine the value of consumption now, and consumption in the future.

b) What is the effect of a decrease in PVR on C1 ?

c) Does it matter whether the decrease in PVR today results from a decrease in current and/or future
disposable income?

d) Suppose the individual receives an announcement today that income in the future might be lower than
previously expected. What is the effect on consumption today?

Problem 3.4
A representative household maximizes its strictly concave utility by smoothing consumption between now
(C1 ), and the future (C2 ). Financial wealth at the beginning of period 1 is W1 , and at the end of period 2
is W2 . Its labor income in periods 1 and 2 are Y1 and Y2 respectively. It’s utility function is defined as:
36 CHAPTER 3. THE SHORT RUN MACROECONOMY

1
U = u(C1 ) + 1+ρ u(C2 ),

where ρ is the subjective discounting factor, and the prevailing interest rate is r.

a) Setting up the household’s maximizing problem of smoothing consumption between periods 1 and 2.

b) Derive the impact of a change in the present value of resources on current consumption, C1 .

Answer Rewrite the Euler equation as:

0 0
U (C1 )(1 + ρ) = U (C2 )(1 + r),

and take notice of the budget constraint: C2 = (P V R − C1 ) · (1 + r), taking the total differentiation of (3.23)
to get:
00 00 ∂C2 ∂C2 ∂C2 0
(1 + ρ)U (C1 )dC1 = U (C2 )[ dC1 + dP V R + dr] + U (C2 )dr
∂C1 ∂P V R ∂r
00 00 0
(1 + ρ)U (C1 )dC1 = (1 + r)U (C2 )[−(1 + r)dC1 + (1 + r)dP V R + (P V R − C1)dr] + U (C2 )dr.
Rearranging terms:
00 00 00 00 0
[(1 + ρ)U (C1 ) + (1 + r)2 U (C2 )]dC1 = (1 + r)2 U (C2 )dP V R + [(P V R − C1 )(1 + r)U (C2 ) + U (C2 )]dr
00 00 0
(1 + r)2 U (C2 ) (P V R − C1 )(1 + r)U (C2 ) + U (C2 )
dC1 = dP V R + dr (3.7)
(1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 ) (1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 )
Set dr = 0, we can solve for the impact of a change dPVR on current consumption C1 as:
00
dC1 (1 + r)2 U (C2 )
=
dP V R (1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 )
00
By assumption, U (C) < 0, which means that: dPdCV1R > 0. The result also indicates that dC1
dP V R < 1.
c) Derive the impact of a change in the interest rate on current consumption, C1 .

Answer
The impact of a change in the interest rate r on current consumption can be decomposed into two separate
effects:

dC1 dC1 ∂C1 ∂P V R

= |dP V R=0 + ,
dr |dr {z } ∂P V R ∂r
∂C1
∂r

the first effect, dC dC1 P V R

dr |dP V R=0 is known as the substitution effect, while dP V R dr is called the income effect,
1

i.e. because r changes, PVR changes, thus changing the consumption behavior. To compute dC dr |dP V R=0 ,
1

dC1
we first set dP V R = 0 in (3.7) and solve for dr to get:
0 00
dC1 U (C2 ) + (P V R − C1 )(1 + r)U (C2 )
|dP V R=0 = . (3.8)
dr (1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 )

Different from the case of dPdCV1R , we cannot say unambiguously the sign of dC1
dr |dP V R=0 . To proceed, we
separate the right hand side of (3.8) into two different components:
0 00
dC1 U (C2 ) (P V R − C1 )(1 + r)U (C2 )
|dP V R=0 = + .
dr (1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 ) (1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 )

Take notice that:

0 0 00
U (C2 ) U (C2 )(1 + r)2 U (C2 )
00 00 =
2
(1 + ρ)U (C1 ) + (1 + r) U (C2 ) [(1 + ρ)U (C1 ) + (1 + r)2 U 00 (C2 )](1 + r)2 U 00 (C2 )
00

0 0
U (C2 ) U (C2 ) dC1
=
(1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 ) (1 + r)2 U 00 (C2 ) dP V R
3.5. EXERCISES 37

and
00
(P V R − C1 )(1 + r)U (C2 ) (P V R − C1 ) dC1
=
(1 + ρ)U 00 (C1 ) + (1 + r)2 U 00 (C2 ) (1 + r) dP V R
so
0
dC1 U (C2 ) dC1 (P V R − C1 ) dC1
|dP V R=0 = +
dr (1 + r)2 U 00 (C2 ) dP V R (1 + r) dP V R
0
Now consider the elasticity of the marginal utility of C2 , U (C2 ), with respect to C2 :
00
U (C2 )C2
C 2
U 0 (C2 )
= ,
U 0 (C2 )

which shows how the marginal utility of consuming one more unit of C2 decreases as the amount of C2
increases, and define:
0
U (C2 )
σ = − 00
U (C2 )C2
dC1
as the elasticity of inter-temporal substitution, dr |dP V R=0 becomes:

dC1 P V R − C1 σC2 dC1

|dP V R=0 = [ − ] .
dr 1+r (1 + r)2 dP V R

C2
Replacing (1+r) = P V R − C1 into (3.38),

dC1 P V R − C1 σ(P V R − C1 ) dC1

|dP V R=0 = [ − ] .
dr 1+r (1 + r) dP V R

or
dC1 1−σ dC1
|dP V R=0 = (P V R − C1 ) .
dr 1+r dP V R
When PVR is kept constant (no wealth effect), the substitution effect depends on σ. If σ > 1, equivalently,
00
0
C 2
U 0 (C2 )
= U U(C
0
2 )C2
(C2 )
< 1, meaning that U (C2 ) is inelastic with respect to C2 , the household does not have a
problem saving more for the future consumption, an increase in the interest rate r will have a strong negative
impact on current consumption. However, when the household put a high price on future consumption, a
large increase in future consumption will have a small reduction in the marginal utility of consumption (con-
0
sumption in the second period is extremely high), U (C2 ), a change in the interest rate will have a negative
impact on current consumption. In other words, when current consumption is very high, C2 is very low, the
elasticity of marginal utility of future consumption is very large, a change in the interest rate induces the
family to save more and move consumption from now to the future. Thus, the substitution effect dC dr |dP V R=0
1

is ambiguous, it depends on the characteristics of the household and their current level of consumption.

Y2
Recall that P V R = (1 + r)F W1 + Y1 + 1+r − F1+r
W2
, the effect on PVR due to a change in the interest
rate is:
∂P V R Y2 F W2
= F W1 − 2
+
∂r (1 + r) (1 + r)2

If the household is very rich (F W1 is large), or wants to give everything to their children (hence F W2 is
large), ∂P∂rV R > 0. In contrast, if the household is a debtor, (F W1 is negative)and they don’t have a sin-
gle hope of helping their children, (hence F W2 =0), the ∂P∂rV R < 0. As we have in the previous section,
dC1 dC1 P V R
dP V R > 0, thus dP V R dr is also ambiguous.

It is for this reason we concluded during the lecture, graphically, that the impact of a change in the
interest rate on current consumption

dC1 dC1 dC1 P V R

= |dP V R=0 + ,
dr dr dP V R dr

is ambiguous.
38 CHAPTER 3. THE SHORT RUN MACROECONOMY

Problem 3.5
Assume a simple economy with only one representative household that owns a firm and that there is no
capital to begin with (in the metaphor of Robinson Crusoe, there are no trees on the island). The key
decision is: how much investment today (I1 ) (coconuts to plant). This will give rise to productive capital
tomorrow K2 = I1 . Let Yi denotes period i’s labor income, i =1,2, and r denotes the interest rate.

a) Define the budget constraint facing this household.

b) Formulate the maximization problem facing the household so that the present value of resources is
maximized.

Problem 3.6
Consider a firm that operates only for two periods, t = 1, 2. The firm produces output, Yt , with the
production function Yt = Alog(Kt ), t = 1, 2, where A is the level of technology, log denotes the natural
logarithm, and Kt the firm’s capital stock. Suppose that the firm’s revenue and cost in period 1 are given
by Revenue1 = Y 1, and Cost1 = p1 .I1 , where I1 is the firm’s investment in the physical capital stock of
period 2, and K1 is exogenously given. At the end of each period, a fraction δ of the firm’s capital stock
depreciates. The firm discounts future profits at the market rate of interest, denoted r.

a) Given investment I1 and capital stock K1 in period 1, what is the firm’s capital stock in period 2?

Answer: The firm’s capital stock in period 2 includes the capital from period 1, K1 , minus the depreciation
δK1 , and plus the new investment I1 :
K2 = (1 − δ)K1 + I1
b) Determine for general parameter values the period 1 present value of revenue as well as period 1
present value of cost, if the firm can sell its remaining capital after production in period 2 at real price p2
and if the real interest rate between periods 1 and 2 is r.

Answer:

ˆ The present value of costs is:

P V C = p1 .I1

ˆ The present value of revenue is:

Alog((1 − δ)K1 + I1 ) p2 (1 − δ) [(1 − δ)K1 + I1 ]

P V Rev = Alog(K1 ) + +
1+r 1+r

c) Derive for general parameter values the condition for the firm’s optimal investment in period 1.

Answer:

The present value of wealth of the firm is:

P V W = P V Rev − P V C

Alog((1 − δ)K1 + I1 ) p2 (1 − δ) [(1 − δ)K1 + I1 ]

P V W = Alog(K1 ) + + − p1 .I1
1+r 1+r
The firm’s maximization problem is:

Alog((1 − δ)K1 + I1 ) p2 (1 − δ) [(1 − δ)K1 + I1 ]

V ≡ max P V W = Alog(K1 ) + + − p1 .I1
I1 1+r 1+r

The F.O.C is:

∂V A p2 (1 − δ)
= + − p1
∂I1 (1 + r)(1 − δ)K1 + I1 ) 1+r
or
∂V A p2 (1 − δ)
= + − p1 .
∂I1 (1 + r)(1 − δ)K1 + I1 ) 1+r
3.6. THE IS CURVE 39

∂V
Set ∂I1 = 0, we find the optimal level of investment must satisfied:

A p2 (1 − δ)
+ = p1 .
(1 + r)(1 − δ)K1 + I1 1+r

A p2 (1 − δ)
= (1 + r)p1 − .
(1 − δ)K1 + I1 1
A
(1 − δ)K1 + I1 =
(1 + r)p1 − p2 (1 − δ)
A
I1 = − (1 − δ)K1
(1 + r)p1 − p2 (1 − δ)

Problem 3.7
Suppose the following interest parity relationship holds:
1
1 + rt = · [1 + (1 − τ ∗ ) · rt∗ ] · t+1
t
where τ ∗ denotes the rate at which capital returns are taxed (for all financial investors, irrespective of their
nationality) in the foreign economy. Also suppose that purchasing power parity holds for the long-run ahead
value of the real exchange rate. Now consider an increase of τ ∗ . Derive algebraically the impact of this
increase. Provide a detailed economic rationale this has on the current value of the real exchange rate
between the domestic and the foreign currencies.
Answer:

Rewrite the above equation as:

s=∞ 
1 + (1 − τ ∗ ) · rt∗ Y (1 + (1 − τ ∗ )rt+s
∗ 
t =
1 + rt s=1
1 + rt+s

Qs=∞ h (1+(1−τ ∗ )rt+s

∗ i
Given no information on the future interest rates and capital tax, the term s=1 1+rt+s is best
guessed to be a constant, denoted as CONST, CON ST > 0. The equation becomes:

1 + (1 − τ ∗ ) · rt∗
t = CON ST
1 + rt
Take the partial differentiation of t w.r.t τ ∗ and get:
∂t −rt∗
∗
= CON ST,
∂τ 1 + rt
which is unambiguously negative. We can conclude that an increase in the foreign tax on capital will reduce
the real effective exchange rate.

