Workbook

Macroeconomía II

Last updated: March 6, 2022

This document contains the exercises students are expected to try to solve during the
semester. Most exercises are in English, since most of those exercises and solutions were
put together when this course was taught in English. However, the exercises used in the
TA sessions will be translated to Spanish.

Solutions to all exercises but those market as “proposed” are provided. The proposed
exercises will not be solved during the TA sessions—if they were solved, the solutions
would be added to this file, and those exercises would no longer be marked as proposed.
However, if you are having problems to solve any exercise (including the proposed ones)
you are free to ask for guidance.

This file may be updated during this semester, with new exercises and corrections. Make
sure you have the latest version.
Contents

1 Money supply 4
1.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

2 Money demand 16
2.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3 Monetary equilibrium 47
3.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

4 Seignioriage 57
4.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5 Goods and money markets 70

5.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

6 IS-LM 80
 6.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
6.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

7 Mundell-Fleming and exchange rates 109

7.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
7.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

8 Phillips curve 123

8.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
8.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

9 IS-LM-PC 126
9.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
9.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

10 Dynamic inconsistency 140

10.1 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
10.2 Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Chapter 1

Money supply

1.1 Exercises

Exercise 1.1

[Adapted from Mankiw] What are the three functions of money? Which of the functions
of money do the following items satisfy? Which do they not satisfy?

1. A credit card;

2. A painting by Rembrandt;

3. A subway token;

4. Bitcoins.

Exercise 1.2

[Mankiw] An economy has a monetary base of 1,000 $1 bills. Calculate the money
supply in scenarios 1-4 and answer part 5.

1. All money is held as currency.

4
Chapter 1. Money supply Macroeconomía II – 2022/1

2. All money is held as demand deposits. Banks hold 100 percent of deposits as
reserves.

3. All money is held as demand deposits. Banks hold 20 percent of deposits as

reserves.

4. People hold equal amounts of currency and demand deposits. Banks hold 20
percent of deposits as reserves.

5. The central bank decides to increase the money supply by 10 percent. In each of
the above four scenarios, how much should it increase the monetary base?

Exercise 1.3

Suppose an economy has two types of deposits: demand deposits (Dv ) and savings
deposits (Dp ). We define the money aggregate M1 as the sum of demand deposits
and currency (C ), that is, M1 ≡ Dv + C . We define the money aggregate M2 as
M2 ≡ M1 + Dp . The monetary base is given by H ≡ C + R, where R denote total
reserves in the banking sector. The total amount of deposits D is given by D ≡ Dv + Dp .
Assume agents always keep a ratio of currency (C ) to total deposits (D) equal to 1/4,
that is:
C 1
= .
D 4

Moreover, the ratio of demand deposits to total deposits and savings deposits to total
deposits is constant and given by

Dv 3 Dp 1
= and = .
D 4 D 4

5
Chapter 1. Money supply Macroeconomía II – 2022/1

Banks keep a ratio of reserves (R) to total deposits equal to some number θ ∈ (0, 1):

R
= θ.
D

1. Compute the M1 money multiplier (that is M1/H) as a function of θ.

2. Compute the M2 money multiplier (that is M2/H) as a function θ.

3. Find a condition that implies a M1 money multiplier smaller than one.

Exercise 1.4

Answer TRUE or FALSE to the following statements:

1. If during a financial crisis banks restrict the level of credits they grant to the
private sector, we should expect the money multiplier to increase.

2. If banks keep 100% of their deposits as reserves, the money multiplier is zero.

Exercise 1.5

1. El Covid-19 ha afectado economías en todo el mundo. La Figura abajo muestra

la relación de circulante a depósitos a la vista en los bancos comerciales de
EE.UU. Como se puede ver, durante la pandemia esta variable se ha reducido
significativamente. ¿Qué puede explicar dicha reducción en esta variable? Tomando
todo lo demás como constante, ¿como deberíamos esperar que dicho cambio va a
afectar la oferta de dinero en EE.UU? Explique claramente la intuición económica
por detrás de su respuesta.

6
Chapter 1. Money supply Macroeconomía II – 2022/1

Circulante dividido por depósitos a la vista

1.1

1.0

0.9

0.8
4

0
01

01

01

01

01

01

01

01

01

01

01

01

01

02

02
/2

/2

/2

/2

/2

/2

/2

/2

/2

/2

/2

/2

/2

/2

/2
4

1
/0

/1

/0

/0

/0

/0

/0

/0

/0

/0

/1

/0

/1

/0

/1
01

01

01

01

01

01

01

01

01

01

01

01

01

01

01
2. Suponga que como definición de oferta de dinero utilizamos el agregado monetario
M1, que es dado por la suma depósitos a la vista y circulante. Explique por que
el multiplicador monetario, en teoría, puede ser menor que 1.

Exercise 1.6

[Proposed] The graph below shows the M1 money multiplier for the US:
M1 Money Multiplier

2.2

2.0

1.8

1.6
Ratio

1.4

1.2

1.0

0.8

0.6
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018

Shaded areas indicate U.S. recessions Source: Federal Reserve Bank of St. Louis myf.red/g/iU0a

1. Provide one possible (theoretical) explanation for the multiplier being lower than
one.

2. Enter the FRED website (https://fred.stlouisfed.org/) and using the time series
for the monetary base and M2, compute the M2 money multiplier for the US.

7
Chapter 1. Money supply Macroeconomía II – 2022/1

3. Still on the FRED website, find data on:

(a) Excess reserves (e.g., “Excess Reserves of Depository Institutions”);

(b) Reserves (e.g., “Reserve Balances with Federal Reserve Banks”);

(c) Demand deposits (e.g., “Total Checkable Deposits”).

How does the data above helps you explain the fact that money multipliers have
remained very low after the Great Recession?

4. Read the article available on https://tinyurl.com/hd2zsxp and explain why US

banks excess reserves have increased so much in the last years.

Exercise 1.7

[Proposed] Imagine an economy with two types of deposits only: demand deposits (Dv )
and savings deposits (Dp ). Suppose the that the ratio is of reserves to demand deposits
is equal to 0.2, and the ratio of currency to demand deposits is 0.25.

1. What is the money multiplier if we define money as M1?

2. Consider now that ratio of reserves to saving deposits is 0.1 and the ratio of
currency to savings deposits is 0.2. What is the money multiplier if we define
money as M2?

Exercise 1.8

[Proposed, Mankiw] To increase tax revenue, the U.S. government in 1932 imposed a
2-cent tax on checks written on bank account deposits. (In today’s dollars, this tax
would amount to about 34 cents per check.)

1. How do you think the check tax affected the currency-deposit ratio? Explain.

8
Chapter 1. Money supply Macroeconomía II – 2022/1

2. Use the model of the money supply under fractional-reserve banking to discuss
how this tax affected the money supply.

3. Many economists believe that a falling in the money supply was in part responsible
for the severity of the Great Depression of the 1930s. From this perspective, was
the check tax a good policy to implement in the middle of the Great Depression?

Exercise 1.9

[Proposed] Suppose an economy with 10 agents and 10 banks, indexed by 1 to 10. Every
bank keeps 50% of deposits in reserves. The central bank gives $10 of currency to agent
1. Agent 1 deposits the money in the bank 1. Then, bank 1 lends half of it agent 2.
Agent 2 deposits all its loan at bank 2, that lends half of it to agent 3, that deposits
in bank 3, and so on. When the money finally arrives to agent 10, he decides not to
deposit it in bank 10 and keep the amount to himself.

1. Compute the increase in the money supply in this economy.

2. Now assume that agent i only deposits a fraction 1/i of the funds it gets in the
bank (instead of the full amount). To simplify, also assume that agent 4 decides
not to use the bank (instead of agent 10). What is the increase in the money
supply?

9
Chapter 1. Money supply Macroeconomía II – 2022/1

1.2 Solutions

Solution to Exercise 1.1

The three functions of money are: unit of account, medium of exchange and store of
value.
A credit card is a medium of exchange, since you can use it to buy goods and services.
It is usually not used as a store of value, and clearly it is not a an unit of account.
A Rembrandt painting is a store of value, but it not usually used as an unit of
account or a medium of exchange. A subway token may be an unusual store of value (if
you can resell or use it later). It is not usually used as a medium of exchange or unit of
account.
A bitcoin may be accepted as a medium of exchange within very small groups, and
owners almost always have to exchange them for dollars or some other currency before
using them to buy anything. Its volatility makes them a poor store of value. Finally, it
is rarely a unit of account. (Who has ever seen a price in term of Bitcoins?)

Solution to Exercise 1.2

Item 1

Let’s start deriving again the money multiplier. Let R/D = θ ∈ (0, 1) denote the
reserve to deposits ratio. Let c = C /D denote the currency to deposits ratio. From the
definition of M:
M =C +D

10
Chapter 1. Money supply Macroeconomía II – 2022/1

Let H = C + R denote the monetary base. Dividing M = C + D by H = C + R we get

the multiplier:
M C +D C /D + 1 c +1
= = = ≡m
H C +R C /D + R/D c +θ

Hence, M = mH. If all money is held as currency, we have that c → ∞. Therefore:

c +1
M = lim 1000 = 1000
c→∞ c + θ

(You need to use L’Hopital rule in that limit).

Item 2

In that case we have c = 0 and θ = 1. Therefore:

0+1
M= 1000 = 1000
0+1

Item 3

In that case we have c = 0 and θ = 0.2. Therefore:

0+1
M= 1000 = 5000
0 + 0.2

Item 4

In that case c = 1 and θ = 0.2. Therefore:

1+1
M= 1000 = 1666, 66
1 + 0.2

11
Chapter 1. Money supply Macroeconomía II – 2022/1

Item 5

Since M = mH we have that ∆M = m∆H, which dividing by M = mH yields

∆M ∆H
=
M H

Hence, if we want the money supply to increase 10%, we need to increase the monetary
base in 10%. Hence, the central bank should increase the monetary base in $100 in
every case.

Solution to Exercise 1.3

Item 1

The definition of M1 and H are:

M1 = Dv + C (1)

and
H = C + R. (2)

Dividing (1) by (2), manipulating and using C

D
= 14 , Dv
D
= 3
4
and R
D
= θ:

M1 Dv + C Dv /D + C /D 3/4 + 1/4 1
= = = = .
H C +R C /D + R/D 1/4 + θ 1/4 + θ

Item 2

The definition of M2 is:

M2 = M1 + Dp

12
Chapter 1. Money supply Macroeconomía II – 2022/1

Which using the definitions of M1 and D becomes

M2 = D + C (3)

and
H = C + R. (4)

Dividing (3) by (4), manipulating and using C

D
= 1
4
and R
D
= θ:

M2 D +C D/D + C /D 1 + 1/4
= = = .
H C +R C /D + R/D 1/4 + θ

Item 3

To get M1
H
< 1 we need:
1
< 1 ⇒ θ > 3/4.
1/4 + θ

Solution to Exercise 1.4

Item 1

False. As shown in Exercise 1.2, the money multiplier is given by m = c+1

c+θ
, where c
is the currency to deposits ratio, and θ is the reserves to deposits ratio. Hence, when
banks restrict credit, they accumulate more reserves as proportion of the deposits they
receive, meaning that θ increases. This reduces the money multiplier.

Item 2

False. Using again that the money multiplier is given by m = c+1

c+θ
as in the previous
item, we have that θ = 1implies m = 1.

13
Chapter 1. Money supply Macroeconomía II – 2022/1

Solution to Exercise 1.5

Item 1

Una explicación posible es que, debido a la pandemia, los hogares cambiaron sus patrones
de consumo, usando el comercio electrónico para consumir bienes que antes compraban
presencialmente. Como el comercio electrónico en general no se paga con efectivo, esto
puede explicar por qué se está usando menos circulante en la economía americana.
Pueden haber otras explicaciones igualmente buenas, la pregunta es bastante abierta.
Si los agentes usan menos circulante relativo a depósitos, esto debería llevar a
una subida del multiplicador monetario, lo que llevaría a una subida de la oferta
de dinero, tomando la base monetaria como constante. Intuitivamente, cuando los
agentes depositan una fracción mayor de su dinero en los bancos comerciales, los bancos
comerciales terminan por generar más dinero en la economía, una vez que no todo dinero
que es depositado en una cuenta corriente es mantenido como reservas del banco (parte
de los depósitos de uno se convierten en préstamos, que suben las tenencias de dinero
de otros agentes).
Uno puede derivar matemáticamente el multiplicador monetario de la siguiente man-
era para ayudar en la respuesta: Oferta de dinero(M) = Circulante(C ) + Depósitos(D)
Dividiendo todo por la base monetaria H = Reservas(R) + Circulante(C ):

M C +D C /D + 1
Multiplicador Monetario = = =
H R +C R/D + C /D

Suponiendo que la relación reservas-depósitos está constante, una reducción en C /D

lleva a una subida del multiplicador monetario.

14
Chapter 1. Money supply Macroeconomía II – 2022/1

Item 2

Si parte de los depósitos en la economía no son considerados como parte de la oferta

monetaria, puede pasar que cuando el banco central pone 100 pesos en la economía
este dinero se transforme en reservas y depósitos a plazo, sin generar una subida en
depósitos a la vista. Por ejemplo, suponga que una persona deposita 80 de estos 100
pesos en una cuenta de depósitos a plazo y el banco mantiene 100% de este monto en
reservas (para simplificar). Entonces, el cambio total en la oferta de dinero M1 será de
solamente 20 pesos.

15
Chapter 2

Money demand

2.1 Exercises

Exercise 2.1

Consider the simple Baumol-Tobin model where an individual spends uniformly his
annual income Y and makes n withdrawals of equal size to minimize his opportunity
cost (iY /2n) plus the linear cost of withdrawing (Zn), where i > 0 denotes the nominal
interest rate and Z > 0.

1. Write the problem of minimizing the costs and identify the trade-off between the
opportunity cost and the linear cost.

2. What is the most important conclusion of this model and what are its main
assumptions? How can you justify the cost of withdrawing?

3. How do you think the demand for money would be affected if the quantity of
banks in which people can make withdrawals increased?

4. Suppose know that people can keep their funds in a savings account (that pays the

16
Chapter 2. Money demand Macroeconomía II – 2022/1

interest rate i) and access this funds whenever they need it for transactions using
a debt card (let’s call it electronic money). The money is discounted from the
account only at the moment that the transaction takes place, and thus people do
not incur the opportunity cost i. On the other hand, for every dollar discounted
from the account using the debt card, there is a fee τ ∈ (0, 1) that must be
paid. What happens to the demand for currency in this economy? Under which
conditions there is a positive demand for currency in this economy?

5. Suppose the same environment as in the previous item, except that now not all
sellers accept electronic money. Individuals spend a fraction λ of their income on
sellers that accept only currency and a fraction 1 − λ on sellers that accept both
currency and electronic money, where λ ∈ (0, 1). Find the demand for currency
as a function of τ and λ. What happens to the demand for currency when λ → 0?

Exercise 2.2

Answer the questions below according to the Baumol-Tobin presented in class (which is
also presented also Section 15.5.3 of de Gregorio’s book). Are the statements below true
or false? Justify.

1. According to the Baumol-Tobin model, a 10% increase in the interest rate causes
a 5% increase in the demand for money.

2. According to the Baumol-Tobin model, the higher the cost of going to the bank,
the higher the elasticity of the money demand with respect to income.

3. Assume the amount of transactions agents realize in a given year is equal to their
income (as usual in the Baumol-Tobin model). Then, the Baumol-Tobin model
predicts that high-income individuals will hold a larger fraction of their income in
monetary assets than low-income individuals.

17
Chapter 2. Money demand Macroeconomía II – 2022/1

Exercise 2.3

Consider a centralized economy where households produce and consume. The production
function Y (Kt, nt ) is homogeneous of degree one in both capital and labor. Here,
household incomes are production Yt , monetary transfers from the government Tt and
money balances Mt . On the other hand, the outcomes are consumption ct , investment in
capital It and money balance held for the next period Mt+1 . Capital evolves according
to the following expression: Kt+1 = Kt (1 − δ) + It .
Now let’s think of a decentralized economy, where households own the resources
capital Kt and labor nt , which is then rented every period by the firms receiving as
payment interest rate Pt rt for capital and wages Pt wt for labor (hence, the real rental
rate of capital is rt and the real wage is wt ). In this economy, there are transfers Tt and
money Mt+1 that households accumulate in t to use in t + 1.

1. Present the budget constraint for both centralized and decentralized economies in
nominal terms.

2. Present the budget constraint for both centralized and decentralized economies in
real terms.

3. Show that we can recover decentralized budget constraint from centralized one in
this setting assuming perfect competition. (Tip: if a function f (x, y ) is homoge-
neous of degree one, then f (x, y ) = ∂f (x,y )
∂x
x + ∂f (x,y )
∂y
y ).

Exercise 2.4

Consider an economy in which there is a single good that can be used as capital or
for consumption. Time is discrete and indexed by t ∈ {−1, 0, 1, 2, ...}. Consider the
following problem of the representative household:

18
Chapter 2. Money demand Macroeconomía II – 2022/1

∞
X
max β t u (ct )
{ct ,kt ,Bt ,Mt }t≥0
t=0

s.t. Pt ct + Pt kt + Bt + Mt = Pt f (kt−1 ) + (1 + it−1 )Bt−1

+ Mt−1 + (1 − δ)Pt kt−1 + Tt , ∀t ≥ 0

ψPt ct ≤ Mt−1 + Tt , ∀t ≥ 0

k−1 = k > 0, B−1 = B > 0, M−1 = M > 0

lim (Bt /Pt ) = 0,

t→∞

kt , ct , Mt ≥ 0, ∀t ≥ 0.

The notation is standard and is the same used in class: u(c) is the instantaneous utility
function; f (k) is the production function; Pt denotes the price of the good at date t;
Bt is the nominal amount of bonds the household chooses to hold at date t; Mt is the
quantity of money he chooses to carry from date t to date t + 1; it is the nominal interest
rate between dates t and t + 1; ct is the household consumption at date t; kt is the
amount of capital he carries from date t to t + 1; Tt are cash transfers received at date t;
β ∈ (0, 1) is the discount factor and δ ∈ (0, 1) is the depreciation rate of capital. Define
bt ≡ Bt /Pt , mt ≡ Mt /Pt , πt ≡ (Pt − Pt−1 )/Pt−1 ,τt ≡ Tt /Pt and let rt denote the real
interest rate between dates t and t + 1.
The constraint ψPt ct ≤ Mt−1 + Tt is interpreted as a standard cash-in-advance
constraint, except that now only a fraction ψ ∈ (0, 1) of the household consumption is
purchased using cash.
In what follows, assume that u(·) and f (·) satisfy all the usual assumptions that
guarantee an unique interior solution (and thus the non-negativity constraints kt , ct , Mt ≥
0 never bind in the optimal choice and you can ignore them). Moreover, assume the
cash-in-advance constraint binds at all dates in the optimal solution (which is true if
it > 0, for every t). There is no uncertainty and the household takes as given the path
of all exogenous variables.

19
Chapter 2. Money demand Macroeconomía II – 2022/1

1. Rewrite the budget constraint and the cash-in-advance constraint in real units
(that is, in terms of only the real variables {ct , kt , bt , mt , rt , τt }t≥0 and the inflation
rate {πt }t≥0 ). (Tip: use the Fisher equation to get rid of nominal interest rates).

