You are on page 1of 35

American Economic Association

The Missing Motivation in Macroeconomics


Author(s): George A. Akerlof
Source: The American Economic Review, Vol. 97, No. 1 (Mar., 2007), pp. 3-36
Published by: American Economic Association
Stable URL: http://www.jstor.org/stable/30034383
Accessed: 03-09-2016 13:09 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The
American Economic Review

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
Number 108 of a series of photographs of past presidents of the Association

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
I

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
The Missing Motivation in Macroeconomicst

By GEORGE A. AKERLOF*

Macroeconomics changed between the early on the cover of Time.2 Even Milton Friedman
1960s and the late 1970s. The macroeconomics was famously-although perhaps misleadingly-
of the early 1960s was avowedly Keynesian. quoted: "We are all Keynesians now."3 A little
This was manifested in the textbooks of the more than a decade later Robert Lucas and
time, which showed a remarkable unity fromThomas Sargent (1979) had published "After
the introductory through the graduate levels.' Keynesian Macroeconomics." The love-fest was
John Maynard Keynes appeared, posthumously, over.

The decline of the old-style Keynesian e


nomics was due in part to the simultaneou
in inflation and unemployment in the late
t Presidential Address delivered at the one-hundred nine-
teenth meeting of the American Economic Association, and early 1970s. That occurrence was imp
January 6, 2007, Chicago, IL. ble to reconcile with the simple nonacce
* Department of Economics, University of California at
tionist Phillips curves of the time.
Berkeley, 549 Evans Hall, Berkeley, CA 94720 (e-mail:
But Keynesian economics also declined
akerlof@econ.berkeley.edu). This paper is based on a long-
term research program with Rachel Kranton on the impli- causeof a change in economic methodo
The Keynesians had emphasized the de
cations of identity for economic behavior. Our previous
joint papers (Akerlof and Kranton 2000, 2002, 2005) have
dence of consumption on disposable in
explored implications outside of macroeconomics of utility
and, similarly, of investment on current p
functions dependent on people's notions of what ought to
its and current cash flow.4 They posit
be. Some of this paper--especially Section III ("The Miss-
ing Motivation: Norms") and Section IX ("Economic Meth- Phillips
curve, where nominal-rather
odology")-has been directly taken from our jointreal-wage inflation depended upon the
manuscript: The Missing Motivation: Economics Made Hu-
employment rate, which was used as an
man (Akerlof and Kranton 2006). I am especially grateful to
cation of the looseness of the labor market.
Professor Kranton for extending to me the invitation to join
this project, after she had the initial insight in the springThey
of based these functions on their own in-
1996 that concerns regarding identity were missing from trospection regarding how the various actors
in the economy would behave. They also
economic theory. I have also benefited from conversations
with Robert Shiller, with whom I am coauthoring work on
brought some discipline into their judgments
behavioral macroeconomics. In addition, I especially wish
to thank Robert Akerlof and Janet Yellen for invaluable by estimating statistical relations.5
advice. I also want to thank Roland B6nabou, Alan Blinder, But a new school of thought, based on clas-
Louis Christofides, Stephen Cosslett, Ernst Fehr, David
Hirshleifer, Houston McCulloch, John Morgan, George
Perry, Antonio Rangel, Paola Sapienza, Robert Solow, Den-
nis Snower, and Luigi Zingales, and seminar participants at 2 Time, December 31, 1965. His appearance on the cover
the IMF, the World Bank, Ohio State University, Vanderbilt was especially remarkable because Time covers are rarely
University, the University of California at Berkeley, theposthumous. Keynes had died in 1946.
Munich Behavioral Economics Summer Camp, the 2006 3 But in a later disclaimer, Friedman said, almost surely
Macroeconomics and Individual Decision Making Confer-correctly, that he had been quoted out of context. See
ence of the NBER and the Federal Reserve Bank of Boston, http://www.libertyhaven.com/thinkers/miltonfriedman/
and at the Social Interactions, Identity, and Well-Being, andmiltonexkeynesian.html, which quotes Friedman (1968),
Institutions, Organizations, and Growth groups of the CIAR. I Dollars and Sense, 15.
am also grateful to Marina Halac for invaluable research as- 4 The treatment of consumption in The General Theory,
sistance and to the Canadian Institute for Advanced Research as we shall see below, was typical of such thinking. Keynes
and to the National Science Foundation under Research Grant first discusses the dependence of consumption on current
SES 04-17871 for invaluable financial support. income, which he clearly sees as the primary determinant of
SSee, for example, Paul A. Samuelson (1964), Thomas current consumption; but, in addition, he makes a long list
F. Dernburg and Duncan M. McDougall (1967), and Gard- of other factors that will alter the relation between consump-
ner Ackley (1961). The econometric model of Lawrence R. tion and current income.
Klein and Arthur S. Goldberger (1955) provides a useful 5 A good example of this methodology can be seen in
synopsis of the variables that the early Keynesians thought Alban W. Phillips's (1958) mixture of light theory and
most important for a macroeconomic model, and how they statistical analysis in his estimation of the relation between
would be included. wage inflation and unemployment.
5

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
6 THE AMERICAN ECONOMIC REVIEW MARCH 2007

cal economics:to
sical economics, objected that constrained
the casual maximization ways
these folks. New Classical critics
of profit and utility functions of Keynes
is the appropriate
economics insisted instead that these relations microfoundation for macroeconomics. They
be derived from fundamentals. They said that also viewed the neutralities as having a certain
macroeconomic relationships should be derived sort of generality. The neutralities do commonly
from profit-maximizing by firms and from utility- describe equilibria of competitive economies
maximizing by consumers with economic argu-with complete information, irrespective of peo-
ments in their utility functions. ple's preferences-as long as those preferences
The new methodology had a profound effect correspond to economists' typical descriptions
on macroeconomics. Five separate neutrality re- of them. The Keynesians then resurrected some-
sults overturned aspects of macroeconomics but not all-of the Keynesian conclusions by add-
that Keynesians had previously considered in- ing a variety of frictions to the New Classical
contestable. These five neutralities are: the in- model. Those frictions include credit constraints,
dependence of consumption and current income market imperfections, information failures, tax
(the life-cycle permanent income hypothesis); distortions, staggered contracts, uncertainty,
the irrelevance of current profits to investmentmenu costs, and bounded rationality. This formu-
spending (the Modigliani-Miller theorem); the lation preserves many (but not all) Keynesian
long-run independence of inflation and unem- conclusions regarding cyclical fluctuations and
ployment (natural rate theory); the inability of macroeconomic policy.
monetary policy to stabilize output (the rational This lecture will suggest a new stance in
expectations hypothesis); and the irrelevance of regard to each of the five neutralities. Like New
taxes and budget deficits to consumption (Ricar- Classical and New Keynesian economics, it will
dian equivalence).6 These results fly in the facederive behavior from utility and profit maximi-
of Keynesian economics. They undermine itszation. That captures the purposefulness of eco-
conclusions about the behavior of the economy nomic decisions. But this lecture will also
and the impact of stabilization policy. question the generality of the preferences that
The discovery of these five neutrality propo- lead to the five neutralities. There is a sense in
sitions surprised macroeconomists. They hadwhich those preferences are very narrowly de-
not suspected that radically anti-Keynesian con- fined. They have important missing motiva-
clusions were the logical outcome of such seem- tion-since they fail to incorporate the norms of
ingly innocuous maximizing assumptions. the decision makers. Those norms reflect how
the respective decision makers think they and
I. Neutralities and Preferences others should or should not behave, even in the
absence of frictions. Preferences reflecting such
How did macroeconomists react to the dis- norms yield a macroeconomics with important
covery of the five neutralities? On the one hand, remnants of the early Keynesian thinking. They
the New Classical economists viewed their neu- also yield a macroeconomics that, in important
trality results as a telltale: that Keynesian econ- details, cannot be obtained only with frictions.
omists of the previous generation had been We shall see that, with such preferences, even
thinking in the wrong way. In their view, sci- in the absence of frictions, each of the five
entific reasoning was producing a new, leaner, neutralities will be systematically violated. Spe-
more precise economics. cifically:
On the other hand, Keynesian economists, for
the most part, reacted differently. In due course * A realistic norm regarding consumption be-
they came to view the neutralities as logically havior will make consumption directly de-
impeccable. These New Keynesians accepted pendent on current income, in violation of the
the methodological dictums of the New Classi- neutrality of consumption given wealth;
* A realistic norm will make investment di-
rectly dependent on cash flow, in violation of
6 Of course, it took some time for the implications ofModigliani-Miller;
these neutrality results to be fully appreciated. For example,
life-cycle consumption and Modigliani-Miller were initially
* A realistic norm will make wages and prices
considered as nothing more than useful codicils to Keyne-dependent on nominal considerations and
sian thinking. thus violate natural rate theory;

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 7

* A realistic norm will make income and em- neity may sometimes dampen, but will rarely
ployment dependent on systematic monetary nullify, the conclusions of this lecture.
policy, and thus violate rational expectations
theory; and II. The Five Neutrality Results
* A realistic norm will make current consump-
tion dependent on the current generation's For clarity, this section will now give an
social security receipts, in violation of Ricar- overview of each of the five neutrality results.
dian equivalence.
A. Dependence of Consumption on Wealth,
Additionally, insofar as the behavior assumed Not Income
by the early Keynesians differed from the be-
havior that produces the neutralities, there is Standard theory tells us that, under only
likely to be a bias in favor of the Keynesians. somewhat special conditions, consumption de-
The Keynesians based their models on their pends on wealth, which is the value of current
observation of motivations, rather than on ab- assets plus the discounted value of future earn-
stract derivations. If there is a difference be- ings.7 Thus there is no tendency for people to
tween real behavior and behavior derived from make their expenditures conform to the pattern
abstract preferences, New Classical economics of their income receipts (as long as their wealth
has no way to pick up those differences. In is given).
contrast, models with norms based on observa- Changes in the pattern of current income that
tion will systematically incorporate such behav- leave overall wealth constant are neutral in their
ior-although, of course, as with any method, effects on current consumption.
there is the possibility for error.
Inclusion of the "missing motivations in mac- B. The Modigliani-Miller Theorem
roeconomics" then combines the observations
of the Keynesians with the intentionality of One version of the Modigliani-Miller Theo-
economic decisions in New Classical econom- rem says that a firm's investment strategy is
ics. Such a synthesis yields the best of the two
totally independent of its liquidity position.8
approaches. Thus, for example, a corporation with an unex-
pected windfall will not spend any additional
Two Disclaimers.--Before beginning in ear- investment dollars. Instead, it will pass the
nest, let me offer two brief disclaimers. First,windfall on to shareholders or seek other finan-
none of the behavior revealing of the norms thatcial investments, since it will make only those
are introduced in this lecture will be new. Oninvestments whose risk-adjusted rate of return
the contrary, I have purposefully chosen phe- exceeds the rate of return on capital.
nomena that have been emphasized since The Changes in the firm's finances will thus be
General Theory by macroeconomists who have neutral in their effect on current investment.
followed Keynes in voicing their continuing
doubts about classical interpretations of macro- C. Natural Rate Theory
economic behavior.
Second, this lecture will discuss different According to Natural Rate Theory, there is
norms that respectively correspond to the fivesome single rate of unemployment that is the
neutralities. I shall assume that these norms are only level that could be permanently maintained
exogenous. Such assumptions of exogeneity are without ever-increasing inflation or ever-
standard in economic analysis. In a given prob- increasing deflation.9 A fiscal/monetary policy
lem in a given time frame, some terms are mix that sought to maintain employment that
assumed constant, while others are allowed to was any higher would result in permanently
vary. I ask you to withhold your doubts regard- increasing inflation. A fiscal/monetary mix that
ing whether such exogeneity is a correct as-
sumption or not. The incorporation of such
7 See Friedman (1957) and Franco Modigliani and Rich-
endogeneity is the next step-not the first ard Brumberg (1954).
step-in the study of the effect of norms on 8 See Modigliani and Merton H. Miller (1958).
macroeconomics, especially since such endoge- 9 See Edmund S. Phelps (1968) and Friedman (1968).

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
8 THE AMERICAN ECONOMIC REVIEW MARCH 2007

been very narrowly described.


sought to maintain employment thatThey was dependany
lower would result in only on real outcomes. For example,
permanently decreasing in the
inflation. Fiscal/monetary mixes models,
consumption-neutrality that yield
utility dependsdif
on consumption
ferent levels of long-term (steady) and leisure; in Modigliani- w
inflation
thus be neutral in their
Miller, iteffects
depends only on on the long-term
discounted real
unemployment. return to shareholders.
But as early as the beginning of the twentieth
D. Rational Expectations
century, Vilfredo Pareto pointed out that such
characterizations of utility missed important as-
According to Rational Expectations
pects of Theory,
motivation.13 According to Pareto, peo-
a systematic responseple oftypically have opinions as policy
monetary to how they to t
business cycle will have
should,noor howeffect
they should not,on the
behave. Theystabi
ity of the macroeconomy.10
also have views regardingWage and
how others should, or pri
should
setters will foresee the not, behave. Such views
systematic are called
component
the money supply; norms,they will
and they may be raise
individual14 asor lowe
well as
prices and wages exactly proportionally,
social. The role of norms can be easily repre- an
thereby neutralize itssented
effect on demand.
in people's preferences by modifying the
The stability of the utility
economy islosses
function to include thus neutral
in utility inso-
far as they, or others,reaction
with respect to the systematic fail to live up to their
of mo
standards.
etary policy to the business cycle.
Sociology has a further concept that gives an
E. Ricardian Equivalence
easy and natural way to add those norms to the
utility function. Sociologists say that people
According to Ricardian
have an idealEquivalence,
for how they should or shouldunder
not
somewhat special conditions, a ideal
behave. Furthermore, that representati
is often concep-
consumer who receives tualizedain lump-sum
terms of the behavior ofintergen-
someone
erational transfer (for
theyexample, in the
know, or some exemplar whom theyform
do not of
social security payment)
know. Thewill notfunction
standard utility spend a sing
is then mod-
dime extra." Instead, ified
she by adding
will a loss in utility,
pass on dependent
theonwho
the distance of behavior from that
extra income, dollar-for-dollar, to ideal.her heir
who will have to pay the Religion and religious identity
higher tax give bills
us a good nec-
essary to retire the increased example of such norms.debtConsider incurred
the Gospels. i
funding the transfer They to are the most
the sacred texts of Christianity.
previous generatio
The transfer is neutral in its effect on current What do they describe? The life of Christ. How
consumption. should a Christian behave? "His life and con-
versation ought to be worthy of the Gospel of
III. The Missing Motivation: Norms'2 Christ [emphasis added]."'" How is a good
Christian supposed to feel when she has not
Each of the neutralities is based on the as- lived up to her conception of that ideal?
Ashamed.16
sumption that the respective decision makers
are utility maximizers. But in each case the
utility functions of the decision makers have

13 See Pareto (1920). George C. Homans and Charles P.


10 See Lucas (1972), Thomas J. Sargent (1973), and
Curtis (1934) give an excellent summary of Pareto that is
Lucas and Sargent (1979). fully consistent with the emphasis here. Jon Elster (1989)
11 See Robert J. Barro (1974) for the modem reincarna- also presents a similar conception of norms.
tion of these ideas, first discovered by Ricardo. 14 For example, the protagonist of the novel Rice Mother
(Rani Manicka 2002) did not believe she should wear red
12 This section, including much of its exact wording, has
been taken from a joint manuscript with Rachel Kranton with black.
(Akerlof and Kranton 2006). I should emphasize that these15 See http://www.orthodoxytoday.org/articles/StBasil-
Behavior.php.
insights have been developed jointly. The initial instigation
of our project is wholly due to Kranton. It is impossible for16 Of course, there are many interpretations of the Gos-
me to say which ideas or wordings are mine and which are pel, and some of them are even contradictory. But that does
hers. not affect whether the person should be ashamed or not. She