Economic Rationale: Higher tax rate τ ∗ in the foreign country reduces the return on foreign assets.
Financial investors will sell their foreign assets and buy domestic assets until the interest rate parity condition
is restored. The domestic currency appreciates, leading the real exchange rate between the domestic and the
foreign currencies to also decrease.

3.6 The IS Curve

Having gone through so much pain, it is now time bring all components of demand into one bag and name
it Desired Demand or DD.

| DD
{z } = C0 + Cy · (Y − T ) − Cr · r + I0 − Ir · r + G0 + Gy · (Ȳ − Y )
| {z } | {z } | {z }
Desired demand Household Consumption Firms’ Investment Government Spending
40 CHAPTER 3. THE SHORT RUN MACROECONOMY

+ T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Y − T ) + T B [0 − r (r − r∗ )]

| {z }
Trade Balance

As we have emphasized class after class, total supply (what all of us produced) must be equal to total
desired demand (what all of us consume plus the consumption by foreign people and minus the proportion
of consumption imported). Put it differently, Y must be equal to DD, so:

| Y
{z } = | C0 + Cy · (Y{z− T ) − Cr · r }
+
|
I0 − Ir · r
{z }
+
|
G0 + Gy · (Ȳ − Y )
{z }
Output/GDP/Supply Household Consumption Firms’ Investment Government Spending
in the short run

+ T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Y − T ) + T B [0 − r (r − r∗ )]

| {z }
Trade Balance

as we boldly postulate at the beginning of the story of the macroeconomy in the short run. Rearrange
terms to group exogenous and endogenous components together to get:

Y }
| {z = C0 + I0 + G0 + T B0 + T B · 0 − (Cy − T Bim ) · T + Gy · Ȳ + T Bex · (Y ∗ − T ∗ ) + T B · r · r∗
| {z }
Output/GDP/Supply A (autonomous component of desired aggregate demand)
in the short run

+ (Cr + Ir + T B · r ) · r + (Cy − T Bim − Gy ) · Y

| {z } | {z }
Interest rate - sesitive components of aggregate demand Income - sesitive components of aggregate demand

The autonomous components bunch together the exogenous (roughly means: outside the model) sub-
components that the model does not try to determine, rather we take them for granted, as if they are given
to us by a natural superior. We do not know their exact value. Macroeconomists try to collect data from
the real world and make the estimation of these terms. Moving the income-sensitive term to the left yields:

[1 − (Cy − T Bim − Gy )] · Y = A − (Cr + Ir + T B · r ) · r

Dividing both sides by [1 − (Cy − T Bim − Gy )] yields:

A (Cr + Ir + T B · r )
Y = − ·r (3.9)
1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy )

Equation (3.9) is the famous IS curve. We derive this equation by setting aggregate supply equal aggregate
demand, and therefore, the IS curve represents the combination of output, Y, and the interest rate, r, so
that the goods market is in equilibrium.
3.6. THE IS CURVE 41

The IS curve divides the (r, Y) plane into the excess supply and excess demand half-planes.

If for some reason, the central bank judges that there are signs of overheating in the economy, it might
raise the interest rate from R to R1 . The economy suddenly find its self operating at a disequilibrium be-
cause aggregate demand falls to Y1 while supply is still at point Y. There will be excess supply, indicated
0
by the gap E1 E1 , firms cannot sell their product and find their inventory increases. They are forced to cut
42 CHAPTER 3. THE SHORT RUN MACROECONOMY

down the current level of production and output falls from Y to Y1 , the new equilibrium is now point E1 .
Suppose further that the economy is now operating at the new equilibrium E1 thanks to the change in the
interest rate. However, the central bank recognizes that it has made a mistake and that the economy is now
heading to a recession as the interest rate is too high. The central bank is forced to cut the interest rate
from R1 to R2 to save the economy from an unnecessary slowing down due to the central bank’s mistake.
0
At the new interest level, the economy once more find itself operating at an unsustainable pont E2 . At such
a low level of the interest rate, demand is as much as Y2 , however, supply is still at Y1 . Firms will find
that their inventory is depleted quickly, therefore, they will increase their output from Y1 to Y2 . The story
can go on as the central bank again realizes that it has cut the interest rate by too much and the econ-
omy now is operating over its potential output, a new rate hike is called for, and it may make a mistake again.

The above discussion seems to be lighthearted. However, it highlights two important points, one is the
IS does divides the (r, Y) space into excess demand and excess supply half-planes. The other is it is really a
hard job for the central bank to correctly judge the level of actually output relative to its potential. Mistake
is not an abnormal but a norm at many old-style central banks which lack well train economists and experts.

3.6.1 Exercises on the IS curve

Problem 3.8
Assume the following specific functions for C, I , G and TB:

C = 100 + 0.7 · (1 − t) · Y

I = 80 − 500 · r

G = 200

T B = 0.2 · (1 − t∗ ) · Y ∗ − 0.3 · (1 − t) · Y + 20 · 
Furthermore, assume that the interest rate equals 5%, the real exchange rate equals 2 and foreign income
equals 1,000. The tax rates t and t∗ are 0.2 and 0.1, respectively.

a) Derive the analytical expression of Desired Demand as a function of income.

b) Plot Desired Demand as a function of income. How do you interpret the slope?

c) Calculate the Goods Market Equilibrium.

Problem 3.9
The real sector of an economy under fixed prices (for which foreign variables are exogenous) is described by
the following relations:
Y = DD = C + I + G + T B;

C = C0 + Cy [(1 − t) · Y + GT ] ;

I = I0 − Ir · r;

G = G0 ;

T B = T B0 + T Bex · [(1 − t∗ ) · Y ∗ + GT ∗ ] − T Bim · [(1 − t) · Y + GT ] + T B · ;

 = 0 − r (r − r∗ )
where t (t∗ ) denotes the domestic (foreign) rate of income taxation and GT (GT ∗ ) denotes transfer payments
by the domestic (foreign) government to domestic (foreign) private households; all other notation is as in
the lectures on the macroeconomy in the short run. When numerical answers are asked for, you may assume
the following parameter values: C0 = 1.45; Cy = 0.75; t = 0.15; GT = 0.25; I0 = 3.25; Ir = 15; r = 0.03; G0 =
2.8; T B = 0; T Bex = 0.15; t∗ = 0.1; Y ∗ = 11; GT ∗ = 0.05; T Bim = 0.15; T B = 1; 0 = 1; r = 1; r∗ = 0.03.

a) Derive (for general parameter values) the Keynesian multiplier describing how changes of the au-
tonomous component of aggregate output demand affect short-run output.
3.6. THE IS CURVE 43

Answer: First, it is best to expand all the products to separate the endogenous and exogenous components
from each equation:

Y = DD = C0 + Cy · GT + Cy (1 − t) · Y + I0 − Ir · r + G0 + T B0 + T Bex · (1 − t∗ ) · Y ∗ + T Bex · GT ∗

−T Bim · (1 − t) · Y − T Bim · GT + T B · 0 − T B · r · r + T B · r · r∗
Next, group the terms into groups according to exogenous, Y, and r terms:

Y = C0 + I0 + G0 + T B0 + T Bex · (1 − t∗ ) · Y ∗ + T Bex · GT ∗ − T Bim · GT + T B · 0 + T B · r · r∗

+Cy · GT + Cy (1 − t) · Y − T Bim · (1 − t) · Y − Ir · r − T B · r · r.
Then, move the Y terms to the left to get:

[1 − Cy (1 − t) + T Bim · (1 − t)] · Y = C0 + Cy · GT + I0 + G0 + T B0 + T Bex · (1 − t∗ ) · Y ∗ + T Bex · GT ∗

−T Bim · GT + T B · 0 + T B · r · r∗
−(Ir + T B · r ) · r
Finally, the model could be written in the IS form:
C0 + Cy · GT + I0 + G0 + T B0 + T Bex · (1 − t∗ ) · Y ∗ + T Bex · GT ∗ − T Bim · GT + T B · 0 + T B · r · r∗
Y =
1 − Cy (1 − t) + T Bim · (1 − t)
Ir + T B · r
− ·r
1 − Cy (1 − t) + T Bim · (1 − t)

Therefore, the general values of the Keynesian multiplier is:

1
KM =
1 − Cy (1 − t) + T Bim · (1 − t)

b) Derive (for general parameter values) the IS curve. Provide a detailed economic rationale as to how
the slope of the IS curve depends on (i) T Bim and (ii) r

Answer: From question a), the IS curve is defined by the equation:

C0 + Cy · GT + I0 + G0 + T B0 + T Bex · (1 − t∗ ) · Y ∗ + T Bex · GT ∗ − T Bim · GT + T B · 0 + T B · r · r∗

Y =
1 − Cy (1 − t) + T Bim · (1 − t)

Ir + T B · r
− ·r
1 − Cy (1 − t) + T Bim · (1 − t)
c) The short run economic output is 19.65816327.
ˆ The slope of the IS curve is defined as − 1−Cy (1−t)+T
Ir +T B ·r
Bim ·(1−t) . T Bim enters the nominator of the slope
formula with a positive sign, and because (1-t) is positive, the higher the value of T Bim , the steeper
the IS curve.The rational for this is that the multiplier effect decreases when T Bim rises and therefore
a change in the interest rate following a change in the autonomous component has a smaller effect on
output.
ˆ r enters to the nominator of the slope. The higher the value of r , the flatter the IS curve. Output
responses more strongly to a change in the interest rate. The economic rationale for this interpretation
is that the higher values of r means the exchange rate appreciate more strongly in response to an
increase in the interest rate, thus in combination with the decreases in consumption and investment
having a strong negative impact on output.

d) Provide a detailed economic rationale as to how the size of the Keynesian multiplier depends on (i)
T Bim and (ii) T Bex .