2. Denote by {λt }t≥0 the Lagrange multipliers associated to the budget constraint and
by {µt }t≥0 the Lagrange multipliers associated to the cash-in-advance constraint.
Write down the Lagrangian using the budget constraints in real units and derive
the first order conditions for ct , kt , bt and mt .

3. In the optimal solution we get the following Euler equation:

u ′ (ct )
= β (1 + rt ) h(it−1 , it , ψ)
u ′ (ct+1 )

where h(it−1 , it , ψ) is a function of it−1 , it and ψ. Derive the functional form of

h(it−1 , it , ψ).

4. Interpret economically the Euler equation you found in the previous item when
ψ = 1. (Tip: you may find useful to rearrange the equation before your interpret
it.)

5. Show that money is superneutral in this model.

Exercise 2.5

Consider the CIA model in Section 3.3.1 of Walsh’s book (also seen in class) with one
modification: transfers received at date t can not be directly used to purchase goods at
date t, so that the the cash-in-advance constraint takes the form Pt ct ≤ Mt−1 . Assume
that u(c) = ln c. Also, assume that the CIA constraint always binds.

1. Write down the household’s decision problem.

20
Chapter 2. Money demand Macroeconomía II – 2022/1

2. Write down the household’s first-order conditions.

3. Derive the Euler equation.

4. Show that:
Mt
= β (1 + it−1 ) .
Mt−1

5. Explain what happens to the nominal interest rate in 3 scenarios:

(a) Moneys grows at a constant rate: Mt /Mt−1 = 1 + µ;

(b) Money grows at an increasing rate: Mt /Mt−1 = 1 + µt;

(c) Money grows at a decreasing rate: Mt /Mt−1 = 1 + 1/ (µt).

Exercise 2.6

When we presented the CIA model in class, we wrote down the centralized version of
it: a single household decided how much to consume, how much capital, money and
bonds to hold. The household had access to a production function that would transform
this capital into consumption goods. Now we set up a decentralized version of the CIA
model.
Consider an economy in which there is a single consumption good that can be used
as capital or for consumption. Time is discrete and indexed by t ∈ {−1, 0, 1, 2, ...}.
There is a representative household and a competitive firm in this economy. Both the
firm and the household are price takers.

Household. The household owns the firm and chooses consumption and how much
capital, bonds and money to hold. At each date t, the household rents the capital it
brought from the previous period (kt−1 ) to the firm at a rental rate Pt × Rt (hence, the

21
Chapter 2. Money demand Macroeconomía II – 2022/1

rental rate in units of the consumption good is Rt ). Every period, the household also
receives profits Dt from the firm. Hence, the household problem is:

∞
X
max β t u (ct )
{ct ,kt ,Bt ,Mt }t≥0
t=0

s.t. Pt ct + Pt kt + Bt + Mt = Pt Rt kt−1 + (1 + it−1 )Bt−1

+ Mt−1 + (1 − δ)Pt kt−1 + Tt + Dt , ∀t ≥ 0

Pt ct ≤ Mt−1 + Tt , ∀t ≥ 0

B−1 = B > 0, M−1 = M > 0

lim (Bt /Pt ) = 0,

t→∞

ct , Mt ≥ 0, ∀t ≥ 0.

We are assuming that the profits Dt cannot be directly used to buy consumption goods,
and hence that money is not useful for transactions, while the transfers Tt are received at
the beginning of the period (and hence can be used for transactions). That is why only
Tt shows up in the CIA constraint. The notation is standard and is the same used in
class: u(c) is the instantaneous utility function; Pt denotes the price of the good at date
t; Bt is the nominal amount of bonds the household chooses to hold at date t; Mt is the
quantity of money she chooses to carry from date t to date t +1; it is the nominal interest
rate between dates t and t + 1; ct is the household consumption at date t; Tt are cash
transfers received at date t; β ∈ (0, 1) is the discount factor and δ ∈ (0, 1) is the capital
depreciation rate. Define bt ≡ Bt /Pt , mt ≡ Mt /Pt , πt ≡ (Pt − Pt−1 )/Pt−1 ,τt ≡ Tt /Pt ,
dt ≡ Dt /Pt and let rt denote the real interest rate between dates t and t + 1. Household
solve their problem taking the sequence of prices, interest rates, transfers and profits as
given.

Firm. The firm has access to a production function f (k̃). At each date t, the firm
chooses how much capital k̃t to rent from the household. Hence, at each date t the firm

22
Chapter 2. Money demand Macroeconomía II – 2022/1

problem is:
max Πt ≡ Pt f (k̃t ) − Pt Rt k̃t
k̃t

subject to k̃t ≥ 0.

Market clearing. The following market clearing must hold:

Bt = 0, ∀t ≥ 0

k̃t = kt−1 , ∀t ≥ 0

ct + kt = f (kt−1 ), ∀t ≥ 0

Moreover, Tt = Mt − Mt−1 , since the government can only make transfers by issuing
more money.
Assume that u(·) and f (·) satisfy all the usual assumptions that guarantee an unique
interior solution (and thus the non-negativity constraints never bind in the optimal
choice and you can ignore them). Moreover, assume the cash-in-advance constraint
binds at all dates in the optimal solution (which is true if it > 0, for every t). There is
no uncertainty and the household takes as given the path of prices, interest rates and
all exogenous variables.

1. Write down the household problem in real units.

2. Derive the first order conditions of the household problem.

3. Write the first order condition of the firm.

4. Show that in equilibrium Rt = f ′ (kt−1 ) and rt = Rt+1 − δ .

5. Show that the steady state in the decentralized model is the same as in the
centralized model (that is, the model of section 3.3.1 in Walsh’s book, which is

23
Chapter 2. Money demand Macroeconomía II – 2022/1

the one seen in class).

Exercise 2.7

[Proposed] Suppose the representative household chooses consumption of cash goods

({ct }t≥0 ), consumption of credit goods ({dt }t≥0 ), nominal bond holdings ({Bt }t≥0 ) and
hours of work ({nt }t≥0 ) to maximize

∞
X
β t [u(ct , dt ) − γnt ]
t=0

subject to the nominal budget constraint

Pt (ct + dt ) + Mt + Bt = Wt nt + (1 + it−1 ) Bt−1 + Tt + Mt−1

and a cash-in-advance constraint of the form

Pt ct ≤ Mt−1 + Tt .

Where γ > 0, nominal wages are denoted by Wt and the remaining notation is the
standard one. Denote the real wage by wt and real bond holdings by bt . The usual
non-negativity constraint on c and l must hold, as well as the solvency constraint
(limt→∞ bt = 0). Moreover, bonds and money at the beginning of period zero are given.
Assume that:
u(ct , dt ) = α ln ct + (1 − α) ln dt , with α ∈ (0, 1).

1. Write the cash-in-advance constraint and budget constraint in real units.

2. Derive the first order conditions with respect to ct , nt , dt and bt .

3. Compute the steady state (assuming a constant inflation rate, real wage, con-

24
Chapter 2. Money demand Macroeconomía II – 2022/1

sumption, hours of work, bond and money holdings). Does this model exhibit
superneutrality? Explain.

4. Maximizing utility at the steady state, what is the optimal inflation in this
economy?

5. Write down an expression for the share of cash in total consumption as a function
of the nominal interest rate. How does the share of cash goods consumed depend
on the interest rate?

Exercise 2.8

[Proposed] Consider the CIA model in Section 3.3.1 of Walsh’s book (also seen in class)
with one modification: to accumulate capital agents need to have cash in advance. In
other words, the CIA constraint now is given by:

Pt ct + Pt (kt − (1 − δ) kt−1 ) ≤ Mt−1 + Tt

That is, now not only consumption must be purchased with cash, but also capital goods.

1. Present the Lagrangian of this problem, denoting by λt and µt the multipliers for
the budget constraint and the CIA constraint. Derive the first order conditions.

2. Show that:

 
′ ′ 1
u (ct ) = βu (ct+1 ) f ′ (kt ) + βu ′ (ct+1 )(1 − δ)
1 + it+1

3. Assuming a production function f (k) = k α , with α ∈ (0, 1) solve for the stationary
state capital k ss . Is money superneutral in this version of the model? Give some
intuition on your result.

25
Chapter 2. Money demand Macroeconomía II – 2022/1

4. Now consider the same economy but with the usual CIA constraint Pt ct ≤ Mt−1 +Tt .
The steady state capital in that economy, with f (kt ) = ktα , is given by:

1
! 1−α
α
k ss = 1
β
−1+δ

Find the inflation rate that equates steady state capital in both economies and then
explain how this relates to Friedman rule (the Friedman rule says that inflation
should be roughly equal to minus the real interest rate, i.e., π = −r ).

26
Chapter 2. Money demand Macroeconomía II – 2022/1

2.2 Solutions

Solution to Exercise 2.1

Item 1

The agent chooses the number of withdrawals n ∈ R+ to minimize

 
Y
C (n) = i + Zn
2n

Note that for simplicity we are assuming that n is chosen from R+ instead of the being
chosen from the set of non-negative integers {0, 1, 2, ...} (as if the agent could do one
and a half withdrawals, for example).
Notice that the term Y
2n
represents the agent average money holdings.1 Hence, when
the agent increases the number of withdrawals it reduces the opportunity cost of holding
money, represented by i 2nY
, since she holds less cash on average, but needs to pay the

cost of withdrawing (represented by Zn).

1
To see that, notice that an agent that goes n times to the bank will withdraw Y /n each time (so
that the total amount withdrawn after n withdrawals is equal to Y ). When the exercise says that the
individual spends his income uniformly, we mean that dM dt is constant between each withdrawals (where
t

Mt denotes the agent money holdings at some date t, measured in years). Hence, in interval of time
[0, 1/n] (i.e., the interval between the first and the second withdrawal) the agent money holdings M̃t
are a linear function of time that: (i) is equal to Y /n when t = Y /n (since he withdrawals Y /n at
t = 0); (ii) approaches zero when t = 1/n (the moment of the second withdrawal), since she is about
to make a new withdrawal when t = 1/n, and she only withdrawals when there is no more money
holdings. Therefore, M̃t = Yn − Y × t, for t ∈ [0, 1/n]. Hence, the average money holdings in the interval
t ∈ [0, 1/n] is:
R 1/n Y
" 1/n #
n − Y × t dt t 2 
 
D 0 Y Y 1 Y
M = =n t −Y  =n 2 −Y 2 =
1/n n 2 0 n 2n 2n
Since the average money holdings between any two consecutive withdrawals is the same, the average
Y
money is 2n .

27
Chapter 2. Money demand Macroeconomía II – 2022/1

Item 2

Taking the first order condition we find the n that minimizes the cost:

iY
=Z
2n2

And hence, the optimal n, denoted by n∗ , is:

r
∗ iY
n =
2Z

Therefore, the average money holdings (that is, the money demand) is given by

r r r
Y Y 2Z Y 2 2Z YZ
D
M = q = = = (1)
iY
2 2Z 2 iY 4 iY 2 i

Taking logs:
1
ln M D = (ln Y + ln Z − ln 2 − ln i) (2)
2

Therefore, we have two main conclusions: (i) the elasticity of the money demand with
respect to income is equal to 1/2; (ii) the elasticity of the money demand with respect
to the nominal interest rate is equal to −1/2.

Item 3

That would probably reduce the cost of going to the bank (Z ). As equation (1) shows,
that should reduce the money demand.

Item 4

The agent now has two decisions to make. She must choose how much of his total yearly
transactions (Y ) she is going to conduct using currency (denoted by X C ) and how much

28
Chapter 2. Money demand Macroeconomía II – 2022/1

she is going to conduct using electronic money (denoted by X E ). (Note that X E and
X C must be such that X E + X C = Y .) Once she has fixed an amount X C , then she
chooses how many times to go to the bank to achieve that amount of transactions using
currency.
First, let’s suppose an agent has chosen a given level of X C . From our solution
q
of the previous items, we know that the agent will go n = iX2Z times to bank and
∗ C

q
her average money holding is X2 Zi . Therefore, the total cost of spending X C using
C

currency is:

r r
C C XC Z iX C
C (X ) = i +Z
2 i}
| {z 2Z}
| {z
average money holdings # of times she goes to the bank

The total cost incurred for a given X C and X E (after optimally choosing how much to
go to the bank to achieve a given X C ) is given by:

r r
C EXC Z iX C
C (X , X ) = i +Z + XE
τ|{z}
| 2 i {z 2Z}
cost of electronic transactions
C C (X C )

Which simplifies to:

√
C (X C , X E ) = 2iZX C + τ X E

The agent chooses X C and X E to minimize C (X C , X E ) above subject to X E + X C = Y ,

X E ≥ 0 and X C ≥ 0. Hence, replacing X E = Y − X C in C (X C , X E ) we get:

√
C̃ (X C ) ≡ C X C , Y − X C = 2iZX C + τ Y − X C (3)

Therefore, we can simplify the problem: the agent chooses X C ∈ [0, Y ] to minimize (3)
(and then X E is given by Y − X C ). Notice that the function above is concave. This

29
Chapter 2. Money demand Macroeconomía II – 2022/1

implies that we can only have corner solutions: either the agent chooses X C = 0 or
X C = Y . Therefore, the agent chooses X C = Y whenever:

√
τ Y >
|{z} | 2iZY
{z }
C̃ (0) C̃ (Y )

Which yields:

r
2iZ
τ>
Y
q
Hence, when τ > the agent chooses to make all transactions with currency (i.e.,
2iZ
Y
q
she chooses X = Y ). When τ < 2iZ
C
she chooses to only use electronic money (i.e.,
qY
she chooses X C = 0). When τ = 2iZ Y
the agent is indifferent between X C = 0 or
XC = Y.

Item 5

Now the agent solves the same problem as in item 4, with the additional constraint
that X C ≥ λY (since at fraction λ of the income must be spent with currency). Hence,
instead of choosing X C ∈ [0, Y ] to minimize (3), the agent now chooses X C ∈ [λY , Y ]
to minimize (3). As before, we can only have corner solutions to this problem, since the
objective function is concave. Hence, the agent chooses X C = λY if:

√ √
2iZ λY + τ (Y − λY ) < | 2iZY
{z }
| {z }
C̃ (λY ) C̃ (Y )

Which becomes:  √ r
1− λ 2iZ
τ<
(1 − λ) Y
√
(1− λ)
q
Hence, when τ < (1−λ)
2iZ
Y
the agent chooses to make as little transactions as

30
Chapter 2. Money demand Macroeconomía II – 2022/1
√
(1− λ)
q
possible with currency and chooses X C
= λY . Similarly, when τ > (1−λ)
2iZ
Y
, the
agents chooses to make all her transactions with currency and then chooses X C = Y .
√ q
(1− λ) 2iZ
When τ = (1−λ) Y
the agent is indifferent between X C = λY and X C = Y . When
√ q
(1− λ) 2iZ
q
λ → 0 we have that (1−λ) Y
→ 2iZ Y
, which is the same cutoff on τ we found in
item 4. Moreover, limλ→0 λY = 0. Hence, the money demand converges to that of item
4 (which is not surprising, of course).

Solution to Exercise 2.2

Item 1

False. Using equation (2) from our answer of Exercise 4 one can see that the elasticity
of the money demand with respect to the interest rate is constant and equal to −1/2.
Therefore, when i increases 10%, the money demand falls 5%.

Item 2

False. Using (2) one can see that the elasticity of the money demand with respect to
income is constant and equal to 1/2.

Item 3

False. Dividing (1) by Y we get:

r r r
MD 1 YZ 1 YZ Z
= = =
Y Y 2 i Y2 2 i 2Yi

Hence, the share of the income held as money is decreasing in income.

Solution to Exercise 2.3

31
Chapter 2. Money demand Macroeconomía II – 2022/1

Item 1

Note the notation in this exercise is a little different from the one we used in class: We
denote the money the agent decides to carry from date t to date t + 1 by Mt+1 (and
not by Mt as we did in class). Similar change of notation holds for capital (Kt+1 is the
capital decides to accumulate at date t). Also, note that we did not introduce bonds.
The budget constraint of the centralized economy is:

Pt ct + Pt It + Mt+1 = Pt Y (Kt , nt ) + Mt + Tt (1)

(You could replace It by Kt+1 − (1 − δ) Kt in the equation above). The budget constraint
of decentralized economy is:

Pt ct + Pt Kt+1 + Mt+1 = Pt (1 − δ)Kt + Pt wt nt + Pt rt Kt + Mt + Tt (2)

Item 2

Dividing by Pt the budget constraint of the centralized economy:

Mt+1 Mt Tt
ct + It + = Yt + +
Pt Pt Pt

Defining mt ≡ Mt /Pt and τt ≡ Tt /Pt and using that

Mt+1 Mt+1 Pt+1

= = mt+1 (1 + πt+1 )
Pt Pt+1 Pt

We can write it as:

ct + It + mt+1 (1 + πt+1 ) = Yt + mt + τt

32
Chapter 2. Money demand Macroeconomía II – 2022/1

Similarly, for the decentralized economy we have:

Mt+1 Mt T t
ct + Kt+1 + = (1 − δ)Kt + wt nt + rt Kt + +
Pt Pt Pt

ct + Kt+1 + mt+1 (1 + πt+1 ) = (1 − δ)Kt + wt nt + rt Kt + mt + τt

Item 3

The homogeneity of degree one property of Y (K , N) allows us to write:

Y (Kt , Nt ) = Yn (Kt , Nt )nt + Yk (Kt , Nt )Kt

In perfect competition, the maximizing firm would pay each factor according to its
marginal contribution to production:

wt = Yn (Kt , Nt ) rt = Yk (Kt , Nt ) (3)

So the nominal budget constraint (1) in the centralized economy becomes:

Pt ct + Pt It + Mt+1 = Pt (Yn (Kt , Nt )nt + Yk (Kt , Nt )Kt ) +Mt + Tt

| {z }
Y (Kt ,nt )

Replacing It = Kt+1 − (1 − δ) Kt and (3) above we get:

Pt ct + Pt Kt+1 + Mt+1 = Pt (1 − δ)Kt + Pt wt nt + Pt rt Kt + Mt + Tt

Which is identical to (2).

