Economic Perspectives
Economic Perspectives
Economic Perspectives
Articles
Economic Perspectives
A journal of the
American Economic Association
The Journal of
Fall 2013
Perspectives
Symposia
The First 100 Years of the Federal Reserve
The Journal of
Fall 2013
The Journal of
Economic Perspectives
A journal of the American Economic Association
Editor
David H. Autor, Massachusetts Institute of Technology
Co-editors
Chang-Tai Hsieh, University of Chicago
Ulrike Malmendier, University of California at Berkeley
Associate Editors
Katherine Baicker, Harvard University
Benjamin G. Edelman, Harvard University
Raymond Fisman, Columbia University
Gordon Hanson, University of California at San Diego
Anil K Kashyap, University of Chicago
Adam Looney, Brookings Institution
David McKenzie, World Bank
Kerry Smith, Arizona State University
Chad Syverson, University of Chicago
Christopher Udry, Yale University
Managing Editor
Timothy Taylor
Assistant Editor
Ann Norman
Editorial offices:
Journal of Economic Perspectives
American Economic Association Publications
2403 Sidney St., #260
Pittsburgh, PA 15203
email: [email protected]
The Journal of Economic Perspectives gratefully acknowledges the support of Macalester College.
Registered in the US Patent and Trademark Office ().
Copyright 2013 by the American Economic Association; All Rights Reserved.
Composed by American Economic Association Publications, Pittsburgh, Pennsylvania, USA
Printed by R. R. Donnelley Company, Jefferson City, Missouri, 65109, USA
No responsibility for the views expressed by the authors in this journal is assumed by the editors or by
the American Economic Association.
THE JOURNAL OF ECONOMIC PERSPECTIVES (ISSN 0895-3309), Fall 2013, Vol. 27, No. 4. The JEP is
published quarterly (February, May, August, November) by the American Economic Association, 2014
Broadway, Suite 305, Nashville, TN 37203-2418. Annual dues for regular membership are $20.00, $30.00,
or $40.00 depending on income; for an additional $15.00, you can receive this journal in print. E-reader
versions are free. For details and further information on the AEA go to http://www.vanderbilt.edu/AEA/.
Periodicals postage paid at Nashville, TN, and at additional mailing offices.
POSTMASTER: Send address changes to the Journal of Economic Perspectives, 2014 Broadway, Suite 305,
Nashville, TN 37203.
Printed in the U.S.A.
The Journal of
Economic Perspectives
Contents
Symposia
The First 100 Years of the Federal Reserve
Ben S. Bernanke, A Century of US Central Banking: Goals, Frameworks,
Accountability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Ricardo Reis, Central Bank Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Gary Gorton and Andrew Metrick, The Federal Reserve and Panic Prevention:
The Roles of Financial Regulation and Lender of Last Resort . . . . . . 45
Julio J. Rotemberg, Shifts in US Federal Reserve Goals and Tactics for
Monetary Policy: A Role for Penitence? . . . . . . . . . . . . . . . . . . . . . . . . . 65
Barry Eichengreen, Does the Federal Reserve Care about the Rest of the
World? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Martin Feldstein, An Interview with Paul Volcker . . . . . . . . . . . . . . . . . . . . . . . 105
Economics and Moral Virtues
Michael J. Sandel, Market Reasoning as Moral Reasoning: Why Economists
Should Re-engage with Political Philosophy . . . . . . . . . . . . . . . . . . . . 121
Luigino Bruni and Robert Sugden, Reclaiming Virtue Ethics for Economics 141
Articles
Nico Voigtlnder and Hans-Joachim Voth, Gifts of Mars: Warfare and Europes
Early Rise to Riches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Benjamin Marx, Thomas Stoker, and Tavneet Suri, The Economics of Slums in
the Developing World . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Features
Timothy Taylor, Recommendations for Further Reading . . . . . . . . . . . . . . . . . 211
James Poterba, Steven Venti, and David Wise Correction: The Composition and
Drawdown of Wealth in Retirement . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Correspondence: Patents and the Dissemination of Inventions: Dean Alderucci
and William J. Baumol, with Michele Boldrin and David K. Levine . . 223
Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Statement of Purpose
The Journal of Economic Perspectives attempts to fill a gap between the general
interest press and most other academic economics journals. The journal aims to
publish articles that will serve several goals: to synthesize and integrate lessons
learned from active lines of economic research; to provide economic analysis of
public policy issues; to encourage cross-fertilization of ideas among the fields
of economics; to offer readers an accessible source for state-of-the-art economic
thinking; to suggest directions for future research; to provide insights and readings for classroom use; and to address issues relating to the economics profession.
Articles appearing in the journal are normally solicited by the editors and associate
editors. Proposals for topics and authors should be directed to the journal office,
at the address inside the front cover.
Policy on Disclosure
Authors of articles appearing in the Journal of Economic Perspectives are expected
to disclose any potential conflicts of interest that may arise from their consulting
activities, financial interests, or other nonacademic activities.
everal key episodes in the 100-year history of the Federal Reserve have
been referred to in various contexts with the adjective Great attached to
them: the Great Experiment of the Federal Reserves founding, the Great
Depression, the Great Inflation and subsequent disinflation, the Great Moderation,
and the recent Great Recession. Here, Ill use this sequence of Great episodes
to discuss the evolution over the past 100 years of three key aspects of Federal
Reserve policymaking: the goals of policy, the policy framework, and accountability
and communication. The changes over time in these threeareas provide a useful
perspective, I believe, on how the role and functioning of the Federal Reserve have
changed since its founding in 1913, as well as some lessons for the present and for
thefuture.
Ben S. Bernanke is Chairman of the Board of Governors of the Federal Reserve System,
Washington, DC.
http://dx.doi.org/10.1257/jep.27.4.3
doi=10.1257/jep.27.4.3
established to provide a means by which periodic panics which shake the American
Republic and do it enormous injury shall be stopped.1
At the time, the standard view of financial panics was that they were triggered
when the needs of business and agriculture for liquid funds outstripped the available
supply as when seasonal plantings or shipments of crops had to be financed, for
example and that panics were further exacerbated by the incentives of banks and
private individuals to hoard liquidity during such times (Warburg 1914). The new
institution was intended to relieve such strains by providing an elastic currency:
that is, by providing liquidity as needed to individual member banks through the
discount window. Commercial banks, in turn, would then be able to accommodate
their customers. Interestingly, although congressional advocates hoped the creation
of the Fed would help prevent future panics, they did not fully embrace the idea that
the Fed should help end ongoing panics by serving as lender of last resort, as had
been famously recommended by the British economist and writer Walter Bagehot
(1873 [1897]), the source of the classic dictum that central banks should address
panics by lending freely at a penalty rate (see also Willis 1923, p.1407; Carlson and
Wheelock 2012; Bordo and Wheelock 2013). Instead, legislators imposed limits on
the Federal Reserves ability to lend in response to panics, for example, by denying
nonmember banks access to the discount window and by restricting the types of
collateral that the Fed could accept.2
Soon after the Federal Reserve was founded in 1913, its mission shifted to
supporting the war effort and then to managing the unwinding of that support. The
year 1923 was thus one of the first in which the Federal Reserve confronted normal
peacetime financial conditions, and it took the opportunity to articulate its views on
the appropriate conduct of policy in such conditions in the Tenth Annual Report of
the Federal Reserve Board (Board of Governors 1924).
The framework that the Federal Reserve employed in these early years to
promote financial stability reflected in large measure the fact that the United States
was on the gold standard as well as the influence of the so-called real bills doctrine.3
A 1929 book review by the financial editor of the New York Times, making reference both to the idea
of a great experiment and to the broad responsibilities for financial stability of the new central bank,
observed: The Federal Reserve System has from the first necessarily been a great experiment, bound
to adjust its general policies to the requirements of such novel and varying situations as should arise in
the course of our financial history and which could not possibly be foreseen (Noyes 1929). To be sure, the
US Treasury carried out some central banking functions before the creation of the Federal Reserve, and
the First and Second Banks of the United States represented early attempts to establish a central bank. By
1913, however, it had been about 75years since the latter institution had ceased fulfilling that purpose.
Moreover, the Federal Reserve operated somewhat differently from the prior institutions, as well as from
existing central banks abroad, and thus its creation amounted to an experiment.
2
The collateral acceptable to be pledged to the discount window has been expanded significantly over
time; in particular, various pieces of banking legislation in the early 1930s enabled the Federal Reserve
to make advances to member banks so long as the loans were secured to the satisfaction of the Federal
Reserve Bank extending the loan. The Monetary Control Act of 1980 gave all depository institutions
access to the discount window.
3
Humphrey (1982) discusses the historical evolution of the real bills doctrine. He notes that, in its
simplest form, the doctrine contends that banks should lend against short-term commercial paper
Ben S. Bernanke
In the real bills doctrine, the Federal Reserve saw its function as meeting the needs
of business for liquidity consistent with the idea of providing an elastic currency
with the ultimate goal of supporting financial and economic stability. When business
activity was increasing, the Federal Reserve would seek to accommodate the need
for credit by supplying liquidity to banks; when business was contracting and less
credit was needed, the Fed would then reduce the liquidity in the system. The policy
framework of the Feds early years has been much criticized in retrospect. Economic
historians have pointed out that, under the real bills doctrine, the Fed increased
the money supply precisely at those times at which business activity and upward
pressures on prices were strongest; that is, monetary policy was procyclical. Thus,
the Feds actions tended to increase rather than decrease the volatility in economic
activity and prices (Friedman and Schwartz 1963; Humphrey 1982; Meltzer 2003).
As noted, the Federal Reserve pursued its real bills approach in the context of
the gold standard. In the 1920s, Federal Reserve notes were redeemable in gold on
demand, and the Fed was required to maintain a gold reserve equal to 40percent
of outstanding notes. In principle, the gold standard should limit discretion by
monetary policymakers, but in practice US monetary policy did not appear to be
greatly constrained in the years after the Feds founding. Indeed, the large size of the
USeconomy, together with the use of market interventions that prevented inflows
and outflows of gold from being fully translated into changes in the domestic money
supply, gave the Federal Reserve considerable scope during the 1920s to conduct
monetary policy according to the real bills doctrine without much hindrance from
the goldstandard.4
Ive discussed the original mandate and early policy framework of the Federal
Reserve. What about its accountability to the public? When the Federal Reserve was
established, the question of whether it should be a private or a public institution
was highly contentious. The compromise solution created a hybrid Federal Reserve
System. The system was headed by a federally appointed Board of Governors, which
initially included the Secretary of the Treasury and the Comptroller of the Currency.
However, the 12regional Reserve Banks were placed under a mixture of public and
private oversight, including board members drawn from the private sector, and they
associated with real business transactions (as opposed to other activities such as speculative investment). According to this doctrine, central banks should expand the money supply to facilitate this type
of bank lending, by buying commercial paper from banks or accepting as collateral banks holdings of
such paper. Thus, the doctrine implies that the money supply should expand and contract along with
businessactivity.
4
Specifically, the Fed was able to sterilize the effects of gold flows on the domestic money supply through
open market operations the purchase and sale of government securities in the open market. Initially,
the Feds main tools were the quantity of its lending through the discount window and the interest rate at
which it lent the discount rate. Open market operations were discovered when, to generate earnings
to finance its operations, the Federal Reserve began in the 1920s to purchase government securities.
Fed officials soon found that these operations affected the supply and cost of bank reserves and, consequently, the terms on which banks extended credit to their customers. Subsequently, of course, open
market operations became a principal monetary policy tool, one that allowed the Fed to interact with the
broader financial markets, not only with banks (Strong 1926).
were given considerable scope to make policy decisions that applied to their own
districts. For example, Reserve Banks were permitted during this time to set their
own discount rates, subject to a minimum set by the Board of Governors.
While the founders of the Federal Reserve hoped that this new institution
would provide financial and hence economic stability, the policy framework and
the institutional structure would prove inadequate to the challenges the Fed would
soonface.
example, in DeLong 1990). It may be that the Federal Reserve suffered less from
lack of leadership in the 1930s than from the lack of an intellectual framework for
understanding what was happening and what needed to be done.
The Feds inadequate policy framework ultimately collapsed under the weight
of economic failures, new ideas, and political developments. The international
gold standard was abandoned during the 1930s. The real bills doctrine lost prestige after the disaster of the 1930s; for example, the Banking Act of 1935 amended
section12A(c) of the Federal Reserve Act so as to instruct the Federal Reserve to use
open market operations with consideration of the general credit situation of the
country, not just to focus narrowly on short-term liquidity needs. The Congress also
expanded the Feds ability to provide credit through the discount window, allowing
loans to a broader array of counterparties, secured by a broader variety of collateral.6
The experience of the Great Depression had major ramifications for all three
aspects of the Federal Reserve Iam discussing: its goals, its policy framework, and
its accountability to the public. With respect to goals, the high unemployment of
the Depressionand the fear that high unemployment would return after World
War II elevated the maintenance of full employment as a goal of macroeconomic policy. The Employment Act of 1946 made the promotion of employment
a general objective for the federal government. Although the Fed did not have a
formal employment goal until the Federal Reserve Reform Act of 1977 codified
maximum employment, along with stable prices, as part of the Feds so-called
dual mandate, earlier legislation nudged the central bank in that direction.7 For
example, legislators described the intent of the Banking Act of 1935 as follows:
Toincrease the ability of the banking system to promote stability of employment
and business, insofar as this is possible within the scope of monetary action and
credit administration (US Congress 1935). At the same time, the Federal Reserve
became less focused on its original mandate of preserving financial stability,
perhaps in part because it felt superseded by the creation during the 1930s of the
Federal Deposit Insurance Corporation and the Securities and Exchange Commission, along with other reforms intended to make the financial system more stable.
In the area of governance and accountability to the public, policymakers also
recognized the need for reforms to improve the Federal Reserves structure and
decision-making. The Banking Act of 1935 simultaneously bolstered the legal independence of the Federal Reserve and provided for stronger central control by the
Federal Reserve Board. In particular, the act created the modern configuration of
For example, section10B enhanced the powers of the Federal Reserve to lend to member banks, and
sections13(3) and 13(13) enabled the Federal Reserve to provide short-term credit to a wide range of
potential borrowers in specific circumstances.
7
More precisely, the three statutory objectives for monetary policy set forth in the Federal Reserve
Reform Act of 1977 are maximum employment, stable prices, and moderate long-term interest rates.
The dual mandate refers to the first twogoals, and the long-term interest rate goal is viewed as likely
to emerge from the macroeconomic environment associated with achievement of the employment and
price stability goals (Mishkin 2007). Thus, the interest rate goal of the Federal Reserve Reform Act can
be regarded as subsumed within the dual mandate.
the Federal Open Market Committee (FOMC), giving the Board the majority
of votes on the Committee, while removing the Secretary of the Treasury and the
Comptroller of the Currency from the Board. In practice, however, the USTreasury
continued to have considerable sway over monetary policy after 1933, with Meltzer
(2003) describing the Fed as in the back seat. During World WarII, the Federal
Reserve used its tools to support the war financing efforts by holding interest
rates and government borrowing costs low. Even after the war, Federal Reserve
policy remained subject to considerable Treasury influence. It was not until the
1951Accord with the Treasury that the Federal Reserve began to recover genuine
independence in setting monetary policy.