Answer:
ˆ The parameter T Bim represents the propensity to import consumption. The higher the value of T Bim ,
the lower the value of the KM due to import leakage. When T Bim is high, stimulus economic policy
will be less effective.
44 CHAPTER 3. THE SHORT RUN MACROECONOMY

ˆ An increase in the propensity to export T Bex leaves the Keynesian Multiplier unaffected because de-
mand of domestic citizen for home output is not affected by foreign demand decisions.

e) Calculate and compare the short-run output effects of (a) an increase in domestic government transfer
payments to domestic private households, GT, by 0.1 with (b) an increase in domestic government expendi-
ture through an increase of G0 , also by 0.1. Provide a detailed economic rationale for your comparison.

Answer: The short run output following the increase by 0.1 in GT and G0 are 20.36183673 and
20.28020408 respectively. The increase in G0 has a stronger impact on short run output because G0 en-
ter the autonomous aggregate demand directly, while GT enters indirectly via the impact on consumption
and import. Analytically, it could be shown that:
∂Y
∆Y GT = ∆GT
∂GT
∂A
= ∆GT
∂GT
1
= ∆GT · (Cy − T Bim ) ·
1 − Cy (1 − t) + T Bim · (1 − t)
= 0.1 · 1.22 = 0.122
Meanwhile,
∂Y
∆Y G0 = ∆G0
∂G0
∂A
= ∆G0
∂G0
1
= ∆G0
1 − Cy (1 − t) + T Bim · (1 − t)
= 0.1 · 2.04 = 0.204

Problem 3.10
Assume the economy can be described by the following system of linear equations:

C = C0 + Cy · (1 − t) · Y
I = I0 − Ir · r
G = G0
T B = T Bex · (1 − t ) · Y ∗ − T Bim · (1 − t) · Y + T B · 
∗

 = 0 − r · (r − r∗ )
a) Derive the Keynesian Multiplier.

b) If t = 0.25; Cy = 0.6, and T Bim = 0.25, What will be the effect on output Y if the government
increases its expenditure G by 5 billion? Explain the intuition behind your result.

Problem 3.11
Assume
Y = C + I + G + TB
with
C = C0 + Cy · (Y − T ) − Cr · r
I = I0 − Ir · r
G = G0 + Gy · (Ȳ − Y )
T B = T Bex · (1 − t∗ ) · Y ∗ − T Bim · (1 − t) · Y + T B · 
3.7. IS-TR 45

 = 0 − r · (r − r∗ )
a) Derive the Keynesian Multiplier in this framework.

b) Derive how the Multiplier depends on Cy , T Bim and Gy . What does it mean, if, respectively, Cy ,
T Bim or Gy increase? Explain in each case why the Multiplier increases or decreases.

3.7 IS-TR
3.7.1 Monetary Policy Rate
The IS curve above depicts a simple relationship between output in the short run and the interest rate. It’s
now the time for us to discuss how the interest rate is determined in the economy. As noted during the
lecture, the old approach excellently described in Jone’s book says that the interest rate is the result of the
money demand in the economy and the quantity of money supply by the central bank. The central bank
controls the money supply and adjust the quantity of this supply to change the interest rate in the economy.

However, reality shows that it is inherently difficult for the central bank to control the quantity of money
supply. In practice, the central set the interest rate and supply whatever amount of money needed to
maintain that interest rate. Therefore, it is best to describe the interest setting activity by modern central
banks by using the Taylor Rule formula which states that:
 
Y − Ȳ
RM P = MP
r̄|{z} +π e + φy +φπ · (π − πT )
Ȳ |{z}
neutral real rate > 0 | {z } inflation target
(relative) ouput gap

Furthermore, we are only concerned with the short run in this chapter, therefore, price is constant, and
π = π T = π e = 0. Defining rM P = RM P − π e , the real monetary interest rate. The Taylor rule reduces to:
 
Y − Ȳ
rM P = r̄M P + φy (3.10)
Ȳ

3.7.2 The lending rate

Commercial banks are firms. Their ultimate goal is to obtain funding (deposits, loans, equity, ect.) at some
interest rate, say rdeposit , and lend the fund to investors/consumers at the rate of rdeposit + r̄RP , where RP
stands for risk premium and r̄RP > 0. For simplicity, we assume that banks can mobilize the fund at the
rate rM P , and offer loans at a single rate rM P + r̄RP .

3.7.3 The TR curve

The TR curve is, thus, defined as:

 
Y −Ȳ
r = rM P + rRP = r̄M P + φy Ȳ
+ r̄RP

Problem 3.12
Consider two central banks that both use the following Taylor rule to set their monetary policy rate:
 
MP MP Y − Ȳ
R = r̄ + φy
Ȳ

Suppose the parameter φy is larger for the first of these two central banks than for the second one. Provide
a detailed economic rationale as to how the two central banks’ monetary policy rate decisions differ when (i)
Y > Ȳ and (ii) Y < Ȳ . Answer: The parameter φy represents how sensitive the central bank is to a shock
that deviates the economy from its potential output. The higher the value of φy , the stronger the change in
the interest rate in the face of a shock to the economy. When (i) Y > Ȳ , the first central bank will raise the
interest rate by a amount that is higher than the react of the second central bank to a shock of the same
magnitude. Similarly, when Y < Ȳ the first central bank will cut the interest rate more than the second
one.
46 CHAPTER 3. THE SHORT RUN MACROECONOMY

Problem 3.13

Consider the real interest rate prediction of the model of commercial bank lending considered in the lectures:

rM P − ν · ALo
r=
1 + ν · Lor

where
V ar(lrp) · χ
ν=
Eq

How does the real interest rate respond to a decrease of Eq? Derive this effect algebraically. (Hint: Calculate
the partial derivative of r with respect to Eq, and then re-write this partial derivative in such form that
you can invoke the condition that the loan demand must be strictly positive.) Provide a detailed economic
rationale for your result.

Answer:
∂r ∂r ∂ν
=
∂Eq ∂ν ∂Eq

ALo · (1 + νLor ) − Lor (rM P + νALo V ar(lrp)

= ·χ
(1 + νLor )2 Eq 2

rM P + νALo
 
−1 Lo V ar(lrp)
= A − Lor · χ (3.11)
1 + νLor 1 + νLor Eq 2
| {z }
Loan Demand

∂r
It can be seen from 3.11 that ∂Eq is negative and,

∂r
∆r = · ∆Eq.
∂Eq

When ∆Eq is negative, ∆r is, therefore, positive. When Eq is smaller, the ratio of loan to equity also
increases. Therefore, banks start charging a higher interest rate because their operation becomes more risky.

3.7.4 The IS-TR model

The IS curve represents the equilibrium on the goods/services market:

A (Cr + Ir + T B · r )
Y = − ·r
1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy )

The TR curve represents the equilibrium on the financial market where the amount of loans supplied is equal
to the amount of loan demanded:
 
Y − Ȳ
r = rM P + rRP = r̄M P + φy + r̄RP
Ȳ

Together, they determine the equilibrium in the interest rate - production/consumption space:
3.8. EXERCISES ON THE IS-TR MODEL 47

3.7.5 Quantitative analysis of the IS-TR

Given the TR curve, the full macroeconomic model in the short run is described by the IS curve:
A (Cr + Ir + T B · r )
Y = + ·r
1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy )
and the TR curve, rewritten in terms of output as:
 
Ȳ
· r − r̄M P − r̄RP


Y = Ȳ +
φy
Combining this equation with the IS equation yields:
 
Ȳ A (Cr + Ir + T B · r )
· r − r̄M P − r̄RP =


Ȳ + + ·r
φy 1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy )
r̄M P + r̄RP
   
Ȳ (Cr + Ir + T B · r ) A
+ ·r = − Ȳ · 1 −
φy 1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy ) φy
or  
A r̄ M P +r̄ RP
1−(Cy −T Bim −Gy ) − Ȳ · 1 − φy
r= (Cr +Ir +T B ·r )
Ȳ
φy + 1−(Cy −T Bim −Gy )
The interest rate is now determined solely by the model parameters. Having computed the interest rate, the
output can be estimated using the IS equation:
A (Cr + Ir + T B · r )
Y = − ·r
1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy )
 
A r̄ M P +r̄ RP
A (Cr + Ir + T B · r ) 1−(Cy −T Bim −Gy ) − Ȳ · 1 − φy
= − · (Cr +Ir +T B ·r )
1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy ) Ȳ
+
φy 1−(Cy −T Bim −Gy )

3.8 Exercises on the IS-TR model

3.8.1 Problem 3.14
Consider the ISTR model and an economy for which

A (Cr + Ir + T B · r )
Y = − ·r
1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy )
48 CHAPTER 3. THE SHORT RUN MACROECONOMY
 
Y − Ȳ
r = rM P + rRP = r̄M P + φy + r̄RP
Ȳ
with A = 10, cy = 0.7, T Bim = 0.1, Gy = 0.1, Ir = 15, Cr = 4, T B = 1, r = 1, r̄M P = 0 : 02, r̄RP = 0 : 02,
φy = 0.5, Ȳ = 16. (a) Determine the goods market equilibrium for r = 0.05.

(b) Determine the money market equilibrium for Y = 15.

(c) Determine the overall short run macroeconomic equilibrium. What are the values for Y and r?

3.8.2 Problem 3.15

In the standard IS-TR model presented in the subsection 3.7.5, suppose that A = 9.63, Cy = 0.6, Gy = 0.1,
T Bim = 0.1, T Bex = 0.1, Cr = 4, Ir = 16, Ȳ = 15, r̄M P = 0.02, r̄RP = 0.01, φy = 1, T B = 1, r = 1.

a) Determine the values of r and Y using the parameter values.

b) Determine the impact of an increase in foreign income by 5, i.e. ∆Y ∗ = 5.

3.8.3 Problem 3.16

On the IS - TR model: An economy under fixed prices (for which foreign variables are exogenous) is described
by the following model:

Y = DD = C + I + G + T B;
C = C0 + Cy · (Y − T ) − Cr · r;
I = I0 − Ir · r;
G = G0 − Gr · r;
T B = T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Y − T ) + T B · ;
 = 0 + r · (r − r∗) 
Y −Ȳ
r = r̄M P + φy · Ȳ
+ r̄RP .