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 9

A. Importance of Norms in theMotivation:


money; it is also about living up to an ideal
Some Examples about who they think they should be.
Such belief regarding how people should be-
have,many
But religion is only one of the and their behavior in accordance with
realms
such To
where people have such an ideal. belief, goes beyond the workplace. It af-
appreciate
the ubiquity of norms in motivation, it areas,
fects disparate is useful
from playing golf to life in
to see some further examples.the family.examples
Those Betty Friedan's Feminine Mystique
gives what
will demonstrate that people tend to bemayhappy
be as good a description of
when they live up to how they norms
think and their should
they impact on people's lives as can
be; and they are, correspondingly, unhappy
be found anywhere-in this case regarding the
norms norms.
when they fail to live up to those for middle-class women of the previous
For the audience for this lecture,
generation. most
Here is aof
brief sample of her de-
whom are professors, teaching provides an es-
scription:
pecially familiar example. We have a view of
what it means to be a good teacher. On lived
"Millions of women ourtheir lives in
lucky days, when we live up tothe our standards
image of those pretty pictures of the
and our classes go well, we tend American suburbanon
to be happy; housewife, kissing
our off days, when something goes
their awry
husbands in in front of the
goodbye
picture window, depositing their station-
class, we may even feel quite miserable.
wagonsful
Such motivation in the workplace of children
is the rule, at school, and smil-
rather than the exception. Mosting as they ran like
workers, the new electric waxer
over the spotless kitchen floor .... Their
teachers, care about the conduct of their jobs.
only dream was to be perfect wives and
Randy Hodson (2001), who surveyed ethnogra-
mothers; their highest ambition was to
phies of the US workplace, found
have fivethat most
children and a beautiful house,
employees care about their dignity
their only at
fightwork.
to get and keep their hus-
They want to conceive of what bands they....do They as gloried
use- in their role as
ful. And they feel a lack of dignitywomen, and wrote are
if they proudly on the census
blank: "Occupation,
thwarted, either by their own actions or by the housewife" (Friedan
actions of others. Those who are 1963, 18).
unable to get
such satisfaction are likely to show their dis-
pleasure by acting up in someMost way women or other.
lived up to these norms. Some of
Studs Terkel's Working (1972) thesecaptures
were dissenters,in like Friedan
a herself, who
single volume much of the ethnographic
disagreed with them, but find-
felt compelled, never-
ings summarized by Hodson. Terkel interviews
theless, to follow a norm with which they dis-
people from many different agreed.
occupations
Friedan says they about
suffered from "the
their feelings about their jobs problem
and concludes
without a name." Inthatour terms, they were
people "search for daily meaning as well
losing utility because they wereas failing to live up
daily bread" (1972, xi). Some of the
to what interview-
one part of them thought they should do.
ees are successful in this search: like the stone
We may appeal to religious texts, to work
mason, who cruises his Indiana county and ethnographies, and, like Friedan, to women's
basks in pride as he not infrequently passes his magazines to see the role of norms. But is there
past work. At the opposite extreme is an Illinois yet harder data, some form of natural experi-
steelworker, whose work denies him the dignity ment, that indicates the importance of norms?
he seeks. He takes out his frustration at work byThe sociologist Erving Goffman has found such
being disrespectful, and, after hours, by gettingan example. He observed the behavior of chil-
into tavern brawls. Most workers are some- dren of different ages when they were brought
where between these extremes, but in all cases, to the local merry-go-round. Because appropri-
following Terkel, they have a feeling for how ate activity differs by age, the children should
they should behave at work. It is not just about have predictably different reactions. For the
toddlers, riding a wooden horse is an accom-
plishment. They show their joy at fulfilling what
thinks she should be ashamed if she fails to live up to her they should do with smiles and waves as they
interpretation of the Gospel. pass by. In contrast, for older children, there is

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
10 THE AMERICAN ECONOMIC REVIEW MARCH 2007

a gap between theirthat conception


can also of be how interpr th
norms. Becker's
should behave and riding the theoretical innovation was to
merry-go-roun
However much they may modify plain-vanilla
enjoy economic it, they utility by
alsothe f
introduction of a discrimination
the need to distance themselves from coefficient.
an activ He
defined that as the
that is so age inappropriate. Theyloss in utility incurred by th
manifest
exchange with
distance by riding a frog, someone from
rather thana different race-
a "serio
for example, the loss of athey
animal like a horse; alternatively white from an ex-
show o
change with a black.during
by standing up "dangerously" The natural interpretation
the ride
some way or other they play
is that the thecoefficient
discrimination clown. represents
Behavior at the merry-go-round is,from
the loss in utility for the white of cours
physically
engaging
just the stuff of kids. But in an exchange with a black.
Goffman But this
suppleme
it with a totally serious
representationexample.
of the utility functionIn can surgi
also be
operations, because of their
interpreted inexperience,
in terms of norms. There is a code as m
to how
ical students are given blacks and whites
tasks that should behave ridiculou
are toward
each other.
easy.17 They respond in Thethe
white has same
a view that sheway
should as t
older children at the not deal with a black. She loses utility equal
merry-go-round: to
they a
act the clown.18 the value of the discrimination coefficient-not
In economics, as elsewhere,
from the physical$500 bills
association-but do no
ipso facto
from
just lie on the street. Ifthe violation
living of the code.
up There
to is reason
norms
to believe that such norm-based
such an important motivation, it must interpretation
show
better reflects the nature even
in many economic examples, of discrimination
if than
it is
identified in exactly our
a physical language.
exchange-based Gary
theory. In the pre-
Becker's Economics Civil
of Discrimination
Rights period, when Becker was writing, (195
there
offers an example of can be no doubt that discrimination,
now-standard economiand
the code that upheld it, was stronger in the
South than in the North. Yet exchanges between
17 Goffman (1961) observed the behavior of such stu-blacks and whites were surely much more com-
dents in medical operations. mon in the South than in the North. At least one
18 Another example, the Milgram experiment (Stanley
statistic reflects such a difference: there were
Milgram 1963, 1965) demonstrates the strength of such
motivation-by showing the lengths that people will take tosignificantly lower levels of residential segrega-
do what they think they should be doing. To see thistion by race in the South than in the North.
interpretation of this experiment, which is only one of many
ways of viewing it, it is useful to give a brief description. On
arrival, subjects were told that they were involved in a
B. Summary
learning experiment. They were put in the role of the
"teacher," who should administer shocks to a "learner" Our examples are illustrative of behavior that
whenever he gave a wrong answer. The subjects are led to is pervasive. Sociology is dense in examples of
identify with their role as teacher in this experiment, and
people's views as to how they and others should
feel that they should obey the experimenter. Rather than
being another subject, and, rather than being wired, as it behave, their joy when they live up to those
appeared, actually the learner was an unwired, trained con-standards, and their discomfort and reactions
federate of the experimenter. Subjects were then instructedwhen they fail to do so.
to administer shocks of escalating voltage as the learner We now turn to examining the role of
made errors. A surprising fraction of subjects escalated their
norms in each of the five macroeconomic neu-
shocks to the maximum 450 volts-even though such a
dosage in real life would have been lethal. There are many
tralities.20 In each case we shall ask whether
different versions of the experiment, but the version where
the confederate grunts and moans at 75 volts, asks to be let
out of the experiment at 150 volts, and refuses to give any
more answers at 300 volts, is typical. Here more than 60 19 See Douglas S. Massey and Nancy A. Denton (1993,
table 3.1, 64).
percent of subjects went all the way. Nor is such motivation
limited to the laboratory. The rampage of the Nazi Reserve 20 Some years ago, at a conference in Spoleto, Italy,
Police Battalion #101 in Poland during World War IIEdmund Phelps gave a still-unpublished lecture wondering
(Christopher R. Browning 1999) gives a real-world mirror why the economics of the twentieth century had failed to
of the behavior Milgram obtained in the laboratory. Like discover what was central to most of the arts, which was the
Milgram's subjects, the members of this unit, were justrole of subjectivity. This paper is about the direct relevance
of such subjectivity for macroeconomics. I have very much
Ordinary Men (Browning's title). They were recruited from
the most prosaic civilian occupations. benefitted from enjoyable conversations with Professor

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 11

people's views as to how they should


is the behave
utility of the child. The parent chooses her
will enter their utility function. consumption
In each in case,
period 1weto maximize her utility.
shall see that such views will nullify Whatever the wealthrespec-
remains, she bequeaths to her
tive neutrality result. Indeed, child. we shall also see
that in each case there will be a natural norm Ricardian equivalence takes the following form
in this model. Suppose that the government
broadly consistent with Keynesians' views of
economic behavior.21 gives a transfer, which we shall call a social
security payment, to the parent in period 1; but
IV. Ricardian Equivalence then in period 2 it taxes the child to retire the
debt caused by this transfer.24 In this case, the
We shall begin our detailed discussion with consumption of a parent who maximizes the
Ricardian equivalence. It was chronologically utility function U, and who leaves a bequest to
the last of the neutralities to be appreciated byher child will be unaffected by her receipt of
modern economists. But it is also the simplest.social security.
That makes it the best place to begin.22 If there The logic of this result is simple. With and
is missing motivation in the utility function, it without social security the discounted value of
should be easiest to see here. consumption of the parent and of the child is
A very simple model demonstrates the es- constrained by the discounted value of the fam-
sence of Ricardian equivalence, as it was redis- ily's earnings (plus its initial wealth). Social
covered by Robert Barro after a lapse of almost security leaves that constraint unchanged. If the
two centuries.23 In the model, there are just two parent found (c,, c2) to be the optimal division
periods, periods 1 and 2. There are just two of consumption between herself and her child in
people, a parent and her child. The utility of the the absence of a social security payment, this
parent depends directly upon her own consump- same division of consumption between herself
tion, in period 1; it also depends upon the utility and her child will optimize her utility with a
of her child. That utility depends upon his con- social security payment.
sumption, in period 2. A vast literature explains why such Ricardian
The parent's utility function can be expressed equivalence is unlikely to be empirically de-
simply as U,(cl, U2(c2)), where c1 is the con- scriptive.25 The long list of reasons includes (a)
sumption of the parent, C2 is the consumption of infinite, rather than finite, horizons; (b) strategic
the child, U, is the utility of the parent, and U2 bequests to obtain the attention of one's heirs
while alive; (c) childless families; (d) uncer-
tainty, including bequests made because of un-
Phelps. He has summarized for me the content of that talk certainty about the age of death; (e) differential
in an e-mail.
borrowing rates between the government and
21 For each of the five neutralities we see that the inclu-
the public; (f) growth of the economy in excess
sion of broader preferences, inclusive of norms, will bring
Keynesian behaviors back to life. But, of course, that does
of the interest rate, allowing steady debt issu-
not mean that the competitive forces and the maximizingance; (g) lack of foresight regarding the effect
behaviors responsible for the five neutralities are not im-of social security on future taxes; (h) foreign
portant as well. ownership of debt; (i) tax distortions;26, 27, 28 (j)
22 That appreciation is of course due to Barro (1974).
23 This model is quite close to Ricardo's original discus-
sion. It is a considerable simplification of Barro's model.
His model had a sequence of overlapping generations, each 24 The tax and the transfer are both lump-sum.
of which lived for two periods. Barro's contribution was not 25 The conventional wisdom is, of course, that social
only to show Ricardian equivalence in the two-generationsecurity will affect aggregate savings. Martin Feldstein
model, but also its extension to a sequence of generations(1974) and Feldstein and Anthony Pellechio (1979) act as if
when parents' utility depended only on their own utility andincreases in social security of the current generation will
the utility of their own children. Ricardo's discussion,result in increased consumption, so that the next generation
which is close to the two-generation model here, was then will have a lower capital stock.
subsequently rediscovered. There is no uncertainty, and all 26 I take this list mainly from the review article by John
taxes are lump-sum. This proposition may be generalized,J. Seater (1993).
for example, following Barro, to a model with m overlap- 27 Barro (1989) also gives a careful review of the frictional
ping generations, each of which has different consumptionreasons why Ricardian equivalence may not in fact occur.
when young and old. Each parent derives utility from his 28 In the case of strategic bequests, the bequest is an
own consumption and the utility of his child. unusual form of incentive payment for a service rendered.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
12 THE AMERICAN ECONOMIC REVIEW MARCH 2007

that the parent will make


constraints on the consumption of an parents
equal and opposite (so th
do not leave bequests); (k)
offset of hermyopia ofin the
social security transfer terms of pare
an increased
regarding children's future taxbequest to her child. Something
payments.29
The preceding list gives
must be missing
empirical
from the motivationreasons
in Barro's fo
model; otherwise, it would not
failure of Ricardian equivalence; but, have given rise
lengthy
it is, it still ignores its that
to results theoretical
are so surprising. challen
B. Douglas Bernheim
According to that challenge, under and Kyleeconomis
Bagwell
standard assumptions,(1988)
with give further
perfect evidence suggesting
certainty that an
RicardianRicardian
with perfect foresight, equivalence is such a telltale. They
equivalenc
show how the
will occur. Such a result same logic
had would apply to a
previously be
network of gift-givers. Remarkably, any
unsuspected by economists.30
member of such
Two possible conclusions cana network
be will drawn
be indifferent from
this surprise. On thewhether
one shehand,
receives an extra
wedollar or any
might co
tinue to assume thatother
classical
participant in the
assumptions
network is the recip- de
scribe economic behavior. The five neutralities ient. Such conclusions, suspect as they are,
that are the subject of this paper concern the suggest a problem with the model beyond the
realignment to macroeconomics that occurred lack of realism involved in perfect foresight
as economists gained understanding of the con- and perfect certainty. They also suggest miss-
sequences of classical assumptions from the ing motivation.
mid-1950s to the mid-1970s. James Andreoni (1989) has put his finger on
Economists may have been correct in draw- what that missing motivation might be.31 A
ing the conclusion that the early Keynesian eco- bequest is a type of gift. The parent will receive
nomics was too simplistic and naive. But they utility from giving such a gift. Ricardian equiv-
could have drawn another conclusion from this alence will fail if the parent has utility from
surprise. In this view, Ricardian equivalence is gift-giving. With a social security transfer, more
a telltale: we do not believe, even in the pres- money is hers, and the same consumption allo-
ence of perfect foresight and perfect certainty, cation to herself entails a greater gift to her
child. With declining marginal utility for bequest-
giving, she will then divide an increased social
This argument suggests that a "bequest" is not really what it security transfer between additional consumption
seems. This is an argument where the preferences of the for herself and an additional bequest to her child.32
parent do play a role, but quite different from the type of Andreoni thus describes the utility missing
reason that I think would have surprised the Keynesians. I
from the standard utility function as that arising
want to show that parents who make bequests for the
conventional reasons, because they care about the welfare offrom the "warm glow" from giving. Such a
their children, will still routinely violate Ricardian equiva- characterization may be accurate. It also sounds
lence, even in the absence of most of the commonplace as if it is very close to classical assumptions-
frictions that almost surely invalidate exact Ricardian equiv-
alence.
that there is nothing fundamentally different
29 This was Ricardo's own reason for dismissal of the about this additional motivation. But this seg-
argument. He said that the parent would alter her bequestment of the utility function is, in fact, very
because she would not take into account the added tax different from economists' usual characteriza-
payments of the child (see Gerald P. O'Driscoll Jr. 1977). tion of motivation. We know that the "warm
Uncertainty regarding the size of the future tax payments is
glow" does not come from the utility the parent
different from such myopia, in which the payment is alto-
gether ignored. But, with quadratic utility and expected
utility maximization, uncertainty regarding the child's fu-
ture tax payments will have no effect on the size of the
parent's bequest. 31 See also John Laitner (2002), Laitner and Henry Ohls-
30 For example, Feldstein (1974) and Feldstein and Pel- son (2001), Alan S. Blinder (1975) and Michael D. Hurd
lechio (1979) engage in no theoretical soul-searching re- (1989), who have also modeled the bequest motive as com-
garding the negative effects of social security on currenting from the utility of the parent from giving the bequest.
savings. There is a voluminous literature (see Roberto Ric- 32 Formally, she trades off the marginal utility of her
ciuti 2003) examining the empirical validity of Ricardianown consumption against the marginal utility from gift-
equivalence. Largely because of the problem of endogene- giving and the marginal utility she gets from her child's
ity, it is difficult to come to firm conclusions regarding its consumption. In making this trade-off, she takes due ac-
empirical validity. There are studies with findings both forcount of the fact that one unit of consumption today is
and against such crowding out. traded off against (1 + r) units of consumption next period.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 13