Solution to Exercise 2.4

33
Chapter 2. Money demand Macroeconomía II – 2022/1

Item 1

Dividing the budget constraint by Pt :

Bt−1 Mt−1
ct + kt + bt + mt = f (kt−1 ) + (1 + it−1 ) + + (1 − δ)kt−1 + τt
Pt Pt

Rearranging:

Bt−1 Pt−1 Mt−1 Pt−1

ct + kt + bt + mt = f (kt−1 ) + (1 + it−1 ) + + (1 − δ)kt−1 + τt
Pt−1 Pt Pt−1 Pt

1 + it−1 mt−1
ct + kt + bt + mt = f (kt−1 ) + bt−1 + + (1 − δ)kt−1 + τt
1 + πt 1 + πt

And using the Fisher equation:

mt−1
ct + kt + bt + mt = f (kt−1 ) + (1 + rt−1 ) bt−1 + + (1 − δ)kt−1 + τt
1 + πt

Similarly, we divide the CIA constraint by Pt and rearrange:

Mt−1
ψct ≤ + τt
Pt

Mt−1 Pt−1
ψct ≤ + τt
Pt−1 Pt
mt−1
ψct ≤ + τt
1 + πt

34
Chapter 2. Money demand Macroeconomía II – 2022/1

Item 2

The lagrangian is given by

∞   
X
t mt−1
L= β u(ct ) − µt ψct − − τt
t=0
1 + πt
 
mt−1
−λt ct + kt + bt + mt − f (kt−1 ) − (1 + rt−1 ) bt−1 − − (1 − δ)kt−1 − τt
1 + πt

The FOC with respect to ct is:

β t [u ′ (ct ) − λt − µt ψ] = 0 (FOC1)

The FOC with respect to kt is:

−β t λt + β t+1 λt+1 [f ′ (kt ) + 1 − δ] = 0 (FOC2)

The FOC with respect to bt is:

−β t λt + β t+1 λt+1 (1 + rt ) = 0 (FOC3)

The FOC with respect to mt is:

β t+1
−β t λt + (µt+1 + λt+1 ) = 0 (FOC4)
1 + πt+1

Item 3

Subtracting (FOC4) from (FOC3) and solving for µt+1 :

β t+1
β t+1 λt+1 (1 + rt ) − (µt+1 + λt+1 ) = 0
1 + πt+1

35
Chapter 2. Money demand Macroeconomía II – 2022/1

µt+1 = λt+1 [(1 + πt+1 ) (1 + rt ) − 1]

Thus, replacing µt = λt [(1 + πt ) (1 + rt−1 ) − 1] on (FOC1) we get:

u ′ (ct ) − λt − ψλt [(1 + πt ) (1 + rt−1 ) − 1] = 0

Using the Fisher equation it becomes:

 
1−ψ
′
u (ct ) = ψλt 1 + it−1 + (1)
ψ

Iterating (1) forward and dividing by (1):

h i
′
u (ct ) λt 1 + it−1 + 1−ψ
ψ
′
= h i
u (ct+1 ) λt+1 1−ψ
1 + it + ψ

But (FOC3) implies λt

λt+1
= β (1 + rt ) and then:

1−ψ
!
u ′ (ct ) 1 + it−1 + ψ
= β (1 + rt ) 1−ψ
(2)
u ′ (ct+1 ) 1 + it + ψ

Thus,
1−ψ
1 + it−1 + ψ
h(it−1 , it , ψ) = 1−ψ
.
1 + it + ψ

Item 4

When ψ = 1 we have:
u ′ (ct ) β (1 + rt ) u ′ (ct+1 )
= .
1 + it−1 1 + it

The LHS is the marginal benefit of one unit of consumption at date t divided by a
function of the opportunity cost of consuming at t (which is the foregone interest it−1

36
Chapter 2. Money demand Macroeconomía II – 2022/1

that the agent has to incur by buying less bonds and carrying more money at t − 1).
The RHS is the marginal benefit of consuming at date t + 1 divided by a function of the
opportunity cost of consuming at t + 1 (which is the foregone interest it that the agent
has to incur by buying less bonds and carrying more money at t). For the agent to be
indifferent between consuming at both dates, this “cost-benefit” ratio must be equal.

Item 5

Since limt→∞ bt = 0 we have that b ss = 0. The Euler equation (2) in steady state implies
1 + r ss = β1 . (FOC2) and (FOC3) imply:

f ′ (kt ) + 1 − δ = 1 + rt

Which in steady state becomes then:

1
f ′ (k ss ) + 1 − δ =
β

Hence, the level of capital in steady state is independent of inflation. Consumption in

steady state must satisfy the budget constraint with bt = 0, ∀t:

mss
c ss = f (k ss ) + + τ ss − δk sss − mss (3)
1 + π ss

But we know that Mt − Mt−1 = Tt , which rearranging becomes:

Mt − Mt−1 Tt
=
Pt Pt

Mt−1 Pt−1
mt − = τt
Pt−1 Pt
mt−1
mt − = τt
1 + πt

37
Chapter 2. Money demand Macroeconomía II – 2022/1

τ ss
Hence, τ ss = mss − 1+π ss
, which plugging in (3) yields:

c ss = f (k ss ) − δk ss ,

which is also independent of inflation.

Solution to Exercise 2.5

Item 1

The problem is identical to the problem we had before, the only differencia is the CIA
constraint. The budget constraint in real terms remains the same, so I will skip the
derivation. Now we write the CIA constraint it as:

Pt ct ≤ Mt−1

Which rearranging becomes:

Mt−1 Mt−1 Pt−1 mt−1

ct ≤ = =
Pt Pt−1 Pt 1 + πt

38
Chapter 2. Money demand Macroeconomía II – 2022/1

Hence, the household problem is:

∞
X
max β t u (ct )
{ct ,kt ,Bt ,Mt }t≥0
t=0

s.t. ct + kt + bt + mt = f (kt−1 ) + (1 + rt−1 ) bt−1

mt−1
+ + (1 − δ)kt−1 + τt , ∀t ≥ 0
1 + πt
mt−1
ct ≤ , ∀t ≥ 0
1 + πt
b−1 = b ≡ B/P−1 > 0, M−1 = m = M/P−1 > 0

lim (bt ) = 0,
t→∞

ct , mt , kt ≥ 0, ∀t ≥ 0.

where u(c) = ln c.

Item 2

The Lagrangian is given by

∞   
X
t mt−1
L= β u(ct ) − µt ct −
t=0
1 + πt
 
mt−1
−λt ct + kt + bt + mt − f (kt−1 ) − (1 + rt−1 ) bt−1 − − (1 − δ)kt−1 − τt
1 + πt

The FOC with respect to ct is:

β t [u ′ (ct ) − λt − µt ] = 0 (FOC1)

The FOC with respect to kt is:

−β t λt + β t+1 λt+1 [f ′ (kt ) + 1 − δ] = 0 (FOC2)

39
Chapter 2. Money demand Macroeconomía II – 2022/1

The FOC with respect to bt is:

−β t λt + β t+1 λt+1 (1 + rt ) = 0 (FOC3)

The FOC with respect to mt is:

β t+1
t
−β λt + (µt+1 + λt+1 ) = 0 (FOC4)
1 + πt+1

Item 3

We can rewrite (FOC1) and (FOC4) as (note that we used (FOC4) lagged):

u ′ (ct ) = λt + µt

βλt−1
= λt + µt
1 + πt

Hence:
(1 + πt ) λt−1
u ′ (ct ) =
β

Dividing the equation above by the same equation forward and multiplying both sides
by 1/β:
1 u ′ (ct ) 1 + πt λt−1
′
=
β u (ct+1 ) 1 + πt+1 λt β

But (FOC3) lagged immplies that λt−1

λt β
= 1 + rt−1 . Hence:

1 u ′ (ct ) 1 + πt
′
= (1 + rt−1 )
β u (ct+1 ) 1 + πt+1

40
Chapter 2. Money demand Macroeconomía II – 2022/1

Multiplying the RHS by (1 + rt )/(1 + rt ):

1 u ′ (ct ) (1 + πt ) (1 + rt−1 )
′
= (1 + rt )
β u (ct+1 ) (1 + πt+1 ) (1 + rt )

Using the Fisher equation we get:

1 u ′ (ct ) 1 + it−1
= (1 + rt )
β u ′ (ct+1 ) 1 + it

Since u ′ (c) = 1/c we can write:

1 + it−1
ct+1 = β (1 + rt ) ct
1 + it

Item 4

Since the CIA constraint binds, we have that ct = mt−1

1+πt
for all t. Plugging that in the
Euler equation we have:

mt 1 + it−1 mt−1
=β (1 + rt )
1 + πt+1 1 + it 1 + πt

Mt 1 + it−1 1 + πt+1 Mt−1

=β (1 + rt )
Pt 1 + it 1 + πt Pt−1
1 + it−1 1 + πt+1
Mt = β (1 + rt ) Mt−1
(1+π
t)

1 + it  1 + πt
 


1 + it−1
Mt = β (1 + πt+1 ) (1 + rt ) Mt−1
1 + it

Using the Fisher equation:

1 + it−1
Mt = β (1+i
t )Mt−1
1 + it
  


41
Chapter 2. Money demand Macroeconomía II – 2022/1

Mt
= β (1 + it−1 )
Mt−1

Item 5

In scenario (a) we have that:

β (1 + it−1 ) = 1 + µ

Hence, interest rates remain constant. In scenario (b) we have that:

β (1 + it−1 ) = 1 + µt

Hence, interest rates increase as time passes. Finally, in scenario (c) we have that:

β (1 + it−1 ) = 1 + 1/ (µt)

Hence, interest rates decrease as time passes.

Note that in scenario (a) even though the money supply is increasing, the interest
rates remain constant. In scenario (b), the money supply is growing at a increasing rate,
and the interest are increasing. Note that this is the opposite one would expect from a
simple static model of money demand and money supply. This a common feature in this
kind of model. In a stochastic version of the CIA model, one can often find that interest
rates increase after a random monetary expansion (this is called the liquidity puzzle).

Solution to Exercise 2.6

42
Chapter 2. Money demand Macroeconomía II – 2022/1

Item 1

Dividing the budget constraint by Pt :

Bt−1 Mt−1
ct + kt + bt + mt = Rt kt−1 + (1 + it−1 ) + + (1 − δ)kt−1 + τt + Dt
Pt Pt

Rearranging:

Bt−1 Pt−1 Mt−1 Pt−1

ct + kt + bt + mt = Rt kt−1 + (1 + it−1 ) + + (1 − δ)kt−1 + τt + dt
Pt−1 Pt Pt−1 Pt

1 + it−1 mt−1
ct + kt + bt + mt = Rt kt−1 + bt−1 + + (1 − δ)kt−1 + τt + dt
1 + πt 1 + πt

And using the Fisher equation:

mt−1
ct + kt + bt + mt = Rt kt−1 + (1 + rt−1 ) bt−1 + + (1 − δ)kt−1 + τt + dt
1 + πt

Similarly, we divide the CIA constraint by Pt and rearrange:

Mt−1
ct ≤ + τt
Pt

Mt−1 Pt−1
ct ≤ + τt
Pt−1 Pt
mt−1
ct ≤ + τt
1 + πt

43
Chapter 2. Money demand Macroeconomía II – 2022/1

Hence, the household solves:

∞
X
max β t u (ct )
{ct ,kt ,Bt ,Mt }t≥0
t=0

s.t. ct + kt + bt + mt = Rt kt−1 + (1 + rt−1 ) bt−1

mt−1
+ + (1 − δ)kt−1 + τt + dt , ∀t ≥ 0
1 + πt
mt−1
ct ≤ + τt , ∀t ≥ 0
1 + πt
b−1 = b ≡ B/P−1 > 0, M−1 = m = M/P−1 > 0

lim (bt ) = 0,
t→∞

ct , mt , kt ≥ 0, ∀t ≥ 0.

Item 2

The Lagrangian is given by

∞   
X
t mt−1
L= β u(ct ) − µt ct − − τt
1 + π t
 t=0 
mt−1
−λt ct + kt + bt + mt − Rt kt−1 − (1 + rt−1 ) bt−1 − − (1 − δ)kt−1 − τt − dt
1 + πt

The FOC with respect to ct is:

β t [u ′ (ct ) − λt − µt ] = 0 (FOC1)

The FOC with respect to kt is:

−β t λt + β t+1 λt+1 [Rt+1 + 1 − δ] = 0 (FOC2)

44
Chapter 2. Money demand Macroeconomía II – 2022/1

The FOC with respect to bt is:

−β t λt + β t+1 λt+1 (1 + rt ) = 0 (FOC3)

The FOC with respect to mt is:

β t+1
t
−β λt + (µt+1 + λt+1 ) = 0 (FOC4)
1 + πt+1

Item 3

The first order condition of the firm is:

 
f ′ k̃t = Rt (FOC5)

Item 4

Market clearing implies that k̃t = kt−1 . Hence, (FOC5) implies that:

f ′ (kt−1 ) = Rt

Combining (FOC2) and (FOC3) we get:

β t+1 λt+1 [Rt+1 + 1 − δ] = β t+1 λt+1 (1 + rt )

rt = Rt+1 − δ

45
Chapter 2. Money demand Macroeconomía II – 2022/1

Item 5

All the FOCs are the same as in the model seen in class, except for (FOC2). Plugging
f ′ (kt−1 ) = Rt in (FOC2):

−β t λt + β t+1 λt+1 [f ′ (kt ) + 1 − δ] = 0 (FOC2’)

Note that (FOC1), (FOC2’), (FOC3) and (FOC4) are identical to the first-order
conditions in the centralized model seen in class. Hence, we should arrive at the same
steady state, since we are departing from the exact same set of equilibrium conditions.

46
Chapter 3

Monetary equilibrium

3.1 Exercises

Exercise 3.1

[Mankiw] In the country of Wiknam, the velocity of money is constant. Real GDP grows
by 3 percent per year, the money stock grows by 8 percent per year, and the nominal
interest rate is 9 percent. What is the growth rate of nominal GDP, the inflation rate
and the real interest rate?

Exercise 3.2

[Mankiw] Suppose a country has a money demand function M d /P = κy , where κ is a

constant parameter and y is real income. The money supply grows by 12 percent per
year, and real income grows by 4 percent per year.

1. What is the average inflation rate?

2. How would inflation be different if real income growth were higher? Explain.

47
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

3. How do you interpret the parameter κ? What is its relationship to the velocity of
money?

4. Suppose, instead of a constant money demand function, the velocity of money in

this economy was growing steadily because of financial innovation. How would
that affect the inflation rate? Explain.

Exercise 3.3

Consider the basic Cagan’s model in which the log of the price level (pt ) and the log of
the money supply (mt ) satisfy the following difference equation:

mt − pt = −η (pt+1 − pt ) (1)

where η is a constant larger than zero. For simplicity, suppose that mt is constant and
equal to m̃, for every t. We know that a solution to (1) is given by

∞  i
1 X η
pt = mt+i = m̃
1 + η i=0 1 + η

The solution above is called the fundamental solution, but we know that there are other
solutions to (1). Answer the questions below:

1. Propose another (non-fundamental) solution and show that it also satisfies (1).

2. Discuss the economic intuition behind the non-fundamental solutions in which

prices increase over time, even though the money supply is constant. (Tip: you
may want to start with “Suppose everyone expects prices to increase a lot in the
future. Then...”)

48
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

Exercise 3.4

Consider the basic Cagan’s model in which the log of the price level (pt ) and the log of
the money supply (mt ) satisfy the following equation:

mt − pt = −η (pt+1 − pt ) (1)

where η is a constant larger than zero. Assume that mt is bounded.

 T +1
1. Assuming that limT →∞ η
1+η
pt+T +1 = 0, show that:

∞  j
X η mt+j
pt =
j=0
1+η 1+η

Interpret your results.

2. Suppose mt = m̃, for every date t (where m̃ is an exogenous constant). Using the
solution for pt proposed in item 1, answer: What is the inflation rate in this case?
Interpret your results.

3. Suppose now that at date 0 the central bank announces that the money supply
will be equal to m̃ from date 0 to date T − 1, and that at date T the money supply
will increase to m̃′ > m̃ (and then will remain at the new level forever). Using the
solution you found in item 1, show the evolution of the price level as a function of
time in a graph. Interpret the results.

4. Are there solutions to (1) that depend on variables other than the money supply
and the parameter η (i.e., non-fundamental solutions, bubbles)? If your answer is
yes, provide one example and show that it satisfies (1). If your answer is no, show
that there is no bubbles.

49
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

3.2 Solutions

Solution to Exercise 3.1

According to the quantitative equation (the notation is the same used in class):

Mt V = P t y t

Mt V
Pt =
yt

Note that we did not add time subscripts to the velocity of money since it is constant.
Taking log’s we can write:

ln Pt = ln Mt + ln V − ln yt (1)

ln Pt−1 = ln Mt−1 + ln V − ln yt−1 (2)

Subtracting (2) from (1):

ln Pt − ln Pt−1 = ln Mt − ln Mt−1 − (ln yt − ln yt−1 )

Let ∆Xt ≡ Xt − Xt−1 Since ln X − ln Y ≈ (X − Y ) /Y we can write:

∆Pt ∆Mt ∆yt

πt = ≈ −
Pt−1 Mt−1 yt−1

Since ∆Mt
Mt−1
= 0.08 and ∆yt
yt−1
= 0.03 we have that inflation is πt ≈ 0.08 − 0.03 = 0.05. To
get the growth rate of nominal GDP (Yt ) we depart from:

Yt = Pt yt

50
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

lnYt = ln Pt + ln yt

Iterating it backwards and subtracting from the equation above we get:

∆Yt ∆Pt ∆yt ∆yt

≈ + = πt +
Yt−1 Pt−1 yt−1 yt−1

Therefore, the growth rate of nominal GDP is 0.05 + 0.03 = 0.08 (alternatively, one
could use Mt V = Pt yt = Yt to see that the growth rate of nominal GDP is equal
to the growh rate of Mt ). Using the Fisher equation, we get a real interest rate of
r = 0.09 − 0.05 = 0.04.

Solution to Exercise 3.2

Item 1

Let M denote the money supply. We can then write P = M/ (κy ). Therefore, adding
time subscripts and taking the log’s:

ln Pt = ln Mt − ln κ − ln yt

Iterating it backwards and substracting from the equation we get the same relationship
we got in Exercise 3:
∆Pt ∆Mt ∆yt
πt = ≈ − (1)
Pt−1 Mt−1 yt−1

Therefore, since ∆Mt

Mt−1
= 0.12 and ∆yt
yt−1
= 0.04, we get that the inflation rate is 8% per
year.

51
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

Item 2

If the real grew at a higher rate, then inflation would be lower, as (1) makes clear. A
higher growth of real GDP increases the demand for real balances, which puts a positive
pressure on the real money supply, lowering prices (everything else constant).

Item 3

The parameter κ captures how sensitive is the demand for real balances to real GDP. In
κ is very large, any increase in real income implies a large increase in the demand for
money. Writing M κ1 = Py , one can see that 1/κ can be interpreted as the velocity of
money. If the velocity of money is very low, any increase in the amount of transactions
people want to make must be fulfilled by a large increase in the real money supply, since
money does not change hands a lot of times.

Item 4

Taking log’s from the quantitative equation we can write:

ln Pt = ln Mt + ln Vt − ln yt (2)

ln Pt−1 = ln Mt−1 + ln Vt−1 − ln yt−1 (3)

Subtracting 3 from 2:

ln Pt − ln Pt−1 = (ln Mt − ln Mt−1 ) + (ln Vt − ln Vt−1 ) − − (ln yt − ln yt−1 )

Let ∆Xt ≡ Xt − Xt−1 Since ln X − ln Y ≈ (X − Y ) /Y we can write:

∆Pt ∆Mt ∆Vt ∆yt

πt = ≈ + −
Pt−1 Mt−1 Vt−1 yt−1

52
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

Hence, if ∆Vt > 0 because of financial innovation, we should expect the inflation rate to
increase. Intuitively, if the velocity of money is increasing, people demand less money for
their transactions (since the same dollar bill is being used a lot of times for transactions).
For a given nominal money supply, prices will have to increase to guarantee that the
supply of real balances falls, matching the lower demand.

Solution to Exercise 3.3

Item 1

Guess a solution of the form (as seen in class, we know that it is a good guess):

pt = m̃ + b0 g t .

Plugging into (1):

−b0 g t = −η b0 g t+1 − b0 g t

Solving for g :

1+η
g=
η

Thus,
 t
1+η
pt = m̃ + b0
η

is a solution to (1).