When numerical calculations are asked for, assume the following parameter values: C0 = 1.6; Cy = 0.55; Cr =
3; T = 2.5; I0 = 3.25; Ir = 15; G0 = 2.8; Gr = 1; T B0 = 0; T Bex = 0.1; Y ∗ = 10; T ∗ = 3; T Bim = 0.15; T B =
1; 0 = 1; r = 1; r∗ = 0.03; r̄M P = 0.02; φy = 1; Ȳ = 15; r̄RP = 0.01.
a) Derive (for general parameter values) the IS - curve. Provide an economic rationale as to how the
slope of the IS - curve depends on the model parameters.

b) What is the value of output in the short-run macroeconomic outcome?

c) Calculate the effects on domestic short-run output of a decrease in the foreign interest rate, r∗ , by
0.01? Decompose the overall output effects into its components (that is, the Keynesian multiplier induced
effect and the domestic interest rate induced crowding out-in effects). Make sure to provide, in addition to
numerical results, a detailed economic rationale as to how the various effects arise.

Answer:

a) Again, the first step is to expand all the products to separate the exogenous from the endogenous.

Y = C0 + Cy · Y − Cy · T − Cr · r + I0 − Ir · r + G0 − Gr · r + T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · Y +
T Bim · T + T B · (0 + r r∗ ) − T B · r · r.

Next, group terms according to three categories: Autonomous, Y, and r:

Y = C0 + I0 + G0 + T B0 + T Bex · (Y ∗ − T ∗ ) + T B · (0 + r r∗ ) − Cy · T + T Bim · T + Cy · Y − T Bim ·

Y − Cr · r − Ir · r − Gr · r − T B · r · r.

Y · (1 − Cy + T Bim ) = A − (Cr + Ir + Gr + T B · r ) · r.

A Cr + Ir + Gr + T B · r
Y = − · r. (3.12)
1 − Cy + T Bim 1 − Cy + T Bim
3.8. EXERCISES ON THE IS-TR MODEL 49

Where A = C0 + I0 + G0 + T B0 + T Bex · (Y ∗ − T ∗ ) + T B · (0 + r r∗ ) − Cy · T + T Bim · T . The general IS

1−Cy +T Bim
curve is described by equation 3.12. The slope of the IS curve is Cr +Ir +G r +T B ·r
, which depends positively
on T Bim ; negatively on Cy , Cr , Ir ,T B , r and Gr .

b) What is the value of output in the short-run macroeconomic outcome?

The TR curve is given as:

 
Y −Ȳ
r = r̄M P + φy · Ȳ
+ r̄RP

Plugging r according to the Taylor rule to equation 3.12 yields:

   
A Cr + Ir + Gr + T B · r MP Y − Ȳ RP
Y = − · r̄ + φy · + r̄ .
1 − Cy + T Bim 1 − Cy + T Bim Ȳ
   
Y = IS0 − IS1 r̄M P + φy · Y Ȳ−Ȳ + r̄RP .

Y = IS0 − IS1 r̄M P − φy + r̄RP − IS1 · φy YȲ .

φy

Y (1 + IS1 Ȳ
) = IS0 − IS1 r̄M P − φy + r̄RP .

IS0 − IS1 r̄M P − φy + r̄RP
Y = φy
. (3.13)
1 + IS1 · Ȳ

Plugging the parameters into 3.13, the short run outcomes are:

ˆ Y = 14.36896552
 
ˆ r = r̄M P + φy · Y Ȳ−Ȳ + r̄RP = −0.012068966

c) Calculate the effects on domestic short-run output of a decrease in the foreign interest rate, r∗ , by
0.01? Decompose the overall output effects into its components (that is, the Keynesian multiplier induced
effect and the domestic interest rate induced crowding out-in effects). Make sure to provide, in addition to
numerical results, a detailed economic rationale as to how the various effects arise.

Recall that

A = C0 + I0 + G0 + T B0 + T Bex · (Y ∗ − T ∗ ) + T B · (0 + r r∗ ) − Cy · T + T Bim · T ,

therefore,

∆A = ∂A
∂r ∗ · ∆r∗ = T B · r ∆r∗ = 1 · 1 · −0.01 = −0.01

ˆ Total change in output, Y:

∂Y ∂Y ∂r
∆Y = ∂A · ∆A + ∂r · ∂A · ∆A

∂Y 1
∂A = 1−Cy +T Bim = 1.666666667

∂r
Note that we need to solve for r as a function of A so that ∂A can be computed:

 
Y −Ȳ
r = r̄M P + φy · Ȳ
+ r̄RP
r = r̄M P + φy · ȲY − φy + r̄RP
φy · ȲY = r − r̄M P − r̄RP + φy
Y = φȲy · r + Ȳ − φȲy · (r̄M P + r̄RP )
50 CHAPTER 3. THE SHORT RUN MACROECONOMY

Plugging this result into the IS equation 3.12 yields:

Ȳ Ȳ MP A Cr +Ir +Gr +T B ·r

φy · r + Ȳ − φy · (r̄ + r̄RP ) = 1−Cy +T Bim − 1−Cy +T Bim ·r
Ȳ C +I +G +T B · A Ȳ MP
φy · r +
r r r
1−Cy +T Bim
 r
· r = 1−Cy +T Bim + −Ȳ + φy · (r̄ + r̄RP )
A Ȳ MP RP
1−Cy +T B +−Ȳ + φy ·(r̄ +r̄ )
r= Ȳ
im
+ Cr +I r +Gr +T B ·r
φy 1−Cy +T B
im
1
∂r 1−Cy +T Bim
∂A = Ȳ Cr +Ir +Gr +T B ·r
φy + 1−Cy +T Bim

Plugging the parameters into this expression of r yields:

∂r
∂A = 0.034482759

Finally,

r +Gr +T B ·r
∂Y
∂r = − Cr +I
1−Cy +T Bim = −33.33333333.

Therefore,

∂Y
∆Y = ∂A · ∆A + ∂Y ∂r
∂r · ∂A · ∆A = 1.666666667 · −0.01 + −33.33333333 · 0.022550033 · −0.01 = −0.00517

ˆ Notice that income is lower due to the decrease in foreign interest rate. The central bank react to this
negative shock by lowering the
 interest
 rate, rM P following the Taylor rule in the short run, leading to
a change in r = r̄M P + φy · Y Ȳ−Ȳ + r̄RP , which is computed as:

∂r
∆r = ∂A ∆A = 0.034482759 · −0.01 = −0.000344828

ˆ Impacts through channels:

– A decrease in the foreign interest rate lead to a change in the autonomous components by ∆A =
T B · r ∆r∗ , through the KM effects, the contribution to ∆Y due to a change in ∆A is:
∂Y ∂A ∗ 1 ∂A ∗
∂A ∂r ∗ ∆r = 1−Cy +T Bim · ∂r ∗ ∆r = 1.666666667 · −0.01 = −0.01666666667

– Crowding in of consumption: A reduction in the interest rate leads to an increase in consumption

by: ∆C = − ∂C ∂r · ∆r = −Cr · ∆r, through the Keynesian Multiplier effects, resulting in the change
in output as:
1
1−Cy +T Bim · −Cr · ∆r = 1.666666667 · −3 · −0.000344828 = 0.001724138

– Similarly, the crowding-in of investment:

1
1−Cy +T Bim · −Ir · ∆r = 1.666666667 · −15 · −0.000344828 = 0.00862069

– Crowding-in of government spending

1
1−Cy +T Bim · −Gr · ∆r = 1.666666667 · −2.85 · −0.000344828 = 0.000574713

– Crowding-in of trade balance

1
1−Cy +T Bim · −T B · r · ∆r = 1.666666667 · −1 · 1 · −0.000344828 = 0.000574713

Economic Rationale: The decrease in the foreign interest rate leads to an increase
in the interest rate differential r − r∗ , whereby, following the uncovered interest rate
parity, the domestic currency appreciates against the foreign currency, leading to
less exports and therefore less output, amplified by the KM effects; the central bank
reacts to this decrease in output by lowering the monetary policy rate, whereby
there are crowding-in effects in consumption, investment, government spending and
the trade balance
3.9. THE EXTENDED IS-TR MODEL 51

3.8.4 Problem 3.17

In 2001, then U.S. President George W. Bush was concerned about the sluggish U.S. economy. To help
stimulate the U.S. economy, President Bush proposed a tax cut (suppose for this question that this tax cut
would occur in the form of an income tax cut). Using the IS - TR model, derive graphically the effects
on short-run U.S. output of the policy President Bush proposed. Make sure to provide, in addition to the
graphical analysis, a detailed economic rationale as to how the output effects arise.

Answer: The decrease in taxes leads to an increase in the autonomous component, A, shifting the IS-
curve outwards, as aggregate demand increases. This effect is amplified by the KM (movement from A to
B in the graph). As the output gap has widened,the central bank will increase the monetary policy rate
following its TR. This will lead to crowding-out effects in consumption, investment, government spending as
well as the trade balance via the exchange rate channel (movement from B to C in the graph). Overall, the
tax cut leads to an increase in output and higher interest rates in the economy.

3.9 The Extended IS-TR Model

3.9.1 The Model
In the previous
 section, the interest rate charged by the commercial bank on its customer isr = r̄M P +
Y −Ȳ
φy Ȳ
+ r̄ , which is the sum of the interest rate charge by the central bank r = r̄M P +φy Y Ȳ−Ȳ , plus
RP

a risk premium to account for the fact that commercial banks face the problem of information asymmetry,
i.e. they cannot know 100% sure how risky their customers are. The risk premium during normal time is
assumed to be a constant due to fierce competition in the banking market.

Following the Taylor rule, the central bank cut the interest rate when the economy is operating below
its potential output, and raise the interest rate when there are signs of overheating. However, during deep
economic crisis like the 2008 Global Recession, the centrals cut the interest rate to or very close to 0 but
the interest rate charged by commercial banks remained high. The link between central interest rate and
commercial bank lending rate broke down, and together with it, gone our then conventional wisdom that
the central could always use monetary policy tools to manipulate the economy.