what housewives
derives from her own consumption; nor,should
yetdo, parents who leave
bequests
more tellingly, does it derive from the derive a warm
utility ofglow from bequests
because
her child (as the child's utility that is what
depends on they
its think they should do
own consumption). It enters the forutility
their children. Ricardian equivalence then
function
as a separate term. illustrates how odd neutralities can occur in
What, then, could account formodels that fail glow"?
a "warm to take such norms into account.
Parent-to-child bequests are a form A of comment
gift. Ifby David Romer (2001,
there 539)
tells us where
is any type of economic transaction that we is
should
gov-venture next. He has
remarked
erned by norms, it is the giving of that Parent-
gifts.33 "quantitatively important" viola-
to-child bequests also occur tions
within families.
of Ricardian equivalence and of the per-
Therefore, they should also bemanent
affected
income/life-cycle
by the hypothesis occur for
norms of family life. We havethe same reasons.
already seen Ricardian
one equivalence is not
important
example of such norms (Friedan's for us of
portrait as anthe
empirical aspect of mac-
roeconomics.
proper place of women in the early There are so many reasons other
1960s).
than constant.
The norms of family life are not the role of norms for its violation. But it
They
vary by culture. They also changedoes give
overus antime.
initial window
As on the type of
motivation
the nature of the ideal family has missingso
shifted, in has
classical macroeconom-
ics.have
the ideal bequest. Actual bequests Inclusion of such motivation
changed in will give us a
tandem. For example, the ideal new perspective century
sixteenth on the consumption function. It
Anglo-Saxon family was dynastic. allows usThe lineage
to return to a view in which consump-
tion willFathers
passed from father to oldest son.34 depend on current
then income, just as its
left the bulk of their estates to their
inclusion
oldest
makes it
sons.
naturalIn
to believe that social
the twenty-first century, in thesecurity
ideal transfers
family,will affect savings and con-
sib-
lings are equal. Most bequests are now
sumption, evenly
even in a world without frictions.
divided between them.35
V. Consumption and Current Income
Summary.-Economic outcomes, such as the
Thisutility
consumption of the parent and the takes us to
ofthe second neutrality. Ac-
the
child, are one determinant of cording bequests. But other
to this result, an- than its contribution
other possible determinant is parents' to a consumer's
viewswealth,re-current income has no
garding how they should behave independent
toward effect
their on the consumption of a
children. Just as Friedan's suburban housewives utility-maximizing consumer.
waxed their floors, because they thought that is Milton Friedman (1957) derived such
consumption-income neutrality in the two-
period model of Irving Fisher. In this model, the
consumer chooses her consumption between
33 The literature on gift-giving is of course replete with
two periods. She maximizes her intertemporal
the notion that gift-giving will be determined by what assets
people consider to be theirs and how much of those assets utility function, given by the function U(cl, C2):
should be given to others (Ruth Benedict 1946), rather than c, denotes her current consumption in the first
by the final utility outcomes for the gift-giver and for the period; c2 denotes consumption in the second
gift-receiver. Theodore Caplow (1984) describes the im-
period.36 If she maximizes U(c1, c2), a dollar of
plicit rules for Christmas gift-giving in "Middletown." Peo-
ple believe that the gifts they should give, and receive, should
income earned today will have the same effect
be given according to these rules. Caplow suggests that one on her current consumption as a discounted
might consider these "rules" as norms for gift-giving. dollar earned in the next period. Thus, her con-
34 For the history of the Anglo-Saxon family and the sumption will depend only on the discounted
change of its conception from dynastic to nuclear, see value of her current and future income and the
Lawrence Stone (1977).
35 Using tax data, Mark D. Wilhelm (1996) found that only rate of interest. This proposition is easy to
10 percent of estates differed by more than 5 percent from prove. It generalizes to many different com-
equality between bequests to siblings. His data are only for modities and to many different time periods,
bequests from estates larger than the federal minimum for
taxation. For a more general population, Jere R. Behrman and
Mark R. Rosenzweig (2004) have examined the difference in
bequests to twins. Once measurement error is taken into ac- 36 She receives income of Y, in period 1, income Y2 in
count, they find no significant differences in the bequests. period 2, and she can borrow and lend at the rate of interest r.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
14 THE AMERICAN ECONOMIC REVIEW MARCH 2007

and, with quadratichisutility,


followers, believed
to thatuncertain
current income i
comes.37 In standard terminology,
played an especially important rolethe value
in the de-
her discounted income is called
termination her wealth; t
of current consumption.
amount of that wealth that can be spent with
its depletion is called "The
permanent income.38
fundamental psychological law [em- A
alternative expression phasis
of added],
Friedman's hypothe
upon which we are entitled
is that consumptionto depends on permane
depend with great confidence both a
rather than on current income.39
priori from our knowledge of human na-
The permanent income hypothesis
ture and from the detailed factsmay
of expe- be in
accordance with most standard economic mod- rience, is that men are disposed, as a rule
els. Nevertheless, it contradicted prior thinking and on the average, to increase their con-
about the consumption function. Keynes, and sumption as income increases, but not by
as much as the increase in income" (Key-
nes, The General Theory, 1936, 96).
37 The simple proof is that her utility-maximizing con-
sumption will depend upon the intercept and the slope of the It is true that The General Theory discussed a
budget line. The budget line states that the present dis- long list of other factors that could affect con-
counted value of consumption is the present discounted sumption. The list was sufficiently rich to in-
value of her future income, which is what Friedman calls
clude not only current income, but also all the
her wealth. The intercept of the budget line is her wealth.
That is how much she could consume today if she consumed other determinants of wealth, such as expected
nothing tomorrow. And the slope of the budget line is future income and the rate of interest. But that
determined by the rate of interest r: on the budget line for does not make Keynes's theory identical to
every unit of c, she gives up (1 + r) units of c2. Her Friedman's. In the Keynesian theory, consum-
consumption will be on the highest attainable utility indif-
ers are more sensitive to current income than to
ference curve. That will be the indifference curve that is just
tangent to the budget line. As a result, we see that, given the other changes in income that have similar effect
utility function, c1 will be a function of W and r. Note that on the consumer's wealth.
current income does not come into this expression.
38 Formally, permanent income is the product of the rate
of interest and wealth.
A. Empirical Results and Their Explanation
39 The permanent income hypothesis also generalizes to
currently popular models of present bias. In these models A large number of tests have demonstrated the
consumers have present bias in the form of "hyperbolic dis- excess sensitivity of consumption to current in-
counting," which means that they put extra weight in their come, in concert with the Keynesian consumption
utility functions on their current consumption. In this case, the
typical consumer's plans will not be consistent, but they can be
function. For example, John Y. Campbell and
analyzed as if she has multiple selves. Her self today decides N. Gregory Mankiw (1989) nested both Fried-
on how much to consume today and then passes on the re- man's view that consumption depends solely on
maining assets to her self tomorrow. There is an exact analogy wealth and the simplified Keynesian view that
to the parent's maximization in Barro's model of bequests. In
consumption depends solely on income. They
that model, today's consumer passes on assets to her child in
the next generation; in consumer theory, today's consumer suppose that a fraction of consumers A are pure
passes on assets to her new self in the next period. Since the Keynesians, while a fraction (1 - A) behave ac-
standard model of intertemporal consumption and Barro's cording to the permanent income hypothesis; they
model of consumption are exactly isomorphic, Ricardian estimate A from the extent to which consumption
equivalence then tells us that current consumption-which is
the consumption of the initial self-depends only on the con-
overreacts to changes in income that would be
sumer's wealth. David I. Laibson (1997) thus shows that predictable from past changes in income and con-
consumption with forward-looking consumers with hyperbolic sumption. Usefully, then, A gives a natural mea-
discounting will balance the marginal utility of present con- sure of the departure from the permanent income
sumption out of wealth against the marginal utility of future
hypothesis. The estimates of A are significant sta-
consumption according to an Euler condition. Such a condition
is wealth-based. It is the generalization of the tangency of the tistically and also of significant magnitude eco-
utility indifference curve to the budget line in the two-period nomically: between 40 and 50 percent (depending
model of Irving Fisher. Both Friedman and Laibson obtain upon whether three or five periods are used to
consumption that is determined solely by current income if
predict the change in current income).
there is a constraint on current borrowing, and consumers'
desires for current consumption exceed their current income.
Other studies corroborate such excess depen-
There is nothing inherent in the preferences in either case that dence on current income: John Shea (1985), for
causes current consumption to be based on current income. union members whose contracts specified their

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 15

think for
future wages; David W. Wilcox (1989), they should consume. Second,
social
ple think
security recipients who had been earlier they
notified of should consume can
viewedJonathan
changes in cost-of-living adjustments; either as entitlements or as o
Finally,
Parker (1999), for payers of social in taxes
security turn, current income is o
major
with predictable inter-year changes; determinants
Nicholes S. of these entitlem
obligations.
Souleles (1999), for changes in disposable income
net of tax refunds; and James Banks, Richard
Blundell, and Sarah Tanner (1998), and Bern-
Sociology of Consumption.-The m
heim, Jonathan Skinner, and Steven
emphasized
Weinberg
by sociologists for consu
(2001), for retirees. very different from that in the life-c
Textbooks explain such excess Sociologists
sensitivity by describe consumption
a variety of frictions, particularly borrowing
determined by the norms regarding w
constraints. For example, Rudiger should consume. These norms, in tur
Dornbusch
and Stanley Fischer (1987) say: "Given
pendent thatupon the the individual's situati
permanent income hypothesis is who she thinks she
correct [sic],is.
Two examples
there are two possible explanations."40 They illustrate
are such dependence on
norms.and
liquidity constraints for consumers Following
myopia Pierre Bourdieu (1984), peo-
in their projections of future income.
ple's consumption of cultural goods-the liter-
Thus, we see the realignment ature
that occurred
they read, the music they hear, and the art
they buy-reflects
because of the life-cycle permanent income hy- not just their individual
pothesis: excess sensitivity may tastes.
occur, The but
upperonlyclass should not make lower-
class choices.
in the presence of credit constraints Correspondingly, the lower class
or myopia.
should avoid
Such a view cannot have been adopted appearing above their station.42
because
of its empirical support. Few studies have
The epithet tested
"lace curtain Irish" illustrates. To
this proposition, but those that the
dousers
have rejected
of this phrase, those lace curtains were
it. For example, credit constraints cannot
indicative of thoseex- violating their social place.
plain the reduction in consumption Weber's analysis of the relation between re-
of retirees.
And, neither myopia nor creditligion
constraint can
and savings further reflects the role of
people's views regarding
explain the reduction in union members' con- who they should be.
sumption at the time of wage declines scheduled
In The Protestant Ethic and the Spirit of Capi-
in their union contracts (Shea 1995,
talism,43 996).
Weber describes Calvinists as aspiring
The adoption of the permanent income/
to be "worldly ascetics." He concludes that
life-cycle hypothesis then must "economic
rest on acquisition
theoret-is no longer subordi-
ical, not empirical, reasons. But nated
the to man as thefails
theory means for satisfaction of his
to take into account norms regarding what peo-
material needs.""44 Here the purpose of saving is
ple think they should, or should not,
to live consume.
up to an ideal. The Calvinists are thrifty
Such a norm-based theory will nest
because Keynes's
they think they should not be consum-
psychological law. Consumption-income
ing. That turns theneu-
motivation of the life-cycle
trality will occur only in a singular special case.

B. Consumption and the Role of Norms41


42 Bourdieu views this as important because of the role
Why should consumption be overly sensitive
of such differential consumption in the transmission of class
to income? This section presents structure
an argument in to the next. The focus on
from one generation
consumption
three steps. First, sociology gives as a reflection of who people want to be can be
motivations
seen throughout the sociology of consumption. On the low-
for consumption that are very different from the
brow-highbrow scale, a study by Ian Woodward (2003) is at
reasons for it in the life-cycle model.
the Aofmajor
opposite end the spectrum from Bourdieu: Woodward
determinant of consumption asked
is what people
Australian housewives about the reasons for their
choice of furniture. Some went for comfort; others, for
aesthetics. But they also indicated, with a surprising degree
of moral fervor, that their choices reflected who they wanted
40 See Dornbusch and Fischer (1987, 284). to be.
41 I am extremely grateful to Robert Akerlof for help in 43 See Weber (1958).
formulating the argument of this section. 44 Weber (1958, 53).

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
16 THE AMERICAN ECONOMIC REVIEW MARCH 2007

model on its head. There people save only


of respondents beca
to a quest
of their desire for consumption in retirement.
buy a $10 theater ticket if
Luigi Guiso, Paola Sapienza, and
to see a play Luigi
and Zin
found
gales (2003, 2006) have
bill.statistically affirm
In contrast, only 46
buy religion
Weber's hypothesis that a new $10 ticket i
is correlate
both with attitudesthey
towardhad savings and w
lost a previous
actual savings. In addition, Tversky they and have Kahneman
more ge
erally affirmed the quantitative ence by "mental significance accoun
culture for savings and in terms consumption; of entitlement in the
regressions, variables sky reflectingand Kahneman culture have say t
much power as variables the derived$10 bill from do notthe conne lif
cycle hypothesis in In explaining
their mental cross-countaccount,
savings ratios.45 those who have lost the ticket see themselves as
paying for it twice. In their mental account, its
Consumption Entitlements cost is $20. Those with and the lost ticket then tend
Obligations.
While sociology is useful to opt out, because
in givingthey see $20 as too us much
the to ge
eral insight that consumption pay to see the play. But the depends
difference in behav-on cu
tural norms, we need ior to
for those
be who more
lost the ticketspecific.
and those who Wh
is the nature of those norms?
lost the $10 bill could also have They can f
been interpreted
quently be described inof two
in terms entitlements.ways:
Most peopleas want entit
to
ments and, also sometimes, think of themselves asas obligations
responsible human be-
spend. Again some examples ings. When they losewill the ticket, illustrate.
they do not feel
First, oddly, people entitled tohave obligations
just buy another one. That is not the t
spend. Social history type is offullperson they
of aspire
the to be. obligation
keep up appearances. Most We should also observe that
Wall it is not coinci-banke
Street
for example, do not dental live thatlike
the lost ticket
mothers paradox could be on ex- we
fare. They do not want plained to.bothBut,
by mental accounting
even ifand by
they d
it would occasion gossip. norms. Formally, It anyismodel notof mental what
account- th
should do. History is ing can be translated
replete into a model
with of norms: justof t
stories
debt of aristocrats struggling replace the rules of mental accounting as the th
to maintain
social obligations.46norms As thatjust one
people think example,
they should follow.48 t
debts to British merchants But even though
bynorms and mental accounting
Southern plante
who were keeping up maywith
be equivalent,the interpretations
Joneses in terms of of th
eighteenth century, norms are are important for this lecture.
considered Mental ac-
a signific
factor underlying the Southern
counting has the connotation, support
whether rightly or of
American Revolution.47 wrongly, of being a heuristic for quick decisions.
In addition to obligations Such a heuristicto will, of course, sometimes
spend, thereresult ar
also entitlements. The lost-ticket
in cognitive error. Whether rightly or paradox
wrongly,
Amos Tversky andmost Daniel
economists wouldKahneman
dismiss cognitive error as (198
457) gives an illustration. unimportant.Eighty-eight
Why? because in their view people perce
are smart about what they want, and their deci-
sions are also very purposeful. But norms cannot
45 Guiso, Sapienza, and Zingales (2006, 39) report re- be dismissed so easily. As I argued earlier, people
gressions of savings ratios on GDP growth, dependency feel strongly about adherence to them. Their
ratios, and responses to the question: "Do you consider it
especially important to encourage children to learn thrift
and savings?" A one-standard-deviation difference to GDP
growth and to attitude toward thrift both produce a 1.8- 48 But it turns out that there is quite possibly a substan-
percentage-point difference in the savings ratio. (A one-tive difference between the two interpretations. With the
standard-deviation difference in the dependency ratio,mental accounting interpretation the losers of the ticket
which could be the result both of cultural differences and ofcould be induced to buy one, if only a wise friend would
life-cycle considerations, produces a 3.2-percentage-point make them aware of the logical problems of their reasoning.
difference.) In contrast with the norms interpretation the friend cannot
46 See, for example, David Cannadine (1977). be so helpful. Buying a new ticket is a departure from the
47 See Woody Holton (1999). person's norm, and she loses utility by it.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 17