Ben S. Bernanke
The consequence of the monetary framework of the 1970s was two bouts of
double-digit inflation during that decade. Moreover, by the end of the decade, lack
of commitment to controlling inflation had clearly resulted in inflation expectations becoming unanchored, or unstable, with high estimates of trend inflation
embedded in longer-term interest rates.
Under the leadership of Chairman Paul Volcker, the Federal Reserve in 1979
fundamentally changed its approach to the issue of ensuring price stability. This
change involved an important rethinking on the part of policymakers. By the end of
the 1970s, Federal Reserve officials increasingly accepted the view that inflation is a
monetary phenomenon, at least in the medium and longer term; they became more
alert to the risks of excessive optimism about the economys potential output; and
they placed renewed emphasis on the distinction between realthat is, inflationadjusted and nominal interest rates (for discussion, see Meltzer 2009b). The
change in policy framework was initially tied to a change in operating procedures
that put greater focus on growth in bank reserves, but the critical changethe willingness to respond more vigorously to inflationendured even after the Federal
Reserve resumed its traditional use of the federal funds rate as the policy instrument
(Axilrod 1982). The new regime also reflected an improved understanding of the
importance of providing a firm anchor for the inflation expectations of the private
sector, secured by the credibility of the central bank.8 Finally, it entailed a changed
view about the dual mandate, in which policymakers regarded achievement of price
stability as helping to provide the conditions necessary for sustained maximum
employment (Lindsey, Orphanides, and Rasche 2005).
10
analysis on variants of the so-called Modigliani and Miller (1958) theorem, which
shows that under a number of restrictive assumptions the value of a firm is
not related to how that firm is financed.9 Influenced by the logic of Modigliani
Miller, many monetary economists and central bankers concluded that the details
of the structure of the financial system could be largely ignored when analyzing the
behavior of the broadereconomy.
An important development of the Great Moderation was the increasing emphasis
that central banks around the world put on communication and transparency, as
economists and policymakers reached consensus on the value of communication
in attaining monetary policy objectives (Woodford 2005). Federal Reserve officials,
like those at other central banks, had traditionally been highly guarded in their
public pronouncements. They believed, for example, that the ability to take markets
by surprise was important for influencing financial conditions (for example,
Goodfriend 1986; Cukierman and Meltzer 1986). Although Fed policymakers of the
1980s and early 1990s had become somewhat more explicit about policy objectives
and strategy (Orphanides 2006), the same degree of transparency was not forthcoming on monetary policy decisions and operations. The release of a post-meeting
statement by the Federal Open Market Committee, a practice that began in 1994,
was therefore an important watershed. Over time, these statements were expanded
to include more detailed information about the reason for the policy decision and
an indication of the balance of risks (Lindsey 2003).
In addition to improving the effectiveness of monetary policy, these developments in communications also enhanced the public accountability of the Federal
Reserve. Accountability is, of course, essential for continued policy independence
in a democracy. Moreover, central banks that are afforded policy independence in
the pursuit of their mandated objectives tend to deliver better economic outcomes
(Alesina and Summers 1993; Debelle and Fischer 1994).
One cannot look back at the Great Moderation today without asking whether
the sustained economic stability of the period somehow promoted the excessive
risk-taking that followed. The idea that this long period of relative calm lulled investors, financial firms, and financial regulators into paying insufficient attention to
risks that were accumulating must have some truth in it. Idont think we should
conclude, though, that we therefore should not strive to achieve economic stability.
Rather, the right conclusion is that, even in (or perhaps, especially in) stable and
prosperous times, monetary policymakers and financial regulators should regard
safeguarding financial stability to be of equal importance asindeed, a necessary
prerequisite formaintaining macroeconomic stability.
Specifically, Modigliani and Miller (1958) argue that, under certain conditions, firms will be indifferent
between obtaining funds via equity finance and obtaining funds via debt issue. As noted in the text, some
researchers have taken their result as implying that detailed modeling of the financial sector may not be
central for understanding private sector decisions or the effects of monetary policy. However, as also noted
in the text, Modiglianis and Millers result depends on restrictive assumptions, including no effects of taxes
on financing choices, no bankruptcy costs, no agency problems, and no asymmetricinformation.
11
12
extent that risks remain, however, the Federal Open Market Committee strives to
incorporate these risks in the costbenefit analysis applied to all monetary policy
actions (Bernanke 2002).
What about the monetary policy framework? In general, the Federal Reserves
policy framework inherits many of the elements put in place during the Great
Moderation. These features include the emphasis on preserving the Feds inflation
credibility, which is critical for anchoring inflation expectations, and a balanced
approach in pursuing both parts of the Feds dual mandate in the medium term.
We have also continued to increase the transparency of monetary policy. For
example, the Federal Open Market Committees communications framework now
includes a statement of its longer-run goals and monetary policy strategy. In the
statement issued January 25, 2012, (http://www.federalreserve.gov/newsevents
/press/monetary/20120125c.htm), the Committee indicated that it judged that
inflation at a rate of 2 percent (as measured by the annual change in the price
index for personal consumption expenditures) is most consistent over the longer
run with the FOMCs dual mandate. FOMC participants also regularly provide estimates of the longer-run normal rate of unemployment; those estimates currently
have a central tendency of 5.2 to 6.0percent. By helping to anchor longer-term
expectations, this transparency gives the Federal Reserve greater flexibility to
respond to short-run developments. This framework, which combines short-run
policy flexibility with the discipline provided by the announced targets, has been
described as constrained discretion (for example, as discussed in Bernanke and
Mishkin 1997, in this journal). Other communication innovations include early
publication of the minutes of FOMC meetings and quarterly post-meeting press
conferences by theChairman.
The framework for implementing monetary policy has evolved further in recent
years, reflecting both advances in economic thinking and a changing policy environment. Notably, following the ideas of Svensson (2003) and others, the Federal
Open Market Committee has moved toward a framework that ties policy settings
more directly to the economic outlook, a so-called forecast-based approach. In a
forecast-based approach, monetary policymakers inform the public of their mediumterm targets say, a specific value for the inflation rate and attempt to vary
the instruments of policy as needed to meet that target over time. In contrast, an
instrument-based approach involves providing the public information about how the
monetary policy committee plans to vary its policy instrumenttypically, a short-term
interest rate, like the federal funds interest ratein response to economic conditions.
In particular, the FOMC has released more detailed statements following its meetings
that have related the outlook for policy to prospective economic developments and
has introduced regular summaries of the individual economic projections of FOMC
participants (including for the target value of the federal funds interest rate). The
provision of additional information about policy plans has helped Fed policymakers
deal with the constraint posed by the effective lower bound on short-term interest
rates; in particular, by offering guidance about how policy will respond to economic
developments, the Committee has been able to increase policy accommodation, even
Ben S. Bernanke
13
when the short-term interest rate is near zero and cannot be meaningfully reduced
further (as elaborated in Yellen 2012). The Committee has also sought to influence
interest rates of securities that mature farther into the future (that is, farther out on
the yield curve), notably through its securities purchases. Other central banks in
advanced economies that also confronted the situation that short-term interest rates
had been lowered to their effective lower bound of near-zero percent have taken
similarmeasures.
In short, the recent crisis has underscored the need both to strengthen monetary policy and financial stability frameworks and to better integrate the two. We
have made progress on both counts, but more needs to be done. In particular, the
complementarities among regulatory and supervisory policies (including macroprudential policy), lender-of-last-resort policy, and standard monetary policy are
increasingly evident. Both research and experience are needed to help the Fed
and other central banks develop comprehensive frameworks that incorporate all
of these elements. The broader conclusion is what might be described as the overriding lesson of the Federal Reserves history: that central banking doctrine and
practice are never static. We and other central banks around the world will have to
continue to work hard to adapt to events, new ideas, and changes in the economic
and financial environment.
This paper is a revised version of remarks presented at The First 100 Years of the Federal
Reserve: The Policy Record, Lessons Learned, and Prospects for the Future, a conference
sponsored by the National Bureau of Economic Research in Cambridge, Massachusetts, on
July 10, 2013. I am indebted to Mark Carlson, Edward Nelson, and Jonathan Rose of the
Boards staff for their substantial contributions to the preparation of this article.
References
Ahmed, Shaghil, Andrew Levin, and Beth
Anne Wilson. 2004. Recent U.S. Macroeconomic
Stability: Good Policies, Good Practices, or Good
Luck? Review of Economics and Statistics 86(3):
824 32.
Alesina, Alberto, and Lawrence H. Summers.
1993. Central Bank Independence and Macroeconomic Performance: Some Comparative
Evidence. Journal of Money, Credit and Banking
25(2): 151 62.
Axilrod, Stephen H. 1982. Monetary Policy,
Money Supply, and the Federal Reserves
14
Bernanke, Ben S. 2004. The Great Moderation. Speech delivered at the meetings of the
Eastern Economic Association, Washington, DC,
February20. http://www.federalreserve.gov/board
docs/speeches/2004/20040220/.
Bernanke, Ben S. 2011. The Effects of the Great
Recession on Central Bank Doctrine and Practice.
Speech delivered at the Federal Reserve Bank
of Boston 56th Economic Conference, Boston,
MA, October 18. http://www.federalreserve.gov
/newsevents/speech/bernanke20111018a.htm.
Bernanke, Ben S., and Frederic S. Mishkin.
1997. Inflation Targeting: A New Framework for
Monetary Policy? Journal of Economic Perspectives
11(2): 97116.
Board of Governors of the Federal Reserve
System. 1924. Tenth Annual Report of the Federal
Reserve Board, Covering Operations for the Year 1923.
Washington, DC: Government Printing Office.
Bordo, Michael D., Ehsan U. Choudhri, and
Anna J. Schwartz. 2002. Was Expansionary Monetary Policy Feasible during the Great Contraction?
An Examination of the Gold Standard Constraint.
Explorations in Economic History 39(1): 128.
Bordo, Michael D., and David C. Wheelock.
2013. The Promise and Performance of the
Federal Reserve as Lender of Last Resort
1914 1933. In The Origins, History, and Future of
the Federal Reserve: A Return to Jekylll Island, edited
by Michael D. Bordo and William Roberds, 59 98.
New York: Cambridge University Press.
Calvo, Guillermo A. 1983. Staggered Prices
in a Utility-Maximizing Framework. Journal of
Monetary Economics 12(3): 383 98.
Carlson, Mark A., and David C. Wheelock.
2012. The Lender of Last Resort: Lessons from
the Feds First 100 Years. Working Paper 2012056B, Research Division, Federal Reserve Bank of
St.Louis, November. http://research.stlouisfed.org
/wp/more/2012-056.
Clarida, Richard, Jordi Gal, and Mark Gertler.
2000. Monetary Policy Rules and Macroeconomic
Stability: Evidence and Some Theory. Quarterly
Journal of Economics 115(1): 147 80.
Cukierman, Alex, and Allan H. Meltzer. 1986.
A Theory of Ambiguity, Credibility, and Inflation
under Discretion and Asymmetric Information.
Econometrica 54(5): 1099 1128.
Davis, Steven J., and James A. Kahn. 2008.
Interpreting the Great Moderation: Changes in
the Volatility of Economic Activity at the Macro
and Micro Levels. Journal of Economic Perspectives
22(4): 155 80.
Debelle, Guy, and Stanley Fischer. 1994. How
Independent Should a Central Bank Be? In
Goals, Guidelines, and Constraints Facing Monetary
Policymakers, edited by Jeffrey C. Fuhrer, 195 221,
15
16
Ricardo Reis
tarting with a blank slate, how could one design the institutions of a central
bank for the United States? This question is not as outlandish as it may seem.
As soon as the Iraq war ended in 2003, the first major issue that Coalition
economists confronted: What should be done with the Iraqi dinar? (Foote, Block,
Crane, and Gray 2004, p.60). The economists involved stated that adopting a new
central bank law in March 2004 was one of their first and most important economic
accomplishments, and a similar judgment would hold true when independent
central banks were created in most transition countries as well. Even looking at
high-income economies, in 1992, Europeans had to answer to this question after
they signed the Maastricht Treaty (von Hagen 1997). The US Federal Reserve has
not been an institution set in stone; slowly, and with turns in different directions, its
structure has been molded over 100 years into what it is today.
My goal here is not to describe these historical developments; for those who
would like a detailed history, Friedman and Schwartz (1963) is the classic account
of the history of the Federal Reserve, and Meltzer (2003, 2009a, 2009b) offers a
more recent alternative. Instead, this paper explores the question of how to design
a central bank, drawing on the relevant economic literature and historical experiences while staying free from concerns about how the Fed got to be what it is
today or the short-term political constraints it has faced at various times. The goal
is to provide an opinionated overview that puts forward the trade-offs associated
with different choices and identifies areas where there are clear messages about
Ricardo
Reis is Professor of Economics, Columbia University, New York City, New York. His
email address is [email protected].
doi=10.1257/jep.27.4.17
18
optimal central bank design. Romer and Romer (1997) and Blinder (2006) are
importantprecursors.
Stripped to its core, a central bank is the sole institution in a country with the
power to borrow from banks in the form of reserves while committing to exchange
these reserves on par with banknotes that the central bank can freely issue. More
broadly, the central bank can choose some policy instruments that it controls directly,
as well as a set of announcements about its knowledge of the economy or future policy
intentions. Designing the central bank then consists of specifying three elements:
First is the objective function,, which comes from somewhere or someone, and
includes only a few macroeconomic variables, which serve as goals for the central
bank, potentially at different horizons, matching the small set of instruments at its
disposal. Second, the central bank faces a resource constraint,, limiting both its ability
to distribute dividends as well as the set of policies that it can pursue. Third, there is
a set of equilibrium constraints mapping policy actions and announcements onto the
simultaneous evolution of private agents beliefs and macroeconomic outcomes, so
that commitments by the central bank and transparency about its future intentions
can have an effect right away. In the course of exploring these three broad categories, I will discuss twelve dimensions in central bankdesign.
Ricardo Reis
19
20
Figure 1
Comparing Long-run Nominal Anchors for the United States, 19922012
0.5
Price level (PCE deflator)
Nominal income
Monetary aggregate (M1)
Price-level target
0.4
0.3
0.2
0.1
0
95
12
20
11
20
10
20
09
20
08
20
07
20
06
20
05
20
04
20
03
20
02
20
01
20
00
20
99
19
98
19
97
19
96
19
94
19
93
19
92
19
19
0.1
0.2
0.3
0.4
Source: Author using data from FRED database of the Federal Reserve Bank of Saint Louis.
Notes: Figure 1 uses US data for the last 20 years to plot the price level (the deflator on personal
consumption expenditures), a monetary aggregate (M1), and nominal GDP, subtracting a trend from
each of these last two series so that their value in 2012 is exactly the same as that for the price level. It also
plots a hypothetical price-level target of 2 percent per year.
come up with arguments that are both persuasive and quantitatively large for why
instability in monetary aggregates is costly per se, independent of price stability
(Williamson and Wright 2010; Woodford 2010). As for nominal income, especially
outside the United States over the past century, there was considerable uncertainty
on the long-run rate of economic growth in many countries. A central bank can
do little about this long-run rate of economic growth, but if it sought to achieve a
pre-set nominal income target, this would lead to an unstable price level.1
Figure 1 illustrates this point using US data for the last 20years. It plots the
price level, a monetary aggregate (M1), and nominal GDP, subtracting a trend from
each of these last two series so that their beginning value in 1992 and their ending
value in 2012 is exactly the same as that for the price level. This detrending exercise
generously assumes that, in designing policies that target monetary aggregates or
nominal income, the central bank would know the long-run trends in velocity and
real output. Nevertheless, as the figure shows, while the Federal Reserve policy of
focusing on prices has led to a reasonably steady rate of increase in price level, it has
1
This distinction should not be overstretched; in the short run, a flexible price-level target that responds
to the output gap with a coefficient is equivalent to a nominal income rule with a coefficient 1 on
the outputgap.