The economic rationale for such a break-down between monetary policy rates and commercial bank lend-
ing rate is the fact that commercial banks loose trust on their customers during extremely bad economic
scenarios. During such time, lending is very risky because their are many customers who have solvency
problems that the banks cannot evaluate with certainty. To compensate for this problem, commercial bank
start charging a higher risk premium. The level of this higher risk premium depends on how bad the situation
of the economy is.
52 CHAPTER 3. THE SHORT RUN MACROECONOMY

To illustrate this point, we first set the Taylor monetary policy interest rate to equal zero to find out the
level of output below which the central bank cannot adjust their policy rate further downward because it is
already zero.
 ZLB 
r̄M P + φy · Y Ȳ −Ȳ = 0
ZLB
φy − r̄M P = φy Y Ȳ
Y ZLB = φȲy φy − r̄M P .

Once the economy hits the Y ZLB , commercial banks start raising the risk premium r̄RP , and the premium
is endogenous to how far the economy is below the Y ZLB . The commercial banks’ strategy of setting the
interest rate is formalized for both the normal time and the deep crisis time as:

rRP = r̄RP − ry · Y − Y ZLB · I Y < Y ZLB

(3.14)

The short-run macroeconomic model during deep crisis time has the same IS curve as in the IS-TR model
during the normal time.

C0 + Cy · (Y − T ) − Cr · r
| {z }
Household Consumption

+ I0 − Ir · r
| {z }
Firms’ Investment
Y = (3.15)
|{z} + G0 + Gy · (Ȳ − Y )
Output produced/Supply | {z }
Government Spending

+ T B0 + T Bex · (Y − T ) − T Bim · (Y − T ) + T B [0 − r (r − r∗ )]

∗ ∗
| {z }
Trade Balance

However, the interest rate is now determined by the extended TR curve:

  
MP Y − Ȳ
+ r̄RP − ry · Y − Y ZLB · I Y < Y ZLB



r = max 0, r̄ + φy · (3.16)
Ȳ

Figure 3.3: The Extended IS-TR Model

As illustrated in Figure (3.3), the extended TR curve becomes kinked and start sloping downward when
the economy’s output falls below the zero lower bound value Y ZLB . That means the risk premium is endo-
genized and depends on how serious the situation of the economy is.
3.9. THE EXTENDED IS-TR MODEL 53

Originally, the economy is operating at point A with the upward sloping part of the Extended TR curve
and the IS curve. With a strong shock of the magnitude of the Great Recession, the IS curve shift to the left
IS’ curve. The conventional monetary policy could still have been effective if their is no shock to the banking
system’s confidence. However, with such a strong shock, the banking system increased the independent part
of the risk premium, r̄RP , and start endogenizing the interest rate charged on customers. The Extended
TR curve shift to the Extended TR’ position, and the economy finds itself in a position where the lower the
ouput level, the higher the interest rate.

3.9.2 Problemset Exercises on the Extended IS-TR Model

Problem 3.18
An economy under fixed prices (for which foreign variables are exogenous) is described by the following
model:
Y = DD = C + I + G + T B;
C = C0 + Cy · (Y − T ) − Cr · r;
I = I0 − Ir · r;
G = G0 − Gr · r;
T B = T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Y − T ) + T B · ;
 = 0 − nr (r − r∗ )  o
r = max 0, r̄M P + φy · Y Ȳ−Ȳ



+ r̄RP − ry · Y − Y ZLB · I Y < Y ZLB .

When numerical calculations are asked for, assume the following parameter values: C0 = 1.6; Cy =
0.55; Cr = 3; T = 2.5; I0 = 3.25; Ir = 15; G0 = 2.8; Gr = 1; T B0 = 0; T Bex = 0.1; Y ∗ = 10; T ∗ = 3; T Bim =
0.15; T B = 1; 0 = 1; r = 1; r∗ = 0.03; r̄M P = 0.02; φy = 1; Ȳ = 15; r̄RP = 0.01; ry = 0.005.

a) Graph the Extended TR- curve (for general parameter values). Provide an economic rationale as to
how the slope of the Extended TR - curve depends on the level of output.

b) What is the value of output in the short-run macroeconomic outcome?

c) Calculate the effects on domestic short-run output of a simultaneous increase of the exogenous com-
ponent of the risk premium, r̄RP , by 0.02, and of the foreign real interest rate, r∗ , by 0.01? Decompose the
overall output effects into its components (that is, the Keynesian multiplier induced effect and the domestic
interest rate induced crowding out/in effects). Make sure to provide, in addition to numerical results, a
detailed economic rationale as to how the various effects arise.

a) Graph the Extended TR- curve (for general parameter values). Provide an economic rationale as to
how the slope of the Extended TR - curve depends on the model parameters.

Figure 3.4: The Extended TR Curve

Answer: In order to analyze how the slope of the extended TR-curve depends on the parameters we need
to consider two cases:
ˆ Case 1: Y > Y ZLB

When Y > Y ZLB , the extended TR curve is actually the TR curve. The slope of this segment of the
extended TR curve depends only on the parameter φy , which reflects how aggressively the central bank
reacts to a change in the deviation between actual output Y and long-term output Y . The larger the
54 CHAPTER 3. THE SHORT RUN MACROECONOMY

value of φy , the steeper, that is, the more positive the slope of the extended TR-curve in the range
Y > Y ZLB .

ˆ Case 2: Y < Y ZLB

When Y < Y ZLB , the slope depends only on the parameter ry reflecting how sharply the endogenous
risk premium changes once output changes. The larger the magnitude of ry , the steeper, that is, the
more negative the slope of the extended TR-curve in the range Y < Y ZLB .

b) What is the value of output in the short-run macroeconomic outcome?

Answer: The IS equation is (you have to derive it yourself!)

A Cr + Ir + Gr + T B · r
Y = − · r. (3.17)
1 − Cy + T Bim 1 − Cy + T Bim

Where A = C0 + I0 + G0 + T B0 + T Bex · (Y ∗ − T ∗ ) + T B · (0 + r r∗ ) − Cy · T + T Bim · T . Rewrite the IS

in a more compact form as:

Y = IS0 − IS1 · r

The definition of IS0 and IS1 should be clear from the context. First, we need to compute the zero-lower
bound output:

 
Y ZLB −Ȳ
r̄M P + φy Ȳ
=0
 MP

r̄
Y ZLB = 1 − φy Ȳ
ZLB
Y = 14.70

How to proceed from here? We do not really know if the level of actual output is below or above the Y ZLB .
The strategy here is to assume that initially Y < Y ZLB , using the extended TR-curve:


r = r̄RP − ry · Y − Y ZLB

Plugging r into the IS curve to get:

Y = IS0 − IS1 r̄RP − ry · Y − Y ZLB .


Y = IS0 − IS1 r̄RP + ry · Y ZLB + IS1 ry · Y.


(1 − IS1 ry ) · Y = IS0 − IS1 r̄RP + ry · Y ZLB

IS0 − IS1 r̄RP + ry · Y ZLB
Y = . (3.18)
1 − IS1 ry

Plugging the parameters, IS0 = 13.97; IS1 = 33.33; and Y ZLB = 14.70 into (2.133), the short run outcomes
are:

ˆ Y = 13.42
The results indeed show that the economy is in a deep crisis. The short run interest rate is computed
as:

ˆ r = r̄RP − ry · Y − Y ZLB = 0.0164

c) Calculate the effects on domestic short-run output of a simultaneous increase of the exogenous component
of the risk premium, r̄RP , by 0.02, and of the foreign real interest rate, r∗ , by 0.01? Decompose the overall
output effects into its components (that is, the Keynesian multiplier induced effect and the domestic interest
rate induced crowding out/in effects). Make sure to provide, in addition to numerical results, a detailed
3.9. THE EXTENDED IS-TR MODEL 55

economic rationale as to how the various effects arise.

Answer:
From question b), we know that the economy’s short run output is below Y ZLB . The simultaneous shocks
from r̄RP and r∗ are negative and positive, respectively. Again, using the extended IS-TR system to get:

IS0 − IS1 · r̄RP − IS1 · ry · Y ZLB

Y = (3.19)
1 − IS1 · ry

Plugging the parameter values into (2.134) to get: Y = 12.64. Using this value in (2.136) yields the value of
the new even deeper crisis interest rate as: r = 0.0403.

Now we need to decompose the impact of the composite shock that hit the economy.

To decompose the overall output effects into its components, first notice that:

∂Y ∂Y
∆Y = ∂A ∆A + ∂r ∆r,
∗
where ∆A = ∂r ∂A
∗ ∆r = T B r ∆r∗ and ∆r = ∆r̄RP − ry ∆Y . Further, we could compute the change in A,
and r using the available information as:

∆A = 1.1.0.01 = 0.01
∆r = 0.02 − 0.005(12.64 − 13.42) = 0.0239

The overall effect can now be decomposed as

ˆ The autonomous components, the KM effects:

∂Y
∂A ∆A = KM · ∆A = 1.666666667 · 0.01 = 0.1666666667

where the value of KM is taken from Problem (3.7).

ˆ Crowding-out of consumption:

− 1−CyC+T
r
Bim ∆r = −0.1195

ˆ Crowding-out of investment:

− 1−Cy I+T
r
Bim ∆r = −0.5975

ˆ crowding-out of government spending

− 1−CyG+T
r
Bim ∆r = −0.039833333

ˆ crowding-out of trade balance

1
1−Cy +T Bim · T B r ∆r = −0.039833333

Economic Rationale:
There is a notably rich set of adjustments. An increase of the foreign interest rate results in a right shift of
the IS-curve, reflecting the spot depreciation of the domestic currency, and the resultant improvement of the
trade balance. However, at the same time there is an increase in the exogenous component of the domestic
risk premium, that is, a parallel shift of the extended TR-curve upwards, which (from the calculations above)
more than offsets the effects induced by the increase of the foreign interest rate. As the economy is already
in a crisis and output is below the Y ZLB threshold the central bank cannot mitigate the negative effect of
the increase in the risk premium. As the interest rate at which households, firms and the government can
borrow increases a crowding-out will occur through the consumption, investment, government expenditure
and the exchange rate/trade balance channels. In addition, due to the decrease in output the endogenous
component of the risk premium rises as well which exacerbates the overall negative output effect.
56 CHAPTER 3. THE SHORT RUN MACROECONOMY

Problem 3.19
On the Extended IS-TR model: Suppose an economy is in a deep contraction, and features a large current
account deficit. The costs of borrowing for households and firms in this economy are high due to a high
overall risk premium. The central bank of this economy operates at the zero lower bound, and the fiscal
authorities of this economy do have fiscal space. The economy’s policy makers consider two options: (i) ex-
pansionary fiscal policy; (ii) unconventional monetary policy in the form of purchases of government bonds
at all maturities. Use an Extended IS-TR model of the form given in Problem (3.9) to contrast how the
different policies will affect the economy’s short-run output and current account. Make sure to provide, in
addition to a graphical analysis, a detailed economic rationale as to how the output and current account
effects under the different policy options arise. Can the sign of the output and current account effects be
predicted without knowledge of the numerical values of the model parameters? Explain your reasoning.