Item 2

The idea is for this kind of equilibrium is the following: suppose people expect pt to
rise in the future (b0 > 0). Then, the Fisher equation implies that the nominal interest

53
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

rate will be high, reducing agents incentives to hold money. But then, there is too little
money demand for too much money supply, and prices have to increase to equate the
demand and the supply for real balances. But this confirm agents initial expectations
(a self-fullfiling prophecy).

Solution to Exercise 3.4

Item 1

Solving (1) for pt we get:

1 η
pt = mt + pt+1 (2)
1+η 1+η

And iterating forward we get:

1 η
pt+1 = mt+1 + pt+2 (3)
1+η 1+η

Replacing pt+1 in (2) using the RHS of (3):

 
1 η 1 η
pt = mt + mt+1 + pt+2 (2)
1+η 1+η 1+η 1+η

Which can be rewritten as:

1  j    2
X η mt+j η
pt = + pt+2
j=0
1 + η 1 + η 1 + η

Continuing the iterations, one can see that after T iterations we have:

T  j    T +1
X η mt+j η
pt = + pt+T +1
j=0
1+η 1+η 1+η

54
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

Taking the limit when T → ∞ and using the assumption given, we get:

∞  j  
X η mt+j
pt =
j=0
1+η 1+η

This equation shows that the price level is a weighted average of the present and future
money supply, with lower weights for the money supply of distant periods. (One can
 T +1
verify that the initial assumption that limT →∞ 1+η
η
pt+T +1 = 0 is satisfied in that
solution if mt is bounded).

Item 2

Using the solution found in item 1 we get:

∞  j  
X η m̃ m̃ 1
pt = = η = m̃
j=0
1 + η 1 + η 1 + η 1 − 1+η

In that case, inflation is zero, since the price level is constant. If the money supply is
constant and people do not expect inflation to increase, the price level that makes money
supply equal to money demand is the same in all periods, since nothing changes from
one period to the other. Hence, zero inflation is consistent with such an equilibrium.

Item 3

Using the solution of item 1 we can write:


 j    j  
 T −1−t m̃′
P∞
η m̃ η
if t ≤ T − 1
P

j=0 1+η 1+η
+ j=T −t 1+η 1+η
pt = j 
P  m̃′

 ∞ η
if t > T − 1


j=0 1+η 1+η

Hence, if we are at date t < T − 1 the price is weighted average of money supplies
either equal to m̃ or m̃′ . As we increase t the weight given to money supplies equal

55
Chapter 3. Monetary equilibrium Macroeconomía II – 2022/1

to m̃′ increases and the weight given to the lower money supplies m̃ decreases, until it
becomes zero and the price level becomes m̃′ (which happens first when t = T ). This is
illustrated in the graph below.
pt

m̃0

0 T t

Since agents anticipate an increase in the money supply at date T , they expect
inflation at date T − 1, which reduces their demand for money, leading to an increase in
prices at T − 1 (to guarantee money demand equals money supply). But then, expecting
an increase in prices at T − 1, agents demand less money at T − 2, increasing prices
already at T − 2 and so on. Since the increase in the money supply is expected, prices
increase slowly, until it reaches the new level.

Item 4

Yes. In fact, for any b0 there is a solution

∞  j  t
X η mt+j 1+η
pt = + b0 .
j=0
1+η 1+η η
| {z } | bubble
{z
term
}
fundamental solution

One can verify that it is a solution by plugging that equation in (1) and checking that
it is satisfied (try to do it).

56
Chapter 4

Seignioriage

4.1 Exercises

Exercise 4.1

Consider an economy with a single final good. The seignioriage revenue in real terms in
a given period is given by
∆M
S= ,
P

where ∆M is the increase in the money supply and P is the price of the good. The
demand for money is given by M/P = Ye −αi , where i is the nominal interest rate, Y is
output and α > 0. Using the Fisher equation (and assuming expected inflation is equal
to actual inflation), we can write M/P = Ye −α(r +π) , where r is the real interest rate and
π is the inflation rate. Suppose that in the long run the interest rate and output are
constant (i.e., ∆i = 0 and ∆Y = 0). Find the inflation rate that maximizes S in the
long run.

57
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Exercise 4.2

Consider the following money demand:

Md
= y [1 − (r + π e )]
P

where y is the real output, r is the real interest rate and π e is expected inflation.

1. Derive an expression for the seignioriage revenue assuming y = 250, r = 0.05

and π e = 0.1 (and all those are independent of money growth). Compute the
seignioriage revenue for the following growth rates of the money supply: 25%, 50%
and 75%. What is the relationship between seigniorage and money growth?

2. Assume now that expected inflation is equal to actual inflation (π = π e ), while the
other variables are the same. Compute the seigniorage revenue for the following
growth rates of the money supply: 25%, 50% and 75%. Why are your answers
different from the previous item? Explain.

3. Assuming π = π e , find the rate of monetary expansion that maximizes the

seignioriage revenue.

Exercise 4.3

[Adapted from de Gregorio] Suppose an economy in which agents keep money as currency
and as deposits. The money multiplier is denoted by µ̃. The demand for real balances
in this economy is given by:
L(i, y ) = ay (b − i) ,

where y is real output.

58
Chapter 4. Seignioriage Macroeconomía II – 2022/1

1. Compute the seignioriage if the inflation rate is 10%. What assumptions do you
need to make to be able to compute the seignioriage?

2. Suppose b > r , where r is the real interest rate. Compute the inflation rate that
maximizes the government revenue. What happens with the inflation rate that
you found if the real interest rate increases?

3. Suppose now that in reality the multiplier increases. How does this change your
answer in item 1?

Exercise 4.4

[Adapted from de Gregorio] The money demand is given by:

M
= α − βi + γy .
P

The notation is the usual one.

1. Find the seignioriage (S), assuming π = π e , and discuss how π relates to S. In

order for a revenue maximizing seigniorage to exist, do we need some restriction
on parameters?

2. If it exists, compute the inflation rate that maximizes seignioriage.

Suppose now that in this economies output grows at a rate g .

3. Write the seignioriage as a function of the parameters α, β, γ, the real output y ,

the growth rate g , the inflation rate π and the nominal interest rate i. Use the
Fisher equation.

4. Find the inflation rate that maximizes seignioriage. Hows does it compare to item
2?

59
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Exercise 4.5

[Proposed, adapted from de Gregorio] Consider the following money demand function:

Mt e
= mt = yt e −απt , (1)
Pt

where M is the nominal quantity money, P is the price level, m is the real quantity of
money, y is real output (which is normalized to 1 hereafter), π e is expected inflation
and α is a positive constant.
Suppose the government wishes to finance a real deficit d by increasing the nominal
money supply at a rate of σ. The seignioriage is Ṁt /Pt (you can omit the subscript t).

1. Write the government budget constraint as a function of σ and π e and plot on the
planer (π e , σ). Using equation (1) (take derivatives), determine the steady state
and find the maximum value of d that you can finance in steady state through
seignioriage (denote it by d M ). Suppose d < d M . How many steady states exist?
Show your results in the graph.

2. Suppose expectations are adaptive:

π̇ e = β (π − π e ) . (2)

Explain that equation. Differentiate equation (1) and using (2) to replace replace
the inflation rate, show what is the dynamic of expected inflation in the graph,
and find two steady states. Show which steady state is stable and which one is
unstable (assume that βα < 1).

3. Suppose there is an increase in the deficit from d to d ′ , both being smaller than
d M . Show the dynamics of the adjustment (remember that σ can jump, but π e

60
Chapter 4. Seignioriage Macroeconomía II – 2022/1

adjusts slowly). Finally, suppose that d increases some d ′′ above d M and show
that we get an hyperinflation.

61
Chapter 4. Seignioriage Macroeconomía II – 2022/1

4.2 Solutions

Solution to Exercise 4.1

Multiplying and dividing the right-hand side of S = ∆M

P
by M we get:

∆M M
S= (1)
M P

We know that M = Pye −αi . Since we assumed that Y and i are constant, changes in M
come only from changes in P and therefore, applying differentials:

∆M = ye −αi ∆P

Dividing the right-hand side by M and the left-hand side by PYe −αi (since both are
equal):
∆M ∆P
= =π (2)
M P

Plugging (2) into (1):

M
S= π
P

Using M/P = ye −α(r +π) we get

S = ye −α(r +π) π (3)

Taking the first order condition and solving for π:

dS
= Ye −α(r +π) − αye −α(r +π) π = ye −α(r +π) (1 − απ) = 0 (4)
dπ

1
π=
α

62
Chapter 4. Seignioriage Macroeconomía II – 2022/1

From (3) it is easy to see that π ≤ 0 does not maximize seignioriage, since it yields a
zero or negative S (and any π > 0 yields a positive S). From (4), one can easily see
that dS/dπ > 0 for π ∈ [0, 1/α) and dS/dπ < 0 for π > 1/α. Therefore, the first order
condition characterizes the maximum and the revenue maximizing seignioriage is 1/α.

Solution to Exercise 4.2

Item 1

Multiplying and dividing the right-hand side of S = ∆M

P
by M we get:

∆M M ∆M
S= = y [1 − (r + π e )] (1)
M P M

Now, we are going to compute the seignioriage revenue assuming exogenous values for r ,
π e , y and ∆M/M. We assume y = 250, r = 0.05 and π e = 0.1. Then, the seignioriage
revenue for different ∆M/M is given by the table below:

∆M/M Seigniorage

25% 53.125

50% 106.25

75% 159.375

Therefore, as we increase the money growth, the higher the seignioriage. Notice that it
is implicit here that the increase in the inflation rate is not affecting π e , and that is why
we get this monotonic effect (instead of the usual Laffer curve).

63
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Item 2

Following the same steps we did in Exercise 1 (do it yourself again to make sure you
understand, assuming a SS with constant i and y ), you can show that:

∆M ∆P
= =π
M P

Hence, using π e = π the seignioriage revenue is given by

  
∆M ∆M
S= y 1− r + = πy [1 − (r + π)]
M M

We assume y = 250, r = 0.05. Then, the seignioriage revenue for different ∆M/M is
given by the table below:

∆M/M Seignioriage

25% 43.75

50% 56.25

75% 37.5

Therefore, we get the usual hump shaped thing. That is because now, as inflation
increases, people adjust their inflation expectations, demanding less real balances. Thus,
while the “tax rate” is increasing, the “tax base” is decreasing.

Item 3

We need to maximize:
S = πy [1 − (r + π)]

One can easily see that this function is strictly concave. The FOC is:

dS
= y [1 − (r + π)] − πy = 0
dπ

64
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Hence, the revenue maximizing seignioriage is:

1−r
π∗ =
2

If r increases π ∗ falls.

Solution to Exercise 4.3

Item 1

We have to be very careful here. So far, for simplicity we have implicitly assumed that
the government had the monopoly of creating money. In other words, we have assumed
a money multiplier of 1 (M = monetary base). But when people use deposits, we know
it is not true (banks also create money). Let M = C + D = µ̃H (the notation is the
standard one). Then, we have:
∆Ht
St =
Pt

Note that the the increase in money is ∆Mt = µ̃∆Ht , but only a fraction of this increase
becomes government revenue (just the increase in the monetary base becomes revenue).
Therefore, we have:
∆Mt 1 ∆Mt Mt
St = =
µ̃Pt µ̃ Mt Pt

Using i = r + π e and the money demand function (and dropping the time subscripts):

1 ∆M
S= ay (b − r − π e ) (1)
µ̃ M

But, applying differentials to the money demand M = Pay (b − i) we have (assuming

we are in a SS with y and i constant):

∆M = ay (b − i) ∆P

65
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Dividing it by M = Pay (b − i)

∆M ∆P
= =π
M P

Hence, (1) becomes:

1
S= πay (b − r − π e )
µ̃

Finally, assuming that π = π e we get and π = 10%:

1
S= 0.1ay (b − r − 0.1)
µ̃

To get to this expression we had to assume π = π e . We also make the assumption that
we were in a SS with constant i, π and y , but we could do it assuming a constant output
growth, for instance (try it).

Item 2

Using the results and assumptions from the previous item, we need to maximize:

1
S= πay (b − r − π)
µ̃

This is clearly strictly concave. The FOC is:

dS 1
= {ay (b − r − π) − πay } = 0
dπ µ̃

Which implies:
b−r
π∗ =
2

66
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Item 3

When the money multiplier increases, S decreases, everything else constant. That is
because now, for a given inflation rate, a smaller fraction of the increase in the money
supply is taken by the government (since banks “contribute” to part of that increase in
the money supply).

Solution to Exercise 4.4

Item 1

Following the same steps and making the same assumptions we did in Exercise 2 (do it
again) we get:
M
S =π
P

Hereafter we assume that inflation equals expected inflation. Thus:

S = [α − β (r + π) + γy ] π

If β > 0, S is a concave parabola, and so a revenue maximizing seignioriage obviously

exists. If β < 0 it is a convex convex parabola, thus a maximum will not exist. If β = 0,
it is a linear an increasing function, thus a maximum will not exist either. Therefore we
need β > 0.

Item 2

The first order condition is:

−βπ + [α − β (r + π) + γy ] = 0

67
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Which yields:
1
π∗ = [α − βr + γy ]
2β

Item 3

We will still assume that ∆i = 0, but now ∆y /y = g . Then applying differentials to

M = P [α − βi + γy ]:
∆M = [α − βi + γy ] ∆P + Pγ∆y

Dividing it by M = P [α − βi + γy ]:

∆M ∆P γ∆y ∆y
= + =π+γ
M P α − βi + γy M/P

Hence:

∆M M
S=
M P 
∆y M
= π+γ
M/P P
M
= π + γ∆y
P
∆y
= π (α − βi + γy ) + γ y
y
= π (α − βi + γy ) + γgy

Using the Fisher equation i = r + π e , with π = π e :

S = π [α − β (r + π) + γy ] + γgy

68
Chapter 4. Seignioriage Macroeconomía II – 2022/1

Item 4

It is easy to see that the inflation that maximizes seignioriage is the same, since we have
just summed a constant (γgy ) on the seignioriage expression. The only difference, is
that now for a given inflation rate, the seignioriage revenue will grow in time, since y is
growing.

69
Chapter 5

Goods and money markets

5.1 Exercises

Exercise 5.1

Consider the basic model of equilibrium in the goods market, which has the following
assumptions: (i) all firms produce the same good that can be used for consumption,
investment or by the government; (ii) the economy is closed; (iii) firms are willing to
supply any amount of the good at a given price level.
Consumption (C ) is given by

C = c0 + c1 (Y − T ),

where Y denotes the income, T denotes taxes (minus government transfers) and c0 > 0
and c1 ∈ (0, 1) are parameters. Investment is exogenous and fixed at some level I and
the same applies to government spending, denoted by G .

1. What is the equilibrium level of output Y ?

70
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

2. Only on this item, suppose that there is some regulation that forces the government
to run a balanced budget. That is, G must always be equal to T . Therefore, any
increase in government spending must be compensated by an equal increase in
taxes. Can the government still affect the level of output using fiscal policy (i.e.,
changing G and T , but keeping the budget balanced)? Explain.

3. Now suppose that T depends on Y , according to

T = t0 + t1 Y , with t0 ≥ 0, and t1 ∈ (0, 1).

Compute the equilibrium level of output Y . Compare the government spending

multiplier you got in Item 1 with the one you got here. Which multiplier is larger?

Exercise 5.2

Suppose that when the nominal interest rate i is larger than zero the money demand
(M d ) is given by
M d = $Y (0.25 − i),

where the nominal income $Y is $100. When i = 0, agents will demand at least 0.25×$Y
units of money, but are indifferent between holding any amount of money equal or larger
than 0.25 × $Y . Suppose the supply of money is $20.

1. What is the equilibrium interest rate?

2. What is the equilibrium interest rate when the money supply is $30?

3. What happens to the equilibrium nominal interest rate when the central bank
increases the money supply from $30 to $40?

71
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

4. Suppose the money supply is $25 and the nominal income $Y increases from $100
to $200. Will the equilibrium nominal interest rate increase? Show what happens
using a demand and supply graph.

Exercise 5.3

Consider the model of Chapter 3 in Blanchard, where investment and government are
fixed at I and G , respectively, and consumption is given by

C = c0 + c1 (Y − T )

where Y denotes real output and T are taxes, and c0 > 0 and c1 ∈ (0, 1). Answer true
or false to the statements below, justifying your answers.

1. An increase of 500 units in government expenditure has the same effect on equilib-
rium output as a decrease of 500 units in taxes.

2. An increase of 10% in G always leads to an increase of 1/(1 − c1 )% in output.

3. The higher the marginal propensity to consume, the higher the increase in GDP
after a marginal decrease in taxes.

4. Suppose taxes are a fraction of total income, i.e., T = τ Y . The government

spending multiplier is larger than when T is fixed.

5. Suppose the government runs a balanced budget always, i.e.,T = G . An increase

of one unit in G causes an increase of one unit in output.

6. Suppose there are two group of people in a country (of equal size and each earns
half of the total income). One group has a marginal propensity to consume higher
than others. A policy that taxes the group with the low propensity to consume

72
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

and transfer the amount to the group with the high propensity to consume will
increase output.

Exercise 5.4

[Proposed] Answer the questions below.

1. Consider a bond that promises to pay $100 in one year. What it is the interest
rate on the bond if its price today is $80, $95 and $110?

2. Suppose there are no costs associated with storing currency. Explain why the
nominal interest rate cannot go below zero.

73
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

5.2 Solutions

Solution to Exercise 5.1

Item 1

Equating demand and supply for goods:

Y = c0 + c1 (Y − T ) + G + I

Solving the equation above for Y we get the equilibrium level output:

1 
(1)

Y = c0 + G + I − c1 T
1 − c1

Item 2

Replacing T by G (since T = G ) in (1):

1  
Y = c0 + G + I − c1 G
1 − c1
1  
= c0 + I + G
1 − c1

The equation above clearly shows that the government can still increase the equilibrium
output by increasing G (and consequently T ). It happens, because as (1) shows, the
negative impact of reducing taxes in one unit is smaller than the positive impact of
increasing expenditure in one unit.

74
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

Item 3

Equating demand and supply for goods and solving for Y :

Y = c0 + c1 [Y − (t0 + t1 Y )] + G + I

1  
Y = c0 + G + I − c1 t0
1 − c1 (1 − t1 )

The government spending multiplier here is 1

1−c1 (1−t1 )
, while in item 1 it was 1
1−c1
, and
thus the latter is larger. The intuition for the smaller multiplier with taxes that increase
with output is that part of the increase in income after an increase in government
spending translate into larger taxes, reducing the increase in available income.

Solution to Exercise 5.2

Item 1

Equating demand and supply (and assuming and later veryfing that the equilibrium
interest rate is larger than zero):

100(0.25 − i) = 20

i = 5%

Note that if i = 0, the demand would be 0.25 × 100 = 25 dollars, which is larger than
supply.

75
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

Item 2

Equating demand and supply (and guess and later veryfing that the equilibrium interest
rate is larger than zero):

100(0.25 − i) = 30 ⇒ i = −5%

Our guess failed since we found a negative solution for i. The equilibrium interest is
i = 0%, since when it happens agents will be willing to demand any amount larger
than 0.25 × 100 = 25 dollars, and therefore they may as well demand $30, clearing the
market.