21
Here is a numerical example: Imagine you have a 0percent inflation target, but this year some shock
hits such that inflation ends up being 1percent. Then the price level goes from 100 to 101. The next year,
the policy says aim for 0percent inflation again, so prices stay at 101 the following year, so the price level
is 101 forever after. If instead your policy is to have the price-level target rise 0percent, after the price
goes to 101, your policy would have you get back to 100 resulting in 1 percent inflation the next year.
You would get back to 100 right away and forever.
22
while with an inflation target, they realize that it requires indexation to have wages
and prices keep up with past inflation. Therefore, because fewer firms and workers
choose to index their prices and wages, a price-level target frees these prices to
become more flexible to react to other shocks (Amano, Ambler, and Ireland 2007).
Fifth, price-level targeting results in a lower cost of capital for the economy relative to inflation targeting. With inflation targeting, the price level follows a random
walk: since surprises are never reversed, as one looks further into the future, prices
can drift further from the starting point and the variance becomes unbounded. The
real payment on nominal assets becomes riskier, which raises their risk premium
and therefore the cost of capital in the economy (Fischer 1981).
Sixth, and particularly relevant today, a price-level target is an effective way to
guarantee that if a shock pushes the economy into low inflation and zero nominal
interest rates, then the central bank automatically commits to higher future inflation escaping from the liquidity trap (Eggertsson and Woodford 2003).
In spite of all of these theoretical virtues, price-level targets have only very
rarely been adopted by actual central banks. While each theoretical mechanism
above has some evidence to back it, the policy of adopting a price-level target
has not been tested. One common objection is that the central bank would have
trouble communicating the ever-changing goal for the inflation rate that comes
with a price-level target (where positive inflation surprises must come with negative inflation adjustment), to a public that is accustomed nowadays to focusing on
2percent inflation every year. Yet, over the past few decades, the Federal Reserve
has shifted from using targets for monetary aggregates, to targets for the federal
funds rate, to targets for inflation. People adapted. In the last few years, the
Federal Reserve has offered more frequent speeches, policy announcements about
the quantitative easing bond purchases, and forward guidance about commitments to a future path for the federal funds interest rates. Again, people adapted.
Price-level targets do not seem like such a radical change, in comparison. Another
objection is that if agents form expectations of future inflation adaptively as a function of past inflation, price-level targeting will increase instability (Ball 1999). But
this begs the question of why agents, even backward-looking ones, would use past
inflation instead of the past price level to form their expectations in an economy
with a price-leveltarget.
The main reason why the Fed has not discussed adopting a price-level instead
of an inflation target is probably empirical. Figure1 plots what a 2percent annual
rise in the price-level would look like, and it is very close to the actual evolution of the price level. The distinction between inflation targeting and price-level
targeting may therefore seem empty of empirical substance. This conclusion
would be incorrect. In modern macroeconomics, policy targets and rules affect the
expectations and choices of private rational agents, so that even if by chance the
path of the price level under policies of price-level and inflation targeting is
the same, the sixchannels described above will lead to higher welfare and more
stable real activity under a price-level target. Moreover, modern econometrics
teaches us that it would take a great deal more data than 20years to distinguish
Ricardo Reis
23
between stable inflation at 2percent and a stable price level growing at 2percent a
year. A more intriguing possibility revealed by Figure1 is that the Feds actions may
already be close to a price-level target even as it describes its actions as following
an inflationtarget.
Whether a central bank chooses the inflation rate or the price level, it still
faces a question of how to measure the price level. Most measures of inflation are
strongly correlated at low frequencies, like a decade or more, but they can differ
from each other substantially over shorter periods of several years. Having to wait
more than a decade for feedback would obviously make it difficult to assess the
central banks performance. Another difficulty is that even if a broad measure of
changes in the cost of living captures social welfare, its variation is dominated by
relative-price changes associated with structural changes to potential GDP. If the
central bank accepts that monetary policy does not affect potential GDP, then
these long-run changes in relative prices are beyond its control. This last argument suggests two ways to construct an appropriate yearly measure of long-run
target inflation: look at the change in prices that are by construction uncorrelated
with output at low frequencies such as a decade or more (Quah and Vahey 1995),
or find a measure of pure inflation that filters out all relative-price movements
and captures only the changes in the unit of account that the central bank can
affect (Reis and Watson 2010).
Finally, since real outcomes are what matter to people, it is tempting to suggest
that the central bank should also have a real long-run target. However, there is
almost a consensus around the FriedmanPhelps claim that the long-run Phillips
curve is vertical, meaning there is no permanent trade-off between inflation and
real activity.3 This implies that the central bank cannot use its power over the price
level to affect output or employment in the long run, so there is no point in asking
it to focus on a long-run real target. Moreover, even if the central bank had such a
target, if we do not understand reasonably well the specifics of the long-run tradeoff
between prices and output, setting monetary policy in a way that seeks to achieve a
real long-run target could have the undesired consequences of inflation or deflation. These twoarguments have convinced most central banks not to have a real
long-run target, and the large bulk of the literature supports this choice.
However, it is worth remembering that the empirical evidence that there is
zero association between the rate of inflation and the rate of economic growth
and employment is quite weak. If inflation goes well into the two-digits, the data
seem to suggest that there is a negative association with growth and employment.
For inflation rates below 10percent, the failure to reject the null hypothesis of no
association involves confidence intervals wide enough that this failure should not
be confused for positive evidence that the long-run Phillips curve is truly vertical.4
For recent theoretical arguments for why the long-run Phillips curve may instead be upward or downward
sloping, see Berentsen, Menzio, and Wright (2011) and Akerlof, Dickens, and Perry (2000),respectively.
4
See Bruno and Easterly (1998) for the long-run evidence, and Svensson (2013) for a recent empirical
argument for a non-vertical Phillips curve in Sweden.
24
25
centurythe Great Depression of the 1930s and the Great Recession that started in
2007were associated with financial crises. Similar to the question of real targets
discussed above, if financial stability is to be included as a separate goal for the
central bank, it must pass certain tests: 1)there must be a measurable definition
of financial stability, 2)there has to be a convincing case that monetary policy can
achieve the target of bringing about a more stable financial system, and 3)financial
stability must pose a trade-off with the other two goals, creating situations where
prices and activity are stable but financial instability justifies a change in policy that
potentially leads to a recession or causes inflation to exceed its target.
Older approaches to this question did not fulfill these three criteria, and
thus did not justify treating financial stability as a separate criterion for monetary
policy. Before the Fed was founded, seasonal and random changes in the demand
for currency and reserves led to wide fluctuations in interest rates and to occasional bank failures and panics. The Fed was in part founded to supply an elastic
currencythat is, to adjust the supply of money to accommodate these demand
shocks. Yet the volatility of interest rates in these cases almost always comes with
volatile inflation and real activity, so financial stability was aligned with the other
goals, and in that sense did not seem to merit separate consideration. Moreover,
deposit insurance and financial regulation conducted outside of the central bank
already address many of the stability concerns related to shifts in the demand for
liquidity. Another approach to defining financial stability was in terms of large asset
price movements. Yet, at most dates, there seems to be someone crying bubble at
one financial market or another, and the central bank does not seem particularly
well equipped to either spot the fires in specific asset markets, nor to steer equity
prices (Blinder and Reis 2005; Blinder 2006).
A more promising modern approach begins with thinking about how to define
financial stability: for example, in terms of the build-up of leverage, or the spread
between certain key borrowing and lending rates, or the fragility of the funding
of financial intermediaries (for example, Gertler and Kiyotaki 2010; Crdia and
Woodford 2010; Adrian and Shin 2010; Brunnermeier and Sannikov, forthcoming;
among many others). This literature has also started gathering evidence that when
the central bank changes interest rates, reserves, or the assets it buys, it can have a
significant effect on the composition of the balance sheets of financial intermediaries as well as on the risks that they choose to take (Kashyap, Berner, and Goodhart
2011; Jimenez, Ongena, Peydro, and Saurina 2012). As a result, even for fixed
output and prices, changes in the funding structure of banks, in their net worth, or
in their perception of tail risk, can create a misallocation of resources that significantly lowers social welfare. While it is not quite there yet, this modern approach to
financial stability promises to be able to deliver a concrete recommendation for a
third mandate for monetary policy that can be quantified and implemented.
Dimension 4: The Choice of Central Banker(s)
Society can give a central bank a clear mandate with long and short-run goals,
but eventually it must appoint individuals to execute that mandate, and they will
26
always have some discretion. Choosing the central banker is a complementary way
to pick an objective function for the central bank. For example, Romer and Romer
(2004a) argue in this journal that different chairmen of the Federal Reserve chose
very different policies, in spite of an essentially unchanged legal mandate, mostly
due to different views on the role and effects of monetary policy.
Most countries do not pick a single person to have absolute power over the
central bank, but prefer a committee of several people. A committee has several
advantages including the ability to pool information, the gains from having a diversity of views that must be argued for and against, the checks it provides against
autocratic power, and the experimental evidence that committees make less volatile
decisions (Blinder 2004). For these potential virtues to be realized requires that the
committee members have different perspectives, supported by independent staffs,
while sharing a common framework to communicate effectively and to come to
agreements (Charness and Sutter 2012).
When a committee makes decisions, there needs to be a rule to aggregate
the separate preferences of individuals. There is a long literature on voting rules
that have some desirable properties, and there is little about the Federal Open
Market Committee that requires special treatment (Vandenbussche 2006). A
more interesting question is who should have a vote in the committee if the goal is
to elicit talent and bring together different information. For example, is it useful to
draw at least some of the membership of the committee from different regions of
the country? On the twelve-member Federal Open Market Committee, the seven
Washington-based members of the Board of Governors are joined by fiveheads of
the existing twelve regional Federal Reserve banks. These regional Federal Reserve
banks are not just local offices of the central bank, spread around the country to
interact with and provide services to local communities, but actually have some
autonomy and a say in monetary policy. The locations of the regional banks, and
even the fact that there are twelve districts, resulted from delicate political equilibriums that only partly reflected economic considerations (Hammes 2001). The
Federal Reserve Act leaves vague how the twelveregional banks should interact and
work together (Eichengreen1992).
In considering how regional interests are represented at the Fed, one should
note there is evidence that US states share most of their risks (Asdrubali, Sorensen,
and Yosha 1996). So even if regional governors had only the consumption of people
in their region in mind, this fact would justify a focus on eliminating aggregate risk
and ignoring idiosyncratic regional shocks. Might regional governors bring additional information that originates from or pertains to their region? Looking at the
forecast performance for key macroeconomic variables, the members of the Fed
Open Market Committee seem to add little value to the forecast produced by the
staff at the Board of Governors (Romer and Romer 2008). Therefore, the case for
having regional governors relies more strongly on promoting different perspectives
and stimulating original thinking. Geographical distance and separate staffs and
budgets may help to cultivate competition in the market for ideas in interpreting
the data and arriving at policy proposals (Goodfriend 1999).
Ricardo Reis
27
Monetary policy not only responds to economic shocks, but it can also be a
source of aggregate risk that agents cannot insure against and that induces redistributions of wealth.5 In a representative democracy, different age cohorts or business
sectors may legitimately ask to be represented when these decisions are made. There
are twocounterarguments to such a request. First, the literature has so far not been
able to determine the systematic direction in which monetary policy redistributes
wealth across industries or stable groups in the population. While some people may
be hurt in each decision to raise or lower the interest rate, if there is no persistent
conflict, then it is hard to defend that some groups should permanently have a say
when monetary policy decisions are made. Instead, policymakers can, and perhaps
should, take into account the redistributions of wealth that their policies induce
without having some members of the Federal Open Market Committee designated
to stand for the interests of one group. Second, fiscal policy is a more targeted tool
when it comes to distributing resources. Even if redistribution is a side effect of
monetary policy, fiscal policies may be a preferable tool to undo its effects on the
distribution of income, wealth, or consumption.
See Bullard and Waller (2004) for some theory applied to central bank design, and Doepke and
Schneider (2006), Berriel (2013), and Coibion, Gorodnichenko, Kueng, and Silvia (2012) for evidence
onredistribution.
28
There is a clearer way to state this important wisdom. As part of its activities,
the central bank will generate resources, which have threeproperties: First, these
resources, in present value, come exclusively from the seignorage arising from
money creation: that is, the resources arise because the central bank pays less-thanmarket interest on some of its liabilities in exchange for the service that they provide
and at the same time earns market interest rates on the assets backed by these
liabilities (Reis 2013b). Second, seignorage depends primarily on the level of inflation, since higher inflation taxes the holders of currency by lowering the value of
this government liability relative to the goods it can buy; but the generation
of substantial revenue requires very high inflation (Hilscher, Raviv, and Reis, in
progress). Third, if the central bank pays out its net income every period, then its
budget constraint will be respected regardless of the monetary policy that is chosen
(Hall and Reis 2013). Governments will always, under fiscal stress, be tempted to
demand that the central bank generate more resources and transfer them to the
Treasury. Combining the three properties above, we know that 1) the resources
come from seignorage, 2)which requires higher inflation, and 3)the central bank
can feasibly make the transfers desired by the Treasury. This suggests that to keep
prices stable in the long run, central bank design should allow the bank to resist
fiscaldemands.
This lesson does not preclude considering the interaction between monetary
and fiscal policy in determining inflation (for example, Sims 2013). It also does not
deny that it may be optimal in some states of the world to generate fiscal revenues
via inflation (Sims 2001; Chari, Christiano, and Kehoe 1991). It simply distinguishes
between seignorage revenues, which are small and require high expected inflation, from the fiscal benefits from unexpected inflation that arise, for instance, by
lowering the real value of public debt outstanding.
Dimension 6: The Importance of Fiscal Backing for the Central Bank
In conventional times, the Federal Reserve mostly holds government bonds
of short maturities and implements monetary policy by buying and selling them
from banks in exchange for reserves (Friedman and Kuttner 2010). Under this oldstyle central banking, using open market operations, the assets and liabilities of the
Fed are close to riskless and they are matched in their maturity, so net income will
almost always be positive (Hall and Reis 2013).
However, if the central bank pays interest on reserves and, especially, if it holds
other assets that create a riskmaturity mismatch with its liabilities, sometimes the
net income of the central bank will be negative. This is true of the Federal Reserve
today as it has embraced a new-style central banking where long-term securities
that are either issued by the Treasury or implicitly guaranteed by it (agency debt
and mortgage-backed securities) now dominate its assets, as shown in Figure2. The
figure shows that in 2007, almost all Federal Reserve assets were in the form of
Treasury securities, mainly of short maturities. By 2013, a large share of Federal
Reserve assets were in the form of mortgage-backed securities and agency debt, and
the Fed primarily held long-term securities.
