Copyrighted Material

1
Introduction

1.1 Dynamic General Equilibrium versus

Traditional Macroeconomics
Modern macroeconomics seeks to explain the aggregate economy using the
ories based on strong microeconomic foundations. This is in contrast to the
traditional Keynesian approach to macroeconomics, which is based on ad hoc
theorizing about the relations between macroeconomic aggregates. In modern
macroeconomics the economy is portrayed as a dynamic stochastic general
equilibrium (DSGE) system that reects the collective decisions of rational indi
viduals over a range of variables that relate to both the present and the future.
These individual decisions are then coordinated through markets to produce
the macroeconomy. The economy is viewed as being in continuous equilibrium
in the sense that, given the information available, people make decisions that
appear to be optimal for them, and so do not knowingly make persistent mis
takes. This is also the sense in which behavior is said to be rational. Errors, when
they occur, are attributed to information gaps, such as unanticipated shocks to
the economy.
A distinction commonly drawn is between short-run and long-run equilibria.
The economy is assumed to be always in short-run equilibrium. The long run, or
the steady state, is a mathematical property of the macroeconomic model that
describes its path when all past shocks have fully worked through the system.
This can be either a static equilibrium, in which all variables are constant, or,
more generally, a growth equilibrium, in which in the absence of shocks, there
is no tendency for the economy to depart from a given path, usually one in
which the main macroeconomic aggregates grow at the same rate. It is not,
therefore, the economy that is assumed to be in long-run equilibrium, but the
macroeconomic model. The (short-run or long-run) equilibrium is described as
general because all variables are assumed to be simultaneously in equilibrium,
not just some of them, or a particular market, which is a situation known as
partial equilibrium.
Individual decisions are assumed to be based on maximizing the discounted
sum of current and future expected welfare subject to preferences and four con
straints: budget or resource constraints, endowments, the available technology,
and information. A central issue in DSGE macroeconomics is the intertemporal

Copyrighted Material

1. Introduction

nature of decisions: whether to consume today or save today in order to con

sume in the future. This entails being able to transfer todays income for future
use, or future income for todays use. These transfers may be achieved by hold
ing nancial assets or by borrowing against future income. The dierent deci
sions are then reconciled through the economy-wide market system, and by
market prices (including asset prices). Much of the focus of modern macro
economics, therefore, is on the individuals responses to shocks and how these
are likely to aect multiple markets simultaneously, both in the present and in
the future. For example, the business cycle is attributed to such shocks.
Shocks are often treated as random variables unpredictable from the past.
Consequently, dynamic general equilibrium (DGE) models are often referred
to as dynamic stochastic general equilibrium models. As it is often simpler to
carry out our analysis without explicitly including the stochastic features of a
model, we will occasionally refer to DGE rather than DSGE models until we for
mally introduce stochastic elements in our discussion of the theory of nance
in chapter 11.
Three main types of decision are taken by economic agents. They relate
to goods and services, labor, and assets: physical assets (the capital stock,
durables, housing, etc.) or nancial assets (money, bonds, and equity)each
has its own economy-wide market. It is convenient to consider the decisions of
individuals according to whether they are acting, in economic terms, as a house
hold, a rm, or a government. Broadly, the decisions of the household relate to
consumption, labor supply, and asset holdings. The rm determines the sup
ply of goods and services, labor demand, investment, productive and nancial
capital, and the use of prots. Government determines its expenditures, tax
ation, transfers, base money, and the issuance of public debt. Financial rms,
including the banks, coordinate the borrowing and savings decisions of these
three agents via the nancial markets.
A convenient starting point for the study of DSGE macroeconomic models
is a small general equilibrium model, but one that includes the main macro
economic variables of interest. It is based on a single individual who produces
a good that can either be consumed or invested to increase future output and
consumption. It is commonly known either as the Ramsey (1928) model or as
the representative-agent model. This is a surprisingly useful characterization
of the economy as it permits the analysis of a number of its key features
consumption and saving, saving and investment, investment and dividend pay
ments, technological progress, the intertemporal nature of decisions, the nature
of economic equilibrium, the short-run and long-run behavior of the economy
(the business cycle and economic growth), and how prices, such as real wages
and real interest rates, are determinedbut without having to introduce them
explicitly.
The basic Ramsey model can be roughly interpreted as that of a closed econ
omy, without a market structure, in which the decisions are coordinated by a
central planner. A rst step toward greater realism is to allow decisions to be