Item 3

Nothing happens. The interest rate remains at zero.

Item 4

Yes, the interest rate will increase. You can easily see drawing the demand and supply
before and after the increase in output, as in the graph below.
i money supply

money demand
12.5% B (M d ) with $Y = 200
money demand
(M d ) with $Y = 100
A
0
25 50 M

76
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

Solution to Exercise 5.3

Item 1

False. Equilibrium output is given by

Y = c0 + c1 (Y − T ) + I + G

1
Y = (c0 − c1 T + I + G ) (*)
1 − c1

Thus an increase in G of 500 increases Y in 1

1−c1
500, while a reduction of 500 in taxes
increases output in c1
1−c1
500. Thus, the decrease in taxes have a lower effect. Intuition:
part of the decrease in taxes is saved, not spent.

Item 2

False. Suppose initially we have Y = 1000 and G = 50 and c1 = 0.5. Using (*), an
increase of 5 units in G (10%), leads to an increase of 1
1−0.5
∗ 5 = 10 in GDP, an increase
of 1% in GDP.

Item 3

True. The marginal increase in GDP after a marginal decrease in taxes is given by

c1
f (c1 ) ≡
1 − c1

Note that

c1 1 c1 + 1 − c1 1
f ′ (c1 ) = 2
+ = 2
= > 0 ∀c1 ∈ (0, 1)
(1 − c1 ) 1 − c1 (1 − c1 ) (1 − c1 )2

77
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

Item 4

False. Now equilibrium output is:

Y = c0 + c1 (Y − τ Y ) + I + G

Y = c0 + c1 (1 − τ )Y + I + G

1
Y = [c0 + I + G ]
1 − c1 (1 − τ )

The multiplier is now 1

1−c1 (1−τ )
, which is lower than 1
1−c1
.

Item 5

True. Using (*), the equilibrium real output is:

1
Y = (c0 − c1 G + I + G )
1 − c1

1
Y = (c0 + (1 − c1 )G + I )
1 − c1
1
Y = (c0 + I ) + G
1 − c1

Item 6

True. Let A denote the group with high propensity to consume and B the group with
low propensity to consume.
Y = CA + CB + G + I

   
Y Y
Y = c0A + c1A − TA + c0B + c1B − TB +G +I
2 2

78
Chapter 5. Goods and money markets Macroeconomía II – 2022/1

cA cB
 
1− 1 − 1 Y = c0A − c1A T A + c0B − c1B T B + G + I
2 2
1  A
c0 − c1A T A + c0B − c1B T B + G + I

Y = c1A c1B
1− 2
− 2

Since c1A > c1B , a increase in T B followed by a decrease of the same magnitude in T A
increases output.

79
Chapter 6

IS-LM

6.1 Exercises

Exercise 6.1

Consider the basic IS-LM model where the central bank fixes the money supply. Suppose
country A is at the liquidity trap, as described in the graph below.

Interest
rate, i

LM

Equilibrium
0

IS

Output, Y

Now consider the following idea:

80
Chapter 6. IS-LM Macroeconomía II – 2022/1

“Imagine that the Fed were to announce that, a year from today, it would pick a
digit from zero to 9 out of a hat. All currency with a serial number ending in that digit
would no longer be legal tender. Suddenly, the expected return to holding currency would
become negative 10 percent.
That move would free the Fed to cut interest rates below zero. People would be
delighted to lend money at negative 3 percent, since losing 3 percent is better than losing
10.” Gregory Mankiw, NY Times, April 18, 2009
(http://www.nytimes.com/2009/04/19/business/economy/19view.html)

Assume that country A initially has notes with 10 different colors and the total value
of the currency of a given color is the same across colors. Based on the idea above,
the central bank implements the following policy. Every year the central bank chooses
randomly one color (with equal probability for each) and all notes with that color lose
its value. Every time the central bank picks the currency color that will have its legal
tender status withdrawn, it immediately puts the same amount of money back in the
economy in notes created with a color not yet used (think that money is thrown from a
helicopter). Thus, the money supply is fixed. People in that country only use currency
to make transactions.

1. How would that policy affect the equilibrium output in this economy? You can
use your intuition to answer it, or you can base it on your answer to item 3 below.

2. Draw the money demand curve of this economy before and after the policy.

3. In the same graph, draw the IS and LM curves before and after the policy. (Obs:
the IS and LM curves before the policy are already drawn in the graph above, but
draw it again to compare with the new equilibrium).

81
Chapter 6. IS-LM Macroeconomía II – 2022/1

Exercise 6.2

In a graph, represents what happens in the IS-LM after each of the shocks below.
Explain how the transition between the old equilibrium and the new equilibrium occurs,
plotting in a graph the path of the nominal interest rate i and output Y . To derive the
transitions, assume that money markets are always in equilibrium, while goods markets
can be not in equilibrium during the transition. Assume that the bank central fixes the
money supply (assumption A1 in class).

1. A permanent increase in G .

2. A permanent increase in M.

Exercise 6.3

Answer the questions below considering the following modifications of the standard IS-
LM model seen in class in which t the central bank fixes the money supply (assumption
A1).

1. Derive the IS curve assuming that the investment does not depend on the interest
rate. What is the impact of monetary and fiscal policy in this case?

2. Derive the LM curve assuming that the money demand does not depend on the
interest rate. What is the impact of monetary and fiscal policy in this case?

3. Derive the LM curve assuming that the demand for real balances does not depend
on income. What is the impact of monetary and fiscal policy in this case?

4. Suppose M d /P = aY , where a > 0. Derive the LM curve when a → ∞.

82
Chapter 6. IS-LM Macroeconomía II – 2022/1

Exercise 6.4

Answer the questions below using the IS-LM model for a closed economy in which the
central bank fixes the money supply.

1. Suppose a country wants to reduce its fiscal deficit, but at the same time, it does
not want to reduce output. Suggest a policy (or a mix of policies) that could
achieve those objectives. Explain it using the IS-LM diagram.

2. Suppose now that a government decides to increase spending (G ), but to keep the
fiscal deficit constant it also increases a lump-sum tax (T ) in the same amount.
Using the IS-LM diagram, explain the effect of these policies.

3. Suppose the central bank increases the money supply in a permanent way at
some date τ . Represent the new equilibrium in the IS-LM diagram, as well as the
transition for the new equilibrium. To show the transition, plot output and the
nominal interest rate as a function of time. When analyzing the transition, assume
that money markets are always in equilibrium, while goods markets adjust slow.

Exercise 6.5

Consider a closed economy in which the central bank fixes the money supply. Short
term economic activity is characterized by the following equations:

C = c0 + c1 (Y − T )

I = I0 − I1 r + I2 Y

G = Ḡ

T = τY

83
Chapter 6. IS-LM Macroeconomía II – 2022/1

Ms M̄
=
P P
Md
= l0 − l1 i
P

where C represents the aggregate consumption function, Y denotes income, I aggregate

investment, G is total public expenditure, T represents the level of taxes (note that we
Ms
are assuming taxes are proportional to income and not lump-sum as usual), P
and
Md
P
denote money supply and demand in real terms, and P represents the price level.
Finally c0 , I0 , I1 , I2 , Ḡ , M̄, l0 are positive, and τ , c1 ∈ (0, 1).

1. Define the IS curve. Determine its functional form at this economy.

2. Obtain the government spending multiplier and explain intuitively it represents.

3. Define the LM curve and determine its functional form.

4. Find an expression for the equilibrium output and nominal interest rate in this
economy, indicating step by step how you obtained it.

5. Using derivatives, quantify the total effect on income of an increase in money

supply. Give the intuition of your results.

6. Why do we generally use without distinction i or r in this model?

Exercise 6.6

[Adapted from de Gregorio] Consider an economy in which the central bank fixes the
money supply described in the following set of equations:

C = 1 + 0.8(1 − t)Y

I = 2 − 0.4i

84
Chapter 6. IS-LM Macroeconomía II – 2022/1

G =1

t = 0.2

Ms = 6

M d = 0.75Y − 1.5i

The notation is the standard one. Note that income taxes are proportional to income,
instead of lump sum.

1. Find Y and i in equilibrium.

2. Assume that autonomous investment decreases by 10%. What level of government

spending restores previous income level?

3. Assume now that private investment is taxed proportionally to the investment

rate. In particular, assume that:

I = 2 − 0.4(1 + τ )i (1)

where τ = 0.25. Compute new equilibrium. Compare your results with those at
item 1.

4. Assuming the new investment function, what level of money supply restores income
to the level of item 1? Explain your results.

Exercise 6.7

[Adapted from de Gregorio] Consider a closed economy in which the central bank fixes

85
Chapter 6. IS-LM Macroeconomía II – 2022/1

the money supply characterized by the equations below:

Y =C +I +G

C = C̄ + c1 (Y − T )

I = Ī − αr

M d = kY − θi

i = r + πe

where Y represent income, C consumption, I denotes investment, G denotes government

expenditure, T represents taxes, i is nominal interest rate, r is real interest rate, π e
denotes expected inflation and M d is the real money demand (note that this implicitly
assumes a price level of 1). Finally c1 , α, Ī , k, θ are positive parameters. Denote the
money supply my M s .

1. Assume initially that expected inflation equals π0 . Obtain equilibrium equations

in money and goods markets. Graph each of them in the plane (r , Y ).

2. Obtain expressions for equilibrium income Y , and real and nominal interest rates.

3. A change in economic expectations triggers a drop of expected inflation to π1 < π0 .

Analyze how this change affects the economy.

4. What monetary policy should be implemented to stabilize income?

5. Assuming C = 5, I = 1, T = 2, G = 2, M s = 3, c = 0.8, α = 0.8, k = 0.25, θ =

0.75, π0 = 3, π1 = −5, obtain equilibrium variables before and after expectations
adjust.

86
Chapter 6. IS-LM Macroeconomía II – 2022/1

6. Given numerical results in the previous answer, which constraint restricts the
central bank when stabilizing income? How can the fiscal authority help to
overcome it?

Exercise 6.8

Answer the questions below using the IS-LM model for a closed economy in which the
central bank fixes the nominal interest rate.

1. Suppose (only on this item) a country in which the congress passed a law that
forces the government to run a balanced budget (that is, total taxes T must
always be equal to government spending G ). Is fiscal policy still effective in
increasing output in this economy? Explain representing the effect of an increase
in government spending in the IS-LM diagram.

2. Assume now (only on this item) that consumption in this economy is a linear
function of available income, with a marginal propensity to consume equal to 0.9.
Suppose this country has been running large fiscal deficits for many years. Some
economists argue that running a balanced budget is not enough, and they suggest
the government should commit to a policy in which government spending should
be always equal to 80% of total taxes, so that over the years the total debt of the
government will fall. In other words, they are suggesting that the government
should be forced to always set G /T = 0.8 (where G is government spending and T
represents taxes). Suppose this policy is implemented and the government has to
follow it. What happens in equilibrium when the government decides to increase
spending? Explain using the IS-LM diagram.

3. Now consider (only on this item) an extension of the IS-LM model in which firms
do not borrow at the policy rate i fixed by the central bank. Instead, assume

87
Chapter 6. IS-LM Macroeconomía II – 2022/1

that whenever firms borrow to invest, they borrow at the policy rate plus a risk
premium (that is, they borrow at a rate i + x, where x is an exogenous risk
premium). During the 2008 crisis, the risk premium increased, increasing the cost
of borrowing. Using a standard IS-LM diagram, explain the effect of an increase
in the risk premium.

4. Suppose the central bank set a nominal interest rate of i > 0. Suppose there is
an exogenous shock in the demand for money, that shifts to the right the money
demand (that is, people are demanding more money after the shock). How should
the money supply react to guarantee the interest rate will not change? Explain
using a graph of money demand and money supply.

Exercise 6.9

[Proposed] The government of a country, after years running large fiscal deficits, is
forced to reduce its deficit in a given year. Increasing taxes takes a lot of time and it is
not an available option in the short run. Using the IS-LM model where the central bank
fixes the nominal interest rate, explain how this country could avoid a fall in output in
the short run.

Exercise 6.10

[Proposed] Take two countries that are identical except for the behavior of the central
bank: in country A the central bank fixes the nominal interest rate (and adjusts the
money supply to achieve it), while in country B the central bank fixes the money supply.
Initially, both countries are at an equilibrium in which the interest rate and output are
the same. Suppose both countries experience the same reduction in goverment spending.
Which country will suffer a larger output loss? Explain your answer using the IS-LM

88
Chapter 6. IS-LM Macroeconomía II – 2022/1

model.

89
Chapter 6. IS-LM Macroeconomía II – 2022/1

6.2 Solutions

Solution to Exercise 6.1

Item 1

As shown in item 3 below, output increases.

Item 2
Nominal interest rate, i

Money demand before the policy

0
Real money balances, M/P

Money demand after the policy

Because of the policy, bonds become more attractive and money less attractive. The
money demand shifts down. Moreover, when the interest rate is 0%, a decrease in
interest rates will still cause the demand for money to increase (now, when the nominal
interest rate is 0%, bonds still pay a interest rate higher than money, and therefore a

90
Chapter 6. IS-LM Macroeconomía II – 2022/1

reduction in i will make people substitute bonds for money). Nominal interest rates can
go negative after the policy, but they certainly can’t go below -10% (otherwise no one
would demand bonds).

Item 3

The IS curve does not shift after the policy. Consumption is still a function of available
income, and that did not change. For a given level of Y and i, firms will be willing to
invest the same as before. G did not change as well. Thus, the only thing that shifts is
the LM curve. Take some interest rate i > 0. For that interest rate, the level of output
Y consistent with equilibrium in the money market will have to be higher after the
policy to make people demand a given quantity of money (to see that, go back to item
2 and draw the equilibrium in the money market for different values of Y before and
after the policy, assuming a constant money supply). Obs: one could always say that
the IS shifts for reasons that are not in the model. Here, you are asked to analyze this
policy through the lens of the basic IS-LM model.
In the new equilibrium output will be higher. The economy can leave the liquidity
trap, or enter a new liquidity trap with negative interest rates (depending on how much
the LM curve shifts). The two cases are represented below, but in both cases, output
increases (note where the origin is positioned).
Interest Interest
rate, i rate, i

LM(before policy) LM (before policy)

LM (after policy)
Equilibrium LM(after policy) Equilibrium
(before policy) (before policy)

0 0
Equilibrium
Equilibrium (after policy)
IS (after policy) IS

Output, Y Output, Y

91
Chapter 6. IS-LM Macroeconomía II – 2022/1

Solution to Exercise 6.2

Item 1

Let’s start with the new equilibrium. The shock in G shifts the IS curve to the right,
and the equilibrium goes from point A to point A′ in the graph below.
i
LM

A0
A ∆G > 0 IS 0

IS
Y

Now we look at the transition. Since we are assuming that money market is always
in equilibrium during the transition, we must never leave the LM curve (remember that
the LM curve gives us all combinations of i and Y such that the demand for money
equals the supply of money). Hence, we must move on the LM curve, as represented
below.
i i
LM

time
0
A
Y
A ∆G > 0 IS 0

IS
time

92
Chapter 6. IS-LM Macroeconomía II – 2022/1

In words what happens is the following. The increase in government spending

increases demand, which increases firms production and income. The higher income
make people demand more money. Since the money supply fixed, as income increases,
interest rates increase to ensure equilibrium in money markets, until production fully
adjusts and we reach the new equilibrium.

Item 2

Again we start with new equilibrium. The increase in M shifts the LM curve to the
right and the equilibrium goes from point A to A′ .
i
LM

LM 0
∆M > 0

A
A0
IS
Y

Again, we must always be on the LM curve. Hence, the representation in the graph
below. Initially, the economy jumps to point B and then slowly converges to A′ . Interest
rate fall at point B, which makes output increase gradually, since the much lower interest
rate increases demand for goods. As output increases, the demand for money increases,
which forces interest to go up a bit to reduce the demand for money.

93
Chapter 6. IS-LM Macroeconomía II – 2022/1

i i
LM

LM 0
time
∆M > 0
Y
A
A0
B IS
time

Solution to Exercise 6.3

Item 1

If investment does not depend on the interest rate, the IS curve is vertical. To see that,
notice that equilibrium in the goods markets is not affected by an increase in the interest
rate (the demand curve Z of the model of equilibrium in the goods markets seen in class
does not change when we change i). If the IS curve is vertical, fiscal policy still affects
output and interest rates, but monetary policy only affects interest rates, as illustrated
below.
i IS IS 0 i IS
LM LM

A0
LM 0
∆M > 0
A A
∆G > 0
A0

Y Y

94
Chapter 6. IS-LM Macroeconomía II – 2022/1

Item 2

If the money demand does not depend on i (but still depends on Y ), the LM curve is
vertical. This is so because, for a given money supply, there is an unique Y that clears
the money markets. But notice that when M s /P increases, the level of Y that clears
the money market increases (since now people need to demand more money), and hence
the LM curve shifts to the right. Hence, fiscal policy is not effective to increase output,
but it increases interest rates. Expansive monetary policy, on the other hand, increases
output and lowers interest rates.
i LM i LM LM 0
A0

A A

∆G > 0
A0
0 ∆M > 0
IS IS IS
Y Y

Item 3

If the demand for real balances does not depend on Y then, the interest rate that clears
money markets is the same for any Y . Hence, we get an horizontal LM curve. When
M s /P increases, the interest rate that clears money markets decreases, shifting LM
down. Expansive monetary leads to lower i and higher Y , while expansive fiscal policy
increases output without raising i.

95
Chapter 6. IS-LM Macroeconomía II – 2022/1

i i

LM LM
A A0 A ∆M > 0

LM 0
A0
∆G > 0

IS IS 0 IS
Y Y

Item 4

Since the LM curve does not depend we know that the LM is vertical at some Y ∗ such
that:

M 1M
= aY ∗ ⇒ Y ∗ =
P aP

Hence, as a → ∞, Y ∗ → 0, implying the LM curve becomes vertical at the origin.

Intuitively, since the LM curve does not depend on i, the income is entirely determined
by the equilibrium in the money markets. As the demand for money explodes for a
given level output, the only way to reach an equilibrium in money markets is to have an
extremely low income.

Solution to Exercise 6.4

Item 1

There are (at least) two ways to do it. First, the country could reduce government
spending (or increase taxes) and then expand the money supply. The IS would shift to
the left and the LM to the right. By choosing the right amount of monetary expansion
one can ensure output will not change. This is represented in the graph below.

96
Chapter 6. IS-LM Macroeconomía II – 2022/1

i
LM

A
LM 0

A0 IS

IS 0

Another way to achieve that is to increase both taxes and government spending.
Remember that equilibrium in the goods markets (the IS curve) requires (I am assuming
a linear consumption function just to simplify, this is not needed for the argument, see
the next item)
1
Y = [c0 − c1 T + G + I (i, Y )]
1 − c1

Hence, if taxes increases in ∆T > 0, spending must increase only increase only
∆G = c1 ∆T to keep the IS relation identical. Hence, since c1 ∈ (0, 1), ∆T > ∆G and
therefore the fiscal deficit improves, without changing the IS curve.

Item 2

The IS relation is given by:

Y = C (Y − T ) + G + I (i, Y )

Therefore an equal and infinitesimal increase in T and G changes the RHS in

dRHS = −C ′ (Y − T ) + 1. Since the marginal propensity to consume C ′ (Y − T ) is
smaller than one, dRHS > 0 and therefore the IS shifts to the right.