29
Figure 2
The Assets of the Federal Reserve by Maturity and Type: Old-Style
(December 31, 2007) versus New Style (September 26, 2013)
90%
2013
80%
70%
60%
50%
Treasury securities
Mortgage-backed securities
and agency debt
2007
40%
2007
30%
2007
2013
20%
10%
0%
2013
Less than 1 year
1 to 5 years
Maturity
Source: Author using data from the Federal Reserve statistical release, table H.41.
Most central banks have a rule, more or less explicit, of handing over their
positive net income to the Treasury, and the Fed has done so in every year of its
existence. However, if there is no transfer in the other direction when income is
negative, then the budget constraint of the central bank will not hold (Hall and
Reis 2013). Something must give. One plausible consequence is that inflation rises
above target so that seignorage is higher and net income does not become negative.
Even if this event is rare, expectations of higher inflation can set in even while net
income is positive.
Preventing this outcome requires giving fiscal backing to the central bank. One
design principle that achieves this backing is to commit the Treasury to transfer
resources to the central bank if net income is negative. An alternative is to allow
the central bank to build a deferred account against the Treasury when net income
is negative, which is then offset against future positive income. Such steps require
strict audits of the Feds accounts, limits to the risks it can take, and an upper bound
on this backing, none of which are easy to specify.
30
Ricardo Reis
31
way to bypass it. The means comes because, if markets are quite illiquid, then even
the relatively small-scale purchases by central banks can significantly raise security
prices and lower their yields (Krishnamurthy and Vissing-Jorgensen 2011). Finally,
the ambition is that, if a combination of illiquidity and limits to arbitrage leads
relative prices of financial assets to be distorted, then there will be a misallocation
of resources that the central bank may be able to correct.
On the other side, there are several objections to a central bank engaging in
credit policies. The central bank may realize significant losses, a risk which is greatly
magnified with credit policies. Furthermore, if the markets are illiquid enough for
the central banks purchases to make a difference when buying, they are also potentially likewise illiquid enough for it to have trouble selling when it wants to at least
without incurring large losses. Moreover, even when the central bank lends against
strong collateral to failed banks, if this keeps nonviable entities operating with
growing losses and deposits, it increases the potential losses that deposit insurance
will eventually have to bear (Goodfriend 2011). It is also tempting for the central
bank to become overconfident about its ability to detect and correct financial
market mispricings and to jeopardize the focus on its macroeconomic objectives.
Moreover, correcting market distortions is typically the domain of tax and regulatory policy, not central banking.
A final objection is that aggressive credit policy exposes the central bank to
legitimate political questions of why some firms, markets, or securities were chosen
for support and not others. While conventional buying and selling of government
bonds does not clearly benefit onefirm or sector, credit policies have clear redistributive effects. At the same time, they also expose the central bank to lobbying pressure
from financial market participants. Both will likely get in the way of the central
banks goals (Reis 2013a). A different type of pressure and temptation may come
from within the central bank. Without a clear policy rule forbidding the bailing out
of systemically important financial institutions, it will always be optimal to do so to
avoid a larger crisis; however, the expectation of a bailout may create incentives for
banks to become larger, take on more risk, and correlate their exposure so that they
become systemically important and thus prime candidates for bailouts (Goodfriend
1994; Stern and Feldman 2004; Farhi and Tirole 2012; Chari and Kehoe 2013).
Given so many virtues as well as objections to credit policy, thoughtful design
of a central bank likely puts some restrictions on the assets that the central bank
can buy. At one extreme, the policy could be the one that the Federal Reserve
faced in 2007, of having to justify unconventional policies to Congress as being
due to unusual and exigent circumstances, which is a fairly vague standard and
thus not difficult to meet. At the other extreme, if we judge that there is too much
of a temptation for the central bank to find a way to get around the rules, then a
strict buy only Treasuries rule may be the answer (Goodfriend 2011). Even in this
case, the central bank would still be able to shift between short-term and long-term
government bonds. These quantitative easing policies expose the central bank to
maturity riskwhen policy becomes contractionary and markets start expecting an
upward-sloping path for short-term interest rates, long-term bond prices will fall,
32
inducing capital losses on the Feds portfoliobut most empirical estimates of this
risk come up with relatively small losses in worst-case scenarios that could easily
be written off against a few future years of positive earnings (Hall and Reis 2013;
Carpenter, Ihrig, Klee, Quinn, and Boote 2013; Greenlaw, Hamilton, Hooper, and
Mishkin2013).
Between these twoextremes, many alternatives are plausible. Oneconcrete
restriction would be to prevent the central bank from taking part in adhoc interventions targeted at specific institutions: that is, the central bank would have to
stick to a general policy that is applied uniformly at arms-length across the entire
financial sector. This would prevent the Federal Reserve from being able to resolve
a particular financial institution, as happened in the bailouts of Bear Stearns in
March 2008 and AIG in September 2008. A tighter restriction would require the
central bank to purchase only securities for which there is a market price, with
enough market participants that compete for the central banks funds. A stronger
version of this rule would prevent the Federal Reserve from intervening in any
over-the-counter financial markets. A weaker version could draw from the experience in industrial organization and require the central bank to run a reverse
auction, with even a small set of institutions, designed to ensure that its purchases
are allocatedefficiently.
Dimension 8: The Payment of Interest on Reserves
When a central bank pays interest on the reserves deposited by banks, it can
use quantitative policy to satisfy the liquidity needs of the economy. By choosing
both the interest on reserves and the federal funds rate, the central bank can at the
same time set the short-term interest rate that will determine inflation, as well as
affect the amount of liquidity held in the banking sector (Kashyap and Stein 2012).
Separately from its interest-rate policy, the Fed can have a large balance sheet, like it
does at present, if society wants to keep a larger share of wealth in money-like investments, or the balance sheet can quickly shrink to the pre-crisis levels, all without
consequences or dangers for the rate of inflation. Most central banks around the
world have the authority to pay interest on reserves, and the Fed joined them in
October2008.
The central bank could go one step further along these lines (Hall 1986).
Ageneral principle of economic efficiency is that the marginal cost of producing
a good should equal its marginal benefit to society. In monetary economics, this
principle leads to the Friedman rule which has been reaffirmed repeatedly in a
wide variety of models of the demand for money (Lucas 2000; Chari and Kehoe
1999; Lagos and Wright 2005). Applied to reserves, note that it costs nothing to
add an extra unit to a banks reserves balances at the Fed, and that the benefit, or
opportunity cost, of holding reserves is the overnight federal funds rate at which
banks could lend these funds to other banks. Therefore, the Friedman rule dictates
that the central bank should pay an interest rate on overnight reserves equal to
the overnight federal funds rate, thus satiating the market with as many reserves as
it wants. This floor policy would make the interest rate on reserves the primary
33
instrument of monetary policy and, unlike the federal funds rate, it is perfectly set
and controlled by the central bank (Goodfriend 2002; Woodford 2003). There is a
strong case for requiring the central bank to not just pay interest on reserves, but
also to always follow the Friedman rule via a floor policy.
34
Chari and Kehoe (2006) associate the adoption of clear rules with addressing the time-inconsistency
problem, Svensson (2003) explains targeting rules, Giannoni and Woodford (2010) provide a very
general theoretical treatment, and Bernanke and Mishkin (1997) early on defined inflation targeting as
a broad framework where communication and transparency are central.
Ricardo Reis
35
from providing forward guidance about future monetary policy (Woodford 2005;
Blinder, Ehrmann, Fratzscher, De Haan, and Jansen 2008). Moreover, there is a
prima facie argument for public institutions to be open in order to be democratically legitimate. The question should therefore be put backwards: is there any strong
argument for the central bank not to reveal everything it knows?
It is arguably appropriate for the central bank to keep to itself the private information it receives from banks it regulates. It may also lead to a more productive
internal discussion if the central bank does not reveal every step of its deliberative
process too soon after monetary policy decisions. But both of these points are minor
exceptions to the general rule of openness, and there is as much risk of these exceptions being violated as there is of them being overstretched.
Of greater concern is whether central bank announcements foster confusion
rather than better understanding. A small literature uses models where agents have
cognitive or informational limitations that can lead them to misinterpret public
information. If the central bank reveals signals about the state variables that agents
use to make decisions but it does so in a manner that buries the information in
statistical noise, or if it announces the information too soon before it becomes
relevant, or if it focuses on variables that are too far from the policy targets, then it is
possible to lower the precision of private actions and achieve worse outcomes (Reis
2011; Eusepi and Preston 2010; Gaballo 2013). Moreover, public signals may lead
agents to collect less private information, making the price system less efficient and
inducing an overreaction of expectations to noisy public signals (Morris and Shin
2002, 2005; Amador and Weill 2010). But while the literature has developed theoretical arguments for why less information might raise welfare in a model, it has not
convincingly shown that these effects are likely to be present (Roca 2010), quantitatively important (Svensson 2006), or empirically significant (Crowe 2010) in reality.
Moreover, in these models, what is usually better than revealing less information is
to optimize the form and timing of announcements. The work of national statistical
agencies is subject to the same caveats, and they respond by working harder at being
informative and clear, not by embracing obscurantism.
Dimension 11: Picking the Channel(s) of Communication
The Federal Reserve has a particular decentralized structure with seven
members of the Board of Governors in the center and twelveFederal Reserve Bank
presidents as independent poles. Having this many actors in monetary policy poses
challenges for making public announcements. First, a decentralized structure
makes it difficult to have model-based monetary policy. There is an economic model
in Washington,DC, that is used to make staff forecasts, but the district presidents
have no input into it. In turn, each of the district presidents has his or her own
model and set of predictions. It is hard to explain monetary policy decisions, and
especially to announce and commit to future policy and targets, when so many decision makers are partially revealing their views and plans (Ehrmann and Fratzscher
2007). Second, many voices raise the danger of confusing disagreement with uncertainty, in spite of the two being conceptually distinct and empirically only weakly
36
related (Mankiw, Reis, and Wolfers 2004; Zarnowitz and Lambros 1987). Third,
the decentralized structure makes it harder for agents to coordinate on the public
signals provided by policy. Some research has suggested that to aid coordination,
the central bank could have fewer speeches, which would be more precise and
targeted at different groups in the population (Chahrour 2013; Morris and Shin
2007; Myatt and Wallace, forthcoming).
While none of these problems can be completely solved, all of them are ameliorated with more information, including requiring each member of the Federal
Open Market Committee to justify his or her views and to report the numerical
forecast distributions that support these views. The literature offers few objections
to giving the central bank a general mandate to be as transparent as possible while
leaving policymakers some discretion on how to implement thismandate.
Dimension 12: The Accountability of the Central Bank
Transparency is a, or perhaps the,, way of achieving accountability. If the central
bank is open about its objectives, its procedure, and its views of the future, that will
go almost all the way towards being accountable in its missions to society as a whole
(Blinder2004).
Political oversight goes hand in hand with accountability. The sevenmembers
of the Board of Governors of the Federal Reserve are appointed by the President,
confirmed by Senate, and periodically answer to Congress. In that sense, both the
executive and legislative powers, and the public that elected them, are represented.
The overlapping terms for the governors ensure that different waves of those holding
political power have an influence, which research has suggested reduces the likelihood
of the central bank becoming captured by partisan governors (Waller 1989,2000).
The regional structure of the Federal Reserve makes power more diffuse, so it is
in principle harder for the central banks actions to be taken over by one particular
interest group (Friedman and Schwartz 1963). The 12 presidents of the regional
Federal Reserve banks each answer to a board of ninemembers: three appointed by
the Board of Governors, three from the local community, and three from the banks in
their district. After the passage of the DoddFrank Act of 2010, banks no longer have
a vote appointing the president. An interesting open question is whether banks from
that district should be singled out, either positively in terms of having threereserved
seats in the board, or negatively in terms of having no vote.
Conclusion
This paper has discussed 12dimensions of central bank design. Table1 summarizes the recommendations, together with the questions it left unanswered, and an
assessment of the Federal Reserve System at present. Threebroad issues have been
pervasivethroughout.
The firstissue is central bank independence. While many have defended the
virtues of central bank independence in general for preventing the tendency of
37
Table 1
Dimensions of Central Bank Design
Dimensions
Suggestions
Open questions
Federal Reserve
Clear on main
goals, otherwise give
discretion
Adopt numerical or
qualitative targets?
Vague
Price-level target as
nominal anchor
To provide a nominal
anchor
Friedman rule at
present, future to be
seen
Increasing role through
forward guidance
Rapidly improving,
revealing more and
sooner?
Rapidly improving,
frequent and clear
speeches
Strong political oversight, peculiar role of
banks
Policymakers with
long-term mandate
and publish inflation
reports
Be as transparent as
possible
All committee members should report
their views
Be transparent, have
overlapping terms of
office
democratic politicians towards ever-higher inflation, looking at more specific questions led to a more mixed message. Even if there is a case for central banks to
independently conduct the operations of monetary policy, democratic principles
would imply that society would still choose the goals of monetary policy. Committing to a stable long-run nominal anchor may reduce the costs of price uncertainty,
but that is not the same as having a fanatic central banker committed to 2percent
38
inflation at all times, and research shows that a flexible price-level target may be
able to lower the variance of inflation and real activity. In turn, releasing the central
bank from the duty to raise seignorage to make transfers to fiscal authorities does
not imply that the central bank can assume large risks through unchecked credit
policies. Moreover, even if central bankers are appointed to long terms that are
independent from political pressure, so that they will not be tempted by the siren
lure of unexpected easy money, the goal of avoiding monetary policies that are
inconsistent over time also requires that the policymakers are politically accountable and transparent.
The secondbroad topic was the level of decentralization of the central bank,
and in particular of the Federal Reserve. There are reasons to be skeptical of the
ability of the Feds regional structure to reconcile different business interests or to
produce new information, and having so many voices raises difficulties for effective
communication. At the same time, a decentralized structure makes different actors
accountable and fosters the competition of ideas and perspectives. It is harder to
argue persuasively that this decentralization should be tied to geography and very
hard to justify the current structure of the Federal Reserve System as optimal if one
were starting from scratch. The best structure to maximize advantages and minimize disadvantages remains an open question.
The final broad topic was the use of unconventional policy. During a financial
crisis, possibly including being stuck in a liquidity trap, the economics literature has
put forward arguments that support price-level targets, forward guidance in setting
interest rates, paying interest on reserves, allowing the Feds balance sheet to grow,
or changing the maturity of the Feds holdings of government securities. Yet there
are strong objections to letting the Fed hold any type of assets, especially as the
risks that comes with them exposes the central bank to potentially large losses of
resources, as well as to pressure and scrutiny by those who benefit or lose from those
purchases. Moreover, because controlling inflation requires fiscal backing from the
Treasury, there must also be limits on the risks to the central banks net income.
More generally, institutional design rules that do not cover exceptional times are
incomplete, and the analysis above suggested principles that apply during crises and
normaltimes.
There are many other design issues that were not addressed, especially
concerning financial regulation (as discussed by Gorton and Metrick, this issue;
Blinder 2010). The broader message of this paper is that designing a central bank
need no longer involve a resort to hunches, old aphorisms, or vague platitudes.
Diverse tools and models, drawn from different branches of economics, can come
together in informing this particular application of mechanism design.
I am grateful for comments from Alan Blinder, Anil Kashyap, David Romer, and Mike
Woodford, and especially from the editors of this journal. An earlier draft of paper was written
for the National Bureau of Economic Research conference The First 100 Years of the Federal
Reserve, which took place in Cambridge, Massachusetts, on July10, 2013.