Copyrighted Material

1.2. Traditional Macroeconomics

decentralized. This requires us to add marketswhich act to coordinate deci

sions, and thereby enable us to abandon the device of the central plannerand
nancial assets. Subsequent steps are to include a government, and hence scal
policy, to introduce money, and hence a distinction between real and nominal
variables, and to allow a foreign dimension (the current account, the balance of
payments, and real and nominal exchange rates). At each stage the economy is
analyzed as a general equilibrium system and the signicance for the economy
of each added feature can be studied. Because it highlights individual behavior,
one of the main attractions of this approach is that it provides a suitable frame
work for analyzing economic policy through respecting the deep structural
parameters of the economy, which are not usually changed by policyunless,
of course, they are policy parameters that are changed. This too is in contrast
to traditional macroeconomic models, like the Keynesian model, which are not
specied in terms of the deep structural parameters but by coecients that
may be changed by policy in a manner that is often unspecied or unknown.

1.2 Traditional Macroeconomics

Views on how best to analyze macroeconomic variables such as aggregate con
sumption, total output, and ination have changed much in the last twenty-ve
years. Under the inuence of Keynesian macroeconomics, the emphasis was on
the short-run behavior of the economy, why the economy seemed to persist
in a state of disequilibrium, and how best to bring it back to equilibrium, i.e.,
how to stabilize the economy. In studying these issues, it was common for each
macroeconomic variable to be modeled one at a time in separate equations; only
then were they combined to form a model of the whole economy. The Brook
ings macroeconometric model was constructed in exactly this way (Duesenberry
1965): in the rst stage individual aggregate variables were allocated to sepa
rate researchers and then, in the second stage, their equations were collected
together to form the complete macroeconometric model. As a result, macro
economics tended to focus on the short-run behavior of the economy and did
so using a partial-equilibrium approach that led to a compartmentalization of
thought in which it was dicult to acquire an overall view of how the system as
a whole was likely to behave in response, for example, to a change in an exoge
nous variable, such as a policy instrument, or to a shock. Consequently, it was
sometimes dicult to take into account the wider and longer-term eects of
policypolicies that may even have been designed to combat the shock. There
was, therefore, a tendency for policy to be too narrowly conceived and analyzed.
The study of macroeconomics was prompted in large part by the Great
Depressionthe worldwide recession of the 1930s. One of Keyness original
objectives in The General Theory (Keynes 1936) was to understand how such
sustained periods of high unemployment could occur. From the beginning,
therefore, interest was directed not so much to how the economic system
behaved in long-run equilibrium, but to why it seemed to be misbehaving in

Copyrighted Material

1. Introduction

the short run by generating periods of apparent disequilibriumin partic

ular, departures from full employmentand to what, if anything, could be
done about this. Until the last few years macroeconomic theory (and espe
cially macroeconomic textbooks) has focused mainly on constructing models
to explain this so-called disequilibrium behavior in the economy with a view to
formulating appropriate stabilization policies to return the economy to equilib
rium. The grounds for stabilization policy are that the economic system departs
from equilibrium and, left to itself, would not return to equilibrium, or would
do so too slowly. The aim of the policy intervention is to restore equilibrium,
or to return the economy to equilibrium (or close to equilibrium) more quickly.
This disequilibrium approach to macroeconomics tends to focus on individual
markets and not the system as a whole. It also emphasizes the demand side
of the economy. The outcome was a piecemeal, partial-equilibrium approach
to macroeconomics. A corollary of the emphasis on disequilibrium was that
equilibrium came to be regarded as a special, and less important, case.