Answer:

ˆ Expansionary Fiscal Policy:

Figure 3.5: Expansionary Fiscal Policy

Fiscal Policy: An increase of government expenditure through an increase of G0 , thus shifting the
IS-curve to the right, would be particularly effective as long as output remains below its zero lower
bound level, Y ZLB , as then:

- No crowding out through the consumption, investment, government ex- penditure or exchange rate
channels takes place, and

- The endogenous component of the risk premium falls as output increases (reflecting lower rates of
default and increasing values of collateral as the economy is beginning to recover).

Output following an increase in government expenditure thus overall increases in such a setting even
more than the Keynesian multiplier would suggest.
ˆ Unconventional Monetary Policy:

Lowering the cost of borrowing for households, rms and the government through lowering the exogenous
component of the domestic risk premium, rRP , corresponds to a parallel shift of the extended TR-curve
3.9. THE EXTENDED IS-TR MODEL 57

downwards. Unconventional monetary policy will be even more effective as long as output remains
below its zero lower bound level, Y ZLB , as then the endogenous component of the risk premium falls
as output increases (reflecting again lower rates of default and increasing values of collateral as the
economy is beginning to recover), and the monetary policy rate stays at zero.

Figure 3.6: Expansionary Fiscal Policy

58 CHAPTER 3. THE SHORT RUN MACROECONOMY
Chapter 4

The Macroeconomy in the Medium

Run

4.1 The Supply Curve

Figure 4.1: A Recall of the Economy in the Short-run

In the previous chapter, we defined the short run as the short enough period so that price remain constant.
Firms do not change their price, workers do not bargain to change their wages, and suppliers do not negotiate
to change the price of inputs. In this setting, supply responds to demand and firms provide what ever quantity
demanded by driving down or building up their inventories.

ˆ By the medium run, we consider the time frame long enough for firms to change their price, workers to
demand higher wages, and supplier to negotiate a new deals. When all prices and wages fully adjust, the
economy fully recover from the shocks and settles at the potential (by potential we mean most comfortable
output level. That is also when the economy comes back to its long-run output level we are so familiar with
in the Solow Growth Model.

ˆ I feel imperative to put a note here on difference between macroeconomic analysis in the short and
medium run, with the macroeconomic analysis in the long run. Through out discussion, it seems to me
that some of you are disappointed that after tireless rounds of adjustments, the economy comes back to the
potential output, without any further progress. There is no contradiction here. Recall that the potential
output is the most comfortable level of production and the economy should operate at that point. In the
short and medium runs, the economy is shocked by unexpected factors beyond our control and could enter
costly rounds of bounce and busts and we need the insights from the short and medium run models to devise
counter measures to keep the economy stable, to stay on the long-run potential path of development.

59
60 CHAPTER 4. THE MACROECONOMY IN THE MEDIUM RUN

4.1.1 The Firm’s Pricing Problem

In the medium run, firms can adjust price and have to accept changes in wages. So the firm’s maximization
problem is:

Π ≡ max P · Y = W · L − r|k{z
· K} , (4.1)
Y |{z} |{z}
Given nomila wages Labor hours Treated as fixed

The F.O.C is:

∂Π ∂P ∂L
∂Y =P +Y · ∂Y −W · ∂Y
∂Π
Setting ∂Y = 0 yields:
∂P ∂L
P +Y · ∂Y − W · ∂Y =0
Y ∂P W


P 1+ P · ∂Y = ∂Y
∂L

∂y ∂x ∂x
Note that in this step we have used the property of the derivative of an inverse function ∂x · ∂y = ∂x = 1.
 
1 W
P 1− = (4.2)
η λ
|{z}
| {z }
marginal revenue marginal cost

where η = −(∂Y /∂P )/(P/) is the price elasticity of aggregate demand and λ is the marginal product of
labor. Recall from the lecture that the degree of pricing power of the firms, that is, the extent to which they
can charge mark-ups (which we will denote by θ) on their goods and services is inversely related to η. From
equation (3.2), we have:
P
P− η = MC
P
P − MC = (4.3)
| {z } η
(Absolute)mark−up
P
Denoting the mark-up as θ = M Cη , we obtain the following expression for the pricing of goods and services:

W
P = (1 + θ) · M C = (1 + θ) · (4.4)
λ
Again, taking the logarithm of both sides and take the total differentiation of the resulting equation to get:
∆P ∆θ ∆W ∆λ
P = 1+θ + W − λ

or in terms of the inflation rate:

∆θ ∆W ∆λ
π= + − (4.5)
1+θ W λ

4.1.2 Wage bargaining

Workers negotiate their wages based on their expectation of future inflation, labor productivity, and how
tight the labor market is. The situation of the labor market is determined by the state of the economy
relative to the long-run potential output. Thus:
 
∆W e ∆λ Y − Ȳ
=π + + |{z}
ω (4.6)
W λ Ȳ
ω>0

Substituting (3.6) into (3.5) yields:

 
∆θ ∆λ Y −Ȳ ∆λ
π= 1+θ + πe + λ + |{z}
ω Ȳ
− λ
ω>0
 
∆θ Y − Ȳ
π= + π e + |{z}
ω (4.7)
1+θ Ȳ
ω>0

Besides, the economy constantly faces supply shocks which can be formalized by including a term s into (3.6)
as:  
∆θ e Y − Ȳ
π= +π +ω +s (4.8)
1+θ Ȳ
4.2. THE DEMAND CURVE 61

4.1.3 The Supply Curve

Notice in equation (3.8) that price (π) still depends on two other unknowns. One way to move forward is to
work out how firms/workers form their inflation expectation. In this context, we assume adaptive inflation
expectation π e = πt−1 . The supply curve in the medium run becomes:
 
∆θ Y − Ȳ
π = πt−1 + +ω +s (4.9)
1+θ Ȳ

Figure 4.2: The Medium Run Supply Curve

4.2 The Demand Curve

The TR curve in the medium run is different from the TR curve in the short run by the inclusion of the
inflation term:  
Y − Ȳ
r = r̄M P + φy + φπ π − π T + r̄RP


(4.10)
Ȳ
And the IS curve is the same as the short run IS curve:
A (Cr + Ir + T B · r )
Y = + ·r (4.11)
1 − (Cy − T Bim − Gy ) 1 − (Cy − T Bim − Gy )
| {z } | {z }
= IS0 = IS1

Combining (3.10) and (3.11) yields:

   
MP Y − Ȳ RP
Y = IS0 − IS1 · r̄ + φy + φπ (π − πT ) + r̄ (4.12)
Ȳ
  φ
Y = IS0 − IS1 · r̄M P − φy − φπ πT + r̄RP − IS1 Ȳy · Y − IS1 φπ π
  φ
IS1 φπ π = IS0 − IS1 · r̄M P − φy − φπ πT + r̄RP − (IS1 Ȳy + 1) · Y
62 CHAPTER 4. THE MACROECONOMY IN THE MEDIUM RUN
 
IS1 · φy + φπ πT − r̄M P − r̄RP + IS0 Ȳ + φy IS1
πt = − · Yt (4.13)
IS1 φπ IS1 φπ Ȳ
This is the demand curve of the economy in the medium run.

Figure 4.3: The Demand Curve in the Medium Run

Why is it the case that the medium run demand curve slope downward? As the rate of inflation increases,
the central bank, following the Taylor rule, raises the monetary policy rate, in turn causing (for a given
risk premium) the real interest rate at which households and firms can borrow to increase. Through the
consumption, investment and exchange rate channels, this lowers output.

Figure 4.4: The Reason Medium Run Demand Curve Slope Downward

4.3 The AD-AS model in the normal time

So armed with the supply:
 
∆θ Y −Ȳ
π= 1+θ + πe + ω Ȳ
+s
4.3. THE AD-AS MODEL IN THE NORMAL TIME 63

and demand curves:

IS1 ·[φy +φπ πT −r̄ M P −r̄ RP ]+IS0 Ȳ +φy IS1
π= IS1 φπ − IS1 φπ Ȳ
·Y
We are in a very good position to discuss how the economy response to shocks and how the central bank
behaves. Let’s do some problems:

4.3.1 Problem 4.1

On the AS-AD model: An economy with sticky prices (for which foreign variables are exogenous) is described
by the following model:
Yt = DDt = Ct + It + Gt + T Bt ;
Ct = C0 + Cy · (Yt − T ) − Cr · rt ;
It = I0 − Ir · rt ;
Gt = G0 − Gr · rt ;
T Bt = T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Yt − T ) + T B · t ;
t = 0 + r · (r − r∗ )  
Y −Ȳ
RtM P = r̄M P + π e + φy · Ȳ
+ φπ (πt − π T );
 
rt = RtM P − π e + r̄RP ; πt = π e + ω · YtȲ−Ȳ + s.
When numerical calculations are asked for, assume the following parameter values: C0 = 1.6; Cy = 0.55; Cr =
3; T = 2.5; I0 = 3.25; Ir = 15; G0 = 2.8; Gr = 1; T B0 = 0; T Bex = 0.1; Y ∗ = 10; T ∗ = 3; T Bim = 0.15; T B =
1; 0 = 1; r = 1; r∗ = 0.03; r̄M P = 0.02; φy = 1; Ȳ = 15; r̄RP = 0.01; π e = 0.02; φπ = 2; φy = 1; π T =
0.02; ω = 0.05; s = 0.

a) Derive the AD - curve. Provide detailed economic reasoning for the AD - curve derived, namely, how
an increase in the medium-run rate of inflation,πt , affects medium-run output, Yt

Answer: Recall from Problem (3.7):

A Cr + Ir + Gr + T B · r
Y = − · r. (4.14)
1 − Cy + T Bim 1 − Cy + T Bim
Where A = C0 + I0 + G0 + T B0 + T Bex · (Y ∗ − T ∗ ) + T B · (0 + r r∗ ) − Cy · T + T Bim · T . However, it is
more convenient to go one step further to rewrite the IS curve in the form:
Yt = IS0 − IS1 · rt , (4.15)
where the definition of IS0 and IS1 should be clear from the context.