Ricardo Reis
39
References
Adrian, Tobias, and Hyun Song Shin. 2010.
Financial
Intermediaries
and
Monetary
Economics. Chap. 12 in Handbook of Monetary
Economics, Vol.3A, edited by Benjamin M. Friedman
and Michael Woodford, 601650. Elsevier.
Akerlof, George A., William T. Dickens, and
George L. Perry. 2000. Near-Rational Wage and
Price Setting and the Long-Run Phillips Curve.
Brookings Papers on Economic Activity 31(1): 1 60.
Alesina, Alberto, and Andrea Stella. 2010. The
Politics of Monetary Policy. Chap.18 in Handbook
of Monetary Economics, Vol.3B, edited by Benjamin
M. Friedman and Michael Woodford, 10011054.
Elsevier.
Alesina, Alberto, and Guido Tabellini. 2007.
Bureaucrats or Politicians? Part I: A Single Policy
Task. American Economic Review 97(1): 169179.
Amador, Manuel, and Pierre-Olivier Weill. 2010.
Learning from Prices: Public Communication
and Welfare. Journal of Political Economy 118(5):
866 907.
Amano, Robert, Steve Ambler, and Peter
Ireland. 2007. Price-Level Targeting, Wage Indexation and Welfare. Unpublished paper, Bank of
Canada.
Asdrubali, Pierfederico, Bent E. Sorensen,
and Oved Yosha. 1996. Channels of Interstate
Risk Sharing: United States 19631990. Quarterly
Journal of Economics 111(4): 10811110.
Backus, David, and John Driffill. 1985. Inflation and Reputation. American Economic Review
75(3): 530 38.
Ball, Laurence. 1999. Efficient Rules for Monetary Policy. International Finance 2(1): 63 83.
Ball, Laurence, N. Gregory Mankiw, and Ricardo
Reis. 2005. Monetary Policy for Inattentive
Economies. Journal of Monetary Economics 52(4):
703 725.
Barro, Robert J., and David B. Gordon. 1983.
Rules, Discretion and Reputation in a Model of
Monetary Policy. Journal of Monetary Economics
12(1): 101121.
Benati, Luca. 2009. Long-Run Evidence on
Money Growth and Inflation. European Central
Bank Working Paper 1027, March.
Berentsen, Aleksander, Guido Menzio, and
Randall Wright. 2011. Inflation and Unemployment in the Long Run. American Economic Review
101(1): 371 98.
Bernanke, Ben S. 2012. The Federal Reserve
and the Financial Crisis. March. Lectures delivered at George Washington University.
Bernanke, Ben S., and Frederic S. Mishkin.
1997. Inflation Targeting: A New Framework for
40
41
42
Ricardo Reis
43
44
Gary Gorton is the Frederick Frank Class of 1954 Professor of Finance and Andrew Metrick
is the Deputy Dean & MichaelH. Jordan Professor of Finance and Management, both at
the Yale School of Management,New Haven, Connecticut. Both authors are also Research
Associates, National Bureau of Economic Research, Cambridge, Massachusetts. Their email
addresses are [email protected] and [email protected].
doi=10.1257/jep.27.4.45
46
also backed by long-term assets, while allowing investors who desire liquidity
to withdraw their funds, or more generally not renew their short-term investment, in a much shorter time horizon. During a financial crisis, the negative
externalities of liquidity demand are manifested when investors race to withdraw
their liquid assets; in normal times, negative externalities occur when each
additional liquid claim does not incorporate in its price its contribution to the
risk of such a crisis.
To mitigate the risk of a liquidity-driven crisis, the United States has a
financial sector safety net with two key pillars: the Federal Reserve as a lenderof-last-resort and the Federal Deposit Insurance Corporation (FDIC) as a
guarantor of bank deposits. The existence of this safety net then alters the incentives of regulated financial institutions: in particular, they can take greater risks
when their depositors and investors know that this safety net is in place. Thus,
the existence of the safety net provides the rationale for close supervision and
regulations that limit the scope, risk-taking, and leverage of these institutions.
If the safety net is too large, then banks lack incentives to manage risks in a
socially optimal way; if the safety net is too small, then failure of a large institution could have major spillovers to the whole financial system; and if only the
largest institutions are thus given the most protection, then the private incentives will be for every institution to grow too big to fail. This dynamic presents
a complex problem for the Fed as the lender of last resort and regulator of the
largestinstitutions.
This paper traces the Feds attempts to address this problem from its founding.
We will discuss how the effectiveness of the lender-of-last-resort function was eroded
in the 1920s, which in turn contributed to the banking panics of the Great Depression and indeed has hampered its lender-of-last-resort efforts to the present day. We
consider the regulatory changes of the New Deal, including deposit insurance and
the centralization of Fed decision-making power in the Board of Governors, which
by some combination of luck and design contributed to a quiet period of nearly
50years in the US financial system. Indeed, during this time bank supervision was
only peripheral to the Feds priorities, which moved steadily towards a focus on
price stability using interest-rate policy as its main instrument, and the Fed rarely
needed to even think about the lender-of-last-resort function. The late 1970s saw the
beginning of a transformation of the banking sector, with a rise of nonbank financial intermediaries and then regulatory adjustments so that banks could compete
with these nonbank firms, which has continued to the present day. The financial
crisis of 20072009 shook bank supervision efforts out of their slumber, made the
lender-of-last-resort function central again, and led to a significant shift for the Fed
back to its financial-stability roots. Indeed, the Feds efforts in the recent financial
crisis can largely be viewed as attempts to expand the lender-of-last-resort function
beyond its traditional institutions and markets. We conclude by bringing the story to
the present day with a discussion of the evolving role of the Federal Reserve in the
context of the changes under the DoddFrank Wall Street Reform and Consumer
Protection Act of 2010.
47
In the modern era with the presence of central banks, the links between financial crises and recessions
are similar. For example, Demirg-Kunt and Detragiache (1998, p.83) examine the period 1980 1994
and find that low GDP growth, excessively high real interest rates, and high inflation significantly
increase the likelihood of systemic problems in our sample.
48
1984; Gorton 1984, 1985; Gorton and Mullineaux 1987; Gorton and Huang 2006).
Clearing houses, with one in each large city, were coalitions of member banks.
Ostensibly set up to efficiently clear checks, they assumed a central banklike role
in crises, even though they were private associations.
A panic would trigger clearing house members to act as one large bank, issuing
special liabilities clearing house loan certificates for which they were jointly
responsible. At the outset of the crisis, the clearing house would prohibit the publication of bank-specific information, which was required during noncrisis times.
Also, the amounts of clearing house loan certificates issued to individual member
banks were kept secret, preventing those banks from being targeted for bank runs.
Following the Panic of 1907, Congress passed the AldrichVreeland Act, which
among other provisions created a system for national banks to issue emergency
elastic currency in a panic.
However, these responses of the clearing house member banks were only
triggered by the panic itself. The ability of the clearing houses to issue loan certificates and AldrichVreeland emergency currency did not prevent panics and their
associated real effects. William Ridgely (1908, p.173), the US Comptroller of the
Currency from 1901 to 1908, put the issue this way: The real need is for something
that will prevent panics, not for something that will relieve them; and the only way
to attain this is through the agency of a Governmental bank.
Thus, the idea behind the establishment of the Federal Reserve System was
that it could do something that the clearing houses and the AldrichVreeland
Act could not do. It could establish a credible emergency mechanism in advance..
When the Federal Reserve System was founded, the main focus was on the potential
benefits of a bills market that is, a market for bankers acceptances, which are
a documented promise by a bank to make a payment at a future time. The Federal
Reserve would participate in this market by purchasing bankers acceptances. In
addition, banks would be able to use their holdings of commercial paper and other
marketable securities as collateral to borrow at the discount windowthus in effect
exchanging private debt for currency.
Moreover, being a (quasi-)government entity, the Federal Reserve System could
be expected to be solvent and would always be able to lend to banks. By contrast,
the coalitions of clearing house banks might not be solvent, so expectations that the
clearing house would act did not fully deter panics. Indeed, currency premia on
the certified checks, which were joint clearing house liabilities, were positive during
crisis periods (in other words, it took more than $1 of certified checks to buy $1 of
currency), reflecting uncertainty about clearing house solvency. The AldrichVreeland
emergency currency was issued with bank loans as collateral, not US Treasury bonds.
Again, there was uncertainty about the outcomes.
There is an important difference between providing the reassurance that can
prevent bank runs and responding to a crisis once it has happened. Once a financial event is seen to be systemic and the lender of last resort begins lending, these
actions take time and the process of exchanging private bank assets for government
assets (whether money or Treasury debt) can be costly and painful.
49
It was widely believed that the discounting authority of the Federal Reserve
would prevent banking panics. Banks needing cash could take bankers acceptances
(that is, their promise to pay at a near-term date) which were discounted from par
to the Feds discount window, where the Fed would buy it at a further discount
rediscounting it. Representative Carter Glass (1927, p.387), who sponsored the
Federal Reserve Act in the House of Representatives, wrote that the most important
accomplishments of the legislation were to remove seasonals in interest rates
and to prevent panics. Senator Robert Owen (1919, p. 99), sponsor of the bill
in the Senate, said that the Federal Reserve Act gives assurance to the business
men of the country that they never need fear a currency famine. It assures them
absolutely against the danger of financial panic . . . Congressman Michael Phelan
of Massachusetts, Chairman of the House Committee on Banking and Currency,
argued (as quoted in Hackley 1973, p.10): In times of stress, when a bank needs
cash, it can obtain it by a simple process of rediscounting paper with the Federal
reserve [sic] banks. Many a bank will thus be enabled to get relief in time of serious
need. Businessmen and regulators agreed. Magnus Alexander, the president of the
National Industrial Conference Board announced (quoted in Angly 1931, p. 12)
that there is no reason why there should be any more panics. The Comptroller
of the Currency (1915, p.10) announced that, with the new Federal Reserve Act,
financial and commercial crises, or panics, . . . with their attendant misfortunes
and prostrations, seem to be mathematically impossible. The Federal Reserve
Systems (1914, p.17) first Annual Report states that its duty is not to await emergencies but by anticipation to do what it can to prevent them.
The 1920s
The establishment of the Federal Reserve System did change the expectations
of depositors about systemic banking crises.2 Gorton (1988) creates a leading indicator of recessions for the earlier US National Banking Era from the Civil War
up to 1913, and finds that panics arose when the unexpected component of this
leading indicator of recession exceeded a threshold. During the National Banking
Era, no panic occurred without this threshold being exceeded, and there are no
cases where it was exceeded without a panic. This model predicts that there should
have been a panic in June 1920 (and another panic in December 1929). Thus, the
192021 recession can be viewed as the first test of the ability of the Federal Reserve
to prevent bankruns.
As dated by the National Bureau of Economic Research, there was a business
cycle peak in January 1920 and a trough in July 1921. Banks started to fail in 1920;
There is some evidence that seasonal swings in short-term interest rates were eliminated, although the
point is controversial. For a sampling of the evidence that the Fed did eliminate seasonal swings, see
Miron (1986) and Mankiw, Miron, and Weil (1987). For the alternative view, see Shiller (1980), Clark
(1986), Fishe and Wohar (1990), and Fishe (1991).
50
505banks failed in 1921, and the number of failures continued to rise, averaging
680 per year from 1923 to 1929. The peak was 950 in 1926 (Alston, Grove, and
Wheelock 1994). Hamilton (1985, p. 585) observes that the failed banks were
overwhelmingly small banks in small rural communities: National banks were only
13percent of the failures and only 17percent were members of the Federal Reserve
System. In other words, for the most part the banks that failed did not have access
to the Federal Reserve discount window.
Though many small banks failed, there was no panic. As many contemporary
commentators noted, depositors did not run on banks. For example, Henry Parker
Willis (1923, p. 1406, emphasis added), who received a PhD in economics from
the University of Chicago and was later the first Secretary of the Federal Reserve
System,wrote:
In previous panics or periods of stringency, difficulty had grown out of the
fact that doubts arose concerning the ability of given institutions to meet
their obligations, owing to the fact that their loans were frozen or that public confidence had resulted in withdrawing an undue amount of cash from
them. On such occasions relief was obtained by the banks banding together
for the purpose of supporting any of their number which had sound assets.
In the depression of 1920 1921, the federal reserve system [sic] was in the
position of a clearing house association, already organized in advance and able
to assist the community . . .
Perhaps predictably, the Federal Reserve Annual Report (1921, p.99) took a
similar view that the creation of the Federal Reserve had prevented a panic:
Other nations, such as Great Britain and France, with their great central banking institutions, have always had their years of prosperity and their periods
of depression, although they have been free from the money panics which
we formerly had in this country as a result of our inadequate banking system
and which we would, no doubt, have had in the most aggravated degree a
year or so ago but for the efficiency and stabilizing influence of the Federal
ReserveSystem.
If bank depositors did not run because they expected banks to have access to the
discount window, then it might not be necessary for banks to have actually borrowed
from the discount window. But in fact, national banks did use the discount window,
as shown in Figure 1. Tallman (2010, p. 104) also notes this use of the discount
window over the years 1914 27. In 1921, discounts and advances as a proportion
of Federal Reserve credit was at its peak of 82 percent with about 60 percent of
member banks borrowing. It was not uncommon, evidently, for hundreds of banks
to be continuously borrowing amounts in excess of their capital and surplus (Shull
1971, p. 37). Notably, there was no evidence that borrowers from the discount
window experienced any particular stigma in credit markets.
51
Figure 1
Federal Reserve Credit Extended, 19171935
4
Banker's acceptances
Discount window borrowing
Government securities held
Federal Reserve credit
3.5
Billions of US $
3
2.5
2
1.5
1
0.5
0
35
19
34
19
33
19
32
19
31
19
30
19
29
19
28
19
27
19
26
19
25
19
24
19
23
19
22
19
21
19
20
19
19
18
19
17
19
19
One reason that banks borrowed so much from the discount window was that
the discount rate was below the market interest rate. During World WarI, the Fed
felt that low discount rates were important. The Board did not believe, during the
war period, that marked advances in rates would be advisable in view of the obvious
necessity of avoiding any policy likely to disturb the financial operations of the Treasury (Harding 1925, p.147). During the steep 1920 21 recession, the low discount
rate may have been fortuitous. As an Assistant Secretary of the Treasury wrote
(Leffingwell 1921, p.35), by permitting rates to remain below the open market rates
and credit to be expanded during the period of deflation of prices, it has prevented
the present business depression from degenerating into an old-fashioned panic. But
over time, of course, freely available discount lending at below-market interest rates
was bound to bring tensions.
Indeed, unbeknownst to the wider world, Fed policy on discount window
lending was fundamentally altered in the mid-1920s. As Shull (1993, p. 20, with
quotations from Keynes, 1930, pp. 239 40) explains: A set of non-price rationing
rules, limiting use of the discount window to short-term borrowing for unanticipated outflows of funds, were developed; banks were encouraged to be reluctant
to borrow; i.e., the Fed turned to gadgets and conventions . . . without any overt
alteration of the law. Creating a reluctance to borrow can informally come about
through possible implicit threats to examine the borrowing bank more frequently
and intensively, ostensibly to determine whether such borrowing is warranted.