absence from utility constitutes a task.the


Those missing
who were asked for the dona-
motivation of macroeconomics. tion afterward were more likely to keep the
money than those who were asked before-
The Link of Entitlements and Obligations to hand. Those who had completed the task felt
Current Income.-It remains to relate current that they had earned the money and were thus
spending to current income. Norms may beentitled to keep it for themselves.49
complex. But a web of evidence still reveals a* The mental accounting model by Hersh M.
strong association between current income andShefrin and Richard H. Thaler (1988) is espe-
entitlements and obligations to spend. Such cially
a useful in our quest for a Keynesian con-
link, in turn, produces the excess sensitivity ofsumption function. Norms take many forms, so
consumption on current income in Keynes'stheir formal model is not unique.50 But it does
Psychological Law. illustrate a possible link between consumption
A few examples follow. and current income. In this model, people have
three separate mental accounts: current income,
* It is common practice in the United States for current assets, and future income combined
parents, even for rich ones with no budget with pension wealth. As consumers exhaust one
constraint, to expect their children to assumeof these accounts and begin to use the next one
financial independence after their graduationfor their current consumption, they incur a dis-
from college. They are indicating their beliefcontinuous "penalty." Those penalties are psy-
in the norm that the child is entitled to spendchological in nature-this is a model of mental
what she earns. (Most parents, of course, giveaccounting-and they take the form of a loss in
their children a helping hand as they seekutility.51 Corresponding to Shefrin and Thaler's
their independence. But that does not mean assumptions regarding the nature of these costs,
that they do not also strongly believe that theiras consumption rises, consumers will first fi-
children should live on their earnings, since thatnance it wholly from current income; then,
norm is only one of their motivations.) from current assets; and, finally, from future
* In a thought experiment, consider a woman income and retirement wealth.
living on $50,000 a year who learns that As we discussed earlier, it should be no sur-
her uncle will die in one year leaving herprise that there is an exact translation of such a
$2,000,000. Even if she has considerable sav- model into one with norms regarding entitle-
ings in the bank, it would be unseemly for her ments to consume. The rules of mental account-
to run down her savings in anticipation of theing become the norms regarding how money
bequest. She is not entitled to do so. She should be spent. The basic norm is that con-
should stick to spending from her current sumption should come from current income.
income. This gives another example in which
norms regarding entitlements to spend are
related to current income, in violation of the49 These are the results for females. The men gave al-
most nothing so their differentials are irrelevant. The
life-cycle hypothesis. women gave on average about 10 percent of their earnings.
* People's expenditures are supposed to reflect
Those who were asked to donate before the task gave twice
their stations in life, and those stations usu-
as much as those who were asked afterward. The task lasted
40 minutes and was to highlight phrases in a manuscript to
ally reflect their earnings. Thus, for example,
be used in making an index.
college students with little earnings are sup-
50 Shefrin and Thaler themselves are explicit about the
posed to live that way-like college students.
possibility of other models.
Their current spending is supposed to reflect51 We should also note that the Shefrin-Thaler model
their current earnings, not what they will has
be elements not discussed in the text. In general, the
discontinuous penalties from mental accounting are one
earning in the future. (At the other extreme,
reason why consumption might be at a corner solution in
as an obligation, the college president is often
one of the three mental accounts. Shefrin and Thaler have
expected to live in the presidential mansion.)another reason. They view saving as taking willpower,
* Preliminary results from an experiment by which entails a cost in terms of lost utility. The less
John Morgan and myself illustrate another people save the less of this costly willpower they need to
expend. This gives another reason why consumption
relation between entitlement and earnings. might
In be on one of the boundaries of the mental accounts.
this experiment, subjects were asked to do- It is useful to remember that at one of the boundaries,
nate to a charity before and after completing consumption will conform to current income.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
18 THE AMERICAN ECONOMIC REVIEW MARCH 2007

And the discontinuous again wepenalties


see that the current versions of the
correspond t
the losses of utility due life-cycle to
hypothesis have left out missing mo-
respective deviation
from that norm. In particular, tivation that easily justifiesShefrin
the excess sensitiv-and Tha
ler assumed that there is no such cost at all if ity of consumption to income in Keynes's
consumption comes only from current income. psychological law.
That means that current income can be consid-
ered as consumers' entitlement to spend, since VI. Investment and Cash Flow
any consumption that is less than current in-
come entails no deviation at all from the norm The debate concerning investment has been
regarding the account that should finance it. surprisingly close to the debate about consump-
* Shefrin and Thaler give an impressive array tion. The early Keynesians emphasized two
of econometric facts in support of their variables as determinants of investment: current
model. Insofar as these facts support their cash flow (with profits as a major component),
mental accounting model, they also equally and the firm's current holdings of liquid assets.
well support its reinterpretation-with the Each of these variables is a measure of funds
norm that current income is an entitlement to available to firms for investment without seek-
spend. Those facts include: differential sav- ing outside finance.56 In contrast, the later liter-
ings out of windfall and current income;52 a ature denied any special role of liquidity in the
less than one-to-one displacement of discre- investment function.
tionary saving by employee pension contri- The first such questioning came from
butions;53 undersaving for retirement;54 and a Modigliani and Miller, who assumed that man-
marginal propensity to consume out of fully agers maximize shareholder value and that mar-
anticipated bonuses that is much greater than kets are frictionless and competitive. In this
the marginal propensity to consume out of case, a firm's financial position plays no role in
monthly income.55 the value of the firm. The argument for this
* Retired people are commonly believed to independence proceeds as follows. By construc-
tailor their consumption to a concept of tion, Modigliani and Miller show how a com-
income rather than to the value of their petitive equilibrium changes if a firm increases
assets. Shefrin and Statman (1984) have its debt and buys back shares. In the new equi-
viewed this as another form of mental ac- librium, investment will be unchanged, and
counting. They also present considerable shareholders will offset the increase in the
evidence regarding such behavior. firm's debt by a compensating increase in the
bonds in their respective private portfolios. The
C. Summary reason the equilibrium changes in this way is
straightforward: if the markets for debt cleared
Considerable evidence suggests that people's in the old equilibrium, they will again clear in
views regarding what they are entitled to spend the new. If managers' choice of investment
play a major role in their consumption choices. maximized shareholder value in the old equilib-
It also suggests strongly that current income rium, the same choice of investment maximizes
plays a special role in those entitlements. She- it in the new. Investment is therefore indepen-
frin and Thaler have explained such patterns by dent of the firm's current financial position,
mental accounting. A reinterpretation of their including its current liquidity position and its
model shows that they also could have ex- current cash flow.
plained this behavior in terms of norms. Once The advent of q-theory similarly questioned a
special place for current variables, such as cash
flow and liquid asset holdings in the investment
52 Shefrin and Thaler (1988, 619-20). decision. In the original version of the theory,
53 Shefrin and Thaler (1988, 622-24). James Tobin (1969) suggested that a firm's op-
timal investment strategy arbitrages between
54 Shefrin and Thaler (1988, 626-27). Especially, they
say that there would be vast undersaving in the absence of
social security and forced private pensions to prevent it.
There is some ambiguity regarding whether there is under-
saving in the presence of these institutions to counteract it. 56 See, especially, John R. Meyer and Edwin Kuh
55 Shefrin and Thaler (1988, 633). (1957).

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 19

the value at which it can sell a tive


unit to available
of its liquidity.59
capital But, as with credit-
and its investment costs to produce
constraint a new unit
explanations of consumption, empirical
of capital. In this case, firms should
evidence, invest up is,
such as there torejects this hypothesis.
the point where the marginal cost Stevenof N. aKaplan
newand Zingales (1997) analyzed
unit
of capital is the valuation of thesuch
subsample a unit
of firmsof that Fazzari, Hubbard,
capital in the stock market. That and valuation
Petersen had considered
is the most likely to be
market value of the firm's shares creditdivided
constrained.by Theyits
find credit constraint to
capital stock, called the q-ratio. If markets
be rare. Furthermore,are they also found that those
efficient, q is also the expected firms
discounted
with the least value
constraint had the greatest
sensitivityper
of current and expected future profits to cash
unitflows.60
of
capital.57 Since q-theory says that There is, thus,should
firms remarkable similarity between
invest in capital up to the point the consumption
where the cost function
of and the investment
an extra unit of capital stock is equal
function. In both to the
cases, economic theory sug-
gested rejection
present discounted value of the stream of earlier views regarding the
of earnings
from a unit of capital, again, role
as ofincurrent flow variables-current income
Modigliani-
Miller, investment is independent in theofcasetheof consumption,
firm's cash flow in the
finance decision.58 case of investment. In both cases, empirical
The empirical testing of q-theory also has a investigation showed the existence of excess
striking parallel to the empirical testing of the sensitivity to the current flow variable. In both
consumption function. Just as Campbell and cases, these rejections support the previous
Mankiw showed that there was excess sensitiv- Keynesian theory. In both cases, economists
ity to current income in the consumption func- have sought to explain the divergence between
tion, Steven M. Fazzari, R. Glenn Hubbard, and practice and theory by the presence of credit
Bruce C. Petersen (1988) showed that invest- constraints. In both cases, the empirical evi-
ment depends not just upon q, but also upon the dence, such as it is, does not support the case
current cash flows. Furthermore, as in the stan- that credit-constraint explanations explain the
dard explanation of excess consumption sensi- theoretical anomaly.
tivity, Fazzari, Hubbard, and Petersen similarly
suggest that credit constraints are responsible A. Theory of Excess Sensitivity of Investment
for the dependence of investment on cash flow. to Cash Flow
They continue with the Modigliani-Miller/
q-theory assumption that managers maximize Whatever the similarities, consumption and
stockholder value. But they posit that the dif- investment differ in one major respect. In the
ference in information between managers and case of investment, economists are already
financiers results in a wedge between the cost of
aware of a fundamental reason why investment
internal and external financing. This is clearest
will depend on current cash flow. Modigliani-
for firms that are credit constrained-so that Miller and q-theory both assume that managers
credit-constrained firms will be especially sensi-

59 See, for example, Fazzari, Hubbard, and Petersen


(1988). Stewart C. Myers and Nicholas S. Majluf (1984)
also argued that cash flow would affect investment when
57 See Andrew B. Abel (1979), Lawrence H. Summers
managers had information not available to investors.
(1981), and Fumio Hayashi (1982). 60 An examination of the investment spending of firms
58 This should not be a surprise, because the assumptionswith cash windfalls from winning or settling lawsuits sup-
of this version of q-theory are in accord with Modigliani- ports this finding (Olivier J. Blanchard and Florencio
Miller: competitive financial markets and investment that Lopez-de-Silanes 1993). These firms had no problems re-
maximizes shareholder value. Thus, the firm's current fi- garding credit constraints; yet they invested in projects they
nancial position should play no role in investment. In q-would not have otherwise pursued. Another striking finding
theory, current profits are just one component of the stream also shows excess sensitivity of investment to cash flow. In
1986, when the price of oil declined dramatically, non-oil
of current and future profits that determine the value of q. In
this sense, they play no special role in the determination of subsidiaries of oil companies cut their investment relative to
investment. This de-emphasis of current cash flow (and thus the median in their industry (Owen Lamont 1997). But
current profits) in investment is analogous to the denial of because this study examines the investment implications of
any special role of current income in the permanent income afall, rather than of a rise, in the price of oil, it is not useful
hypothesis. in resolving the role of credit constraint.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
20 THE AMERICAN ECONOMIC REVIEW MARCH 2007

maximize shareholder value. In the now-standard many instances of lax corporate oversight in favor
theory of the firm, the interests of the shareholders of management interests. For example, he has
and the interests of the managers are viewed as cited the excess exploration and drilling opera-
different. The managers are only the agents of the tions of oil companies when retained earnings
owners, and accordingly they maximize their own were high, from 1975 to 1981,62 and the mainte-
interests instead. Such incentives are said to turn nance of low-return operations in many US indus-
the managers into "empire-builders,"61 who will tries, as in the investments of General Motors
use the resources they control to increase their throughout the 1980s.63 In Jensen's views, share-
own domains. holders would have fared better if profits had been
Empire-building can result from two types of returned to them, giving them the option of invest-
motivations. On the one hand, managers may ing at a higher rate of return, or perhaps if profits
have only strict economic interests in mind: had been used for takeovers outside the industry.
they care only about their take-home pay, and To cure what he calls the "failure of corporate
their effort on the job. Such managers, for ex- internal control," Jensen has also suggested that
ample, will be biased in favor of investments firms should issue large amounts of debt, perhaps
whose operation or construction enhances their even by going private. In that case, the added debt
firm-specific human capital, and thereby in- obligations act as a brake on excess investment.
creases their bargaining power. Regarding investment behavior, Jensen is then on
On the other hand, empire-building may be the same page as Keynesian economists such as
pursued as a goal of its own, for its own sake. Klein and Goldberger. They refer to "the prefer-
We saw earlier that most workers have views ence of many businessmen for internal as opposed
regarding how they should or should not per-to external financing" (1955, 12-13) and also con-
form their jobs. Accompanying such views,sider it the major reason for the dependence of
most managers and workers will have the fur-investment on cash flow.
ther view that the firm should be investing in
those jobs. For this reason, the agents making B. Sociology of the Corporation
the investment decision are likely to engage in
empire-building. We can represent such moti- Once again, we have seen a neutrality result
that depends on the goals of the respective de-
vation by adding a term to the utility function of
the agent-decision maker. Her utility functioncision makers. Accordingly, the norms of cor-
will not only depend on her own pecuniaryporate decision makers are central to the
returns and her expenditure of effort. It will alsosociology of the corporation. For example, Dirk
include an additional term reflective of her M. Zorn (2004) has examined how the locus of
control has changed in large US firms over the
norms. She will lose utility insofar as the firm's
investment fails to live up to her ideal of whatpast 40 years. He has shown how this control
has shifted away from those with a production
she thinks it should be. In this case, the typical
norm is that she thinks that the firm should or a sales orientation to those with a financial
engage in investment that will enhance her job orientation.64 Empirically, this is seen in the rise
performance. of the chief financial officer. Prior to the 1960s,
Following the logic of Michael Jensen (1986,corporate finances were handled by corporate
1993), empire-building, accompanied by thetreasurers, whose duties were mainly restricted
abdication of corporate oversight in favor of to keeping the accounts and producing the bud-
management interests, explains a correlation be- gets. Now, most large corporations have re-
tween investment and cash flow. Furthermore, placed them by a CFO. With the change in title
this correlation will occur regardless of the moti-has come a change in function. CFOs are typi-
vation for the empire-building, whether for purely cally central to major decisions. Such a change
economic reasons as in the principal-agent model, affects investment decisions. If they are com-
or, instead, because of managers' norms for how mitted to their missions, managers with sales or
they think they should behave. Jensen has given
62 Jensen (1986, 327).
63 See Jensen (1993, 853).
61 Empire-building is especially emphasized by Jeremy 64 That distinction was emphasized earlier, for example,
C. Stein (2003), following Jensen (1986, 1993). by Neil Fligstein (1990).