1.3 Dynamic General Equilibrium Macroeconomics

The traditional approach to macroeconomics may be contrasted with the con
ception of DSGE macroeconomics that economic agents are continuously reopti
mizing, subject to constraints, with the result that the macroeconomy is always
in some form of equilibrium, whether short run or long run. According to this
view, the short-run equilibrium of the economy may dier from its long-run
equilibrium but, if stable, the short-run equilibrium will be changing through
time and will over time approach the long-run equilibrium; but the only sense
in which the economy can be in disequilibrium at any point in time is through
basing decisions on the wrong information. From this perspective, even the
view sometimes expressed that disequilibrium is a special case of DSGE macro
economics is misleading. DSGE models assume that ex ante the economy is
always in equilibrium.
Although for most economies macroeconomics has retained a focus on shortrun behavior and stabilization, inspection of the path followed by gross domes
tic product (GDP) shows that the dominant feature is the growth in the trend of
potential output; the loss of output due to recessions is almost trivial by com
parison. This suggests that it is far more important to raise the rate of growth
of potential output through supply-side policies than to move the economy
back toward the trend path of potential output by demand-side stabilization
policies.
The origins of DSGE macroeconomics lie in the work of Lucas (1975), Kyd
land and Prescott (1982), and Long and Plosser (1983) on real business cycles.
Their aim was to explain the dynamic behavior of the economy (notably the
autocovariances of real output and the covariances of output with other aggre
gate macroeconomic time series) based on a competitive rational-expectations
equilibrium model that took its inspiration from models of economic growth.

Copyrighted Material

1.3. Dynamic General Equilibrium Macroeconomics

The initial focus was on the role of technology shocks in generating the busi
ness cycle. The model used by Kydland and Prescott was, in essence, the model
of Ramsey; that of Lucas included, in addition, government expenditures and
money. Subsequent work extended the model in various ways in order to exam
ine the eects of other types of shocks. We consider the principal results of
this research in chapter 16. These issues are not, however, the sole concern of
DSGE macroeconomics, or of this book.
A frequent motivation for constructing macroeconomic models, and one of
the rst questions usually asked of a model, is what it implies for economic
policy. (A recent discussion of the usefulness of DSGE models in formulating
policy is Chari and Kehoe (2006).) Nonetheless, it is important to realize that
the aim of macroeconomics is not just to study policy issues. There are prior
questions that should be asked, such as how the economy might behave if it
were in equilibrium, and how it responds to changes in exogenous variables and
to shocks. Finding the answers to these questions is a sucient reason to study
macroeconomics. Although it is common to ask what the policy implications of
a macroeconomic theory are, with the implicit assumption that we are forever
seeking to interfere in the economy, it is not always necessary to search for
policies that alter the equilibrium solution, especially if we are unclear what the
broader consequences might be. Arguably, simply trying to understand how the
economy behaves is sucient justication.
Much of our analysis will be on considering what sort of factors might disturb
an economys equilibrium, and how the economy responds to these. They may
be changes in exogenous variables or shocks. They may be permanent or tempo
rary, anticipated or unanticipated, real or nominal, demand or supply, domestic
or foreign, and the response of the economy may be dierent in each case. The
conclusions we reach are often very dierent from those based on traditional
macroeconomic models. Shocks may be serially uncorrelated, but the intrinsic
dynamic structure of the economy may result in them having persistent eects
on macroeconomic variables. This is the cause of short-term uctuations in the
economy, and hence the basis of business-cycle theory.
DSGE models are forward looking and hence intertemporal. Current decisions
are aected by expectations about the future. As a result, we use intertemporal
dynamic optimization, in which people are treated as if they rationally process
current information about the future when making their decisions. There is a
premium on obtaining the correct information and on deciding how best to use
it. These concerns are linked to the concept of rationalityanother key feature
of DSGE macroeconomics. The willingness of macroeconomists to make the
assumption of rationality rapidly divided professional opinion into two camps.
Traditional macroeconomics was based on the assumption of myopic decision
making in which mistakes, even when realized, were often persisted in. The
central idea behind rational expectations is that people do not make persistent
mistakes once they are identied. This does not necessarily imply, as is often
assumed, that people have more or less complete knowledge, for the mistakes