ˆ The TR curve in the medium run is the defined as:

 
Yt − Ȳ
rt = RtM P − π e + r̄RP = r̄M P + φy + φπ (πt − π T ) + r̄RP (4.16)

Using the formula of r given by the TR curve in the IS curve yields:

   
Y − Ȳ
Y = IS0 − IS1 · r̄M P + φy + φπ (π − πT ) + r̄RP (4.17)
Ȳ
 MP  φ
Y = IS0 − IS1 · r̄ − φy − φπ πT + r̄RP − IS1 Ȳy · Y − IS1 φπ π
  φ
IS1 φπ π = IS0 − IS1 · r̄M P − φy − φπ πT + r̄RP − (IS1 Ȳy + 1) · Y
 
IS1 · φy + φπ πT − r̄M P − r̄RP + IS0 Ȳ + φy IS1
π= − ·Y (4.18)
IS1 φπ IS φ Ȳ
| {z } | 1{zπ }
≡ AD0 ≡ AD1
Economic Rationale: The AD-curve features a negative slope. An increase in inflation in-
duces the central bank to set a higher nominal policy rate. For given inflation expectation
and risk premium, this raises the interest rate faced by households, firms and government,
corresponding to crowding-out via consumption, investment, government and exchange rate
channels. This reduces aggregated demand and thus output. There is a negative relationship
between inflation and output.
64 CHAPTER 4. THE MACROECONOMY IN THE MEDIUM RUN

ˆ b) Obtain the medium-run macroeconomic outcome.

The AD curve is given by:
π = AD0 − AD1 · Y
And the AS curve is given by:
 
e Yt − Ȳ ω ω
πt = π + ω · + s. = π e − ω + s + · Yt = π e − ω + s + ·Yt (4.19)
Ȳ Ȳ | {z } Ȳ
≡ AS0 |{z}
≡ AS1

Combine both equations:

AD0 − AD1 · Yt = AS0 + AS1 · Yt

output is obtained as:
AD0 −AS0
Yt = AD1 +AS1

The resulting value of output Yt is 14.40967742. Plugging this value in the AD equation to get the value of
πt as: 0.018032258.

c) Calculate the effects on medium-run output and inflation of a fiscal stimulus in the form of a decrease
of (domestic) income taxes, T, by 0.25. Provide a detailed economic rationale as to how the medium-run
output and inflation effects arise.

Answer:
ˆ The impact of a change in taxes on output is defined as: ∆Yt = ∂Y
∂T · ∆T . ∆T is given. Our next task
is to compute ∂Y
∂T .

AD −AS
∂Y ∂ AD0 +AS0
∂T = 1
∂T
1

∂AD0 1
= ∂T · AD1 +AS1

∂IS0 1 1
= ∂T φπ ·IS1 · AD1 +AS1

∂A 1 1 1
= ∂T 1−Cy +T Bim φπ ·IS1 · AD1 +AS1

1 1 1
= (T Bim − Cy ) 1−Cy +T Bim φπ ·IS1 · AD1 +AS1

Plugging in the relevant parameters yields ∂Y /∂T = −0.193548387. The change in out put is given
by: ∆Y = ∂Y /∂T · dT = −0.193548387 · (−0.25) = 0.048387097
ˆ The change in inflation due to the change in government spending is defines as ∂πt /∂T · ∆T
∂πt ∂AS0 ∂AS1 ∂Yt
∂T = ∂T + ∂T · Yt + ∂T · AS1

∂πt ∂Yt
∂T = ∂T · AS1

∂πt ∂Yt
Thus, the change in inflation due to a change in T is given by: ∆πt = ∂T ∆T = ∂T · AS1 · ∆T =
0.00016129
Economic Rationale: Lower domestic income taxes increase disposable income of households.
This increases consumption and decreases the trade balance, enforced by the Keynesian multi-
plier. In the short-run, the higher desired demand increases output. The central bank increases
interest rates, which leads to a partial crowding-out of consumption, investment, government
expenditure and trade balance (movement from A to B). In the medium-run, the higher output
translates into higher wage mark-up, increasing inflation. The central bank increases interest
rates even more, leading to a further crowding-out of aggregate demand (movement from B
to C). Overall, we end up with higher output and inflation than before.
4.4. THE AD-AS MODEL DURING DEEP CRISIS 65

Figure 4.5: The Impact of a Decrease in Income Taxes

4.4 The AD-AS Model During Deep Crisis

The discussion is done during class sessions. So we will focus on the problem set questions only.

4.4.1 Problem 4.2

On the extended AS-AD model: An economy with sticky prices (for which foreign variables are exogenous)
is described by the following model:

Yt = DDt = Ct + It + Gt + T Bt ;
Ct = C0 + Cy · (Yt − T ) − Cr · rt ;
It = I0 − Ir · rt ;
Gt = G0 − Gr · rt ;
T Bt = T B0 + T Bex · (Y ∗ − T ∗ ) − T Bim · (Yt − T ) + T B · t ;
∗
t = 0 + r · (r
n t − rt )   o
Y −Ȳ
RtM P = max 0, r̄M P + π e + φy · Ȳ
+ φπ (πt − πT ) ;



rt = RtM P − πt−1 + r̄RP − ry · Yt − Yt ZLB
|πt · I Yt < YtZLB |πt ;
πt = π e + ω · YtȲ−Ȳ + st .

When numerical calculations are asked for, assume the following parameter values: C0 = 1.6; Cy = 0.55; Cr =
3; T = 2.5; I0 = 3.25; Ir = 15; G0 = 2.8; Gr = 1; T B0 = 0; T Bex = 0.1; Y ∗ = 10; T ∗ = 3; T Bim = 0.15; T B =
1; 0 = 1; r = 1; r∗ = 0.03; r̄M P = 0.02; φy = 1; Ȳ = 15; r̄RP = 0.01; π e = 0.02; φπ = 2; φy = 1; π T =
0.02; ω = 0.05; s = 0; ry = 0.07.

ˆ a) Derive the Extended AD-curve. Provide detailed economic reasoning for the Extended AD-curve
derived, namely, how an increase in the medium-run rate of inflation, πt , affects medium-run output, Yt .

ˆ Answer: Combine monetary policy rule (with zero lower bound) and the risk premium to obtain the
Extended TR-curve:
66 CHAPTER 4. THE MACROECONOMY IN THE MEDIUM RUN

rt = RtM P − πte + rtRP

n   o
+ φy · Y Ȳ−Ȳ + φπ (πt − π T ) + r̄RP − ry · Yt − YtZLB |πt · I Yt < YtZLB |πt



= max −π e , r̄M P

ˆ Normal business cycle outcome, RM P > 0: The Extended TR-curve collapses to the standard TR-
curve. We obtain the same AD-curve as before: πt = AD0 − AD1 · Yt . There is a negative relationship
between output and inflation.
ˆ During crisis time, central banks keep cutting the interest rate to encourage demand, thus pushing
the economy back to the equilibrium, i.e. potential ouput level. However, when the crisis is too deep,
RM P could be cut to, or close to 0. At this point, the central runs out of ammunition, and cannot cut
RM P further, at least too much below the zero level. We face the problem of zero lower bound, the
situation in which output falls below the threshold potential output level so much that below it, the
Taylor rule dictates RM P ≤ 0. We name this threshold as YtZLB |πt , which can be solved as:

YtZLB |πt −Ȳ

 
r̄tM P + πte + φy · Ȳ
+ φπ · (πt − π T ) = 0

YtZLB |πt
 
r̄tM P + πte − φy + φy · Ȳ
+ φπ · (πt − π T ) = 0

YtZLB |πt
 
r̄tM P + πte − φy + φy · Ȳ
+ φπ · (πt − π T ) = 0

YtZLB |πt
 
φy · Ȳ
= φy − r̄tM P − πte − φπ · (πt − π T )

Ȳ


YtZLB |πt = φy φy − r̄tM P − πte − φπ · (πt − π T ) .

Facing this situation, commercial banks become wary of the fact that their customers have become
more risky and face a higher probability of default. To compensate for this risk, commercial banks
endogenize the risk premium rRP that they would charge their customer. This risk premium increases
as the actual level of out put falls deeper below the zero lower bound output level, YtZLB |πt . This
insight is formalized by specifying the risk premium as:


rtRP = r̄RP − ry · Yt − YtZLB |πt

where ry > 0. The real interest rate facing banks’ customers is:


rt = rM P + r̄RP − ry · Yt − YtZLB |πt .

rtM P = RtM P − πte = −πte because RM P = 0. Therefore,

rt = −πte + r̄RP − ry · Yt − YtZLB |πt .

In this exercise, taking the assumption that inflation expectation is adaptive, and current level of in-
flation expectation is equal to last period’s inflation,

rt = −π t−1 + r̄RP − ry · Yt − YtZLB |πt .

This is the extended TR curve during crisis time in the medium run. Plugging the expression of
YtZLB |πt into the extended TR curve yields:
n
o
Ȳ
rt = −π t−1 + r̄RP − ry · Yt − φy φy − r̄tM P − πt−1 − φπ · (πt − π T ) .

Ȳ Ȳ
rt = −π t−1 + r̄RP − ry · Yt + ry · φy (φy − r̄tM P − πt−1 + ·φπ · π T ) − ry · φy · φπ · πt .

Plugging this result into the IS curve yields:

4.4. THE AD-AS MODEL DURING DEEP CRISIS 67
n o
Ȳ Ȳ
 
Yt = IS0 − IS1 · −πt−1 + r̄RP − ry · Yt + ry · φy φy − r̄M P − πt−1 + φπ π T ) − ry · φy φπ πt

n o
Ȳ
φy − r̄M P − πt−1 + φπ π T ) + IS1 · ry · Yt + IS1 · ry · φȲy φπ πt

Yt = IS0 − IS1 · −πt−1 + r̄RP + ry · φy

n o
Ȳ Ȳ

Yt (1 − IS1 · ry ) = IS0 − IS1 · −πt−1 + r̄RP + ry · φy φy − r̄M P − πt−1 + φπ π T ) + IS1 · ry · φ y φπ πt

IS1 · ry · φȲy φπ πt =
h n oi
− IS0 − IS1 · −πt−1 + r̄RP + ry · φȲy φy − r̄M P − πt−1 + φπ π T )

+ Yt (1 − IS1 · ry )

n o
Ȳ

IS0 − IS1 · −πt−1 + r̄RP + ry · φy φy − r̄M P − πt−1 + φπ π T ) (1 − IS1 · ry )
πt = − Ȳ
+ Ȳ
·Yt
IS1 · ry · φy φπ IS1 · ry · φ y φπ
| {z } | {z }
≡ AD
]0 ≡ AD
]1

πt = −AD ]1 · Yt
]0 + AD

Figure 4.6: The extended AD curve

Economic Rationale: Technically in the deep crisis scenario, the AD-curve is upward-sloping. A
decrease in inflation increases the level of output at which the central bank hits the zero lower bound.
Thus, the endogenous risk premium matters at higher levels of output than before, and is higher for any
given Y < YtZLB |πt . This causes crowding-out of aggregate demand. There is a positive relationship
between in ation and output.