97
Chapter 6. IS-LM Macroeconomía II – 2022/1

i
LM

A0

IS 0

IS

Item 3

We start with new equilibrium. The increase in M shifts the LM curve to the right and
the equilibrium goes from point A to A′ .
i
LM

LM 0
∆M > 0

A
A0
IS
Y

Given that money markets are always in equilibrium, we must always be on the LM
curve. Hence, the representation in the graph below. Initially, the economy jumps to
point B and then slowly converges to A′ . Interest rate fall at point B, which makes
output increase gradually, since the much lower interest rate increases demand for goods.
As output increases, the demand for money increases, which forces interest to go up a
bit to reduce the demand for money.

98
Chapter 6. IS-LM Macroeconomía II – 2022/1

i i
LM

LM 0
time
∆M > 0
Y
A
A0
B IS
time

Solution to Exercise 6.5

Item 1

The IS curve represents those pairs of interest rate and output that are consistent
with equilibrium in the goods market. To determine this curve we depart from market
equilibrium and then we solve for Y or r :

Y = c0 + c1 (Y − τ Y ) + I0 − I1 r + I2 Y + Ḡ

Y (1 − c1 (1 − τ ) − I2 ) = c0 + I0 − I1 r + Ḡ

c0 + I0 − I1 r + Ḡ
Y =
1 − c1 (1 − τ ) − I2

Using the Fisher equation we have that r = i − π e and therefore the IS curve becomes:

c0 + I0 − I1 (i − π e ) + Ḡ
Y =
1 − c1 (1 − τ ) − I2

Item 2

In the model of equilibrium in the goods markets, we took the interest rate as given. In
general, in the IS-LM model where the central bank fixes the money supply (like the

99
Chapter 6. IS-LM Macroeconomía II – 2022/1

one presented in this exercise), when we increase G we also change the interest rate.
But here, since we assumed the money demand does not depend on output, the LM
curve will be flat (as will be clearer when we solve item 3). Hence, interest rates will
not move after we change G . Therefore, at this stage, I will simply compute the change
in output after a change in G , assuming the interest rate does not move. Hence, we get:

dY 1
=
dG 1 − c1 (1 − τ ) − I2

If c1 (1 − τ ) + I2 > 1 then an increase in government spending increases output (again,

assuming the interest rate would not move). The increase government spending expands
aggregate demand, and firms respond by increasing production, which increases income
and further increases demand (which further increases production, income and so on).
(Remark: If c1 (1 − τ ) + I2 < 1 this model makes little sense, because for a real interest
rate of zero, there would be no positive level of output consistent with equilibrium in
the goods markets. Hence, we will keep that parametric assumption in the rest of the
solution).

Item 3

The LM curve represents all the pairs of i and Y that are consistent with equilibrium in
the money (financial) markets. To determine it we impose equilibrium in the money
market (M d = M s ) and solve for i (Y ) as a function of Y (i):

M
= l0 − l1 i
P

100
Chapter 6. IS-LM Macroeconomía II – 2022/1

Since in this exercise the money demand is independent from income (Y ), the LM curve
will be horizontal on the plane (Y , i):

M
l0 − P
i=
l1

Item 4

Now we will look for the pair (i, Y ) that equilibrates both, money and goods markets at
the same time, by intersecting both curves IS and LM curves:

M
l0 −
i∗ = P
l1
 
l0 − M e
c0 + I0 − I1 l1
P
−π + Ḡ
∗
Y =
1 − c1 (1 − τ ) − I2

Item 5

To answer this question, we take the derivative of the equilibrium Y ∗ we found on the
last item with respect to M:

dY ∗ I1 /(l1 P)
=
dM 1 − c1 (1 − τ ) − I2

Assuming 1 − c1 (1 + τ ) − I2 ) > 0 this is positive, and therefore an expansionary monetary

policy reduces interest rate promoting investment and short term growth.

Item 6

The IS-LM model analyzes the short term: the time horizon in which prices are fixed.
Hence, we can completely abstract from inflation, and for simplicity we usually set
π e = 0, which implies i = r by the Fisher equation.

101
Chapter 6. IS-LM Macroeconomía II – 2022/1

Solution to Exercise 6.6

Item 1

From the equilibrium conditions at goods and money markets we get:

Y = 4 + 0, 64Y − 0, 4i

6 = 0, 75Y − 1, 5i

Hence, Y = 10 and i = 1.

Item 2

Using Y = 10 in goods market equilibrium we can solve for G :

10 = 1 + 0, 64 × 10 + 1, 8 − 0, 4 × 1 + G

We then get that G should get to 1.2 increasing by 20% from initial level.

Item 3

As in item 1 we have now:

Y = 4 + 0, 64Y − 0, 5i

6 = 0, 75 − 1, 5i

And then we get Y ≈ 9, 8 and i ≈ 0, 92. This new tax increases the sensitivity of
investment to interest rate. At the previous interest rate investment and output, which
pushes interest rates down, since there is too little demand for money.

102
Chapter 6. IS-LM Macroeconomía II – 2022/1

Item 4

The exercise consists of solving the system for M s imposing equilibrium values from
question 1:
10 = 4 + 0, 64 × 10 − 0, 5i

M = 0, 75 × 10 − 1, 5i

We obtain M = 6.3. The money supply expands to decrease interest rates, compensating
the negative effect of the tax on investment.

Solution to Exercise 6.7

Item 1

Equilibrium in the goods markets imply:

Y = C̄ + c1 (Y − T ) + Ī − αr + G

Which gives the IS curve below (in term of the real interest rate instead of nominal):

1 
r= C̄ + (1 − c1 ) Y − c1 T + Ī + G
α

Equilibrium in the money market implies:

kY − θ (r + π e ) = M s

Which implies the following LM curve in term of the real interest rate:

1
kY − M S − π e


r=
θ

103
Chapter 6. IS-LM Macroeconomía II – 2022/1

Hence, we get the usual graph:

r
LM

IS

Item 2

Equilibrium income is given by:

1  1
kY − M S − π e


C̄ + (1 − c1 ) Y − c1 T + Ī + G =
α θ
−1 
Ms
 
∗ k 1 − c1 e 1 
Y = − +π + C̄ − c1 T + Ī + G
θ α θ α

To obtain equilibrium r we simply plug Y ∗ that into the LM curve:

1
kY ∗ − M S − π e


r=
θ

Item 3

This shifts the LM curve to the left (as if there was a negative shock in the demand
for money). The equilibrium goes from A to A′ . Output falls and the real interest rate
increases.

104
Chapter 6. IS-LM Macroeconomía II – 2022/1

r LM 0 LM

A0

IS

Item 4

By increasing the money supply, monetary policy can offset the initial shift to the left
of the LM curve, keeping output constant and the economy at the same equilibrium.

Item 5

Equilibrium values before the change in expectations are Y = 25.7, r = 1.6, i = 4.6.
After inflation expectations fall, new equilibrium values are Y = 12, r = 5, i = 0.

Item 6

Nominal interest rates are at its minimum possible level (they cannot be negative).
Therefore expansionary monetary policy loses effectiveness, since interest rates cannot
fall. Expansionary fiscal policy is still effective. An increase in G displaces the IS curve
to the right stabilizing the output after a negative shock.

Solution to Exercise 6.8

105
Chapter 6. IS-LM Macroeconomía II – 2022/1

Item 1

The equilibrium condition in the goods markets becomes (the notation is the standard
one):
Y = C (Y − G ) + G + I (i, Y )

Note that we replaced T inside the argument of the consumption function by G , because
of the assumption of a balanced budget (alternatively, you could also replace G by T ).
Since the marginal propensity to consume is smaller than one, an increase in G leads to
an increase in the RHS of the equation above (note that dRHS/dG = 1 − C ′ (Y − G ) > 0),
which implies that for a given interest rate the output Y consistent with equilibrium
in the goods markets increases. Therefore, the IS curve still shifts to the right when
G increases (Remark: here, you could also have used a linear consumption function to
make the point that IS shifts to the right when G increases).
The equilibrium before an increase in G is represented by point A and the equilibrium
after the increase in G is represented by point A′ in the graph below. Fiscal policy is
still effective and an increase in G increases equilibrium output.
i

A0
i LM
A
IS 0

IS
Y

106
Chapter 6. IS-LM Macroeconomía II – 2022/1

Item 2

The equilibrium condition is (the notation is the standard one):

Y = c0 + 0.9 (Y − T ) + G + I (i, Y )

Using that T = G /0.8 we get (alternatively, you could replace G by 0.8 × T ):

 
G
Y = c0 + 0.9 Y − + G + I (i, Y )
0.8

Note that dRHS

dG
0.9
= − 0.8 + 1 < 0 on the equation above. Hence, when G increases, the
output consistent with equilibrium in the goods markets must decrease. Therefore, under
the fiscal rule proposed, an increase in G leads to a fall in output (for a given interest
rate), and therefore the IS shifts to the left. Hence, after G increases, equilibrium output
falls, as represented by point A′ in the graph below.
i

A
i 0
LM
A
IS

IS 0
Y

Item 3

The IS relation now is given by (the notation is the standard one):

Y = C (Y − G ) + G + I (i + x, Y )

107
Chapter 6. IS-LM Macroeconomía II – 2022/1

Since investment is a decreasing of the borrowing rate i + x, an increase in x shifts the

IS curve to the left. This decreases equilibrium output, as represented by point A′ in
the graph below.
i

A
i LM
A0
IS

IS 0
Y

(Alternatively, one can make a graph with i + x on the vertical axis. In that case,
the IS curve remains unchanged but the LM curve shifts upwards, leading, of course, to
a reduction in equilibrium output as well.)

Item 4

The money supply should increase, to guarantee the interest rate will not increase, as
illustrated in the graph below.
0
i MS MS

A0
i
A
0
MD

MD
Y

108
Chapter 7

Mundell-Fleming and exchange rates

7.1 Exercises

Exercise 7.1
1+rt∗ e
1. Show that qt = q
1+rt t+1
(the notation is the standard one).

2. Express the Marshall-Lerner condition in terms of elasticities. You may assume

that the country has a zero trade deficit (X = qM).

Exercise 7.2

Consider an open economy characterized by the following equations:

C = c0 + c1 (Y − T )

I = d0 + d1 Y

IM = m1 Y

109
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

X = x1 Y ∗

where, C denotes consumption, Y denotes the domestic income, I denotes investment,

IM and X are the quantity of imports and exports, respectively. The parameters m1
and x1 are the propensities to import and export. Assume that the real exchange rate
is fixed at a value of 1 and treat foreign income, Y ∗ , as fixed. Also assume that taxes,
T , are fixed and that government purchases, G , are exogenous (i.e., decided by the
government). Note that differently from the standard Mundell-Fleming model seen in
class, here we assume that investment does not depend on the interest rate.

1. Write the equilibrium condition in the market for domestic goods and solve for Y .

2. What is the government spending multiplier? (Assume that 0 < m1 < c1 + d1 < 1)

3. What is the change in net exports when government purchases increase by one
unit?

4. Suppose the government decides to close the economy, not allowing trade with the
rest of the world (i.e., IM = X = 0). Find the government spending multiplier. Is
it larger or smaller than the multiplier you found in item 2? Explain, in words,
why the two multipliers are different.

Exercise 7.3

Suppose a country operates under a fixed exchange rate regime and has perfect capital
mobility. At date t, the central bank spends 100 units of local currency buying domestic
bonds in an open market operation. Suppose that between dates t + 1 and t financial
investors had time to fully react to the open market operation.

1. Is the monetary base at t + 1 larger, equal or smaller than the monetary base at
date t?

110
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

2. Did the central bank balance sheet change between dates t + 1 and t? If yes, how
did it change?

Exercise 7.4

To answer the questions below use the Mundell-Fleming model of an open economy.

1. Suppose a country has perfect capital mobility and operates under a flexible ex-
change rate regime. Suppose the country is in a severe recession, and policymakers
are looking for a policy (or a mix of policies) to increase output. Moreover, they
would like to ensure net exports will not be affected by those policies. Changing
the exchange regime from flexible to fixed is not a posssibility. How could this
country increase output without affecting net exports? Explain using graphs and
words.

2. Suppose now a country that has perfect capital mobility and operates under a
fixed exchange rate regime. Suppose there is a sudden exogenous increase in the
demand for exports (say that the income in the rest of the world increased). What
will be the effect of this higher demand for exports on the amount of foreign
currency reserves of the central bank? Explain using graphs and words.

Exercise 7.5

To answer the questions below use the Mundell-Fleming model of an open economy.
You must always assume that Marshall-Lerner condition holds.

1. Consider a small open economy that has perfect capital mobility and operates under
a flexible exchange rate regime. Suppose the government increases government
spending. How does this shock affects equilibrium output, nominal exchange rates
and net exports? Explain using graphs and words.

111
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

2. Suppose the Federal Reserve of the United States decides to increase the US
nominal interest rate. An economist argues that this could trigger a recession in
Chile. Is this prediction consistent with the Mundell-Fleming model? Explain
using graphs and words. In your answer, consider that Chile is a small open
economy that has perfect capital mobility and operates under a flexible exchange
rate regime.

3. Suppose now Chile operates under a fixed exchange rate regime, with perfect
capital mobility. The domestic currency is the Chilean peso (CLP) and the foreign
currency is the US dollar (USD). Moreover:

• The exchange rate is fixed at one (1 CLP costs 1 USD);

• Initially the monetary base in Chile is 100 CLP;

• Initially the Chilean central bank holds 50 USD of (dollars) reserves (those
are foreign exchange reserves held by the Central Bank of Chile).

Explain why the Central Bank of Chile has no control of monetary policy under
this fixed exchange rate regime. You must do it by explaining what happens after
the Central Bank of Chile decides to buy 10 CLP worth of domestic bonds in an
open market operation (say, in an attempt to increase the monetary base). In
your answer you must indicate what is the level of the monetary base and the
dollar reserves of Central Bank of Chile after the open market operation and after
all adjustments have taken place (you should provide the exact number of those
variables).

4. Consider a small open economy that has perfect capital mobility and operates
under a fixed exchange rate regime. Suppose the government increases government
spending. How does this shock affects equilibrium output? Explain using graphs

112
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

and words.

Exercise 7.6

[Proposed] Answer the following questions according to the Mundell-Fleming model with
perfect capital mobility and flexible exchange rates seen in class (de Gregorio, chapter
20).

1. How does an increase in government spending affects output, interest rates and
the exchange rate? Explain using graphs and words.

2. How does an increase in the money supply affects output, interest rates and the
exchange rate? Explain using graphs and words.

3. How does an increase in the international interest rate affects output interest rates
and the exchange rate? Explain using graphs and words.

4. Answer items 1, 2 and 3 again assuming a fixed exchange rate regime instead of
flexible exchange rate regime.

5. Still assuming a fixed exchange rate regime and perfect capital mobility, what
happens when the central bank increases the fixed exchange rate? Explain using
graphs and words.

Exercise 7.7

[Proposed] Suppose that at date t you can buy or sell USD in future markets at a price
ft+1 . In other words, agents agree at t that they will buy/sell dollars at t + 1 at the
agreed price ft+1 .

1. Suppose all agents in the economy are risk neutral. Show that 1 + it = (1 + i ∗ ) ft+1
et
.

2. Do you expect that equation to hold in practice?

113
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

Exercise 7.8

[Proposed] Suppose inflation in the US is always 3%, while inflation in Chile is always
2%. Suppose also that the nominal exchange rate between CLP and USD is fixed. What
do you expect to happen to the nominal interest rate in Chile?

114
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

7.2 Solutions

Solution to Exercise 7.2

Item 1

Equilibrium in the goods markets is given by

Y = c0 + c1 (Y − T ) + d0 + d1 Y + G + x1 Y ∗ − m1 Y

1
Y = [c0 − c1 T + x1 Y ∗ + d0 + G ]
1 − c1 − d1 + m1

Item 2

The government spending multiplier is

1
1 − c1 − d1 + m1

Item 3

Let NX = X − IM = x1 Y ∗ − m1 Y (remember that the real exchange rate is one). The,

since output increases in 1
1−c1 −d1 +m1
when government spending increases in one unit,
the change in NX is  
 1 
∆NX = −m1 
 
|{z}  1 − c1 − d1 + m1 

dNX /dY | {z }
dY /dG

115
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

Item 4

The equilibrium condition becomes

Y = c0 + c1 (Y − T ) + d0 + d1 Y + G

And therefore
1
Y = [c0 − c1 T + d0 + G ]
1 − c1 − d1

And then the multiplier is 1

1−c1 −d1
. Since m0 > 0, it is larger than the multiplier in item
2. The reason is that in the open economy part of the increase in demand caused by
the increase in G (and the subsequent increases in income) do not become demand for
domestic goods, but demand for foreign goods.

Solution to Exercise 7.3

Item 1

The monetary base is the same. When the central spends $100 buying bonds, it initially
increases the monetary base, putting pressure to lower domestic interest rates. Hence
domestic investors start to exchange domestic currency for foreign currency to invest
abroad. This forces the central bank to sell its reserves, decreasing the monetary base
back again. He keeps selling reserves until the pressure on interest rates are totally
offset, which happens when the monetary base return to its previous level. Since all
this adjustment is quick and happens between t and t + 1, the monetary base does not
change between those dates.

Item 2

It changed. Now the central bank has less reserves and more bonds.

116
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

Solution to Exercise 7.4

Item 1

The idea here is the following: you do expansive monetary policy and then you cancel
the change in net exports with fiscal policy (since with perfect capital mobility and
flexible exchange rate ∆G = −∆XN). Below we explain it in more detail.
Let’s start looking first at the effect of increasing the money supply alone. An
increase in M shifts the LM to the right. Since interest rates must be equal i ∗ , domestic
residents have too much money and they start to buy foreign bonds. This leads the
exchange rate to depreciate, shifting the IS curve to the right (IS ′ ). The economy would
move from point A to point B. At point B, net exports may be have changed, since e
increased (which causes an increase in net exports) and Y increased (which causes a fall
in net exports). Denote by ∆B the impact on net exports of the increase in the money
supply (it does not matter if it is positive or negative).

i e LM ∗ LM ∗
0

LM IS ∗
0
IS ∗
LM 0 B

A A0
∗
i B, A0 IS 00
A

IS 0

IS

Y Y

Now consider we are at point B and let’s see what happens when the government does
fiscal policy. Say the government increases G (which would shift IS ′ to IS ′′ ). The higher
demand for goods, increases the demand for money. As residents sell their domestic

117
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

bonds, foreign capital enters, appreciating the exchange rate, pushing the IS curve back
to IS ′ . The exchanges rate keeps appreciating, until the change in demand is offset,
that is, ∆G = −∆XN. Hence, by changing government spending in ∆G = ∆B the
government ensures net exports will not change.