Why was the policy on discount lending changed? There seem to be several
reasons. First, it became clear that hundreds of banks were borrowing from the Fed
for extended periods of time. Shull (1971, p.35) reports that as of August31,1925,
52
588 banks had been borrowing continuously for at least a year; 239 had been
borrowing since the start of the recession in 1920; and 122 had been borrowing
continuously since before 1920. In addition, 259 national member banks had
failed since 1920, and a guess was made that at least 80 per cent had been habitual
borrowers prior to their failure. Thus, the Federal Reserve Annual Report of 1926
(p.4) stated that the funds of the Federal Reserve banks are ordinarily intended to
be used in meeting temporary requirements of members, and continuous borrowing
by a member bank as a general practice would not be consistent with the intent of
the Federal Reserve Act.
In addition, by the latter part of the 1920s, the Fed became concerned with
trying to distinguish between speculative security loans and loans for legitimate
business. In other words, was discount window credit being used to pump up stock
market values (Anderson 1966)? Was it leading to high growth in real estate prices,
labeled a bubble by some (White 2009)? The Fed sought to restrain credit growth
through moral suasion that would deter member banks from borrowing for speculative purposes, while at the same time trying to maintain a preferential discount rate
for legitimate borrowing (Friedman and Schwartz 1963, p.22526). But the Fed
decided that attempting to influence the economy via the discount window was not
going to work. In short, the purpose of the discount window changed. It would no
longer serve to provide an elastic currency. While contemporary observers noted
that there had been no banking panics in the 1920s, there appears to have been no
understanding of the details of how freely available lending through the discount
window had avoided the panic. The Feds new policy of creating a reluctance to
borrow based on nonpecuniary measures, and an emphasis that such lending
should be only temporary, meant that a bank that did borrow from the discount
window must be in trouble. This was the creation of stigma, which has complicated lender-of-last-resort policy ever since.
53
they were quite hesitant to do so. As shown in Figure1, discount window borrowing
from 1929 to 1931 was much lower than in the 1920s, and after peaking in 1932, it
declines slightly. Apparently, banks feared the stigma the Fed policies had created
in the mid-1920s on discount lending.
When the Great Depression started in 1929, there were no bank runs. As
mentioned earlier, Gortons (1988) calculations looking at how unexpected
movements in leading indicators had predicted financial crises in the pre-Fed era
suggested that, in the Great Depression, there should have been bank runs starting
in December 1929. Similarly, Wicker (1980, p. 573) noted: Historically, banking
panics in the United States usually developed shortly after a downturn in economic
activity. The banking crisis in NovemberDecember 1930, however, was unlike
previous banking collapses: there was little or no discernible impact on the central
money market, and the panic lagged the downturn by eighteen months.
Bank runs did not happen in the Great Depression until late in 1930. As
Richardson (2007, p.40) notes: Before October 1930, the pattern of [bank] failures
resembled the pattern that prevailed during the 1920s. Small, rural banks with large
loan losses failed at a steady rate. In November 1930, the collapse of correspondent
networks triggered banking panics. Runs rose in number and severity after prominent financial conglomerates in New York and Los Angeles closed amid scandals
covered prominently in the national press. There is some dispute over which bank
collapse loomed largest. Friedman and Schwartz (1963) argue that the failure of
the Bank of United States on December 11, 1930, was especially importantin part
because of the banks name. Wicker (1980, p.581; 1996) disputes the importance
of that bank failure, and instead cites the collapse of Caldwell and Company in
mid-November as the trigger of the panic. Caldwell was large; it controlled a large
chain of banks in the South.
A second wave of bank runs began in March 1931. There were runs, for
example, on Chicago-area banks that were followed by a 40 percent increase in
postal savings deposits (Wicker 1996, p.85; for additional discussion, see Calomiris
and Mason 1997). Finally, there was the Panic of 1933, actually in the last quarter
of 1932 and early 1933, which led to President Roosevelt declaring a four-day bank
holiday in March 1933, during which banks and the stock exchange were closed
and forbidden to do any business without special government permission.
During this time, although the Federal Reserve was not engaging in much discount lending, the Reconstruction Finance Corporation, established in January 1932
under President Hoover, had started lending to banks in February 1932. The Reconstruction Finance Corporation action was needed because the Fed took no positive
action to intervene directly to keep open troubled banks. No direct assistance was
offered other than to discount eligible paper of the [Federal Reserve] member banks
(Wicker 1996, p.85). There were 17,000banks in existence just prior to Roosevelts
March 1933 banking holiday. Only 12,000 survived, and half of those were borrowing
some or as much as all of their capital from the Reconstruction Finance Corporation
(Todd 1992). Ironically, the chairman of the Reconstruction Finance Corporation was
Eugene Mayer, who was also chairman of the Fed.
54
At first, there was apparently no stigma attached to borrowing from the Reconstruction Finance Corporation until the clerk of the House of Representatives revealed
the names of borrowers in July 1932 (Butkiewicz 1995, 1999; see also Friedman and
Schwartz 1963, p.331). Figure2 illustrates the scale of loans from the Reconstruction Finance Corporation to banks as well as to other institutions like state and local
governments, railroads, and mortgage institutions. Prior to the revelation of borrower
names beginning in July 1932, total Reconstruction Finance Corporation borrowing
had reached approximately $1billion, with about half of this total going to banks.
Following the name revelation, net bank borrowing flattened out and was below
$500million fouryears later, even though nonbank borrowingwhere stigma is far
less of an issuerose to more than $2billion of thetotal.
The bank runs of the Great Depression were haphazard, chaotic, and spread
out in time, unlike those of the pre-Fed period. Given that there was no bank run in
1929 at the onset of the Depression, the timing suggests that when depositors eventually saw the failures of large banks in the 1930s, they realized that the discount
window mechanism was not working and the bank runs started. What happened?
Friedman and Schwartz (1963, pp. 31819) write: The aversion to borrowing by
banks, which the Reserve System had tried to strengthen during the twenties, was
still greater at a time when depositors were fearful for the safety of every bank and
were scrutinizing balance sheets with great care to see which banks were likely to be
the next to go . . . Wheelock (1990, p.424) provides some evidence for this:
This study also finds evidence of a downward shift in borrowed reserve demand
during the Depression. Financial crises made banks cautious and less willing
to borrow reserves. The Feds failure to recognize this change in bank willingness to borrow contributed to its failure to interpret monetary conditions
accurately. Fed officials continued to believe that low levels of bank borrowing
signaled easy money.
The problem was that the expectations of depositors that banks could and
would avail themselves of the discount window when in trouble were not (widely)
realized. Large banks failed and depositors then ran on the banks.4
We are not making any claims here about the effectiveness of the Fed as a lender of last resort when
banks actually did borrow. For example, Richardson and Troost (2009) contrast the policies of tworegional
Federal Reserve Banks (St. Louis and Atlanta) with regard to their responses to bank troubles in
Mississippi during the Great Depression. Atlanta aggressively assisted banks and the bank failure rate was
lower than in the part of Mississippi in the St. Louis district. The interesting question here is how Atlanta
managed to overcome (or avoid) the stigma that depressed borrowing in other districts.
55
Figure 2
Reconstruction Finance Corporation Loans Outstanding
3.5
3.0
Billions of dollars
Grand total
2.5
2.0
1.5
1.0
Banks
.5
g.
Au
r.
Ap
n.
35
19
35
19
35
19
93
34
93
19
.1
ct
Ja
n.
93
.1
ar
Ju
33
19
33
19
33
93
32
19
.1
ec
p.
Se
ay
b.
Fe
.1
ov
32
19
32
19
93
19
r.
l.
Ju
Ap
n.
.1
ct
Ja
represented the last major set of changes in financial regulation until the 1970s.5
The Banking Acts of 1933 and 1935 amended the Federal Reserve Act to establish the Federal Deposit Insurance Corporation. The advent of deposit insurance
rendered mootfor a timethe mistake of developing the policy of reluctance to
borrow, and there was no discussion or realization of the problem that had been
created by the discountrate policies of the 1920s. Over the subsequent 75years,
the original insurance cap of $2,500 per bank account would be raised many times,
finally reaching $250,000 in the aftermath of the recent financial crisis.
The Banking Acts also had a profound influence on the power and structure
of the Fed. The balance of power between the Board and the regional Reserve
Banks was tipped in favor of the center, with a Board-dominated Federal Open
Market Committee established in 1935. Far more obscure at the time was a small
We do not attempt anything close to a review of all financial regulation during this time period. For
a comprehensive treatment of regulatory and competitive changes in the key 19791994 period, see
Berger, Kashyap, and Scalise (1995). For a discussion of changes since the 1990s leading to the rising
share of nonbank financial intermediaries, see Gorton and Metrick (2010).
56
amendment to Section13 of the Federal Reserve Act, granting the Fed the power
to greatly expand its lending programs under unusual and exigent circumstances.
These powers were invoked often in the recent crisis, as discussed later in this paper.6
The Banking Act of 1933 is often known by the last names of its sponsors, Glass
and Steagall, and by the provision of the law that enforced the separation of deposittaking and securities underwriting. This separation of banking and securities was
coincident with significant new financial regulation, beginning with the Securities
Act of 1933, which focused on the primary sale of securities, and the Exchange Act
of 1934, which created the Securities and Exchange Commission and focused on the
secondary trading markets. The SEC was granted further powers to regulate market
intermediaries in the Investment Company Act of 1940 (for mutual funds and other
investment companies) and in the Investment Adviser Act of 1940 (which today
covers hedge funds and private equity funds, in addition to traditionaladvisers).
After the New Deal legislation, the most important piece of financial regulation
to affect the Fed during this time period was the Bank Holding Company Act of
1956, in which the Fed was given oversight responsibility over holding companies
that included commercial banks in their structure, with rules codified about the
separation of banking and nonbanking activities. Importantly, this responsibility
gave the Fed insight and access to the largest commercial banks, all of which (over
time) became part of bank holding companies. The role of bank holding companies in the overall financial system has increased steadily, so that today they cover
the vast majority of assets in the US banking system.
The Feds emergency-lending power in Section13(3) was first granted by the Emergency Relief and
Construction Act of 1932, which later received amendments in the Banking Act of 1935 and in Federal
Deposit Insurance Corporation Improvement Act of 1991. As discussed later, these amendments proved
crucial for the lending powers used in the recent crisis (Mehra 2011).
57
7
Our discussion of capital rules and the Basel process focuses on the role played by the Federal Reserve
and the implications for the growth of the shadow banking system. For a more comprehensive treatment, Goodhart (2011) is a definitive history of the Basel process up through 1997, and Hanson,
Kashyap, and Stein (2011) is an accessible survey of the intellectual debate about capital standards in
the post-crisisworld.
8
DeAngelo and Stulz (2013) point out that if banks liabilities, short-term liquid debt, are useful because
of their liquidity, they have a convenience yield (part of the return the holder gets is the benefits of
liquidity) and then banks optimally have high leverage. Kashyap, Stein, and Hanson (2010) point out
that even small increases in the cost of the capital could be sufficient to drive significant flows from banks
into nonbank financial institutions. For the most forceful argument in favor of the ModiglianiMiller
interpretation that raising additional capital would not be costly for banks, see Admati and Hellwig
(2013). Other recent perspectives on this debate include Baker and Wurgler (2013) and Gorton and
Winton(2002).
58
59
60
expanded the Feds role as a supervisor of financial institutions. However, the legislation was drafted and passed during a time when the Fed was under tremendous
political and media pressure for its actions during the financial crisis, and this pressure
led to some restrictions on the Feds discretionary power as a lender of last resort.
From a supervisory viewpoint, the 2010 legislation created the Financial Stability
Oversight Council, a new coordinating body that has the power to designate some
financial institutions (including nonbanks) as being systemically important, with these
institutions then subject to oversight and (additional) regulation by the Fed. Such
designations effectively make the Fed a primary regulator for all large financial institutions, no matter what their main function. Furthermore, the Fed now has an explicit
mandate to set higher capital standards and to give extra scrutiny to these largestfirms.
One motivation of the DoddFrank Act was to end public bailouts of the largest
institutions. Such a promise is complex and somewhat at odds with the lender-of-lastresort function. Specifically, the 13(3) powers that the Federal Reserve used during
the crisis have been restricted by requiring more cooperation with the Treasury,
more disclosure to Congress, and less flexibility to design programs to aid specific
borrowers. In addition to the restrictions on the Feds 13(3) powers, other restrictions were made on Treasurys emergency use of rescue powers such as those used
for money-market funds, and the ability of the Federal Deposit Insurance Corporation to broadly guarantee bank assets without an act of Congress. Taken together,
DoddFrank significantly reduced the flexibility of the executive branch and the
Federal Reserve to act quickly during a financial crisis, while expanding their ability
to act pre-emptively before one.
The DoddFrank Act did little to address the vulnerabilities in the shadow
banking system at the heart of the panic during the crisis. For instance, repurchase
agreements serve as a market for short-term loans and can be a source of troubles
in a crisis when such loans are not rolled over as expected; yet reform of repurchase
agreements was left entirely out of the legislation, with no clear jurisdiction for
any agency to act. Reform of money market mutual funds was left to the existing
statutory powers of the Securities and Exchange Commission, and it is has proved
difficult (so far) to make significant changes to the status quo. Financial securitization received some new rules under which those who originally make loans need to
retain some of the risk, rather than completely passing it on to others, but largerscale reforms were not included. The Financial Stability Oversight Council has
some flexibility to address all of these shadow-banking issues in the future, but the
necessary powers are still untested. Overall, the Fed and other regulators still have
significant limitations for liquidity provision and oversight for many of the shadow
banking markets in which financial runs occurred in 20072008.
Conclusion
The Federal Reserve plays a central role in financial regulation, with responsibility as both a lender of last resort and as a supervisor for the largest institutions.
61
Thanks to David Autor, Doug Diamond, Chang-Tai Hseih, Anil Kashyap, Ulrike
Malmendier, Christina Romer, David Romer, and Timothy Taylor for helpful comments, and
to Ellis Tallman for sharing Figure 1, which appears in Tallman (2010).
References
Admati, Anat, and Martin Hellwig. 2013. The
Bankers New Clothes: Whats Wrong with Banking and
What to Do about It. Princeton University Press.
62
Dang, Tri Vi, Gary Gorton, and Bengt Holmstrm. 2013. Ignorance, Debt and Financial
Crises. Unpublished paper.
DeAngelo, Harry, and Rene M. Stulz. 2013.
Why High Leverage is Optimal for Banks. ECGI
Finance Working Paper no. 356/2013, European
Corporate Governance Institute.
Demirg-Kunt, Asli, and Enrica Detragiache.
1998. The Determinants of Banking Crises in
Developing and Developed Countries. IMF Staff
Papers 45(1): 81109.
Federal Reserve System. 1921, 1923, 1926.
Annual Reports.
Fishe, Raymond P. H. 1991. The Federal
Reserve Amendments of 1917: The Beginning of a
Seasonal Note Issue Policy. Journal of Money, Credit
and Banking 23(3): 308 326.
Fishe, Raymond P. H., and Mark Wohar. 1990.
The Adjustment of Expectations to a Change in
Regime: Comment. American Economic Review
80(4): 968 76.
Fleming, Michael J., Warren B. Hrung, and
Frank M. Keane. 2010. Repo Market Effects of the
Term Securities Lending Facility. Federal Reserve
Bank of New York Staff Reports 426.
Friedman, Milton, and Anna Jacobson
Schwartz. 1963. A Monetary History of the United
States. Princeton University Press.