Copyrighted Material

1. Introduction

are largely the result of unanticipated information gaps or shocks. It is sucient

to suppose that mistakes are not repeated. As forward-looking decisions must
be based on expectations of the future, they may be incorrect; decisions that
seem correct ex ante may not therefore be correct ex post. As previously noted,
only as a result of this type of mistake might it be appropriate to think of the
economy as being in disequilibrium.
Recent research has suggested that a policy of intervention by the govern
ment tends to be most successful when the government has an informational
advantage over the private sector. If the private sector is able to fully anticipate
the intervention, then the policy may be less successful. This is mostly true
when the intervention involves private rather than public goods. As the private
sector would substitute public for private goods, in eect, households would be
paying for these goods from taxes instead of after-tax income. In contrast, the
provision of public goods leads to a net increase in output as total private bene
ts exceed total private costs. These arguments about government expenditures
apply more generally, of course, and will be developed further below.
Another attraction of this intertemporal approach to macroeconomics is that
it resolves a long-standing weakness in Keynesian economics. This concerns the
way equilibrium is dened in dynamic macroeconomics. The problem goes back
to Keyness (1930) A Treatise on Money, in which the concept of equilibrium
used was that of a ow equilibrium (savings equals investment). Keynes real
ized that his formulation of equilibrium was incorrect and wrote The General
Theory partly in an attempt to correct this error (Keynes 1936). In A Treatise
on Money, one variable (the interest rate) had the task of equilibrating two vari
ables (savings and investment). Keyness solution in The General Theory was
to introduce a second variable (income). This allowed savings to be equal to
investment for any possible values of savings and investment. Unfortunately,
this is still an inappropriate concept of equilibrium, namely, a ow equilibrium.
It is suggested by Skidelski (1992) that Keynes was never entirely happy with
The General Theory. Perhaps this was because he realized that he had not fully
solved the problem of macroeconomic equilibrium.
The correct concept of equilibrium is that of stock equilibrium and not ow
equilibrium. Perhaps Keynes focused on the latter because he was principally
interested in the short run and not in long-run equilibrium. In DSGE macro
economic models individual preferences relate to consumption (a ow variable),
but equilibrium in the economy is dened with reference to capital (a stock).
There are an innite number of possible ow equilibriasometimes called tem
porary (or short-term) equilibriabut only one of these is consistent with the
stock equilibrium. The problem is how to obtain this ow equilibrium. (We show
that this is usually the unique saddlepath to equilibrium.) It is common in DSGE
macroeconomics to start by deriving the stock equilibrium.
A crucial feature of the stock equilibrium is that it involves a forward-looking
component and is not just backward looking. This introduces a vital distinc
tion between economics and the natural sciences that in some ways makes

Copyrighted Material

1.3. Dynamic General Equilibrium Macroeconomics

economics harder. Because people look forward when making decisions, there
may be opportunities for others to manipulate strategically for their own ben
et the information on which these decisions are based. The inanimate natural
sciences, such as physics and chemistry, do not have such a forward-looking
component in their dynamic structure, and therefore do not have this strategic
dimension.
Stocks, or capital, consist of physical and nancial capital. This provides a nat
ural link between macroeconomics and nance. The theory of nance is largely
concerned with pricing nancial assets, notably bonds and equity. As nancial
assets play a crucial role in macroeconomics, and are sometimes substitutes in
wealth portfolios for physical assets, asset-pricing theory is an essential com
ponent of macroeconomics and not something that may be omitted as in the
past, or allowed to become a dierent discipline from macroeconomics.
A related issue is the unit of time. It is common, in both macroeconomics
and nance, to conduct the analysis in continuous rather than discrete time
the convention we adopt here. Using continuous time does not prevent the
use of discrete lags but it does imply that decisions are taken continuously.
In part, the choice depends on the frequency of observation of the data. Con
tinuous time is often a better approximation for nancial decisions as events
and much of the available data are high frequency, possibly very high fre
quency, such as minute by minute. Discrete time is better suited to most macro
economics as neither the decisions nor the data are usually high frequency:
they may be monthly, quarterly, or, for national income data, even annual. In
practice, both continuous and discrete time are just approximations. It is there
fore important to bear in mind that the unit of time used in our discrete-time
analysis may relate to dierent units of calendar time, depending on the sub
ject matter. For example, the units of time in growth theory or overlappinggenerations models are longer than those in the analysis of oating exchange
rates.
Prompted in part by the nancial crisis of 2008, the DSGE approach to macro
economics has recently come under heavy criticism. Writing in the New York
Times, Krugman (2009) claims that the macroeconomics of the last thirty years
is spectacularly useless at best and positively harmful at worst. He asserts that
we are living through the dark age of macroeconomics in which the hard-won
wisdom of the ancients has been lost. In his view:
The economics profession has gone astray because economists, as a group,
mistook beauty clad in impressive-looking mathematics, for truth. Not only
did few economists see the current crisis coming, but most important was the
professions blindness to the very possibility of catastrophic failures.