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 21

production orientations will becallempire-builders.


its real price-in excess of unity. A standard
In contrast, the role of the conscientious CFO is
natural rate model illustrates why this occurs. That
to curb those enthusiasms. model
Fifty years
assumes that in have
each period the typical firm
elapsed since the publication
sets of Modigliani-
a desired real price for the following period;
Miller. According to Zorn, when it first
in each period ap- a bargain with its
it also makes
peared, it did not describe labor
theregarding
investment
next period's real wages. Next
decision of large corporations. Now,
period's quite
nominal pricepos-
and nominal wage are then
sibly, changes in corporate respectively
decision-making
set by adjusting this desired real price
since that time make it more and
realistic.65
this bargained real wage according to infla-
tionary expectations. When demand is higher, the
C. Summary desired real price of the representative firm is
higher for two reasons: on the demand side, be-
The investment decision demonstrates once cause the demand for its product is higher, and, on
again that the respective neutrality result dependsthe cost side, because the bargained real wage is
on the objective function of the decision makers. higher. That bargained real wage is higher both
because the typical employee's opportunity costs,
VII. Natural Rate Theory which take into account her chances of being
unemployed, are higher, and because the firm's
We now turn to natural rate theory. Oncedesire for her labor is higher. Since the firm's
again, the debate concerns the behavior of eco-owners, customers, and workers care only about
nomic decision makers. The early Keynesians real prices or real wages, a given level of real
viewed wage setters, and possibly also priceaggregate demand will be associated with a given
setters, as setting nominal wages and prices, real wage bargain between the firm and its work-
respectively, without taking full account of in-ers, and a given desired real price for the firm's
flationary expectations. In contrast, New Clas-product. If unemployment is sufficiently low-
sical revisionists have assumed that wage and below the natural rate-that desired real price will
price setters care only about relative wages orbe in excess of unity. If unemployment is above
prices, and therefore wage and price setting will
the natural rate, it will be less than unity.
fully incorporate inflationary expectations. Such It is now easy to explain the inflationary and
deflationary spirals in natural rate theory. Con-
behavior yields a long-run neutrality result with
severe limits on the ability of monetary and sider what happens when the representative firm
fiscal policy to affect unemployment and out-wishes to set its price above that of other firms.
put. When wage and price setters care onlyIn this case, actual inflation will exceed ex-
about relative wages and relative prices, accel-pected inflation. With such a positive gap be-
erating inflation will occur if unemployment istween actual and expected inflation, inflationary
below a critical level, called the natural rate; expectations will rise, as inflationary expecta-
accelerating deflation will occur if unemploy-tions are adjusted upward to conform to reality.
ment is above it. But the firm's desired real price, and therefore
As we shall see, such spirals occur because, the difference between actual and expected in-
at high levels of demand, the representative firm flation, will be unchanged as long as unemploy-
will wish to set the price of its product relative ment is constant. There will be no abatement in
to the price of other firms' products-which we the rise in expected inflation. Inflationary ex-
pectations will be forever increasing, and infla-
tion will rise with it, as nominal prices and
65 Curiously, the rise of the CFO may have substituted wages adjust the real wage bargains and the
one overenthusiasm (from the point of view of shareholders)
desired real prices for these increasing inflation-
for another. There is considerable division regarding whether
mergers and acquisitions have positive returns to the buyer. ary expectations. By similar logic, if unemploy-
Robert Bruner's meta-analysis (2002) of many different stud- ment is above the natural rate, there will be a
ies concludes that, on balance, the returns to bidders have been deflationary spiral. The natural rate is the only
zero. This is a poor return for an activity that has involved so
sustainable level of unemployment without ac-
much corporate time and initiative. Furthermore, if some op-
portunities can be identified as having positive returns, then, to
celerating or decelerating inflation. It corre-
reach an average return of zero, the marginal merger and sponds to the exact level of demand where firms
acquisition has negative payoff. wish to set a real price of exactly one.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
22 THE AMERICAN ECONOMIC REVIEW MARCH 2007

A. Acceptance of Natural Rate


dard errors of such Theory
estimates are quite large; thus,
they also fail to reject sums whose departure from
Most macroeconomists do not
one is of sufficient just
size to result view
in departures of na
ural rate theory aseconomically
a useful significant null hypothe
magnitude from natural
They also see it as a rate
description
theory. But the standard of reality.
treatment of the Phil- S
a view is revealed in textbook
lips curve presentatio
ignores this inconvenient fact.
Economists accept naturalThe textbooks thus typicallyfor
theory present theoret
natural
and empirical reasons.rate theory as a "just-so" story. It runs as fol-
Theoretically, theylows.viewThe previous
the Keynesian economists had
assumptions o
natural rate theory posited a Phillips curve without
as realistic. A astandard
dependence
terion for an economicon inflationary
model expectations.
is that Friedman (1968)
particip
in the economy careandonly Phelps (1968)about
perceived that
realsuch a outcom
theory
That is the fundamental assumption
could not result from models where theof partic-natu
rate theory. Also, unlike
ipants in the our
economy other neutrali
are concerned only with
results, natural rate real variables. Theyis
theory modified the relationship so to
insensitive
viations due to "frictions,"
that wage and price such as imperfe
equations would be affected
one for one by inflationary
information, taxes, myopia, expectations. Such
or transaction co
As long as these "frictions"
judicious use of economic can
theory be expres
explained the
solely in real terms, the neutrality
otherwise-mysterious finding of the simulta- resul
neous increases
natural rate theory will be in inflation and unemployment
robust.
Empirical considerations have
of the late 1960s/early also
1970s. The theorybeen
is also i
consistent with
fluential in economists' most econometric estimates.of nat
acceptance
rate theory. The original Phillips curve sho
a close fit between the rate of change of nom
wages and the inverseB. Nominal
of the Considerations in Wage Behavior
unemployment
for 97 years of British data, between 1861
We now turn to the
1957. There was no inflation same question regarding in t
adjustment
wages that
equation. In the United we asked concerning
States in the consumption
late 19
and investment.
and early 1970s, however, suchIs there a
"excess sensitivity"inve
simple
relative to the
relation between changes in respective
nominalneutrality? Natural
wages
unemployment broke rate theory
down, is basedason the assumption
both that
price a
wage inflation rose, wages
alongand prices are set only
with thewith real con-
unempl
siderations
ment rate. Natural rate in mind. "Excess
theory sensitivity" herean ex
offered
nation for this occurrence: it explained the
in inflation by the large oil supply shock a
also an increase in about the Phillips curve, including the natural
inflationary rate of un-
expectatio
employment itself, is considered to be estimated with great
both of which shifted the Phillips curve ou
imprecision. Akerlof, William T. Dickens, and George L.
ward; it explained the rise
Perry (2000) show ain unemployment
range of estimates for both wage and
a decline in demand. price equations with many different specifications. These
Furthermore, new estimates of Phillips curvesestimates, particularly when made for periods of low infla-
tion, show considerable variation in the sum of the coeffi-
seemed to show that the theory closely fit the data.
cients on lagged inflation, dependent on the specification.
If inflationary expectations are formed as a simple Another bit of evidence that suggests such estimates will be
lag of past inflation, estimates of Phillips curvessensitive to specification comes from the high standard
should find that the coefficients on past inflationerrors on the natural rate itself (Douglas Staiger, James H.
sum to one. Many Phillips curve estimates fail toStock, and Mark W. Watson 1997); it would be surprising
that the sum of lagged coefficients could be estimated
reject that this sum is equal to one.66' 67 The stan-
precisely if another component of the Phillips curve, the
natural rate, could be estimated only with very low preci-
sion. Gordon's own estimates show very different values for
66 See, for example, Robert J. Gordon (1977, table 3, this sum of coefficients. Of course, there is a theoretical
lines 6 and 7, 260). reason why estimates of such a sum should not be robust.
67 Given the importance of such findings, it is remark- With rational expectations, rather than a simple mechanical
able that their robustness to specifications of time period, theory of formation of inflationary expectations, Sargent
data, and exact specification of the Phillips curve has never (1971) shows that there is no theoretical reason that they
been subjected to tough tests-even though everything else should sum to one.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 23

takes the form that nominal considerations af- shows such a spiral.71 For example, the United
fect real wage or price setting in some way or States experienced rapid deflation from 1929 to
other. 1933, but inflation systematically neither rose nor
Evidence of one form of violation of the fell for the next decade. The predictions of natural
assumptions of natural rate theory is especially rate theory are thus grossly violated. But sticky
stark. That evidence concerns downward wage wages offer a good explanation for such behavior.
rigidity. Such wage behavior can easily be per- For example, a dynamic simulation of the US
ceived statistically by examining distributions economy with money wage rigidity and with
of wage-changes. These distributions are char- Depression-level unemployment fits the data all
acterized by a bunching of wage changes but at exactly (Akerlof, Dickens, and Perry 1996).72
exactly zero; there are some wage changes justNominal wage rigidity may not only be sta-
above zero in these distributions, but almost tistically
no perceptible; it can also be macroeco-
wage changes just below.68 Careful studies have nomically important, even outside of Great
documented such wage stickiness in Australia, Depressions. Nominal wage rigidity imparts a
Canada, Germany, Japan, Mexico, New Zea- long-run trade-off between unemployment and
land, Switzerland, the United States, and the long-run inflation. This trade-off is of sufficient
United Kingdom.69,70 There seems to be no way size that it should deter central banks from
to account for such nominal wage rigidity with targeting very low levels of inflation. For exam-
the basic assumptions underlying natural rate ple, simulations of the US economy (Akerlof,
theory: that participants in the economy care Dickens, and Perry 1996) show that an increase
only about real prices and real wages. of the inflation target from 0 to 2 percent will
Wage stickiness also explains a macroeco- permanently reduce unemployment by 1.5 per-
nomic observation that is an anomaly for natural centage points.73
rate theory. Unemployment was so massive in
the Great Depression that inflation should have Norms as Explanation for Sticky Money
been below inflationary expectations throughout Wages.-It seems to be impossible, or all but
this long period. With any natural-rate adaptive- impossible, to explain the existence of sticky
expectations Phillips curve, such high unemploy- money wages, without relaxation of the basic
ment would have caused a deflationary spiral. assumption that the utility functions of employ-
Data on inflation are available for 12 countries for ees or of employers contain real arguments. A
the Great Depression. Not a single one of them simple and natural amendment to the standard
model explains such sticky money wages: that
employees have a norm for what wages should
68 These distributions have accumulations at zero, and According to that norm, they will lose utility
be.
from
they are also asymmetric: there are more wage changes a money wage decline. Sticky money
above zero than below zero. This suggests that the accumu- wages then result, as the bargains between em-
lations at zero do not occur just because there is a menu cost
ployers and employees reflect the presence of
for changing wages.
this ideal in the utility function.
69 The following studies have all found significant signs
of nominal wage rigidity: Truman Bewley (1999), David Indeed, the study by Bewley (1999) gives
Card and Dean Hyslop (1997), Shulamit Kahn (1997),
David E. Lebow, Raven E. Saks, and Beth Anne Wilson
(1999), and Joseph G. Altonji and Paul J. Devereux (1999)71 See Janet L. Yellen and Akerlof (2006, 12).
for the United States; Pierre Fortin (1996) for Canada; 72 There are other possible reasons for this failure of the
Vincenzo Cassino (1995) and Simon Chapple (1996) for standard predictions from natural rate theory. Inflationary
New Zealand; Jacqueline Dwyer and Kenneth Leong (2000) expectations may not have been adaptive; the failure of
for Australia; Sara G. Castellanos, Rodrigo Garcia-Verdii,
deflation to accelerate could be due to expectations that the
and David Kaplan (2004) for Mexico; Sachiko Kuroda and price level would return to some normal level. In the United
Isamu Yamamoto (2003a, b, c) and Takeshi Kimura and States, the National Recovery Act, which encouraged firms
Kazuo Ueda (2001) for Japan; Ernst Fehr and Lorenz Goette
to increase prices, and unionization, which gave a fillip to
(2003) for Switzerland; Thomas Bauer, Holger Bonin, andwages, could also have affected the trade-off between in-
Uwe Sunde (2003) and Christoph Knoppik and Thomas flation and unemployment. But since unemployment was so
Beissinger (2003) for Germany; Stephen Nickell and very high for so very long, and since the absence of accel-
Glenda Quintini (2001) for the United Kingdom; and Jonas
erating deflation was so universal across countries, this still
Agell and Per Lundborg (2003) for Sweden. seems to be a dog that did not bark. It seems to point to a
70 See, for example, Anthony P. O'Brien (1989) and
problem with natural rate theory.
Christopher Hanes (2000). 73 See Akerlof, Dickens, and Perry (1996, table 4).

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
24 THE AMERICAN ECONOMIC REVIEW MARCH 2007

direct evidence that such a norm exists and is siderations in mind, but desired wage changes
responsible for wage stickiness. His extensive will be truncated insofar as they entail money
open-ended interviews sought to elicit why em- wage decreases. To my mind, such a view en-
ployers failed to cut money wages in the Con- tails a theoretical error. As we have seen, the
necticut recession of 1991-1992. Bewley existence of money wage rigidity occurs be-
concludes that, even though substitute labor wascause workers have a norm, which affects their
easily available, employers were reluctant to cututility function, that their employers should not
wages because of the negative effects of such make such cuts. The message of this finding is
cuts on morale. He says that managers werethat norms in the utility function yield at least
afraid that cuts in money wages would causeone clear violation of natural rate theory. That
workers no longer to "identify" with their com-suggests the further empirical possibility that
panies.74 There might be no immediate conse- workers (and also employers and customers)
quences during the recession. But employersmay also have other norms regarding what nom-
thought that such cuts would cause workers to inal wages (and prices) should be. All such
shirk after the recession had ended. They also violations are exceptions to natural rate theory,
feared that their best workers would be more and yield reasons for long-run trade-offs be-
likely to quit. These stories indicate that work- tween inflation and unemployment.
ers are not thinking about their wages only in Money wage rigidity is then potentially only
real terms, relative to the price level or the the tip of an iceberg. If there is one way in
wages received by others. They also have a which nominal wages enter utility functions,
special aversion to cuts in wages below their because of employees' norms regarding what
current nominal levels.75 their employers should or should not do, there
could also be many other ways.
Norms about Wage Increases.-The motiva- There is another natural way whereby such
tion underlying resistance to money wage cuts norms could enter utility functions: employees
is so obvious, and the facts are so unexception- may not only have a norm that they should not
able, that most macroeconomists accept the pos- take wage cuts. They may also have norms
sibility that money wages are sticky. Even so, regarding the nominal rate of increase of their
they rarely appreciate the broader implications wages or salaries. For example, employees may
of such violation of the assumptions of naturalbelieve that their employer should give them a
rate theory. Their adjusted model is that price nominal raise.
and wage decisions are made only with real con- There is little research on the existence of
such norms. The two questionnaire studies that
have investigated it obtain strong and mutually
reinforcing
74 In more detail, Bewley (1999, 1-2) summarizes his results. Eldar Shafir, Peter Dia-
findings: "Other theories fail in part because they are based
mond, and Tversky (1997) asked respondents to
on unrealistic psychological assumptions that people's abil-
comment on a vignette about two young women
ities do not depend on their state of mind and that they are
who take their first jobs with the same initial
rational in the simplistic sense that they maximize a utility
that depends only on their consumption and working con- income. Specifically they asked respondents
who will be better off: Barbara, who receives a
ditions, not on the welfare of others. Wage rigidity is the
product of more complicated employee behavior, in the face
5-percent raise in the presence of 4-percent in-
of which manager reluctance to cut pay is rational. Worker
behavior, however, is not always rational and completely
flation; or Ann, who receives a 2-percent raise
when inflation is zero; 79 percent of respon-
understandable. A model that captures the essence of wage
rigidity must take into account the capacity of employeesdents
to correctly said that Barbara would be
identify with their firm and to internalize its objectives. This
worse
off than Ann economically. Nevertheless,
internalization and workers' mood have a strong impact on
64 percent of respondents also said that Barbara
job performance and call for material, moral, and symbolic
reciprocation from company leadership." would be happier.76 Such responses are con-
75 Following the argument by Raj Chetty and Adamtrary to the natural rate hypothesis that em-
ployees only care about real returns. But an
Szeidl (2006), some employers may have been concerned
with the fact that their employees had fixed mortgages that
easy explanation for this phenomenon occurs if
they would find difficult to pay with cuts in nominal wages.
This puts the violation of natural rate theory in another
place: why were these financial contracts in nominal rather
than in real terms? 76 Shafir, Diamond, and Tversky (1997, 351-52).