Item 2

The effects are identical to the ones of an increase in G . This higher demand for exports
will shift the IS and IS ∗ , as illustrated below. The initial increase in demand, makes
domestic residents demand more money. They sell their bonds to foreign investors, who
exchange foreign currency by domestic currency. Since exchange rates are fixed, this
forces the central bank to buy reserves, which increases the money supply, shifting the
LM curve as well. This is illustrated in the graph below. In the end, the central bank
will have more reserves.

i e LM ∗ LM ∗
0

LM IS ∗

0
LM 0 IS ∗

A e
i∗ A0 A A0

IS 0

IS

Y Y

Solution to Exercise 7.5

118
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

Item 1

The increase in government shifts IS ∗ to the right and in principle would shift the IS
curve to the right (dashed line below). However, as the demand for goods increase, we
have Money Demand > Money Supply, and domestic residents sell domestic bonds to
foreigners, which in turn demand the domestic currency, leading to an appreciation of the
domestic currency. This appreciation shifts the IS curve back. In the new equilibrium,
output remains unchanged, the exchange rate falls and net exports decrease (in fact,
∆G = −∆XN). The new equilibrium is represented by point A′ in the graph below.

i e LM ∗
LM IS ∗

0
IS ∗

i∗ A and A0 A

IS 0 A0

IS

Y Y

Item 2

The increase in the foreign interest rate i ∗ is represented below (i2∗ is the interest rate
after the increase).

119
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

i e LM ∗ LM ∗
0
0
LM IS ∗

A0 IS ∗
i∗2 A0
A
i∗1

IS 0 A

IS

Y Y

The increase in i ∗ increases domestics interest rates, and shifts the IS ∗ curve to the
left (for a given e, there is less demand for goods), triggering an depreciation of the
Chilean currency. This depreciation shifts the IS curve to the right, increasing output.
Putting it differently (and at the risk of being repetitive), when i ∗ increases, there
is less demand for goods for a given e (shifting IS ∗ to the left). The increase in i ∗
also reduces the demand for money, which leads to Money Demand < Money Supply.
As domestic residents have excess money, they buy bonds abroad, which make them
demand foreign currency. This depreciates the domestic currency, shifting IS to the
right.
Note that this is not consistent with the prediction that this shock will create a
recession, equilibrium output actually in Chile increases according to this model.

Item 3

After, the central bank spends 10 CLP buying bonds the story goes like this:

• The money supply in principle would increase, and since the interest rate is fixed
at the international level, Money Supply > Money Demand.

• Since domestic residents have excess money, they start trying to buy bonds. As

120
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

all domestic residents are trying to buy bonds, they have to buy foreign bonds.

• To buy foreign bonds, they exchange CLP for USD, and therefore the central bank
needs to sell the dollars the residents are demanding.

• As the central bank sell reserves to meet this demand for dollars, the money supply
starts to fall. This process continues until Money Supply = Money Demand again.

This why the central bank has no control of monetary policy (the monetary base remains
unchanged).
In the end, the central bank will be forced to undo the initial attempted increase of
10 CLP in the monetary base by selling 10 CLP worth of dollars reserves. Since the
exchange is one, it implies that he has to sell 10 USD of reserves. Hence, after all, the
monetary base remains at 100 CLP, but the dollar reserves of central bank go down to
40 USD.

Item 4

This increase in G will shift the IS and IS ∗ , as illustrated below. The initial increase
in demand, makes domestic residents demand more money. They sell their bonds to
foreign investors, who exchange foreign currency by domestic currency. Since exchange
rates are fixed, this forces the central bank to buy reserves, which increases the money
supply, shifting the LM curve as well. This is illustrated in the graph below. In the end,
and equilibrium output increases (the equilibrium goes from A to A′ ).

121
Chapter 7. Mundell-Fleming and exchange rates Macroeconomía II – 2022/1

i e LM ∗ LM ∗
0

LM IS ∗

0
LM 0 IS ∗

A e
i∗ A0 A A0

IS 0

IS

Y Y

122
Chapter 8

Phillips curve

8.1 Exercises

Exercise 8.1

In class we have seen three different types of frictions that can generate a positive
relationship between output and inflation: (i) wage rigidities; (ii) price rigidities (Calvo’s
model); and (iii) information rigidities (Lucas’ model). Briefly explain in words how
they can generate a positive relationship between output and inflation (the Phillips
curve). You do not need to solve the models, you should only clearly explain in words
the economic intuition behind it.

Exercise 8.2

[Proposed] Suppose an economy with three different types of firms.

The first type has flexible prices (they can adjust it freely in each date) and they
fix it according to:
pft − pt = κ (yt − y ) , κ>0

123
Chapter 8. Phillips curve Macroeconomía II – 2022/1

where pft denotes the price set by firms with flexible prices; pt is the price level; yt is
output and y is potential output. All variables are expressed in logarithms.
The second type of firms has fixed prices, and they fix their price prt according to
their expectations of the price level and output:

prt − pte = σ (yte − y ) , σ>0

Let’s assume that expected output is always equal to potential output: yte = y for every
t. Let αr denote the fraction of firms with fixed prices
The third type of firms fix their prices at date t according to the past price level
pt−1 , readjusting the price by the inflation rate (indexation). In particular those firms
set their price pit according to:

pit = pt−1 + πt−1

The proportion of those firms is αi . Hence, there is a proportion 1 − αr − αi of firms

with flexible prices. Since all variables are expressed in logarithms, you can use the
approximations: pt − pt−1 = πt and pt − pte = pt − pt−1 − (pte − pt−1 ) = πt − πte , where
the superscripts e denote expectations and πt is the inflation rate.

1. Show that:
1
yt = y + [αi (πt − πt−1 ) + αr (πt − πte )]
λ

where λ = (1 − αr − αi ) κ.

2. Interpret the equation you found in the previous item. Compare it with the
situation in which there are no firms that index their prices (αi = 0).

124
Chapter 8. Phillips curve Macroeconomía II – 2022/1

8.2 Solutions

Solution to Exercise 8.1

Wage rigidities: suppose firms workers have set their wage at some date. Now suppose
right after they set the wage, inflation was much higher than expected. This will reduce
the real wage, making firms use much more labor, reducing unemployment. Hence,
taking as given inflation expectations, higher inflation leads to higher employment, and
therefore to higher output.
Price rigidities: if firms take time to adjust price and supply as much goods as
consumers demand, a higher inflation reduces the real price goods, making consumers
demand more, leading to higher production.
Information rigidities: firms care about the relative price to decide how much to
produce. They do not perfectly observe the price level. Hence, when the price level
increases (inflation), they are uncertain if that is an increase in relative prices or an
increase in the price level, leading them to produce more.

125
Chapter 9

IS-LM-PC

9.1 Exercises

Exercise 9.1

Consider the IS-LM-PC model and suppose there is an increase the price oil. Intepret
this increase in the price of oil as a change in potential output. Assume that before the
shock output is equal to potential output. Compare the response of the economy in two
cases: (i) expected inflation always equal lagged inflation; (ii) inflation expectations are
anchored at some level π. In the medium run, the central bank adjusts interest rates
to keep it as close as possible to potential output. You may assume that initially the
interest rates are very high, so that the zero lower bound will not be binding. Write
a graph with time on the horizontal axis and the path of output and inflation on the
vertical axis, for each of the cases mentioned.

126
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

Exercise 9.2

Consider the IS-LM model where the central bank fixes the real interest rate. Assume
that inflation expectations are anchored at some level π. Suppose the government
decides to reduce government spending in a permanent way. Also, assume interest rates
are sufficiently high, so that the central will not be constrained by the zero lower bound
(therefore, you can simply ignore the zero lower bound).

1. What is the effect on output of this change in the short run? Explain your answer
using the IS-LM model.

2. What is the effect on output of this change in the medium run, assuming the
central bank will try to make inflation equal to π? Explain your answer using the
IS-LM-PC model.

3. In the medium run, the IS-LM-PC model predicts that investment will increase or
decrease? Justify.

Exercise 9.3

You should provide your answer to this question using the IS-LM-PC model seen in
class. Suppose that at the initial date, an economy has inflation, expected inflation and
real interest equal to 0. The natural real interest rate of this economy (i.e., the real
interest rate that would make output equal to potential output) is negative. Expected
inflation is always equal to the inflation of the previous period (i.e., πte = πt−1 ). The
central bank always tries to bring output as close as possible to potential output, but
nominal interest rates cannot go below zero.

1. According to the IS-LM-PC model, after the intial date the real interest rate
will increase, decrease or remain stable? And output will increase, decrease or

127
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

remain stable? Explain (drawing a graph may be very useful for that).

2. Explain, using the IS-LM-PC model: can fiscal policy make output equal to
potential output? How?

Exercise 9.4

Brazil’s policy real interest rate are very high. Many economists argue that part of
that is caused by the fact that there is a development bank in Brazil (called BNDES)
lending at interest rates way below the the market interest rate (for some selected firms).
They argue that it makes monetary policy less powerful, forcing the central bank to
increase the interest rate a lot to reduce inflation. The following simple extension of the
IS-LM-PC model is proposed to check if that intuition survives a more formal treatment
and to try to understand better the mechanisms behind this kind of argument.
Suppose that a fraction 1 − λ of firms in the economy borrow at the policy rate plus
the risk premium (r + x), and a fraction λ borrows at (1 − s) (r + x) from the BNDES,
where s denotes the BNDES subsidy. Each firm has a linear investment function on Y
and on the interest rate they borrow, so that the aggregate investment is equal to:

I = b0 + (1 − λ) [b1 Y − b2 (r + x)] + λ [b1 Y − b2 (1 − s)(r + x)]

The consumption function is standard: C = c0 + c1 (Y − T ), with the usual assumption

that b1 + c1 < 1. Also, inflation expectations are anchored at π.

1. Suppose that initially λ = 0 (no BNDES) and output is equal to potential. At

date t λ becomes positive (the development bank is introduced). How will that
affect the equilibrium in the short and medium run?

2. Now suppose that the government decides to increase λ but at the same time

128
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

increase taxes, so that the short run equilibrium remains the same. What happens
in the medium run?

3. Now consider the case with λ = 0 and the case where the government increases λ
and increases taxes as in item 2. Suppose there is an increase in c0 and compare the
medium run equilibrium in both cases. Does it make sense to say that monetary
policy became less powerful because of the subsidized credit?

Exercise 9.5

Consider a country that has had a very low level of investment for decades. Some
economists argue that that investment is low because government spending is very high.
Suppose that the government of this country decides to reduce government spending
(G ) in a permanent way.
Using the IS-LM-PC model, answer the questions below. Assume that inflation
expectations are anchored at some level π. Also, assume that interest rates are sufficiently
high, so that the central bank will not be constrained by the zero lower bound (therefore,
unless otherwise stated, you can simply ignore the zero lower bound). As usual in the
IS-LM-PC model, you must assume that the central bank will not react to shocks in
the short run, but will adjust real interest rates in order to bring output as close as
possible to potential output in the medium run. Also, assume that before the change in
government spending, the economy was in an equilibrium with output equal to potential
output.

1. What is the effect on output of the reduction in government spending (G ) in the

short and medium run? Explain your answer using the IS-LM-PC diagram.

2. In the short run, is investment larger or smaller after the reduction in G ? In the
medium run is investment larger or smaller after the reduction in G ? In your

129
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

answer, you must assume that investment is an increasing function of output and
a decreasing function of the real interest rate (as usual).

Now we replace some assumptions. First, instead of assuming anchored expectations,

we assume adaptative expectations. Second, we assume that at the initial date inflation,
expected inflation, nominal and real interest rates are all equal to zero (using the
notation used in class i = π = π e = r = 0 at the initial date). Hence, you can no longer
ignore the zero lower bound. The economy still has output equal to potential output at
the initial date. Answer the question below.

3. After a permanent decrease in government spending, what will happen to the real
interest rate in the medium run? Explain using the IS-LM-PC diagram.

130
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

9.2 Solutions

Solution to Exercise 9.1

Before we start, a remark of why makes sense to interpret an increase in the price as
a fall in potential output Yn . Think of potential output as the output level consistent
with long run equilibrium in labor markets with perfect wage rigidity. In case oil prices
increase, firm’s marginal cost increase, shifting the labor demand to left and reducing
the equilibrium employment.
Since the PC curve is given by π − π e = θ (Y − Yn ), a fall in Yn shifts up the PC
curve, as illustrated below for anchored (left panel) and adaptative expectations (right
panel). A′ represents the short run response and A′′ the medium run response, after the
central bank has fully reacted.

r r

A00 A00
LM 0 LM 0
A and A0 A and A0
r LM r LM
IS IS
π Y ∆π Y
P C0 P C0
PC PC
A0 A0

π 0
A00 A Y A00 A Y

Inflation increases in both cases after the shock and then, when the central bank
adjusts the interest rate it falls in the left panel and remains constant in the right panel.

131
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

Output goes back to potential in both cases.

π
Adaptative expectations

π Anchored expectations

Time
Y

Same curve for both

Yn
Yn0
Time

Solution to Exercise 9.2

Item 1

As shown in the graph below, after the IS curve shifts to left, output falls.
r

r A0 A
LM

IS 0 IS
Y

Item 2

The figure below shows the medium run equilibrium before the cut in government
spending (point A) and after the cut in government spending (point A”), using the
IS-LM-PC model. As one can see, output returns to potential output, so there is no
effect on output in the medium run.

132
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

r A0 A
LM
r00
A00 LM 00

IS 0 IS
Y
π
PC

π
A and A00 Y
π0
A0

Item 3

As shown in the graph above, the medium run equilibrium with lower government
spending (point A”) features lower real interest than the medium run equilibrium with
higher government spending (point A). Thus, IS-LM-PC predicts that the medium run
equilibrium with lower government spending will have higher investment, since the real
interest rate is lower.

Solution to Exercise 9.3

Item 1

This the classical case that will generate deflate spirals. The graph below shows the
initial equilibrium of the economy (point A). Since output is initially below potential,
inflation is going down, eventually forcing the central bank to increase the real interest
rate because of the zero lower bound and decreasing output (point A’). As inflation

133
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

continues to go down, the lowest possible real interest continues to increase, forcing
the real interest rate up and output down as in illustrated in point A”. This process
continues. (To save space, I did not draw all LM curves in the graph). Note that
the central bank cannot make inflation stop decreasing. That would require setting
r = rn < 0, which is never possible because of the zero lower bound and since inflation
starts at zero and only decreases.
r A00
A0

0 A
LM
rn
B
IS

Yn Y
∆π
PC

0
B Y
0
A
A
A00

Item 2

Yes. If the government increases spending G or decreases taxes T sufficiently he can

makes the natural real interest rate positive, bringing output to potential output already
in the short run. The graph below illustrates the argument.

134
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

0 A A0
LM
rn
B
IS 0 (for G > G0 or T < T 0 )
IS(for G and T )

Yn Y
∆π
PC

0
B Y

Solution to Exercise 9.4

Item 1

First thing we need to understand is how this will affect the IS curve. Equilibrium in
the goods market imply

Y = c0 + c1 (Y − T ) + b0 + (1 − λ) [b1 Y − b2 (r + x)] + λ [b1 Y − b2 (1 − s)(r + x)] + G

1
Y = [c0 − c1 T + b0 − b2 (1 − λs) (r + x) + G ]
1 − c1 − b1
1
r= [−(1 − c1 − b1 )Y + c0 − c1 T + b0 + G ] − x
b2 (1 − λs)

Note that
dr 1 − c1 − b1
=−
dY b2 (1 − λs)

Thus, as λ increases, the IS curve becomes more inclined and also shifts up.

135
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

r
A00 LM 0

A A0
LM
IS 0
IS
π Y
PC

A0

π
A and A00 Y

Graphically, the equilibrium goes from point A to A’ in the short run, and to A” in
the medium run. The real policy rate is indeed higher with higher λ.

Item 2

The policy rate does not change:

r

A
LM

ISλ>0 ISλ=0

π Y
PC

π
A and A0 Y

136
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

Item 3

Using r = 1
b2 (1−λs)
[−(1 − c1 − b1 )Y + c0 − c1 T + b0 + G ] − x, one can see that when
λ > 0, the vertical shift in the IS is larger, for a given change in c0 . The graph below
shows the change in equilibrium:
r
C

B
A
LM
0
ISλ=0

ISλ>0 0
ISλ>0 ISλ=0

π Y
PC

π
A, B and C Y

Note that interest rates are higher at point C (medium run equilibrium after the
shock when λ > 0), than at point B (the medium run equilibrium after the shock when
λ = 0). Monetary policy needs to react more strongly with λ > 0, so it makes sense to
say it became less powerful.

Solution to Exercise 9.5

Item 1

In the diagram below point A denote the initial date, point A′ the short run equilibrium
after the change G and point A′′ the medium run equilibrium. In the short run output
falls, but since in the medium run the central bank reacts decreasing interest rates,
output goes back to potential in the medium run.

137
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

A0 A LM

A00 LM 00
IS
IS 0
Yn Y

π−π
PC

0
A and A00 Y

A0

Item 2

In the short run, investment is lower, since output is lower and the real interest rate
is the same. In the medium run, investment is higher, since output goes back to its
previous level, but real interest rates are lower.

Item 3

In that case, the reduction in government spending will put the economy in a deflation
trap. Point A illustrates the initial equilibrium, and point A′ is the short run equilibrium
after the decrease in G . Since at A′ we have ∆π < 0 in the PC curve, inflation eventually
decreases once prices adjust (causing deflation), which pushes real interest rates up
because of the zero lower bound: i = r + π e ≥ 0 implies r ≥ −π e . That is, since π e is
going down because of adaptative expectations and falling inflation , r goes up in the
medium run as time goes by (which is denoted by points A′′ and A′′′ ).

138
Chapter 9. IS-LM-PC Macroeconomía II – 2022/1

r
A000
LM 000
00
A
LM 00
A0 A LM
0

IS
IS 0
Yn Y

∆π
PC

0
A Y

A0
A00
A000

139
Chapter 10

Dynamic inconsistency

10.1 Exercises

Exercise 10.1

Suppose the loss function of the central bank is given by

L = π 2 + γ (u − un + κ)2 .

The Phillips curve is given by u = un − η (π − π e ). The parameters γ,η and κ are both
strictly larger than zero.

1. Compute the equilibrium inflation under discretion.

2. Compute the equilibrium inflation under full commitment.

3. Is the central bank loss larger under discretion or under full commitment? Justify
doing the algebra.

140
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Exercise 10.2

Time starts at date 0. The loss function of the central bank at date t is given by

Vt = πt2 + λ (yt − y − κ)2

and the Philips curve is as usual: yt = y + θ (πt − πte ). At each date τ , the central bank
minimizes the discounted sum of Vt :

∞
X
Wτ = β t−τ Vt
t=τ

where β ∈ (0, 1) and λ, θ > 0. We know that under discretion in an one period model, the
central bank would choose π = π q = λθκ, and we also know that under full commitment
the inflation would be zero. Now suppose people forms their expectations according to:

if πτ = 0, ∀τ < t

0

πte =
λθκ otherwise



In other words, as long as the central bank keeps choosing zero inflation, people believe
it will do that in the future. If it chooses some inflation different from zero, people
believe it will choose the inflation consistent with a static equilibrium under discretion,
which is π = π q = λθκ. Assume that at date τ the central has chosen zero inflation for
every t < τ .

1. What is the value of Wτ if the central bank chooses zero inflation forever (that is,
if it chooses πt = 0 for every t ≥ τ )?

2. Suppose the central bank decides to deviate from the strategy assumed in item 1
and chooses an inflation rate different from zero at date τ (but he does the best

141
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

deviation possible). In that case, what is the value of Wτ ?

3. Write a condition on the parameters that makes the value of Wτ you found in
item 2 smaller than the value of Wτ you found in item 1. Interpret.