Skidelski (2009), a biographer of Keynes, expresses a similar view that is worth

describing in more detail. He argues that macroeconomics since Keynes has
become increasingly divorced from reality because it has ignored Keyness fun
damental insight that the future is uncertain. In Skidelskis opinion, uncertainty

Copyrighted Material

1. Introduction

colored much of Keyness macroeconomics and led to the main dierences

between the economics of Keynes and all subsequent macroeconomic theory.
Some of Keyness better-known conclusions emanating from this are that
it makes investment which is forward looking highly volatile, much
more so than consumption, which is based more on current income and
habit;
in times of uncertainty it causes people to prefer holding money to
other nancial assetsKeyness liquidity preference theoryand it makes
monetary policy ineective due to a liquidity trap;
it implies that the emphasis should be put on the short run rather than
the long run, which is subject to much greater uncertainty;
related to this, assumptions should be realistic and appropriate for the
present rather than selected in order to focus on the main factors likely
to be dominant in the long run and to make the analysis more amenable
to a mathematical treatment.
With the nancial crisis in mind, Skidelski emphasizes that uncertainty can
not be reduced to calculable risk, thereby challenging not just modern macro
economics but modern nance theory too.
How justied are these criticisms? While it is undoubtedly true that the
future is uncertain, we still have to take decisions involving the long term,
such as those concerning pensions, durable goods like houses and cars, and,
for businesses, investment in buildings and machinery. These force us to take
a view about the future and hence to make intertemporal decisions under
uncertaintya key feature of modern DSGE macroeconomics that seeks the
best way to do this. Although some events may be unpredictableeven catas
trophic, such as earthquakes (particularly if they are followed by tsunamis)
most shocks to the macroeconomy are open to being modeled as stochastic
processes and hence becoming calculable risks. It is not clear, however, that
the nancial crisis was due to fundamental uncertainty. As argued in chap
ter 15, it was caused by human error, resulting from a failure to interpret and
apply existing theories of economics and nance correctly. And the reason for
using mathematics is to ensure that the analysis is carried out logically and
hence accurately. In this book I have taken the view that, although there may
be uncertainty about the stochastic processes aecting the economy, DSGE
macroeconomics, in combination with modern nancial theory, provides the
best means we possess of trying to understand the macroeconomy.
Various other complaints are sometimes made about DSGE models. First, the
models are far too simple to capture the full complexity of the economy, and
as a result are of dubious value. Second, a comment made by some who are
familiar with engineering, is that in general macroeconomic models are far too
complex to be useful. The rst opinion is easier for most people to sympathize

Copyrighted Material

1.3. Dynamic General Equilibrium Macroeconomics

with than the second. However, engineering has found that it is necessary to
nd a way of simplifying matters in order to make progress. There may be a
lesson in this for macroeconomics. It may provide a justication for the use of
models that are designed to capture key features of the economy while retain
ing their simplicity by abstracting from unnecessary detail. The simplicity of
macroeconomic models, commonly seen as a major weakness, may therefore be
a potential strength. It is true that virtually all macroeconomic policy is based on
a simplied model of the economy, and there is an obvious danger in applying
the conclusions obtained from such models to situations where the simplifying
assumptions are too distorting. This makes it advisable to take care to establish
the robustness of the conclusions to departures from the model assumptions.
Nonetheless, as in engineering, the simple models of macroeconomics can often
be remarkably robust and, therefore, useful. DSGE macroeconomic models tend
to be more complex than Keynesian models, but they are still essentially highly
stylized. Given the complexity of the economy and the high level of abstraction
of theory, perhaps the best that one can hope for from theory is something quite
modest: that it provides the intuition necessary to understand why the economy
behaves as it does and what consequences policy might have. In interpreting
our analysis we should, therefore, bear in mind its limitations: that we are only
using models of the economy and that these models are necessarily simplify
ing because to do otherwise would almost certainly mean making the analysis
intractable. For another view of the relationship between macroeconomics and
engineering, see Mankiw (2006).
An assumption often objected to is that of a representative-agent economy.
An interesting argument against this simplication is that it commits the fal
lacy of composition. This asserts that if each individual attempts to do some
thing, they end up achieving the reverse due to the aggregate consequences.
The best-known illustration of this contradiction is Keyness example of house
holds each trying to save more, with the result that aggregate consumption and
outputand hence individual incomesfall, causing savings to decrease rather
than increase.
The problem with this particular example is that it illustrates the dangers of
using a partial rather than a general equilibrium analysis. One might ask what
caused this sudden desire to save more, and whether the aim is a temporary
or a permanent increase in individual savings. If temporary, then, in a general
equilibrium analysis, the additional supply of savings would be expected to
reduce the cost of borrowing temporarily, thereby stimulating borrowing and
consumption, and deterring an increase in savings. If permanent, then the fall
in the cost of borrowing would make the cost of capital lower and so stimulate
additional investment, raising the optimal level of the capital stock and hence
output and income. A temporary shock would not be expected to aect the
optimal level of the capital stock, and therefore investment. Alternatively, if
the cause of the additional saving was the assumption that in the future a tem
porary negative shock would aect output in the economy, this would induce