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 25

high and therefore also very salient, the answers


Barbara and Ann both think that their employer
should give them a nominal wage increase. to such questionnaires could be very different.
And they could impart a very different shape to
Another study, with a different form of ques-
tionnaire, independently found a similar there- trade-off between macroeconomic demand
sponse. Robert Shiller found that 49 percent and
of steady-state inflation.78 In such cases, peo-
a sample of the general public either fully or
ple may gain satisfaction only from wage and
weakly agreed with the following statement: salary
"If increases that exceed inflation. Such
norms regarding how employers should behave
my pay went up, I would feel more satisfaction
in my job, more sense of fulfillment, even will
if then necessitate higher real wages (to main-
prices went up as much." An additional 11
tain the same level of satisfaction) at higher
percent of the general public were undecided, levels of inflation. The long-run inflation-
while only 27 percent completely disagreed.employment
As relation will then be downward
in the case of Ann and Barbara, such opinions sloping. Such behavior gives a much stronger
are consistent with the view that workers think rationale, even than current rational-expectations
their employers should give them a nominal credibility models (Barro and Gordon 1983; Ken-
wage increase: they will be disappointed when neth Rogoff 1987), why central banks should
it does not occur. Shiller's finding may be sim- maintain price stability. Failure to appreciate this
ilar to the public's view of Ann and Barbara. realistic possibility again may be another case in
But, as he reports, it is also in stark disagree- which the absence of norms from utility functions
ment with the view of professional economists has unduly blinkered macroeconomic thinking.79
that underlies natural rate theory. Ninety per-
cent of economists weakly or strongly disagreed
78 Bankruptcy and financial considerations become es-
with the statement; 77 percent were in complete
pecially important when inflation is very high. It is also
disagreement.77
worth noting, at least parenthetically, that high levels of
Such norms-regarding the wage or salary in- bankruptcy at times of high inflation are themselves a symp-
crease that employees think they should receive-- tom of money illusion. Such bankruptcies reflect the non-
indexation of financial contracts.
can be economically consequential. They cause
79 In addition to the two questionnaire studies I have
the long-run inflation-unemployment trade-off
mentioned, indexed contracts give another indicator for the
to be downward sloping. With such a norm, at existence of nominal notions concerning what wage in-
higher levels of inflation workers will not expe- creases should or should not be. Economists are often sur-
rience disappointment from receiving lower prised at the small fraction of union contracts that are
nominal wage increases than they think they indexed at all. (Louis N. Christofides and Amy Chen Peng
2004, for example, analyzed a sample of almost 12,000
should receive; therefore, at higher inflation, Canadian union contracts from 1976 to 2000. The mean
ceteris paribus, wage bargains will result in length of these contracts was slightly more than two years
lower real wages, which will reduce the relative (25 months). Only 19 percent of these contracts were in-
price that the firm wants to set, and therefore dexed.) But even when such indexation occurs, their form
violates the condition that they were struck with only real
raise the rate of sustainable employment. There
considerations in mind. For an imperfect index such as the
is a need for further research following Shafir, CPI, which reflects both supply shocks and demand shocks,
Diamond, and Tversky and Shiller regarding the optimal COLA adjustment will be less than one, but it
whether workers have norms regarding the nom- will always be (almost) symmetric for positive and negative
inal wage increases they think they should receive. deviations of inflation from a threshold. (See Jo Anna Gray
1978, Ronald G. Ehrenberg, Leif Danziger, and Gee San
1983, and David Card 1986 for the derivation of optimal
High Inflation.-The opinions expressed re- indexation.) But COLA adjustments are only positive. (Card
garding Barbara and Ann, and also the opinions 1986, S146, has expressed this in terms of a formula: w(t) =
of Shiller's respondents, suggest that the long- w"(t) + max{0, a[p(t) - p1}, where w(t) is the nominal
wage, wn(t) is the nominal target, p(t) is the actual price
run trade-off between inflation and employment
level, and p' is the threshold.) Thus, the form of the contract
is upward sloping. These answers were elicited violates optimality. In practice, this violation is also biting.
in the United States and thus are reflective of For example, in roughly one-third of a large Canadian
respondents' views in an environment where sample of indexed contracts, inflation was always below the
threshold (see Christofides and Peng 2004, 11, fn. 19). Thus,
inflation has been low. But if inflation is very
the form of indexed contracts, when they exist, shows that
union wage negotiators think that COLA adjustments
should never be negative. The form of indexed contracts
77 Shiller (1997, 37). gives another robust indicator that, indeed, wage setters

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
26 THE AMERICAN ECONOMIC REVIEW MARCH 2007

C. Prices extensive evidence on price stickiness reveals


violation of the assumptions of natural rate the-
We have just seen that employees' norms ory, and also the existence of norms regarding
regarding nominal wages may affect bargained price change. Like wage changes, price changes
also agglomerate at zero. Dennis Carlton (1986)
real wages, and therefore cause trade-offs between
long-run inflation and long-run unemployment. has shown that prices are often sticky for sig-
nificant periods of time.81 Furthermore, prices
Similarly, customers' norms regarding price levels
and price changes may also cause long-run trade- seem to be especially sticky in customer mar-
offs between output and inflation. kets.82 Alan Kackmeister (2002) has compared
Indeed, models by Katsuhito Iwai (1981),price changes at the end of the nineteenth cen-
Julio J. Rotemberg (1982), and Andrew S. Cap- tury to such changes a bit more than a century
lin and John Leahy (1991) all have long-run later. Price changes of specific goods at retail
trade-offs between inflation and unemployment. stores were recorded from June 1889 to Sep-
Each of these models assumes that there are real tember 1891; Kackmeister revisited the same
commodities and their price change for a com-
costs to nominal price changes. If, instead, there
were real costs to real price changes, the as- parable period, from June 1997 to September
sumptions of natural rate theory would still be1999. Price change in the late twentieth century
satisfied, and no such trade-off would occur. was five times more frequent than a century
These models then pose the question why there earlier. Furthermore, in the nineteenth century,
should be such real costs from nominal price
the average spell of constant price for an indi-
vidual good was very long. It was approxi-
changes. Iwai, Rotemberg, and Caplin and
Leahy all respectively assume that there ismately
a 80 months.83 Such constancy of prices
"menu" cost in making these changes known.80 can easily be explained by customer norms re-
But the physical costs of making such changes,garding price change. The customers have a
as in the printing of new menus, are triviallynotion of the price that they ought to pay at
small. Norms regarding price changes, how- stores where they are continued and knowing
ever, give an alternative reason why these costs customers. Kackmeister suggests that the de-
might--indeed-be of sufficient size to induce cline in long-term customer relationships is one
a significant long-run trade-off between infla-factor responsible for greater frequency of price
tion and unemployment. Customers may think change today.
that firms should not raise prices. In that case, Emi Nakamura and J6n Steinsson (2005)
give an economic reason why customers would
price increases (or increases of greater size) are
likely to induce angry customers to search for have such a norm that firms should not change
alternative suppliers. At higher steady-state in-prices. They view consumer purchases as habit-
flation, firms will be changing their nominalforming. Thus, by buying a particular brand, or
prices more, and therefore will face more elasticpatronizing a particular store, consumers are
demands for their product. Producers' naturalputting themselves in a position where they can
microeconomic response to this increased elas-be exploited. Their loyalty puts the firm in a
ticity-a lower price for their product-will
produce a macroeconomic trade-off between in-
81 See also Blinder and Don Choi (1990) and Blinder et
flation and aggregate demand.
al. (1998).
Just as sticky money wages indicated that 82 The meaning of customer markets was especially ex-
employees have norms regarding wage change, plored by Arthur Okun (1978).
similarly, sticky prices indicate that customers 83 derive this result from Kackmeister's data in the
have norms regarding price change. Thus, thefollowing way. He finds that in the nineteenth century, only
5 percent of items changed their prices per month. This
means that the average spell of constant prices would have
been 20 months (the inverse). But that is a biased statistic
for the average length of time between price changes for an
have notions regarding what nominal wage increases should
or should not be. This, of course, is just one of many item on the shelf. The difference between the average spell
anomalies in the form of indexed contracts. of employment or unemployment and the average spell
80oMarika Karanassou, Hector Sala, and Dennis J. being experienced by an individual suggests a rule of thumb
Snower (2003) find considerable long-run trade-off between ratio for four to one. Using this ratio as a rule of thumb
inflation and unemployment in a model with nominal price suggests that the spell between price changes averaged over
staggering and money growth. the individual items on the shelf would be 80 months.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 27

position where it can take advantage ofExpectations


VIII. Rational the Theory
consumer by raising prices. Firms then make an
implicit contract with their customers: Our discussion
theyof rational
will expectations pig-
not change their prices unjustifiably. gybacks onSince
our previous
such discussion of the nat-
an implicit contract is easier to ural make
rate. (and en-
force) regarding nominal prices According than real to rational
prices, expectations theory, in-
the implicit guarantee is in nominal sofar as the centralNa-
terms. bank changes the money
kamura and Steinsson have also discovered a supply systematically in response to employ-
phenomenon that suggests strikingly that firmsment conditions, the public will foresee that
do behave this way. Goods in store 126 (chosenresponse and change prices and wages exactly
to compensate. The public's anticipation will
for its completeness of data) of Dominicks Finer
Foods chain frequently go on sale; when the then exactly offset the response. Monetary pol-
sale ends, their nominal price returns to the icy is neutral.85
exact same level. Such behavior is consistent There are two key assumptions underlying
this neutrality. The obvious one is rational ex-
with the view that consumers think that prices
should not change (for whatever reason); and pectations. To some, rational expectations re-
that they are also likely to retaliate (change garding the effects of the money supply on
brands) when prices do change. prices and wages would seem to be beyond the
I should also remark that in countries where sophistication of most wage and price takers,
inflation is very high, customers will expect and also of most wage and price setters.
price changes to occur frequently, and possibly Even in the case where all those involved in
be of large magnitude. The inhibitions against buying and selling goods and labor services
price changes when inflation is low are eroded have rational expectations, however, the neu-
at high inflation. Thus, while norms concerning trality results of rational expectations theory
prices give a negative long-run trade-off be- require also that nominal considerations do not
tween inflation and unemployment at low infla- enter into the setting of either wages or prices.
tion, at high inflation that trade-off could very The previous descriptions of the ways in which
well be reversed. nominal wages and prices enter into preference
functions, via employees' views of the wages
D. Summary that ought to be received and consumers' views
of the prices that ought to be paid, give further
To summarize, there is considerable evidence reason why the neutrality results of rational
of violation of the assumptions and predictions expectations will be violated. If prices and
of natural rate theory. Wages and prices are wages are affected by people's notions of what
nominally rigid; there were no deflationary spi- their nominal values should be, monetary policy
rals in the Great Depression; and questionnaire can be effective in stabilizing output-and pos-
respondents act as if they have a positive like for sibly in raising its long-run level-even in the
nominal wage increases.84 This evidence sug- presence of rational expectations.
gests that wage earners and customers have
views on what wages and prices should be. The IX. Economic Methodology
reflection of such views in utility functions pro-
duces trade-offs between inflation and unem- We have seen that the absence of norms plays
ployment. Those trade-offs have significanta key role in each of the five neutralities. Why
implications for economic policy. On the onehave economists made such systematic omis-
hand, central banks should avoid very low tar-sions? The omission of norms from macroeco-
gets for inflation. On the other hand, they shouldnomics, as well as from economics more
avoid high inflation, where the trade-offs be-
tween inflation and unemployment may be
reversed. 85 Empirically there is a theoretical puzzle of excess
sensitivity to monetary shocks (Lawrence J. Christiano,
Martin Eichenbaum, and Charles L. Evans 1998). Christina
and David Romer (1989) have shown that such a response
84 Also, COLA clauses are asymmetrically positive. See occurs with lags that would be surprisingly long if expected
footnote 79 above. monetary shocks were always neutralized.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
28 THE AMERICAN ECONOMIC REVIEW MARCH 2007

generally, models by
can be explained based only on
economists'
herence to positivevariables
economics.86 Friedm
verified only
(1953) essay on positive economics
testing-has severedescri
bias
the methodological implications
economic phenomena of such be w
In particular, he saysSummers
that economic (1986) illus
theori
this bias. The conventional test modeling.
should strive for parsimonious of the efficient A
cording to Friedman, marketstheyhypothesis-that
should stock prices are the for
even
realistic assumptionsexpected
invalue pursuit
of future returns-looks
of such for au- par
mony. tocorrelations
Maximization models of the excesswith
returns on stocks
only ob
tive arguments of utility
relative to bonds. have
Followingbeen
Summers, it define
would
more parsimonious than take approximately
models 5,000 years of data with
where peop
such a test
additionally, lose utility to obtain as much
insofar as as 50 percentor
they,
ers, fail to live up to rejection
their of anstandards.
alternative model where As stock
a res
whatever the empirical prices are validity
more than 30 percent oraway relevanc
from their
such norms, positive economics
fundamentals 35 percent of the has a meth
time. With such
ological bias against lack
their of power, nulls are important. When they
consideration. It p
ileges models without are not rejected, alternative theories, such as
norms.
The prescriptions of those positive
with norms, are noteconomics
even considered. This re
garding the conduct lecture
of has empirical
illustrated such reversion investiga
to norm-
compound the biasless against
nulls. Consumption norms. Fried
behavior, investment
says that economists behavior,
should and wage andnot price pay
behavior-the
heedthree to
stated intentions of most decision
important componentsmakers,of most macro mod- wh
would include their els-all
norms display excessas to how
sensitivity relative tothey
re-
others should behave. spectiveInstead,
neutralities. All ofempirical
these violations could wo
should test only hypotheses
be easily explained by norms. that econom
Yet in each case
consider to be basedeconomists have sought to explain such viola-
on parsimonious mode
If economic tests tionshad great
of classical theory bypower,
norm-less models. then
would be easy, of course, In contrast to toreliance on statistical Friedm
follow testing,
dictum of making more disciplines and
other than economics typically
more refined put t
of hypotheses much
with greater weight on a naturalistic
decreasing parsimon ap-
proach. This
norms really do affect approach involves this
behavior, detailed case
meth
would reject modelsstudies.
withoutSuch observation of the small often
norms and in
course would arrive at the
has been models where
key to the understanding of peo
the
views regarding howlarge.
they To me,should
the most dramatic
behave example aff
of
decision such a relation
But making. between the small tests
economic and the lac
power. All economiclarge occurs in the
models are structure
very of life itself
impre
Francis
in their specification ofCrick the andindependent
James D. Watson87 conjec v
able, the nature of tured correctly
the dependentthat if theyvariables,
could describe th
nature of
and lags, leads
crystalline andofthe
structure a singlenatur
DNA mole-
cule,
residuals. Yet worse, they would
most have unlocked the
economic secret o
proble
involve simultaneitylife.
(asThein
duality between the
supply structure
and demaof the
making establishmentDNA molecule and the way in
of causality which organ-
difficul
almost any isms are
instance, suchgenerated and reproduced
a large numbeis one o
models can be thestatistically
fitted most beautiful findingsthat
of humanit
knowl-
is
tremely hard-and perhaps impossible-to
edge. It indicates the sense in which Crick and
tistically reject allWatson
the variants
were, of
indeed, profoundly mod
correct.
without norms. As What a result, thefor
are the implications program
social science?
positive economics-with its
Positive economics, initial
with its emphasis nulls
on statis-
tical analysis of populations, would suggest tha

86 Some of the thoughts and wording in this section have


been presented in Akerlof (2005). 87 As dramatically described by Watson (1969).