Glass, Carter. 1927. An Adventure in Constructive
Finance. New York: Doubleday, Page and Company.
Goodhart, Charles. 2011. The Basel Committee
on Banking Supervision: A History of the Early Years
19741997. Cambridge University Press.
Gorton, Gary. 1984. Private Bank Clearinghouses and the Origins of Central Banking.
Federal Reserve Bank of Philadelphia Business
Review, January/February, 3 12.
Gorton, Gary. 1985. Clearinghouses and the
Origin of Central Banking in the United States.
Journal of Economic History 45(2): 277 83.
Gorton, Gary. 1988. Banking Panics and Business Cycles. Oxford Economic Papers 40(4): 751 81.
Gorton, Gary, and Lixin Huang. 2006. Bank
Panics and the Endogeneity of Central Banking.
Journal of Monetary Economics 53(7): 1613 29.
Gorton, Gary, Stefan Lewellen, and Andrew
Metrick. 2012. The Safe-Asset Share. American
Economic Review 102(3): 101106.
Gorton, Gary, and Andrew Metrick. 2010.
Regulating the Shadow Banking System. Brookings Papers on Economic Activity, Fall, 41(2): 26197.
Gorton, Gary, and Andrew Metrick. 2012.
Securitized Banking and the Run on Repo.
Journal of Financial Economics 104(3): 425 51.
Gorton, Gary, and Don Mullineaux. 1987.
The Joint Production of Confidence: Endogenous Regulation and Nineteenth Century
63
64
uring its first century, the Federal Reserve has made a substantial number
of changes in the conduct of monetary policy. Figure1 plots a short-term
interest rate, a measure of inflation, and the dates of recessions (shaded
areas), which allows one to separate the 100-year history of policy making at the Fed
into distinct periods. I focus on four of them.
During the first period, starting in the mid-1920s, the Federal Reserve official
policy was to support high-quality bank lending, but not speculative lending. This
goal was set aside once in 1927, in an episode that many observers then blamed for
the economic collapse that followed the financial crash of 1929. The Fed was then
reluctant to increase the funds available to banks through the early 1930s, even as
the Great Depression ravaged the economy. The Feds concern with the volume and
quality of lending in the setting of monetary policy did eventually wither. However,
this only seems to have happened after the publication of Friedman and Schwartz
(1963), a revisionist history of the Great Depression that blamed its depth on the
Feds inappropriate focus on productive lending.
In the second period, after the experience of post-World WarII inflation, the
Federal Reserve in the 1950s was highly concerned with inflation and was willing to
raise interest rates and bring on recessions to nip even modest inflation rates in the
bud. This brought withering criticism for the Federal Reserve on the grounds that
the recessions of 1957 and 1960 had been unnecessary. By the mid-1960s, some Fed
officials seem to have developed an aversion to creating recessions as a method of
Julio J. Rotemberg is William Ziegler Professor of Business Administration, Harvard Business School, Boston, Massachusetts. His email address is [email protected].
doi=10.1257/jep.27.4.65
66
Figure 1
Interest Rate Policy, Inflation, and NBER Recessions
25
20
Inflation
15
10
Short-term
interest rate
5
0
5
10
1915
1920
1925
1930
1935
1940
1945
1950
1955
16
20
12
16
12
Inflation
(left scale) )
4
Short-term
interest rate
(right scale)
0
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
2005
2010
Notes: Due to data availability, the variables (and sources) are not the same for these two panels. In the
1955 2013 panel, the short-term interest rate is the federal funds rate from the Federal Reserve Board
while the inflation rate is the growth rate in the Consumer Price Index (CPI) (from the Bureau of Labor
Statistics) over the past 12months. The data used to construct the 19141955 panel are drawn from the
NBER Macrohistory Database. The short-term interest rate is the call money rate while inflation is given
by the 12-month growth rate in the NBERs estimate of CPI inflation. Periods of NBER recessions are
indicated by shading.
Julio J. Rotemberg
67
fighting inflation, and this aversion may have contributed to the Great Inflation of
the late 1960s and 1970s.
The third period I focus on is Paul Volckers pursuit of disinflation from 1979
to 1982. This involved a change in operating procedures that yielded unparalleled
interest volatility. This too seems responsive to a criticism, in this case that the Feds
focus on interest rates as an intermediate target led to large departures from the
Feds announced paths for the growth of monetary aggregates.
Finally, the fourth period from 19822007 was a time of renewed inflation
intolerance known as the Great Moderation. This period shows that, although the
Volcker-led deflation of the late 1970s and early 1980s was widely viewed as a success,
the Fed continued to change its approach to monetary policy. The federal funds
rate became more stable, for example, though this change was much more gradual
than the change in 1979.
A theme that emerges in these episodes is the tendency of the Fed to alter
its methods and its objectives drastically when critics successfully argue that bad
outcomes are a product of Fed mistakes. The Fed then acts as if it were penitent,
in that it becomes averse to this now vilified pattern of behavior. My discussion draws
on Romer and Romer (2002), in that they too emphasize the role of policymakers
ideas in the determination of Fed policy. However, many of the changes in ideas
emphasized in the existing literature on the Fed are unrelated to the penitence
scheme I propose here.
68
Federal Reserve Goals and Tactics for Monetary Policy: A Role for Penitence?
69
to lend to banks subject to runs. Moreover, the Fed resisted large-scale open-market
purchases to offset the declines in banking, even as the money supply dropped
substantially. Under pressure from Congress, such a program was started in April
1932, though it quickly ended in August.
Meltzer (2003, pp.327328, pp.341, 364) and Romer and Romer (2013) point
out that several Fed officials argued that, because banks were holding excess reserves,
monetary conditions were easy and attempts to loosen monetary policy further
would be ineffective. Meltzer (2003) and Romer and Romer (2013) suggests that this
explains the Fed inaction at the time, but this explanation seems incomplete as an
explanation of the Feds behavior because some Fed members including Chairman
Meyer favored increasing purchases even in 1933. Meyers lack of success presumably
owes something to people who saw expansionary policy not as irrelevant, but as actually detrimental. Negative views of this sort were expressed by Federal Reserve Bank
of Richmond President George Seay, who believed that the dangers of a further accumulation of reserves were greater than those of disposing of some securities (Open
Market Policy Conference Meeting, January 4, 1933). As excess reserves increased
further in the 1930s, this concern became more widespread and reserve requirements
were doubled between 1935 and 1937 (Meltzer 2003, p.509).
A common explanation for the Feds unwillingness to be more expansionary
in this period is that it stuck to the principles of its Tenth Annual Report (1924)
and to the procedures it had adopted in its wake (Calomiris and Wheelock 1998;
Meltzer 2003, p.400). As Friedman and Schwartz (1963, p.411) argued, however,
the expansionary policy of 1927 seems to represent a break from these principles
and procedures. Given that this break was later condemned, it seems possible that
penitence for departing from these principles in 1927 played a role in the 1930s.
If the Fed now viewed the 1927 open market purchases as a mistake because they
increased the liquidity of banks without a clear sense that this would be used for
productive lending, penitence would be consistent with the Feds aversion to excess
reserves during the 1930s.
Of course, other factors contributed to the Feds relatively tight stance. The
1931 increase in discount rates was clearly designed to stem gold outflows, for
instance, so faithfulness to the ideals of the gold standard must have mattered too
(Eichengreen 1992). However, Hsieh and Romer (2006) argue that even before the
gold inflows that followed the devaluation of 1933, the Fed had ample room for
more expansionary policies.
The level and quality of bank loans continued to play a role in Federal Open
Market Committee (FOMC) discussions for some time. In 1953, for example, New
York Federal Reserve President Allen Sproul told the FOMC that bank credit, except
for consumer credit and perhaps mortgage credit, has not moved out of line with
a balanced situation so that the evolution of several classes of bank loans was still
followed closely. This changed after Friedman and Schwartz (1963) published their
landmark study showing that the depth of the Great Depression was attributable to
the Feds concern for productive lending and its lack of attention to monetary
aggregates. Even as late as 1964, Friedman complained that independent central
70
banks inevitably fell under the influence of bankers and thus put altogether too
much emphasis on the credit effects of their policies and too little emphasis on the
monetary effects (US House of Representatives, 1964, p.73). Instead, Friedman
argued: Monetary policy ought to be concerned with the quantity of money and
not with the credit market (p.74).
Eventually, this perspective became dominant and, consistent with penitence
for its pattern of behavior during the Great Depression, members of the Federal
Open Market Committee stopped focusing on the asset side of bank balance sheets.
In the detailed memoranda of the first three meetings of the FOMC in 1970, for
example, there is no substantive discussion concerning the composition of bank
lending. The aggregate behavior of the banking sector, and total bank credit in
particular, were still discussed, though some members explicitly said that they
thought monetary aggregates were more relevant.
One has to wait until after the financial crisis of 2007 to see a resurgence of
the argument that the Federal Reserve should pay attention to the quality of loans
being made by financial institutions. The lead-up to the Great Recession featured
a substantial number of mortgages that ended up in default. The dynamics of the
financial crisis also suggest (as in the formal model of Shleifer and Vishny 1992)
that economic downturns can force banks to sell certain assets at fire sales prices.
As noted by Stein (2012), this means that an increase in one banks risky lending
imposes an externality on other banks because it reduces the fire sale prices at which
these other banks can dispose of their own assets. This externality suggests that the
main institution charged with macroeconomic stabilization should pay some attention to the quality of loans being made and to how they would fare in a downturn.
Friedman and Schwartzs (1963) analysis of the Great Depression also
seems responsible for Ben Bernankes (2002) apology on the Feds behalf for its
Depression-era policies. Consistent with a degree of penitence for these policies,
the Fed responded to the 2007 financial crisis with heroic efforts to prevent bankruptcies among liquidity providers and with dramatic increases in excess reserves.
Such policies were the opposite of the Feds passivity in the face of bank failures and
its reluctance to allow excess reserves to rise during the Great Depression.
Julio J. Rotemberg
71
72
in August 1958, and the Fed had been lowering interest rates since November
1966. Some members of the Federal Open Market Committee had been expressing
concern about inflation for several months. Martin recognized that the simple logic
of the economic situation implied the desirability of changing monetary policy and
then added, [b]ut the overriding need at this point was to get some restraint from
fiscal policy through a tax increase, and in his judgment that would be less likely if
Congress came to believe that adequate restraint was being exercised by monetary
policy (FOMC Minutes, September 12, 1967, p. 78). As Bremner (2004, p. 237)
notes in his biography, it was extraordinary for Martin to trust Congress to take
the initiative against inflation. Nonetheless, monetary easing continued. In August
1968, when the 12-month inflation rate had climbed to 4.5percent, Martin said that
the objective should be disinflation without recession (FOMC Minutes, August13,
1968, p.81). The birth of the Great Inflation may thus be partly explicable by penitence over causing recessions earlier.
The Federal Open Market Committee did set a course for tighter monetary
policy starting with the December 1968 meeting (Romer and Romer 1989). While
Martin was absent from this meeting, he endorsed tight policy from then on.
In the January14, 1969, FOMC meeting, in particular, he said that he thought
monetary policy was now on the right track and that, in his judgment, it would
be better to risk overstaying, rather than understaying, a policy of restraint
(Minutes, January14, 1969, p.73). The rate of money growth fell substantially. In
December 1969, Milton Friedman (1969, p.75) called this policy unduly restrictive and predicted it would lead to a recession. Indeed, a recession would soon
start in November 1969. A short while later, Friedman (1970, p. 68) expressed
satisfaction that his close friend and former teacher Arthur Burns would become
chairman of the Federal Reserve, and urged the Fed to shift promptly to a less
restrictivepolicy.
The Flourishing of Inflation under Arthur Burns
Like many contemporaries, Arthur Burns was openly critical of the Fed actions
that preceded the 1960 recession. Before taking office, he had written: The
abrupt shift in policy proved more restrictive than government officials planned or
expected. Largely as a result of their actions, the economic expansion that started
in April 1958 came to a premature end (Burns 1969, pp. 284 85). Consistent
with this, he was averse to creating recessions and told the Federal Open Market
Committee in 1973 that it was attempting to achieve an objective that had never
been accomplished beforethat of keeping the economy from developing an
inflationary boom but without releasing forces of a new recession (Memoranda of
Discussion, March20, 1973, p.108).
Burns agreed with Friedman that the Fed needed to reduce the volatility of
its own actions if it wanted to avoid unnecessary recessions. Friedman had testified, The chief defect in Federal Reserve policy has been a tendency to go too
far in one direction or the other, and then to be slow to recognize its mistake and
correct it (US House of Representatives 1964, p. 27). Echoing this sentiment,
Federal Reserve Goals and Tactics for Monetary Policy: A Role for Penitence?
73
Burns (1969,p.284 85) had written before becoming chairman we need to make
necessary shifts of economic policy more promptly, so that they may be gradual
instead ofabrupt.
Once Burns joined the Fed, his conviction that smooth changes in monetary
policy were desirable appears to have had two implications. First, he seemed
unwilling to react sharply to the inflation facing him. As he put it in his July 1974
testimony, From a purely theoretical point of view, it would have been possible for
monetary policy to offset the influence that lax fiscal policies and the special factors
have exerted on the general level of prices. . . . But an effort to use harsh policies of
monetary restraint to offset the exceptionally powerful inflationary forces of recent
years would have caused serious financial disorder and dislocation (US House of
Representatives 1974, p.257).
Second, Burns repeatedly expressed his intention to extinguish inflation over
a number of years. His July 1974 testimony, for example, also said that we shall
need to stay with a moderately restrictive monetary policy long enough to let the
fires of inflation burn themselves out. . . . We are determined to reduce, over time,
the rate of monetary and credit expansion to a pace consistent with a stable price
level (US House of Representatives 1974, p.253, 258). Similarly, in July 1977, Burns
said: Weve enunciated a policy and repeated it on every occasion, namely, that we
will gradually move our longer-range [money supply] targets down so that, several
years from now, the monetary basis for general price stability may be restored. Weve
been proceeding slowly, perhaps too slowly, but that is a debatable point (FOMC
Transcript, July 19, 1977, p.32).
However, certain apparent inconsistencies in Burnss statements have allowed
him to be characterized differently. In particular, Nelson (2005), DiCecio and Nelson
(2013), and Romer and Romer (2013) have attributed Burnss general failure to
act against inflation to his conviction that the Fed was somewhat impotent. In a
statement reflecting this conviction, Burns declared at the Federal Open Market
Committee meeting of April7, 1970, that the inflation that was occurringand
that was now being accentuated, how far he could not saywas of the cost-push
variety. That type of inflation, he believed, could not be dealt with successfully from
the monetary side and it would be a great mistake to try to do so (FOMC Memoranda of Discussion, April 7, 1970, p.49).
p. 49). Some members of the FOMC strongly
disagreed with this position.
Nonetheless, Burns continued to make statements of this sort, particularly
in connection with his advocacy of administrative controls to prevent excessive
increases in wages and prices. His July 1971 testimony, for example, stated: In my
judgment, and in the judgment of the Board as a whole, the present inflation in the
midst of substantial unemployment poses a problem that traditional monetary and
fiscal remedies cannot solve as quickly as the national interest demands. That is what
has led me, on various occasions, to urge additional governmental actions involving
wages and prices ((Federal Reserve Bulletin,, August 1971, p.662). According to Wells
(1994, p.72), this testimony was instrumental in pressuring a reluctant President
Nixon to impose wage and price controls less than a month later.