Copyrighted Material

10

1. Introduction

forward-looking households to save more today in order to smooth consump

tion. A general equilibrium analysis, therefore, rather dispels the contradiction
contained in the fallacy.
This does not, however, alter the fact that the assumption of a representativeagent economy is made to make the analysis more tractable. More advanced
treatments of macroeconomic problems often allow for heterogeneity.

1.4 The Structure of This Book

This book is organized as a sequence of steps that extend the basic model in
order to produce, by the end, a general picture of the economy that can be used
to analyze its main features. The sequence starts with the basic closed-economy
model and is followed by the introduction of growth, markets, government, the
real open economy, money, price stickiness, asset price determination, nan
cial markets, the international monetary system, nominal exchange rates, and
monetary policy. The book concludes with a discussion of some empirical evi
dence on DSGE models. In view of the above strictures on the advantages of
using simple models, rather than retaining each added feature for each subse
quent step, thereby nishing up with a complex general model of the economy,
we aim to return to the original model as closely as we can and add each new
feature to that. In this way we aim to provide a toolbox that is suitable for
understanding how the economy works and that is useful for macroeconomic
analysis, rather than a fully specied macroeconomic model so complex that it
can only be studied using numerical methods.
Our starting point, in chapter 2, is the basic dynamic general equilibrium
centralized model for a closed economy expressed in real terms. Its purpose is
to introduce the methodology of DSGE macroeconomics. Although the setting
is simple and we assume perfect foresight, it provides a remarkably powerful
representation of an economy that has been used to study a wide variety of
problems in macroeconomics, including real business cycles. It captures the key
problem of DSGE macroeconomics, namely, the intertemporal decision to either
consume today or invest in order to accumulate capital and produce more for
extra consumption in the future. This simple model illustrates the important
distinction between stock equilibrium and the sequence of ow equilibria that
bring this about subject to suitable stability conditions.
The steady-state solution of the basic model of chapter 2 is a static equilib
rium. In chapter 3 we show how the basic model can be modied so that in
steady state the economy may achieve balanced economic growth. As a result,
we are able to reinterpret the basic static solution as a description of the behav
ior of the economy about its steady-state growth path. As it is simpler to ana
lyze a model with a static equilibrium, in later chapters we ignore growth, where
possible, in the knowledge that we can focus on the deviations of the economy
from its growth path. If we require the full solution, we can add the growth
path to the deviations from it.