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. I AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 29

the intensive study of a single economics.


moleculeIn contrast,
would a more naturalistic ap-
be an all-but-worthless anecdote. In the case of proach would prescribe a different methodol-
DNA, we know that the exact opposite is true: ogy. In this case, economists would observe
because DNA is a template that determines all decision makers as closely as possible, with the
of the cells of the organism, and also its repro- express intent of characterizing their motiva-
duction, one molecule may not tell all, but it tion, and would use such characterization as the
does tell a great deal. Form follows function. basis for modeling of economic structure. In-
Is there some reason to believe that economic deed, sociological and anthropological ethnog-
behavior and economic units are any different? raphers do precisely that: they depict their
Economic decisions may not be as duplicable as subjects' motivation from close observation.
biological processes, but the basic reason why
science intensively studies the microscopic ap- X. Endogeneity of Norms
plies to economics as well. The individual eco-
nomic unit, be it a firm, a consumer, or an It is now time to discuss the endogeneity of
employee, behaves the way it doesfor a reason. the norms. There is a special reason for its
And if these actors behave as they do for a consideration. Robert Lucas discovered that,
reason, we can expect to find those reasons from with endogenous rational expectations re-
the structures that we see in close observation; garding inflation, monetary policy that was
and because of those structures their behavior intended to stabilize the macroeconomy
will also tend to be duplicated. This dualitywould, instead, be exactly neutral. Similarly,
between duplicability and structure explains is it not possible that endogeneity of the
why much of science concerns very close ob- norms, like Lucas's endogeneity of inflation-
servation, as it also explains why the study of ary expectations, will cause the neutralities
even a single part of a single DNA molecule again to hold? We shall discuss this question
will be revealing. regarding all five neutralities. For the most
Standard economic methodology says that part, we find that the type of government
it is impossible to infer motivation of individ- interventions being considered are usually of
ual actors from intensive case studies. An- such frequency, or of such order of magni-
thropologists and sociologists listen carefully
tude, that they should provoke relatively little
change in the norms. Endogeneity of the
to individuals in such studies. When people
norms should have little effect, then, on our
follow the norms, they use them to explain
previous conclusions.
their actions; when, on the other hand, they
violate the norms, they become the subject of
local gossip. Those case studies are revealing A. Ricardian Equivalence
because-like a language, which dictates how
one should speak-the norms are common Let's begin by returning to Ricardian equiv-
knowledge. In this lecture, we have seen one alence, which is still the simplest case. We
prominent example of the use of such knowl-
found that if people have a norm regarding the
edge: Bewley's interviews uncovered the amount of their bequest, then lump-sum trans-
common understanding of the norms regard- fers to an older generation will not be neutral.
ing wage cuts among Connecticut employers There remains the possibility that the source of
in the early 1990s. the warm glow to the older generation is not the
total bequest, but instead the bequest to the
Summary. -Positive economics systemati- younger generation net of the transfer. In this
cally denies that norms can be understood from case, if the transfers change, then the norm
intensive case study. Precedence given to mod- changes. Ricardian equivalence will again be
els without norms because they are by definition valid. While such changes in norms with the
more parsimonious and statistical tests of low size of transfers are a theoretical possibility,
power then jointly create a firewall against con- they also seem highly unlikely. The size of the
sideration that norms play a role in determining transfers involved--especially for those rich
behavior. For these reasons, current economic enough to make large nonaccidental bequests-
methodology inherently has created a biased would seem to be too small to warrant such a

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
30 THE AMERICAN ECONOMIC REVIEW MARCH 2007

powerful evidence
sophisticated calculation. Ourinearlier
favor of normsdiscussi
comes
did discuss at least from
one employees'
change resistance in
to money
the wage nor
cuts and
regarding bequests, but thatcustomers' resistance to from
resulted nominal a ve
price increases.
large change in people's As long as inflation
orientation. It is low, it
result
from changes in theiris conception
doubtful that small changes
of the in inflation
family will
of their own place within affect suchitnorms.
and People
ofseemthe to find
placeit
their heirs. That alsoeasier to think in nominal,
occurred over rather
a than
very in real,
lo
run-over the course of centuries. terms. Indeed the facilitation of such thinking
is one of the benefits of money according to
B. Life-Cycle Hypothesis the textbook mantra on its three uses: for
transactions, as a store of value, and as a unit
Regarding the life-cycle hypothesis, we ar- of account. Money is useful as a unit of
gued that consumption depends upon current account especially if people think in nominal,
income because norms regarding how much rather than in real, terms. As a result, as long
people think they should spend are linked toas inflation is low, people are unlikely to
it. But such a norm Wyould be highly unlikelyforsake making calculations in nominal terms,
to change as a result'of the use of fiscal andespecially regarding the norms of what wages
monetary policy for stabilization. In the first or prices should be. Of course, if inflation
place, such stabilization will make the adher-increases to high levels, the norms for wages
ence to the norm less costly, not more costly, and prices and the method of calculating
in purely economic terms. Furthermore, mac- those norms will change. Exactly how they
roeconomic sources are responsible for only a change-with the possibility that they under-
small fraction of the variation in individual adjust to increases in inflation when it is low
and overadjust when it is high-should be
incomes. As a result, there is further reason
why the role of current income in norms empirically
is investigated.
unlikely to change as a result of macroeco-
nomic stabilization. E. Where Do the Norms Come From?

C. Cash Flow and Investment We do not know the general answer to the
question where norms come from. This lecture
The rise of the CFO suggests that norms has tried to make the case that norms, such as
regarding investment have changed in large US they are, could potentially play an important
role in macroeconomics. Hopefully, then, it has
firms. Quite possibly, this change occurred be-
cause firms realized the need for financial con- added to the motivation for research on their
trols that compared the returns on inside and microfoundations.88
outside options. Such an endogenous response
would make Modigliani-Miller correct. But, XI. Conclusion

following Zorn (2004), this change took 40


years. In the meantime, in the short run, follow-This lecture has shown that the early Keyne-
ing our earlier logic, investments would have sians got a great deal of the working of the
depended on cash flow. And, of course, even in economic system right, in ways that are denied
the long run the CFO, who is only one voice by the five neutralities. As quoted from Keynes
among many in corporate decisions, may not be earlier, they based their models on "our knowl-
fully effective. edge of human nature and from the detailed
facts of experience." They used their intuitions
D. Natural Rate Hypothesis and the Role of regarding the norms of how consumers, inves-
Rational Expectations tors, and wage and price setters thought they

Regarding the natural rate hypothesis and


88 This lecture has been very much influenced by the
also the rational expectations hypothesis, we
insights of the Ph.D. thesis of Robert Akerlof (2006) on
saw that they will no longer hold if norms of preferences for beliefs. His thinking on this subject has
price and wage setting have nominal compo- influenced many of the sections of this paper, especially on
nents. Regarding prices and wages, the most consumption and the endogeneity of norms.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 31

Akerlof,reason
should behave. There is systematic George A.,
whyed. 2005. Explorations in
such knowledge and experiencePragmatic
are likelyEconomics:
to be Selected Papers of
George
accurate: by their nature, norms areA.generated
Akerlof (and Co-Authors). Oxford
and known by a whole community. They
and New York: are
Oxford University Press.
known to those who abide byAkerlof,
them, George
and A.,those
William T. Dickens, and
who observe them as well. George L. Perry. 1996. "The Macroeconom-
We have shown ways in which macroeconomic ics of Low Inflation." Brookings Papers on
variables will be affected by norms. The neutral- Economic Activity, 1: 1-59.
ities say that consumption should have no special Akerlof, George A., William T. Dickens, and
dependence on current income; investment should George L. Perry. 2000. "Near-Rational Wage
be independent of current cash flow; wages and and Price Setting and the Long-Run Phillips
prices should not depend on nominal consider- Curve." Brookings Papers on Economic Ac-
ations. The very construction of those neutralities tivity, 1: 1-60.
denies the possibility that peoples' decisions Akerlof, George A., and Rachel E. Kranton. 2000.
might be influenced by their views regarding how "Economics and Identity." Quarterly Journal
they, and how others, should behave. In practice, of Economics, 115(3): 715-53.
however, the neutralities are systematically vio- Akerlof, George A., and Rachel E. Kranton. 2002.
lated. Insofar as economists have felt it necessary "Identity and Schooling: Some Lessons for
to explain these violations, they have appealed to the Economics of Education." Journal of
a variety of different frictions, such as myopia and Economic Literature, 40(4): 1167-1201.
credit constraint. In so doing, they have failed to Akerlof, George A., and Rachel E. Kranton. 2005.
consider that those violations would occur even in "Identity and the Economics of Organiza-
the absence of those frictions: they will occur tions." Journal of Economic Perspectives,
because of decision makers' norms. 19(1): 9-32.
The incorporation of norms based on careful Akerlof, George A., and Rachel E. Kranton. 2006.
observation imparts an appropriate balance to "The Missing Motivation: Economics Made
macroeconomics. The New Classical research Human." Unpublished.
program was correct in viewing models of the Akerlof, Robert J. 2006. "A Theory of Social
early Keynesians as too primitive. They had notMotivation." Unpublished.
Altonji, Joseph G., and Paul J. Devereux. 1999.
been sufficiently attentive to the role of human
intent in choices regarding consumption, invest- "The Extent and Consequences of Downward
ment, wages, and prices. But that research pro-Nominal Wage Rigidity." National Bureau of
Economic Research Working Paper 7236.
gram itself has failed to appreciate the extent to
which the Keynesians' views of macroeconom- Andreoni, James. 1989. "Giving with Impure
ics were also reflective of reality, since they Altruism: Applications to Charity and Ricar-
were based on experience and observation. dian Equivalence." Journal of Political Econ-
A macroeconomics with norms in decision omy, 97(6): 1447-58.
makers' objective functions combines the best Banks, James, Richard Blundell, and Sarah Tanner.
features of the two approaches. It allows for 1998. "Is There a Retirement-Savings Puzzle?"
observations regarding how people think theyAmerican Economic Review, 88(4): 769-88.
should behave. It also takes due account of the Barro, Robert J. 1974. "Are Government Bonds
purposefulness of human decisions. Net Wealth?" Journal of Political Economy,
82(6): 1095-1117.
REFERENCES Barro, Robert J. 1989. "The Ricardian Ap-
proach to Budget Deficits." Journal of Eco-
Abel, Andrew B. 1979. Investment and the Value nomic Perspectives, 3(2): 37-54.
of Capital. New York: Garland Publishing. Barro, Robert J., and David B. Gordon. 1983. "A
Ackley, Gardner. 1961. Macroeconomic Theory. Positive Theory of Monetary Policy in a Nat-
New York: Macmillan. ural Rate Model." Journal of Political Econ-
Agell, Jonas, and Per Lundborg. 2003. "Survey omy, 91(4): 589-610.
Evidence on Wage Rigidity and Unemploy- Bauer, Thomas, Holger Bonin, and Uwe Sunde.
ment: Sweden in the 1990s." Scandinavian 2003. "Real and Nominal Wage Rigidities
Journal of Economics, 105(1): 15-29. and the Rate of Inflation: Evidence from Ger-

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
32 THE AMERICAN ECONOMIC REVIEW MARCH 2007

dence." In NBER
man Microdata." Institute forMacroeconomics
the StudyAnnual of
Labor Discussion Paper
1989, ed. 959.
Olivier J. Blanchard and Stanley
Becker, Gary S. 1957. The
Fischer, Economics
185-216. of D
Cambridge, MA, and Lon-
crimination. Chicagodon: MIT Press.
and London: Universit
of Chicago Press. Cannadine, David N. 1977. "Aristocratic Indebt-
Behrman, Jere R., and Mark R. Rosenzweig. edness in the Nineteenth Century: The Case
2004. "Parental Allocations to Children: New Re-Opened." Economic History Review Sec-
Evidence on Bequest Differences among Sib- ond Series, 30(4): 624-50.
lings." Review of Economics and Statistics,Caplin, Andrew S., and John Leahy. 1991.
86(2): 637-40. "State-Dependent Pricing and the Dynamics
Benedict, Ruth. 1934. Patterns of Culture. Bos- of Money and Output." Quarterly Journal of
ton: Houghton Mifflin. Economics, 106(3): 683-708.
Bernheim, B. Douglas, and Kyle Bagwell. 1988. Caplow, Theodore. 1984. "Rule Enforcement
"Is Everything Neutral?" Journal of Political without Visible Means: Christmas Giving in
Economy, 96(2): 308-38. Middletown." American Journal of Sociol-
Bernheim, B. Douglas, Jonathan Skinner, and ogy, 89(6): 1306-23.
Steven Weinberg. 2001. "What Accounts for Card, David. 1986. "An Empirical Model of
the Variation in Retirement Wealth among Wage Indexation Provisions in Union Con-
U.S. Households?" American Economic Re- tracts." Journal of Political Economy, 94(3):
view, 91(4): 832-57. S144-75.
Bewley, Truman F. 1999. Why Wages Don't Card, Fall David, and Dean Hyslop. 1997. "Does In-
During a Recession. Cambridge, MA, and flation Grease the Wheels of the Labor Mar-
London: Harvard University Press. ket?" In Reducing Inflation: Motivation and
Blanchard, Olivier Jean, Florencio Lopez-de- Strategy, ed. Christina D. Romer and David
Silanes, and Andrei Shleifer. 1994. "What Do H. Romer, 71-114. Chicago and London:
Firms Do with Cash Windfalls?" Journal of University of Chicago Press.
Financial Economics, 36(3): 337-60. Carlton, Dennis W. 1986. "The Rigidity of
Blinder, Alan S. 1975. Toward an Economic Prices." American Economic Review, 76(4):
Theory of Income Distribution. Cambridge, 637-58.
MA, and London: MIT Press. Cassino, Vincenzo. 1995. "The Distribution of
Blinder, Alan S., Elie R. D. Canetti, David E. Wage and Price Changes in New Zealand."
Lebow, and Jeremy B. Rudd. 1998. Asking Bank of New Zealand Discussion Paper
About Prices: A New Approach to Under- G95/6.
standing Price Stickiness. New York: Russell Castellanos, Sara G., Rodrigo Garcia-Verd6i, and
Sage Foundation. David S. Kaplan. 2004. "Nominal Wage Ri-
Blinder, Alan S., and Don H. Choi. 1990. "A gidities in Mexico: Evidence from Social Se-
Shred of Evidence on Theories of Wage curity Records." National Bureau of
Stickiness." Quarterly Journal of Economics, Economic Research Working Paper 10383.
105(4): 1003-15. Chapple, Simon. 1996. "Money Wage Rigidity
Bourdieu, Pierre. 1984. Distinction: A Social in New Zealand." Labour Market Bulletin, 2:
Critique of the Judgment of Taste. Cam- 23-50.
bridge, MA, and London: Harvard University Chetty, Raj, and Adam Szeidl. 2006. "Con-
Press. sumption Commitments and Risk." Na-
Browning, Christopher R. 1992. Ordinary tional Bureau of Economic Research
Men: Reserve Police Battalion 101 and the Working Paper 12467.
Christiano, Lawrence J., Martin Eichenbau
Final Solution in Poland. New York: Harper-
Collins. and Charles L. Evans. 1998. "Monetary
Bruner, Robert F. 2002. "Does M&A Pay? Aicy Shocks: What Have We Learned an
What End?" National Bureau of Economic
Survey of Evidence for the Decision-Maker."
Journal of Applied Finance, 12(1): 48-68. Research Working Paper 6400.
Campbell, John Y., and N. Gregory Mankiw.Christofides, Louis N., and Amy Chen Peng.
1989. "Consumption, Income, and Interest 2004. "The Determinants of Major Provi-
Rates: Reinterpreting the Time Series Evi-sions in Union Contracts: Duration, Index-