Exercise 10.3

Time starts at date 0. The loss function of the central bank at date t is given by

Lt = ut + γπt2

and the Philips curve is as usual: ut = un − η (πt − πte ). At each date τ , the central
bank minimizes the discounted sum of Lt :

∞
X
β t−τ Lt
t=τ

We know that under discretion in an one period model, the central bank would choose
π = πd ≡ η
2γ
, and we also know that the optimal inflation is zero. Now suppose people
forms their expectations according to:

if πτ = 0, ∀τ < t or πτ ̸= 0 for τ ∈ {t − k, ..., t − 1}

0

πte =
η otherwise


2γ

In other words, as long as the central bank keeps choosing zero inflation, people
believe it will do that in the future. If he chooses some different inflation at some date,
people believe he will choose the inflation consistent with a static equilibrium under
discretion, π e = η
2γ
, for k periods, where k ≥ 1, and then they will believe he will choose
zero again.
Suppose the central bank is at a date τ such that it has chosen zero inflation in all

142
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

previous periods. Is it optimal for the central bank to choose zero inflation in all periods
in the future?

Exercise 10.4

Suppose a central banker loss function (L) depends on inflation (π) and unemployment
(u) and is given by:
L = u + γπ 2 , γ>0

where γ > 0 and π denotes the inflation rate. The Phillips curve in terms of unemploy-
ment is:
u = un − η (π − π e ) , η>0

where u denotes the unemployment rate, un is the natural unemployment and π e is the
expected inflation.

1. Find the equilibrium π and L under discretion.

2. Find the equilibrium π and L under commitment.

3. Suppose the central banker can deviate from the inflation he promised under
commitment. What is its loss L in that case? Which inflation will it choose?

4. Why do we say the central banker is time inconsistent? Explain in words.

5. Suppose that instead of playing the game with discretion, the central bank can
delegate the conduction of monetary policy to a third party who has a loss function
L̃ = u + αγπ 2 , with α > 1 (who will then play a game under discretion). Would
the central banker be willing to do so? Justify your answer. (10%)

143
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

In what follows consider a dynamic version of the problem presented. Time is discrete
and indexed by t ∈ {0, 1, 2, ...}. The central banker loss at function at any date τ is:

∞
X ∞
X
(t−τ )
β (t−τ ) ut + γπt2
 
V = β Lt =
t=τ t=τ

The Phillips curve is as before but with time subscripts:

ut = un − η (πt − πte )

Denote by Ldisc , Lcom and Ldev the losses you found in item 1, 2 and 3, respectively.
Denote by π disc , π com and π dev the inflation you found in item 1, 2 and 3, respectively.

6. Suppose the inflation setting game under discretion is played at every date. Also,
assume that agents form expectations according to:

π com if πt−j , for all j > 0


πte =
π disc otherwise



Write a condition on β that guarantees that the central bank will prefer to cooperate
(i.e., choose π = π com at every future date). Write the conditions in terms of β,
Ldisc , Lcom and Ldev . Interpret the condition.

7. In light of your answer to item 6, do you think that guaranteeing that a central
banker will have a long term in his position alleviates or exacerbates the time
inconsistency problem? Explain in words.

144
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Exercise 10.5

Suppose the central bank loss function (L) depends on inflation (π) and unemployment
(u) and is given by:
L = u + γπ 2 , γ>0

The Phillips curve in terms of unemployment is:

u = un − η (π − π e ) , η>0 (PC)

where u denotes the unemployment rate, un is the natural unemployment rate and π
and π e are inflation and expected inflation, respectively.

1. Find the equilibrium π and L under discretion.

2. Find the equilibrium π and L under commitment.

In what follows, consider the following modification to the model presented. Suppose
the timing is the following:

Stage 1. The central bank announces some inflation π A ;

Stage 2. Agents form their expectations π e about inflation, to minimize the forecast
error (π − π e )2 ;

Stage 3. The central bank fixes the inflation π, and given the inflation chosen
unemployment is determined according to the Phillips curve (PC).

We modify the loss function of the central bank, which is now given by:

2
L̂ = u + γπ 2 + ψ π − π A , γ > 0, ψ>0

2
Notice that we added the term ψ π − π A to the central bank loss function. It captures,

145
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

in a reduced form way, the reputation losses of deviating from the announced inflation
π A . Moreover, notice that we are not saying that central bank necessarily needs to honor
the announced inflation π A .

3. Find the equilibrium π, π A and π e .

4. In equilibrium, does the central bank loss increase or decrease with ψ? Interpret
this result.

Exercise 10.6

[Proposed] A central bank has decided to adopt inflation targeting and is now debating
whether to target 5 percent inflation or zero inflation. The economy is described by the
following Phillips curve:
u = 5 − 0.5 (π − π e )

where u and π are the unemployment rate and inflation rate measured in percentage
points. The social cost of unemployment and inflation is described by the following loss
function:
L = u + 0.05π 2

The central bank would like this loss to be as small as possible.

1. If the central bank commits to targeting 5 percent inflation, what is expected

inflation? If the central bank follows through, what is the unemployment rate?
What is the loss from inflation and unemployment?

2. If the central bank commits to targeting zero inflation, what is expected inflation?
If the central bank follows through, what is the unemployment rate? What is the
loss from inflation and unemployment?

146
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

3. Based on your answers to items 1 and 2, which inflation target would you recom-
mend? Why?

4. Suppose the central bank chooses to target zero inflation, and expected inflation is
zero. Suddenly, however, the central bank surprises people with 5 percent inflation.
What is unemployment in this period of unexpected inflation? What is the loss
from inflation and unemployment?

5. What problem does your answer to part 4 illustrate?

Exercise 10.7

[Proposed] Suppose the loss function of the central bank is given by

V = π 2 + λ (y − y − κ)2 .

The Phillips curve is given by y = y + θ (π − π e ). The parameters λ, θ and κ are both

strictly larger than zero.

1. Compute the equilibrium inflation under discretion.

2. Compute the equilibrium inflation under full commitment.

3. Is the central bank loss larger under discretion or under full commitment? Justify
doing the algebra.

4. What happens when κ = 0? Explain with equations and words.

147
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

10.2 Solutions

Solution to Exercise 10.1

Item 1

Under discretion, first people form their inflation expectations, and then the central
bank chooses the inflation rate. Thus, let’s compute first the choice of the central bank
for each level of expected inflation. Plugging the Phillips curve into the loss function,
we get that for a given π e , the central bank chooses π to minimize:

L = π 2 + γ (−η (π − π e ) + κ)2 (1)

The first order condition is:

2π − 2γη (−η (π − π e ) + κ) = 0

Solving for π we get

γη 2 e γηκ
π= π +
1 + γη 2 1 + γη 2

This is the reaction function of the central bank. In equilibrium, we must π = π e .

Replacing π by π e in the reaction function and solving for π:

γη 2 γηκ
π= 2
π+
1 + γη 1 + γη 2

π = γηκ

This is the equilibrium inflation.

148
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Item 2

Under full commitment, the central bank announces π before agents form their expecta-
tions and cannot change it later. Since agents believe the announcement, π = π e . Thus,
(1) becomes:
L = π 2 + γκ2 (2)

Clearly, the optimal inflation is π = 0.

Item 3

To get the central bank loss under discretion, we replace π = π e = γηκ in (1):

Ldiscretion = (γηκ)2 + γκ2

To get the loss under commitment, we replace π = 0 in (2):

Lcommitment = 02 + γκ2 = γκ2

Thus, the losses are larger under discretion.

Solution to Exercise 10.2

Item 1

If the CB decides to cooperate forever he gets:

∞
cooperate
X V commitment
W = β t−τ V commitment =
t=τ
1−β

149
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Where V commitment denotes the instantaneous loss if the central bank commits to zero
inflation. To find it as a function of parameters, we use the fact that if the central bank
commits to zero inflation, agents will choose π e = 0, and therefore by PC we have y = y .
Plugging that in Vt :
V commitment = λκ2

Item 2

If the central bank deviates at date τ , he minimizes:

  πτe
 2
z}|{
πτ2 + λ θ πτ − 0  + κ

Hence, the FOC is:

2πτ + 2λ [θπτ + κ] θ = 0

Which implies:
λθκ
πτ =
1 + λθ2

Therefore, the instantaneous payoff at date τ (the deviation date) is:

 2  2
deviate λθk λθk
V = +λ θ −k
1 + λθ2 1 + λθ2
 2  2
λθk 1
= + λ −k
1 + λθ2 1 + λθ2
(λθk)2 + λk 2
=
(1 + λθ2 )2
1 + λθ2
= λk 2
(1 + λθ2 )2
λk 2
=
1 + λθ2

150
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

After the deviation (dates τ + 1, τ + 2, τ + 3, ...) agents will choose the inflation rate
consistent with discretion. Since the exercise already says that the equilibrium inflation
rate under discretion is λθκ, we do not need to show it again (but we did it in class).
Besides, we know that π e = π in equilibrium under discretion, and hence π = π e = λθκ,
which implies that the loss of the central bank is after date τ (the deviation date) is:

V discretion = (λθκ)2 + λk 2

His discounted payoff of that deviation is then:

βV discretion
W deviate = V deviate +
1−β

Item 3

We want to check for which parameters:

W deviate < W cooperate

Which becomes:

βV discretion V commitment
V deviate + <
1−β 1−β

Subtracting 1
1−β
V discretion from both sides:

1
V deviate − V discretion < V commitment − V discretion


1−β

Multiplying both sides by -1 and using the fact that V discretion > V deviate (and abbreviating
the words):

151
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

1
V disc − V dev > V disc − V com


1−β

V disc − V com
1−β >
V disc − V dev
 
disc dev disc com
V − V − V − V
 
disc com  
V −V  
β <1− disc
=
V − V dev V disc − V dev
V com − V dev
β< ∈ (0, 1)
V disc − V dev

(It is between zero and one because V dev < V com < V disc ). Remember that V discretion =
λk 2
(λθκ)2 + λk 2 , V commitment = λκ2 and V deviate = 1+λθ2
. Therefore, if β is low deviating is
optimal, and not optimal if β is sufficiently high. This implies that if β is sufficiently
high, cooperation is a equilibrium. Note that the expectations assumed are correct on
the cooperating equilibria (which only exists for β sufficiently high).

Solution to Exercise 10.3

Suppose that the central bank does not deviate from the strategy. His payoff will then
be
∞
X un
W Cooperate = β t−τ un =
t=τ
1−β

If he deviates, he will choose πτ to minimize Lτ . Plugging the Phillips curve into the
loss function and using πτe = 0

Lτ = un − ηπτ + γπτ2

The first order condition is

η
−η + 2γπτ = 0 =⇒ πτ =
2γ

152
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

η2 η2 η2
Which yields a loss function of Lτ = un − 2γ
+ 4γ
= un − 4γ
. For the next k periods
then, he will choose the discretion inflation πτ = η
2γ
, but people will choose π e = η
2γ
,
η2
yielding a loss un + 4γ
. After date τ + k, he gets his loss of cooperation again. Thus,
the payoff of this deviation is

τ +k
η2 η2
  X  
Deviate t−τ
W = un − + β un + + β k+1 V Cooperate
4γ t=τ +1
4γ

Note we can rewrite

τ +k
η2 η2 η2 η2
X        
t−τ 2 k
β un + = β un + + β un + + · · · + β un +
t=τ +1
4γ 4γ 4γ 4γ

η2 β k+1 η2 β − β k+1 η2
     
β
= un + − un + = un +
1−β 4γ 1−β 4γ 1−β 4γ

Thus, the payoff of deviating for k periods

 
W Cooperate
η2 β − β k+1 η2
    z }| {
Deviate
 un 
k+1 
W = un − + un + +β 
4γ 1−β 4γ  1−β 

η2 β − β k+1 η 2 η2 β − β k+1 η 2
 
β un
= un − + un + = − +
4γ 1−β 1 − β 4γ 1 − β 4γ 1 − β 4γ

Remark: of course, we are considering the best deviation he can do (if he deviates for
k + 1 periods). If it is not optimal to deviate k + 1 periods, it will not be optimal to
deviate for longer periods. (If it is not optimal to deviate at date τ , it is not optimal to
deviate at {τ + k, τ + k + 1, ...} since he will face the exact same problem).
The deviation is not profitable when W Deviate > W Cooperate

un η2 β − β k+1 η 2 un
− + >
1 − β 4γ 1 − β 4γ 1−β

153
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

η2 β − β k+1 η 2
− + >0
4γ 1 − β 4γ

2β − β k+1 > 1

Thus, if the condition above is satisfied, there is an equilibrium with cooperation. Note
that as k becomes larger, this condition is more easily satisfied. Intuitively, if agents
punish the central bank for a lot of periods, it is easier to have cooperation.

Solution to Exercise 10.4

Item 1

First, we need to find the reaction of the central bank (that is, the inflation he would
choose for a given π e ). Plugging the Phillips curve on the loss function we get:

un − η (π − π e ) + γπ 2 (1)

The central banker will choose π to minimize this expression. To find the minimum it
suffices to take the first order condition:

−η + 2γπ = 0

Which implies that equilibrium inflation (denoted by π disc ) is:

η
π disc =
2γ

Anticipating the central bank will choose this level of inflation, agents set their expecta-
tions π e equal to π disc . Replacing π and π e by π disc = η/ (2γ) in (1) we get the central

154
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

banker loss under discretion (Ldisc ):

 2
disc η
L = un + γ
2γ
η2
= un +
4γ

Item 2

Under commitment, the central banker chooses inflation to minimize (1) subject to
π = π e . Hence, it minimizes:
un + γπ 2

Clearly the chosen inflation is π com = 0 (and agents set π e = 0). The loss is Lcom = un .

Item 3

If agents have set their inflation expectations equal to zero, the central bank will
minimize:
un − ηπ + γπ 2 (2)

The first order condition is:

−η + 2γπ = 0

Which implies he chooses inflation equal to π dev = η

2γ
. Replacing π by π dev in (2) we
get the following loss:  2
dev η η
L = un − η +γ
2γ 2γ
2 2
η η
= un − +
2γ 4γ
η2
= un −
4γ

155
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Item 4

We say it is time inconsistent because Ldev < Lcom . In other words, after promising zero
inflation, the central bank has incentives to deviate if it has the chance.

Item 5
2
Yes. As one can see Ldisc = un + 4γ
η
is decreasing in γ. Delegating to this authority hence
will yield a payoff
η2 η2
un + < un + = Ldisc
4αγ 4γ

Item 6

If the central bank cooperates at every future date it gets:

Lcom
V cooperate =
1−β

If it deviates at date zero then he get Ldev at that date but then it will get Ldisc in all
subsequent periods. Therefore, the loss of deviating is:

Ldisc
V deviate = Ldev + β
1−β

The condition that guarantees that it will prefer to cooperate is V cooperate < V deviate (if
you put a non strict inequality it is fine), which becomes:

Lcom dev Ldisc

<L +β
1−β 1−β

156
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

We can rearrange it a bit:

Lcom − Ldev β
Ldisc − Ldev


<
1−β 1−β

Lcom − Ldev
<β
Ldisc − Ldev

Therefore, β must be sufficiently large for the central bank to be willing to cooperate.
Intuition: if the central banker is very patient, it is not very willing to give up long
term gains for some short run gain. Hence, it is easier to cooperate with a patient
central banker.

Item 7

By increasing the term of the central banker, the time horizon becomes larger. Hence,
we should expect it to become more patient, which can be captured our simple model
by a larger β. As shown in the previous item, a higher β helps cooperation, and hence
can alleviate the time inconsistency problem.

Solution to Exercise 10.5

Item 1

Plugging the Phillips curve into the loss function we get

L = un − η (π − π e ) + γπ 2 (1)

For a given π e the central bank chooses π to minimize the expression above. The first
order condition is
−η + 2γπ = 0

157
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Which yields:
η
π=
2γ

In equilibrium, agents set their expectations equal to η

2γ
. Hence, in the equilibrium
under discretion π = π e . The central bank loss under discretion is then

2
η2

discretion η
L = un + γ = un +
2γ 4γ

Item 2

Under commitment, the central bank chooses π to minimize (1) subject to π e = π.

Hence, it minimizes:
L = un + γπ 2

Clearly the solution is π = 0. The central bank loss is Lcommitment = un .

Item 3

Plugging the Phillips curve into the equation of L̂ we get

2
L̂ = un − η (π − π e ) + γπ 2 + ψ π − π A (2)

We start solving for stage 3 (backward induction). For a given π A and π the central bank
chooses π to minimize the expression above. The first order condition is

−η + 2γπ + 2ψ π − π A = 0

Which implies that it chooses

η + 2ψπ A
π= (3)
2 (γ + ψ)

158
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Now we solve stage 2. For a given π A agents set expectations so that π = π e – taking
into account the central bank will behave according to (4) in the third stage – and hence

η + 2ψπ A
πe =
2 (γ + ψ)

Now look at stage 1. Here, the central bank chooses π A to minimize (2) subject to
π = π e and (3). Hence, it minimizes:

2 2
η + 2ψπ A η + 2ψπ A
 
L̂ = un + γ +ψ − πA
2 (γ + ψ) 2 (γ + ψ)
2 !2
η + 2ψπ A  A − 2 (γ +  A

η + 2ψπ ψ)
 π

= un + γ +ψ
2 (γ + ψ) 2 (γ + ψ)
2 2
η + 2ψπ A η − 2γπ A
 
= un + γ +ψ
2 (γ + ψ) 2 (γ + ψ)

The first order condition is

η + 2ψπ A η − 2γπ A
   
2ψ 2γ
2γ  − 2ψ
  =0
2 (γ +ψ)  2
(γ + ψ)   2 (γ +ψ)  2
(γ + ψ)
 
  

A A


η
 + 2ψπ − η
 − 2γπ =0

π A (ψ + γ) = 0

πA = 0

Hence, the central bank announces π A = 0 and sets inflation equal to

η
π=
2 (γ + ψ)

Inflation expectations are such that π = π e .

159
Chapter 10. Dynamic inconsistency Macroeconomía II – 2022/1

Item 4

The loss function in equilibrium is

L̂ = un + γ (π ∗ )2 + ψ (π ∗ )2

dπ ∗
where π ∗ ≡ η
2(γ+ψ)
. Note that dψ
< 0. The derivative with respect to ψ is

d L̂ dπ ∗ dπ ∗
= 2γπ ∗ + 2ψπ ∗
dψ dψ dψ

This derivative is smaller than zero iff:

dπ ∗ dπ ∗
2γπ ∗ + 2ψπ ∗ <0
dψ dψ

γπ ∗ + ψπ ∗ > 0

π ∗ (γ + ψ) > 0

Which is always satisfied. Hence, the central bank loss function decreasing in ψ (that
is, increasing ψ leads to lower losses). The intuition is that a larger ψ induces lower
inflation (which is good for the central bank), by working as a partial commitment
device.1

1
Note that a larger ψ also increases the costs of deviating from the the announced inflation (which
increases the central bank loss). But the effect through commitment dominates this effect.

160
References

[Blanchard] Olivier Blanchard, Macroeconomics, 7th edition

[de Gregorio] José de Gregorio, Macroeconomía – Teoría y Políticas, 1st edition
[Mankiw] Gregory Mankiw, Macroeconomics, 9th edition
[Walsh] Carl Walsh, Monetary Theory and Policy, 3th edition

161