Copyrighted Material

1.4. The Structure of This Book

11

We decentralize decision making in chapter 4 and include markets to coor

dinate these decisions. In this way we are able to study the joint decisions of
households and rms and their interactions in goods, labor, and capital mar
kets. We show that various pricesnotably wages and the rate of return on
capitalalthough not included explicitly in the basic model, are nonetheless
present implicitly. We also discuss the relationship between households and
rms, and how the prots of the rm generate the rms total value, the value
of a share, and dividend income to households who are the owners of rms.
We introduce government into the model in chapter 5. We discuss the basis
of government expenditures and how best to nance them with debt and taxes.
In the process we consider optimal debt and taxation policy and the sustainability of the scal stance. This discussion is extended in chapter 6 to cover the
problem of time inconsistency in scal policy, and to introduce the overlappinggenerations model. This is particularly useful for scal and other decisions
involving time periods that are very long, such as that of a generation. We use
this to study the increasingly important issue of how to nance pensions.
In chapter 7 we introduce the foreign sector. This has important conse
quences for the economy as it alters the economys resource constraint. The
economy is no longer constrained to consume only what it can produce itself.
Domestic residents can then borrow from and/or invest abroad. All of this
should result in a welfare improvement for the economy. Making the economy
open introduces many new issues and variables, and so greatly complicates the
basic model. For example, there is the allocation problem between domestic and
foreign goods and services and the determination of their associated relative
price (the terms of trade), the relative costs of living of dierent countries (the
real exchange rate), and the sustainability of current-account decits.
So far all variables have been dened in real terms. This is partly to show
that money plays a minor role in most of the real decisions of the economy. In
chapter 8 we study nominal magnitudesincluding the general price level and
the optimal rate of inationby introducing money into the closed economy.
Our focus here is on what determines the demand for money and why this
might be aected by interest rates. We also discuss the use of credit instead of
money. Although in a partial-equilibrium view of the economy money appears to
impose a real cost, we show that in general equilibrium money is far more likely
to be neutral in its eect on real variables. This chapter paves the way for the
later discussion of monetary policy as it covers a key channel in the monetary
transmission mechanism, namely, how money and interest rates aect other
variables via the money market. Later we consider whether money has real
eects in the short run.
Up to this point it has been assumed that prices are perfectly exible and
adjust so that markets clear each period. This is often regarded as a major weak
ness of DSGE models compared with Keynesian models, which tend to stress
the imperfect exibility of prices, arguing that this causes a consequent lack of

Copyrighted Material

12

1. Introduction

market clearing. In chapter 9 we show how to introduce imperfect price exibil

ity into the DSGE model. We show how monopolistic competition in goods and
labor markets may cause imperfect price exibility and result in a cost to the
economy in terms of lost output. In this way we are able to incorporate price
stickiness yet retain the benets and insights of the DSGE model. Such mod
els are sometimes known as New Keynesian models. The principal remaining
dierence between Keynesian and DSGE macroeconomics is that in the DSGE
framework we continue to assume that the economy is always in equilibrium
albeit a temporary, and not necessarily a long-run, equilibrium. Thus economic
agents always expect to be in their preferred positions subject to the constraints
they face, one of which is the information they possess. In this sense, even prices
are chosen optimally. Having introduced price stickiness, we then consider how
this aects the determination of the aggregate supply function, a key equation
in the determination of ination.
As a result of viewing the economy as always being in equilibrium ex ante (but
not ex post ), it becomes more dicult to explain unemployment that is invol
untary. The obvious implication of being in equilibrium is that unemployment
must be voluntary. In chapter 9a new chapter for the second editionwe
examine whether two well-known theories of unemploymentsearch theory
and eciency-wage theoryprovide a satisfactory explanation of unemploy
ment. We contrast these theories with an alternative explanation: that the
persistence of unemployment is due to price and wage stickiness.
Chapters 11 and 12 are a notable departure from traditional treatments of
macroeconomics as they consider the determination of asset prices, the behav
ior of nancial markets, and the role of the bond market in the transmission of
monetary policy. Once the preserve of nance, these issues are now increasingly
recognized as essential components of economics and, in particular, of DSGE
theory. After adding uncertainty due to stochastic features of the economy, the
same basic model that we have used to analyze consumption, savings, and cap
ital accumulation can be used to determine nancial assets and asset prices.
This provides an explanation of asset prices based on economic fundamentals
as opposed to the usual approach in nance, namely, relative asset pricing.
Consequently, asset prices are determined in conjunction with macroeconomic
variables instead of in relation to other asset prices. The general equilibrium
theory of asset pricing is set out in chapter 11 and it is specialized to apply to
the bond, equity, and foreign exchange (FOREX) markets in chapter 12.
Before reaching chapter 11, we have, for the most part, ignored the fact that
intertemporal decisions involve uncertainty about the future and are based
on forecasts of the future formed from current information. Our analysis has
therefore been conducted using nonstochastic, rather than stochastic, intertemporal optimization. This has allowed us to use Lagrange multiplier analysis
instead of the more complicated stochastic dynamic programming. In general
equilibrium asset pricing, uncertainty about future payos is a central feature
of the analysis. The degree of uncertainty about each asset can be dierent.