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 33

ation, and Non-Contingent Wage Friedman, Milton. 1957. A Theory of the Con-
Adjustment." Unpublished. sumption Function. Princeton: Princeton Uni-
Dernburg, Thomas F., and Duncan M. McDou- versity Press.
gall. 1963. Macroeconomics: The Measure-Friedman, Milton. 1968a. Dollars and Deficits.
ment, Analysis, and Control of Aggregate New York: Prentice Hall.
Economic Activity. 2nd ed. New York, Auck-Friedman, Milton. 1968b. "The Role of Mone-
land, London, and Montreal: McGraw-Hill. tary Policy." American Economic Review,
Dornbusch, Rudiger, and Stanley Fischer. 1987. 58(1): 1-17.
Macroeconomics. 4th ed. New York, Auck-Goffman, Erving. 1961. Encounters. Indianapo-
land, London, and Montreal: McGraw-Hill. lis: Bobbs-Merrill.
Dwyer, Jacqueline, and Kenneth Leong. 2000. Gordon, Robert J. 1977. "Can the Inflation of
"Nominal Wage Rigidity in Australia." Re- the 1970s Be Explained?" Brookings Papers
serve Bank of Australia Discussion Paper on Economic Activity, 1: 253-79.
2000-08. Gray, Jo Anna. 1978. "On Indexation and Con-
tract Length." Journal of Political Economy,
Ehrenberg, Ronald G., Leif Danziger, and Gee San.
1983. "Cost-of-Living Adjustment Clauses 86(1): 1-18.
in Union Contracts: A Summary of Re- Guiso, Luigi, Paola Sapienza, and Luigi Zingales.
sults." Journal of Labor Economics, 1(3): 2003. "People's Opium? Religion and Eco-
215-45. nomic Attitudes." Journal of Monetary Eco-
nomics, 50(1): 225-82.
Elster, Jon. 1989. "Social Norms and Economic
Theory." Journal of Economic Perspectives,
Guiso, Luigi, Paola Sapienza, and Luigi Zingales.
3(4): 99-117. 2006. "Does Culture Affect Economic Out-
Fazzari, Steven M., R. Glenn Hubbard, and Bruce
comes?" Journal of Economic Perspectives,
20(2): 23-48.
C. Petersen. 1988. "Financing Constraints and
Corporate Investment." Brookings Papers Hanes,
on Christopher. 2000. "Nominal Wage Ri-
Economic Activity, 1: 141-206. gidity and Industry Characteristics in the
Fehr, Ernst, and Lorenz Goette. 2003. "Robust-
Downturns of 1893, 1929, and 1981." Amer-
ness and Real Consequences of Nominal
ican Economic Review, 90(5): 1432-46.
Wage Rigidity." University of Zurich Insti-
Hayashi, Fumio. 1982. "Tobin's Marginal q and
tute for Empirical Research in Economics
Average q: A Neoclassical Interpretation."
Working Paper 44. Econometrica, 50(1): 213-24.
Feldstein, Martin S. 1974. "Social Security,Hodson,
In- Randy. 2001. Dignity at Work. Cam-
duced Retirement, and Aggregate Capital Ac- bridge: Cambridge University Press.
cumulation." Journal of Political Economy,Holton, Woody. 1999. Forced Founders: In-
82(5): 905-26. dians, Debtors, Slaves, and the Making of
Feldstein, Martin S., and Anthony Pellechio. the American Revolution in Virginia.
1979. "Social Security and Household Chapel Hill: University of North Carolina
Wealth Accumulation: New Microeconomet- Press.
ric Evidence." Review of Economics and Sta-Homans, George C., and Charles P. Curtis. 1934.
tistics, 61(3): 361-68. An Introduction to Pareto, His Sociology.
Fligstein, Neil. 1990. The Transformation of New York: Knopf.
Corporate Control. Cambridge, MA, and Hurd, Michael D. 1989. "Mortality Risk and
London: Harvard University Press. Bequests." Econometrica, 57(4): 779-813.
Fortin, Pierre. 1996. "The Great Canadian Iwai, Katsuhito. 1981. Disequilibrium Dynam-
Slump." Canadian Journal of Economics. ics: A Theoretical Analysis of Inflation and
29(4): 761-87. Unemployment. New Haven and London:
Friedan, Betty. 1963. The Feminine Mystique. Yale University Press.
New York: W.W. Norton and Co. Jensen, Michael C. 1986. "Agency Costs of Free
Friedman, Milton. 1953. "The Methodology of Cash Flow, Corporate Finance, and Take-
Positive Economics." In Essays in Positive overs." American Economic Review (Papers
Economics, ed. Milton Friedman, 3-43. Chi- and Proceedings), 76(2): 323-29.
cago and London: University of Chicago Jensen, Michael C. 1993. "The Modern Indus-
Press. trial Revolution, Exit, and the Failure of In-

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
34 THE AMERICAN ECONOMIC REVIEW MARCH 2007

Laitner, John,
ternal Control Systems." and Henry Ohlsson.
Journal of 2001.Finance
"Be-
48(3): 831-80. quest Motives: A Comparison of Sweden and
Kackmeister, Alan K. 2002.
the United States.""Has
Journal of Retail
Public Eco- Pr
Behavior Changed since nomics, 79(1):1889?
205-36. Evidence
from Micro Data." PhD Dissertation. Univer- Lamont, Owen. 1997. "Cash Flow and Invest-
sity of California, Berkeley. ment: Evidence from Internal Capital Mar-
Kahn, Shulamit. 1997. "Evidence of Nominal kets." Journal of Finance, 52(1): 83-109.
Wage Stickiness from Microdata." AmericanLebow, David E., Raven E. Saks, and Beth Anne
Economic Review, 87(5): 993-1008. Wilson. 1999. "Downward Nominal Wage
Kaplan, Steven N., and Luigi Zingales. 1997. "Do Rigidity: Evidence from the Employment
Investment-Cash Flow Sensitivities Provide Cost Index." U.S. Board of Governors of the
Useful Measures of Financing Constraints." Federal Reserve System, Finance and Eco-
Quarterly Journal of Economics, 112(1): nomics Discussion Paper 1999-31.
169-215. Lucas, Robert E., Jr. 1972. "Expectations and
Karanassou, Marika, Hector Sala, and Dennis J. the Neutrality of Money." Journal of Eco-
Snower. 2003. "A Reappraisal of the Inflation- nomic Theory, 4(2):103-24.
Unemployment Tradeoff." University of Lon- Lucas, Robert E., Jr., and Thomas J. Sargent.
don Working Paper 479. 1978. "After Keynesian Macroeconomics."
Keynes, John M. 1936. The General Theory of In After the Phillips Curve: Persistence of
Employment, Interest, and Money. New High Inflation and Unemployment, (Proceed-
York: Macmillan. ings of Conference 19). Boston: Federal Re-
Kimura, Takeshi, and Kazuo Ueda. 2001. serve Bank of Boston.
"Downward Nominal Wage Rigidity in Ja- Manicka, Rani. 2002. The Rice Mother. New
pan." Journal of the Japanese and Interna- York: Penguin.
tional Economies, 15(1): 50-67. Massey, Douglas S., and Nancy A. Denton. 1993.
Klein, Lawrence R., and Arthur S. Goldberger. American Apartheid: Segregation and the
1955. An Econometric Model of the United Making of the Underclass. Cambridge, MA,
States 1929-1952. Amsterdam: North-Holland. and London: Harvard University Press.
Knoppik, Christoph, and Thomas Beissinger. Meyer, John R., and Edwin Kuh. 1957. The In-
2003. "How Rigid Are Nominal Wages? Evi- vestment Decision: An Empirical Study.
dence and Implications for Germany." Scandi- Cambridge, MA, and London: Harvard Uni-
navian Journal of Economics, 105(4): 619-41. versity Press.
Kuroda, Sachiko, and Isamu Yamamoto. 2003a. Milgram, Stanley. 1963. "Behavioral Study of
"Are Japanese Nominal Wages Downwardly Obedience." Journal of Abnormal and Social
Rigid? (Part I): Examinations of Nominal Psychology, 67(4): 371-78.
Wage Change Distributions." Monetary and Milgram, Stanley. 1965. "Some Conditions on
Economic Studies, 21(2): 1-29. Obedience and Disobedience to Authority."
Kuroda, Sachiko, and Isamu Yamamoto. 2003b. In Current Studies in Social Psychology, ed.
"Are Japanese Nominal Wages Downwardly Ivan D. Steiner and Martin Fishbein. New
Rigid? (Part II): Examinations Using a Fric- York: Holt, Rinehart, and Winston.
tion Model." Monetary and Economic Stud- Modigliani, Franco, and Richard Brumberg.
ies, 21(2): 31-68. 1954. "Utility Analysis and the Consumption
Kuroda, Sachiko, and Isamu Yamamoto. 2003c. Function: An Interpretation of Cross-Section
"The Impact of Downward Nominal Wage Data." In Post-Keynesian Economics, ed.
Rigidity on the Unemployment Rate: Quan- Kenneth K. Kurihara, 388-436. New Bruns-
titative Evidence from Japan." Monetary and wick: Rutgers University Press.
Economic Studies, 21(4): 57-85. Modigliani, Franco, and Merton H. Miller. 1958.
Laibson, David I. 1997. "Golden Eggs and Hy- "The Cost of Capital, Corporation Finance
perbolic Discounting." Quarterly Journal of and the Theory of Investment." American
Economics, 112(2): 443-77. Economic Review, 48(3): 261-97.
Laitner, John. 2002. "Wealth Inequality and Al- Myers, Stewart C., and Nicholas S. Majluf. 1984.
truistic Bequests." American Economic Review "Corporate Financing and Investment Deci-
(Papers and Proceedings), 92(2): 270-73. sions When Firms Have Information That

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
VOL. 97 NO. 1 AKERLOF: THE MISSING MOTIVATION IN MACROECONOMICS 35

Investors Do Not Have." Journal of Finan-


Adjustment and Aggregate Output." Review
cial Economics, 13(2): 187-221.
of Economic Studies, 49(4): 517-31.
Samuelson,
Nakamura, Emi, and J6n Steinsson. 2005.Paul A. 1964. Economics: An Intro-
"Price-
Setting in Forward Looking Customer ductory Analysis.
Mar- 6th ed. New York, Auck-
kets." Unpublished. land, London, and Montreal: McGraw-Hill.
Nickell, Stephen, and Glenda Sargent, Quintini.
Thomas2001.
J. 1971. "A Note on the 'Ac-
"Nominal Wage Rigidity and celerationist' the Rate of
Controversy." Journal of Money,
Inflation." London School of Economics Dis- Credit, and Banking, 3(3): 721-25.
cussion Paper CEPDP0489. Sargent, Thomas J. 1973. "Rational Expecta-
O'Brien, Anthony Patrick. 1989. "A Behavioral tions, the Real Rate of Interest, and the Nat-
Explanation for Nominal Wage Rigidity Dur- ural Rate of Unemployment." Brookings
ing the Great Depression." Quarterly Journal Papers on Economic Activity, 2: 429-72.
of Economics, 104(4): 719-35. Seater, John J. 1993. "Ricardian Equivalence."
O'Driscoll, Gerald P., Jr. 1977. "The Ricardian Journal of Economic Literature, 31(1): 142-90.
Nonequivalence Theorem." Journal of Polit- Shafir, Eldar, Peter Diamond, and Amos Tversky.
ical Economy, 85(1): 207-10. 1997. "Money Illusion." Quarterly Journal
Okun, Arthur. 1978. Prices and Quantities: A of Economics, 112(2): 341-74.
Macroeconomic Analysis. Washington, DC: Shea, John. 1995. "Union Contracts and the
Brookings Institution Press. Life-Cycle/Permanent-Income Hypothesis."
Pareto, Vilfredo. 1980. Compendium of General American Economic Review, 85(1): 186-200.
Sociology. Abridged in Italian by Giulio Fa- Shefrin, Hersh M., and Meir Statman. 1984. "Ex-
rina from Trattato di Sociologia Generale. plaining Investor Preference for Cash Divi-
Translated by Elisabeth Abbott. Minneapolis: dends." Journal of Financial Economics,
University of Minnesota Press. 13(2): 253-82.
Parker, Jonathan A. 1999. "The Reaction of Shefrin, Hersh M., and Richard H. Thaler. 1988.
Household Consumption to Predictable Changes "The Behavioral Life-Cycle Hypothesis."
in Social Security Taxes." American Economic Economic Inquiry, 26(4): 609-43.
Review, 89(4): 959-73. Shiller, Robert J. 1997. "Why Do People Dislike
Phelps, Edmund S. 1968. "Money-Wage Dynam- Inflation?" In Reducing Inflation: Motivation
ics and Labor-Market Equilibrium." Journal of and Strategy, ed. Christina D. Romer and
Political Economy, 76(4) Part 2: 678-711. David H. Romer, 13-65. Chicago and Lon-
Phillips, Alban W. 1958. "The Relation between don: University of Chicago Press.
Unemployment and the Rate of Change ofSouleles, Nicholas S. 1999. "The Response of
Money Wage Rates in the United Kingdom, Household Consumption to Income Tax Re-
1861-1957." Economica, 25(100): 283-99. funds." American Economic Review, 89(4):
Ricciuti, Roberto. 2003. "Assessing Ricardian 947-58.
Equivalence." Journal of Economic Surveys, Staiger, Douglas, James H. Stock, and Mark W.
17(1): 55-78. Watson. 1997. "How Precise Are Estimates
Rogoff, Kenneth. 1985. "The Optimal Degree of of the Natural Rate of Unemployment?" In
Commitment to an Intermediate Monetary Reducing Inflation: Motivation and Strategy,
Target." Quarterly Journal of Economics, ed. Christina D. Romer and David H. Romer,
100(4): 1169-89. 195-242. Chicago and London: University of
Romer, Christina D., and David H. Romer. 1989. Chicago Press.
"Does Monetary Policy Matter? A New Test Stein, Jeremy C. 2003. "Agency, Information
in the Spirit of Friedman and Schwartz." In and Corporate Investment." In Handbook of
NBER Macroeconomics Annual 1989, ed. the Economics of Finance, ed. George M.
Olivier J. Blanchard and Stanley Fischer, Constantinides, Milton Harris and Rene M.
121-70. Cambridge, MA, and London: MIT Stulz, 111-65. Amsterdam, London, and
Press. New York: Elsevier, North-Holland.
Romer, David H. 2001. Advanced Macroeco- Stock, James H., and Mark W. Watson. 2003. "Has
nomics. New York, Auckland, London, and the Business Cycle Changed? Evidence and
Montreal: McGraw-Hill. Explanations" In Proceedings of a Symposium
Rotemberg, Julio J. 1982. "Monopolistic Price on Monetary Policy and Uncertainty: Adapting

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms
36 THE AMERICAN ECONOMIC REVIEW MARCH 2007

Personal Account
to a Changing Economy, 9-56.of the Kansas
Discovery of the
City
Federal Reserve Bank Structure
of Kansasof DNA. New York: City. New American
Stone, Lawrence. 1977. Library. The Family, Sex a
Marriage in England, 1500-1800. New Weber, Max. 1958. The Protestant Ethic and the
York: Harper and Row. Spirit of Capitalism. New York: Charles
Summers, Lawrence H. 1981. '"Taxation and Cor- Scribner' s.
porate Investment: A q-Theory Approach." Wilcox, David W. 1989. "Social Security Bene-
Brookings Papers on Economic Activity, 1: fits, Consumption Expenditure, and the Life
67-127. Cycle Hypothesis." Journal of Political
Summers, Lawrence H. 1986. "Does the Stock Economy, 97(2): 288-304.
Market Rationally Reflect Fundamental Val-Wilhelm, Mark 0. 1996. "Bequest Behavior and
ues?" Journal of Finance (Papers and Pro- the Effect of Heirs' Earnings: Testing the
ceedings of the Forty-Fourth Annual Meeting Altruistic Model of Bequests." American
of the American Finance Association), 41(3): Economic Review, 86(4): 874-92.
591-601. Woodward, Ian. 2003. "Divergent Narratives in
Terkel, Studs. 1972. Working: People Talk About the Imagining of the Home amongst Middle-
What They Do All Day and How They Feel Class Consumers: Aesthetics, Comfort, and
About What They Do. New York: Pantheon. the Symbolic Boundaries of Self and Home."
Tobin, James. 1969. "A General EquilibriumJournal of Sociology, 39(4): 391-412.
Approach to Monetary Theory." Journal Yellen,
of Janet L., and George A. Akerlof. 2006.
Money, Credit, and Banking, 1(1): 15-29. "Stabilization Policy: A Reconsideration."
Tversky, Amos, and Daniel Kahneman. 1981. Economic Inquiry, 44(1): 1-22.
"The Framing of Decisions and the Psychol-Zorn, Dirk M. 2004. "Here a Chief, There a
ogy of Choice." Science, 211(4481): 453-58. Chief: The Rise of the CFO." American So-
Watson, James D. 1969. The Double Helix: A ciological Review, 69(3): 345-64.

This content downloaded from 202.170.57.242 on Sat, 03 Sep 2016 13:09:12 UTC
All use subject to http://about.jstor.org/terms

You might also like