ˆ Intuitively, inflation often falls during recessions when demand is low, firms cut their output, and
therefore, employment. Cuts in employment lead to the loss of income, thus lowering demand further.
Lower demand leads to lower price, i.e. a further decrease in inflation. When the economy has reached
the trough of the business cycle, i.e. the worst point, ineffective firms are driven out of the market due
to bankruptcy, only efficient firms remain. These efficient firms also cut their prices along the way
( due to lower costs as a result of wage cuts and decreases in input price) to the level that prices are
considered as low, demand starts to pick up. As we discuss in the short run, firms respond to short run
68 CHAPTER 4. THE MACROECONOMY IN THE MEDIUM RUN

increase in demand by changing their inventories. However, as the increase in demand is sustained,
they start to increase production, call back their former employees, and stop cutting prices, and perhaps
they may also start to consider more investment to meet the future demand. With higher employment,
higher investment, demand continues to increase, leading to the first increase in price, i.e. inflation.

ˆ The message of this discussion is that during crisis time, further falls in prices (inflation)
constitutes a signal of a worsening situation, thus raising the risk premium banks want to charge, and
lowering demand. In the reversed situation, increases in inflation signal the recovery from a deep
recession/crisis, enticing banks to cut the risk premium, and therefore, the interest rate they charge,
pushing the demand curve to the right. The demand curve will keep shifting to the right until the
economy is strong enough that the central bank decides to increase its RM P rate, marking the end of
the deep crisis period and bringing the economy back to its normal AD-AS model. Note that this is
also in line with the explanation we have in the case of the Extended IS-TR, the only
difference is now we are in the medium run, having a lot of short run periods (with
different price levels) pieced together in a time line.

b) Obtain the values of output and inflation in the medium-run macroeconomic outcome (suppose that
πt−1 = 0.02)

ˆ From a) the demand curve is given by πt = −AD ]1 · Yt , combining with the supply curve given as
]0 + AD

 
Yt −Ȳ
πt = π e + ω · Ȳ
+s

Yt


πt = π e + ω · Ȳ
−ω+s

ω
πt = π e − ω + s + Ȳ
· Yt

ω
πt = π e − ω + s + ·Yt
| {z } Ȳ
≡ AS0 |{z}
≡ AS1

the medium run macroeconomic outcome is:

]0
AS0 +AD
Yt = ]1 −AS1
AD

Using the available information, AD

]0 = 1.35; AD ]1 − AS1 = 0.11; AS0 = −0.03; AD1 = 0.0033. Plugging
these parameters in the expression of Yt yields:

Yt = 12.37

πt = 0.011

If you take all decimals, Yt = 12.4529148, and πt = 0.011509716.

c) Graphically analyze in the AS-AD diagram the effects of a fiscal contraction in the form of an increase
of (domestic) income taxes, T, under the assumption that the initial medium-run macroeconomic outcome
for output is the one given in part (b). In addition to a graphical analysis, provide a detailed economic
rationale as to how the medium-run output and inflation effects arise.

ˆ Answer: An increase in domestic income taxes reduces disposable income for households, therefore
reducing aggregate demand, enforced by the Keynesian multiplier. The decrease in output triggers an increase
in the risk premium, thus leading to a crowding-out of consumption, investment, government expenditure
and trade balance. In the short-run, we end up in point B.
4.4. THE AD-AS MODEL DURING DEEP CRISIS 69

Figure 4.7: Increase of domestic income taxes during deep crisis time

ˆ In the medium-run, the lower output translates into lower wage mark-ups, thus leading to lower inflation.
The lower inflation increases YtZLB |πt , thus increasing the risk premium, leading to a further crowding-out
of demand (B to C). Due to adaptive inflation expectations, these are revised downwards, such that the real
interest increases, leading to a further crowing-out. The AD-curve shifts to the left, the AS-curve downwards.
We end up with lower output and inflation.

4.4.2 The Non-conditioned π ZLB

We have derived the formula for YtZLB |πt as:
Ȳ


YtZLB |πt = φy φy − r̄tM P − πte − φπ · (πt − π T ) .

It is natural to ask, what is the level of inflation when the RM P hits the zero lower bound? Similarly, we
can set RM P = 0, and solve for the πtZLB |Yt , which satisfies the RM P = 0 condition at different values of Yt .

Figure 4.8: πtZLB |Yt when RM P = 0

 
Yt −Ȳ
r̄tM P + πte + φy · Ȳ
+ φπ · (πtZLB |Yt − π T )

φy
r̄tM P + πte + Ȳ
· Yt − φy + φπ · πtZLB |Yt − φπ · π T = 0
φy
φπ · πtZLB |Yt = φy − r̄tM P − πte − Ȳ
· Yt + φπ · π T
1
  φy
πtZLB |Yt = π T + φπ φy − r̄tM P − πte − Ȳ ·φπ
· Yt

At the kinked point, πt = πtZLB |Yt , and Yt = YtZLB |πt , the extended TR curve sets:
70 CHAPTER 4. THE MACROECONOMY IN THE MEDIUM RUN

rt = r̄M P + rtRP = −π e + r̄RP

because the term Yt − YtZLB |πt in


rt = −πte + r̄RP − ry · Yt − YtZLB |πt .

drops out. Using this result in the IS curve yields:

Yt = IS0 − IS1 · −π e + r̄RP

Therefore,

1  φy
πtZLB |Yt = π T + φy − r̄tM P − πte − · IS0 − IS1 · −π e + r̄RP = π ZLB
 

(4.20)
φπ Ȳ · φπ

Equation (3.20) shows that πtZLB |Yt does not depend on the value of Yt , it is known at the non-conditioned
π ZLB .

4.4.3 Problem 4.3

An economy with fixed prices (for which foreign variables are exogenous) is described by the following model:
Y = DD = C + I + G + T B;
C = C0 + Cy · (1 − τy ) · Y ;
I = I0 − Ir · (1 − τr ) · r;
G = G0 + Gy · (Ȳ − Y ) · I(Y < Ȳ );
T B = T B0 + T Bex · (1 − τy∗ ) · Y ∗ − T Bim · (1 − τy ) · Y + T B · ;
 = 0 − r · (r − r∗ );
r = rM P + r̄RP − ry · (Y −Y ZLB) · I(Y < Y ZLB );
Y −Ȳ
rM P = max{0, r̄M P + φy · Ȳ
}.

Here, τy denotes the rate at which income is taxed in the domestic economy, τy∗ denotes the rate at which
income is taxed in the foreign economies, and τr denotes the rate of interest rate-funded investment tax credit.

a) Derive the IS-curve and graph it. Show all your work. Provide detailed economic reasoning for the
slope of the IS-curve derived 30 points.

b) What is your model-based forecast for the adjustment of the monetary policy rate following a decrease
of the foreign income tax rate, τy∗ ? Thoroughly describe all channels through which the decrease in the
foreign income tax rate affects the monetary policy rate in the domestic economy, according to this model.
In your answer, suppose that the initial level of income is above Ȳ . 30 points

c) If the government is very determined at keeping the output close to its potential level, please prescribe
a fiscal measure such that the potential output could be restored following the shock in question b). Explain
the mechanism of the impact of such measure with the demonstration of suitable graphs. 10 points

d) If, instead, the central bank is very determined at keeping the output close to its potential level,
please prescribe a monetary measure such that the potential output could be restored following the shock
in question b). Explain the mechanism of the impact of such measure with the demonstration of suitable
graphs. 10 points

e) Compare the impacts the fiscal and monetary measures suggested in questions c), and d). In terms
of crowding-out/in effects, which policy measure is more effective? Please draw suitable graphs and explain
how fiscal and monetary policy measures can be combined to stop a severe recession from becoming a deep
crisis. 10 points

f) Using detailed graphical analysis, show the effects of a simultaneous decrease of the exogenous com-
ponent of the risk premium, r̄RP , coupled with an increase of the rate of interest rate-funded investment
tax credit, τr . Your graphical analysis must reflect that prior to the changes in r̄RP and τr , it holds that
rM P = 0. Provide detailed economic reasoning describing the set of adjustments as the economy transitions
from the old to the new short-run macroeconomic outcome. 20 points
4.4. THE AD-AS MODEL DURING DEEP CRISIS 71

4.4.4 Problem 4.4

An economy with price adjustment (for which foreign variables are exogenous) is described by the following
model:
Y = DD = C + I + G + T B;
C = C0 + Cy · Y − Cr · r;
I = I0 − Ir · r;
G = G0 + Gy · (Ȳ − Y );
T B = T B0 + T Bex · (1 − τy∗ ) · Y ∗ − T Bim · (1 − τy ) · Y + T B · ;
∗
 = 0 − r · (r
( − r ); ! )
 
MP MP e Y −Ȳ T
R = max 0, r̄ + π + φy · Ȳ
+ φpi · π − π ;
 

r = RM P − π e + r̄RP − ry · Yt − YtZLB |πt · I Yt < YtZLB |πt ;
 
π = π e + ω · Y Ȳ−Ȳ + s · I(Y > Ȳ );

PART A: In this part, inflation expectation π e = πt−1 .

a) Derive the AS and AD curves of aggregate demand and aggregate supply of this economy in the
medium run. Show your work. 10 points

b) Explain why the AD curve is down-ward sloping while the AS curve is upward sloping. 10 points

c) Solve for the medium run macroeconomic outcomes in terms of the model parameters. 10 points

d) Suppose that the economy is hit by a negative shock due to a drop in foreign demand. Please explain
how the economy will adjust in the short run, the medium run and the long run when there is no counter-
crisis measures from the government and the central bank.10 points

e) Explain both in words and graphically how fiscal and monetary measures could be combined to restore
the equilibrium output of the economy.10 points

PART B: In this part, inflation expectation is assumed to be static π e = π̄.

a) Derive the (non-conditioned) value of π ZLB .10 points

b) Provide detailed economic reasoning as to how π ZLB depends on r̄M P , and r̄RP . 10 points

c) Using detailed (AS-curve and extended AD-curve based) graphical analysis, show how – starting from
a normal business cycle outcome for which Y > Ȳ , a decrease in the marginal propensity to consume, Cy ,
coupled with a negative supply shock, s (s > 0), affects medium-run domestic output and inflation. Provide
detailed economic reasoning describing the set of adjustments as the economy transitions from the old to
the new medium-run macroeconomic outcome. 20 points