Copyrighted Material

1.4. The Structure of This Book

13

It ranges from certain to highly uncertain payos. Assets with certain pay
os have risk-free returns; those with uncertain payos have risky returns that
incorporate risk premia in order to provide compensation for bearing the risk.
A key issue in asset pricing is the problem of determining the size of the risk
premium that is required in order for risk-averse investors to hold a risky asset,
i.e., the expected return on a risky asset in excess of the return on a risk-free
asset.
In our previous discussion of the economy, in eect we treated savings as
being invested in a risk-free asset. It may seem, therefore, that we must rework
many of our previous results in order to allow for investing in risky assets.
This would, of course, greatly complicate the analysis as it would necessitate
the inclusion of risk eects throughout. We show in chapter 11 that this is not,
in fact, necessary as all we need do is risk-adjust all returns, i.e., adjust all risky
returns by subtracting their risk premium. This implies that we can continue to
work with only a risk-free asset and to use nonstochastic optimization. Hence,
most of the time we are able to ignore such uncertainty.
In chapter 7 we treat the nominal exchange rate as given and consider only
the determination of the real exchange rate. In chapter 13 we analyze the deter
mination of nominal exchange rates. As the exchange rate is an asset price (the
relative price of domestic and foreign currency), we must use the asset-pricing
theory developed in chapters 11 and 14. The no-arbitrage condition for FOREX
is the uncovered interest parity condition, which relates the exchange rate to
the interest dierential between domestic and foreign bonds. Macroeconomic
theories of the exchange rate are based on how macroeconomic variables aect
interest rates and, through these, the exchange rate. Before embarking on our
analysis of exchange rates in chapter 13, we discuss the eect of dierent
international monetary arrangements on the determination of exchange rates.
Having covered the principal components of the DSGE model, in chapter 14
we study the use of the model in formulating monetary policy. Our analysis
is based on the New Keynesian model of ination. To bring out its new fea
tures we contrast this with the traditional Keynesian analysis of ination. We
consider alternative ways of conducting monetary policy: via exchange rates,
money-supply targets, and ination targeting. In the process we extend our
discussion of the determination of exchange rates in chapter 13 by developing
a New Keynesian model of exchange rates. We then focus solely on ination
targeting. We examine the optimal way to conduct ination targeting both in
a closed economy and in an open economy with a oating exchange rate. We
conclude our discussion of monetary policy by proposing a simple model of
monetary policy in the eurozone, where there are independent economies but
a single currency, and hence a single interest rate for all economies.
In chapter 15another new chapterwe develop the interconnections be
tween macroeconomics and nance further. We consider several issues: we dis
cuss borrowing constraints and default, the role of the banking and nancial
systems in the nancial crisis of 200811; we examine alternative models of the

Copyrighted Material

14

1. Introduction

banking sector, their ability to account for the nancial crisis, and how best to
incorporate a banking sector in a DSGE model; and we propose a DSGE model
with default risk.
The nal chapter, chapter 16, presents a brief account of how well simple
DSGE models perform in explaining the main stylized facts of the economy, and
tries to identify some of their shortcomings. We base our discussion on a small
selection of studies of the real business cycle that is designed to illustrate the
principal issues rather than to be fully comprehensive. The main focus in this lit
erature is on the ability of these models, whether they are for a closed or an open
economy, to explain the business cycle solely by productivity shocks. We then
examine a DSGE model of the economy that claims to provide a better explana
tion of economic uctuations by including various market imperfections and
frictions that introduce dierent types of shocks, including monetary-policy
shocks. The attraction of this approach is that monetary shocks may then have
persistent real eects. It has been suggested, however, that such models suer
from a lack of identication as these market imperfections are open to more
than one interpretation. We discuss this issue together with the more general
question of whether DSGE models are identied. Given the diversity of DSGE
models considered in this book, we complete the chapter with some reections
on how one might decide which features to include when constructing a DSGE
model.
Finally, we provide a mathematical appendix in which we explain the main
mathematical results and techniques we have used in this discussion of
contemporary macroeconomic theory based on DSGE models.
There are also exercises (with solutions) for students and these are available
on the Princeton University Press web site at
http: //press.princeton.edu/titles/9743.html