74
When these wage and price controls were eventually lifted, inflation rose
considerably, and the Fed became sufficiently concerned to raise interest rates to
the point of causing the 1974 recession. Indeed, interest rates were increased even
as this recession was in progress. As noted by Wells (1994, p.136), Burnss July 1974
testimony alludes to the costs that a fight against inflation would impose, and this
suggests he was aware at the time that he had temporarily departed from gradualism. In any event, the ensuing disinflation brought Burns a great deal of notoriety
and prestige (Wells 1994, p.178).
Alternative Sources of the Great Inflation
The Great Inflation of the 1970s has been attributed to a number of additional
forces. Fed officials may, for example, have felt that they could not be tough on
inflation for fear of the reactions in Congress and the Executive Branch (Burns
1979). What is certain is that Nixon pressured Burns to maintain a high rate of
money growth on the eve of the 1972 election. On the other side, it is difficult to
provide concrete evidence that political pressure for looser monetary policy had
much effect (Mayer 1999, p.64 82); after all, politicians sometimes were extremely
critical of the Fed for having caused inflation..1
Another view emphasizes the influence of the idea that a long-run downwardssloping Phillips curve existed, so that higher inflation would bring down
unemployment (Taylor 1992, p. 13; DeLong 1997). Analyses based on this idea
were common among members of the Council of Economic Advisors in the 1960s
(Romer and Romer 2002, p. 20). However, as far as I know, no one has found a
Fed official arguing for higher inflation on the grounds that this would lower
long-term unemployment. Indeed, several Fed officials went out of their way to
distance themselves from this idea. For example, Martin testified in January 1963
that he thought the Phillips Curve was a fallacy (Federal
(
Reserve Bulletin,, February
1963, p. 124). Indeed, he suggested that the long-run relation between inflation
and unemployment was actually upwards sloping when he said that low rates of
unemployment have been facilitated, and indeed made possible, by the absence
of inflationary expectations on the part of both labor and management (Federal
(
Reserve Bulletin,, December 1965, p.1,678). Similarly, in the hearings conducted by
Congressman Wright Patman in 1974 to pin the blame for inflation on the Fed
(and thereby absolve budget deficits), Burns said the so-called tradeoff between
inflation and unemployment was quite misleading (US House of Representatives
1974, p.252). DiCecio and Nelson (2013) offer extensive additional evidence that
Burns did not think a rise in inflation would lower unemployment.
In a very interesting article, Weise (2012) shows that Federal Open Market Committee discussions were
more likely to mention politicians who desired looser conditions in meetings in which the committee
chose to loosen monetary policy. Note, however, that this correlation may reflect less the effect of outside
pressure than the desire to present all the arguments that come to mind in favor of ones desired course
ofaction.
Julio J. Rotemberg
75
M1 and M2 are measures of the total money supply. While their definitions changed somewhat over
time, M1 always included currency in circulation and most checking accounts, while it always excluded
savings deposits and small time deposits, both of which were always included in M2.
76
alternative suggestion would have been to encourage the Fed to change its interest
rate objectives more vigorously.3 However, Friedman argued that the Federal Open
Market Committee should target the growth of reserves (or the monetary base) and
let all interest rates be determined by the market.
Mayer (1999, p.45) and Clarida, Gal, and Gertler (2000) emphasize the weakness of the response of
the federal funds rate to inflation and output in this period, and this is related to Friedmans complaint
in his November 1975 statement that the Fed did not lower interest rates rapidly enough during the 1974
recession (US Senate, 1975, p.38).
4
The procedures that the Fed adopted were not identical to those recommended by its critics. The
focus on nonborrowed as opposed to total reserves or the monetary base was deemed by Allan Meltzer
to lead to excessively volatile money growth (Rasche, Meltzer, Sternlight, and Axilrod 1982). Moreover,
according to Friedman (1982), the requirement that banks hold reserves on the basis of their past (rather
than their current) deposits also complicated the control of money. One reason the Fed may have settled
on nonborrowed rather than total reserves might have been to stabilize interest rates somewhat.
Federal Reserve Goals and Tactics for Monetary Policy: A Role for Penitence?
77
Figure1 shows that interest rates did become substantially more volatile after
these procedures were instituted. The average of the absolute value of monthly
changes in the federal funds rate from October 1979 to November 1980 was
145basis points. For the twelvemonthly changes from September 1978 to October
1979, it had been only 42basis points. More generally, the volatility of interest rates
immediately after October 1979 was both historically unprecedented and contrary
to a key goal of the founding of the Federal Reserve (Strong 1922 [1989]). Consistent with that goal, the creation of the Fed had stabilized seasonal fluctuations in
interest rates (Mankiw, Miron, and Weil 1987).
The procedures also had a mixed record in terms of keeping money growth
rates within their announced ranges. In the period between October6, 1979, and the
Federal Open Market Committee meeting of January8, 1980, money growth rates
were close enough to their targets that Governor Partee considered the procedures
to have been a successful experiment in the latter meeting (Transcript, p.14). On
the other hand, monthly money growth rates proved quite volatile under the new
procedures (McCallum 1985). The standard deviation of monthly M1growth rates
was 9.3percent from November 1979 to November 1981, whereas it had been only
4.6percent from September 1977 to September 1979 inclusive.5 Not surprisingly,
Volcker complained that we got criticized by the bankers when they were here the
other day for having too much volatility in the money supply growth and too much
volatility in interest rates (Transcript, September16, 1980, p.9).
Moreover, there were long periods in which money growth exceeded its official target. In particular, the growth in M1 equaled 11 percent in the 11 months
from May 1980 to April 1981, and this led the Fed to be severely criticized by some
Reagan administration officials (Greider 1987, p. 378). One potential reason for
this growth was that money market mutual funds and checking accounts that paid
interest (NOW accounts) grew in this period. Financial innovation of this sort led
Governor Morris to exclaim, we simply dont have any basis for measuring what
transactions balances are anymore (FOMC Transcript, July7, 1981, p.24).
These failures to meet money targets should not be taken to mean that the
procedures failed to have an effect on policy. Perhaps the most telling evidence that
they mattered is that Volcker complained about their role in the October5, 1982,
meeting in which these procedures were at least partially jettisoned. Volcker was
unhappy at the interest rate that had resulted from the previous meetings decision
concerning nonborrowed reserves and said: What we did last time was unacceptable to me. I just want to make that plain. I think we made a mistake last time . . .
[I]ts unfortunate that we ended up at this meeting with the federal funds rate and
private rates about 1percentage point higher than they were at the time of the last
These figures and those below are based on current measures of seasonally adjusted M1. In December
1980, before all these data became available, twoFederal Reserve economists presented a paper at the
AEA annual meetings saying that money growth over longer periods of time was close to its targets under
the new procedures (Axilrod and Lindsey 1981).
78
meeting because we had a high M1figure in September. That was the only reason it
happened (FOMC Transcript, October5, 1982, p.32).
Rationales for the October 1979 Procedures
Istart with the rationales that were given when the October 1979 procedures
were first instituted and then discuss the reasons why they remained in place even
after they had quite clearly failed to stabilize money growth. Volcker seemed an
unlikely champion for these new procedures because he had stated, for example in
a 1978 Journal of Monetary Economics article, that the demand for money was sufficiently unstable in both the short run and the long run that fixing money growth
rates would lead to undesirable movements in interest rates. In Volcker (1978), he
also seemed somewhat uncertain of the Feds ability to hit its money growth targets
by setting the level of reserves.
Nonetheless, Volcker gave an argument for these procedures in October 1979,
namely that their announcement would lower inflation expectations. As he explained
in Greider (1987, p.111), What I hoped was that there would be a strong reaction in
the markets. . . . The sign of psychological success was whether long-term rates would
stabilize and start coming down. This did not happen right away; long-term rates rose
alongside short-term rates immediately after the October 1979announcement.
Meltzer (2009, pp.1040, 1064, 1075, and 1093) suggests that, more generally,
the 1979 procedures had only a modest effect on inflation expectations, and that
these fell mainly when economic activity slowed. After the procedures had been
operating for a year, Volcker himself seemed to doubt that they mattered for inflation expectations. In December 1980, he said If we, in effect, go to the brink or let
some of these things happen that we have not allowed to happen during the entire
postwar period, people are not expecting that and they are not going to be very
happy if and when it happens. And Im not at all sure that we can change inflationary
expectations without it happening (FOMC Transcript, December19, 1980, p.62).
Governor Partees initial support may have been based in part on his view
during the September 1979 meeting: I think its important, very important, that
we try to keep the aggregates within the ranges that we specify (FOMC Transcript,
September18, 1979, p.26). Partee recalled a different reason for his approval in his
interview in Greider (1987, p.112). There, he declared that the new procedures dealt
with the Feds past tendency of sticking stubbornly with a strong position too long
and causing more damage to the economy than it had intended and that in recessions, particularly in the 1974 75 recession there [was] also a hesitancy to reduce
interest rates once they have been raised. As it happens, this hesitancy to lower rates
may have had some benefits. While interest rates rose substantially when the new
procedures were instituted, the decline in rates when the 1980 recession started was
so dramatic that the recession was over almost immediately, and the reduction in
inflation to acceptable levels had to wait until the arrival of the 1981 82 recession.
Partee did not mention any concern he might have had with sticking stubbornly
to the 1979 procedures themselves if velocity shifted. Such velocity shifts did, in fact,
eventually lead to difficulties with the procedures.
Julio J. Rotemberg
79
As the procedures were being abandoned, two arguments for keeping them
became prominent. The first was that the procedures provided political shelter
for raising rates to fight inflation (FOMC Transcript, February 8 9, 1983, p.24,
26, 29, and 30). The procedures may have diminished the criticism of the Fed, but
they certainly did not eliminate it. Indeed, the high rates of interest of 1982 had
led to a strong movement in Congress to reduce the Feds independence (Greider
1987, p.474).
A second argument for keeping the procedures intact was made at the
October5, 1982, meeting in which the Federal Open Market Committee decided to
announce that it would pay less attention to M1. Federal Reserve of St. Louis President Lawrence Roos, an ardent supporter of monetary targets, argued that reducing
the official importance of the growth rate of M1 would imperil the Feds credibility
and would be misconstrued by the markets (FOMC Transcript, October5, 1982,
p.48). In fact, the reduction in short-term interest rates that followed this meeting
was accompanied by a reduction in long-term rates.
It would seem, then, that the arguments that were given for initiating and maintaining these procedures were not very strong. This suggests another possibility, namely
that these procedures embodied a form of penitence for the pre-1979 procedures,
which critics had successfully associated with the Great Inflation. Roos emphasized this
association at the October5, 1982, Federal Open Market Committee meeting when
he argued that the high interest rates that prevailed at the time were the ultimate
consequence of irresponsible monetary policies throughout the world and to a wellmeant effort on the part of the Federal Open Market Committee . . . to try to do just
what were doing today, and that is to lean against interest rate movements. I think that
contributed in a major way to inflation (Transcript, October5, 1982, p.48).
Extreme concern with the possibility of uncontrolled money growth if interest
rates were stabilized even at very high levels was also on display at the earlier
meeting of July1, 1982, when Partee noted that he seemed to have shocked quite
a number of people with my suggestion that we ought to put a cap on the funds
rate. He had proposed that the federal funds rate should not be allowed to rise
above 15percent. Since the rate that day was equal to 14.73percent, this cap was
perceived as being potentially binding. At the same time, the unemployment rate
was 9.8 percent and the growth rate in the Consumer Price Index over the last
12months had been 6.5percent, so a 15percent federal funds rate would have
been likely to be associated with a high real interest rate. This led Partee to argue
that this would give us an upper limit that is not unreasonable.
Nonetheless, Partee was asked by Governor Henry Wallich, in apparent disbelief,
But if it got there, we would provide unlimited reserves? and by Roos how would
that differ from the pre-1979 practices of our Committee? When Partee answered
it would be similar on the top side, Federal Reserve Bank of Atlanta President
William Ford said Are you implying that there wasnt a change in October 79?
If Iunderstood you, you said it would be similar to pre-October79that there is
precedent for it (FOMC Transcript, July1, 1982, p.55). One reason for the aversion to returning to the pre-October 1979 may have been that, as Volcker and others
80
suggested, some members of the Federal Open Market Committee may have been
afraid of losing their self-discipline if they were not constrained by rules (FOMC
Transcript, December21, 1982, p.29, 38, and 43).
Federal Reserve Goals and Tactics for Monetary Policy: A Role for Penitence?
81
the federal funds rate much earlier. Indeed, as Thornton (2006) documents, some
Federal Open Market Committee members openly suspected Volcker of doing so
as early as 1983.
Unlike what happened in October 1979, the public was not told that a change
in the conduct of monetary policy had taken place. No target for the federal funds
rate was announced throughout the 1980s or into the early 1990s. Rather, just as
had been true since 1983, the press releases continued to suggest that the federal
funds rate would remain within a 4percent range until the next meeting. Meanwhile, the Fed continued to publish its expected ranges for the growth in monetary
aggregates, though it softened its commitment to these ranges.
Even in February 1993, many members of the Federal Open Market Committee
expressed apprehension about releasing their federal funds target (Transcript,
February 2, 1993, p. 62 67). But by then, movements in velocity of M1 and M2
had become so large that the Feds plans regarding the growth in these aggregates
were not very informative. After this point, its statements started explaining the
federal funds rate changes that the FOMC had instituted in the past. Still, as late as
March 1997, when FOMC members voted to raise the federal funds rate from 5.25
to 5.5percent, the public minutes only commented on the past rate of 5.25percent.
This lack of transparency would finally end in August 1997, when the intended
federal funds rates started to be published in the official minutes, although this
was accompanied by a statement that the operating procedures of the Fed would
not change. After this, the Fed gradually expanded the amount of information it
released about its intentions concerning future policy (Woodford 2005). The Fed
managed to stop supplying any monetary targets whatsoever when the legislation
requiring these expired in 2000.
One of the most striking aspects of US monetary policy in this period is that
the simple rule proposed by Taylor (1993) in which the suggested federal funds
rate is a function of inflation (as measured by the Consumer Price Index) over the
last year and of the distance between current real GDP and trend GDPleads to
a federal funds rate that is remarkably close to the actual one for the period 1987
to 2000. This too was the result of a gradual evolution. Even though the correspondence is weaker before 1987, the relatively fast rise in the federal funds rate in 1983
and early 1984, as well as its subsequent decline were consistent with the Taylor
rule. As Kahn (2012) demonstrates, discussion of the implications of variants of the
Taylor rule for the federal funds rate quickly became part of the fabric of meetings
of the Federal Open Market Committee. Nonetheless, the FOMC drifted towards
applying the coefficients of the Taylor rule to their anticipations of future values of
inflation rather than to the past values (FOMC Transcript, January27, 2004, p.76).
Conclusion
This paper has suggested that some of the changes in the Feds approach to
monetary policy are consistent with a form of penitence, where this penitence is the
82
end result of a three-step process. First, there are some deplorable economic results
such as those in the initial 1930 downturn, the full Great Depression, the recessions of
1957 and 1960, or the Great Inflation. Second, critics attribute these results to patterns
of Fed behavior that are interpreted as having been mistaken. Third, the Fed acts as if it
implicitly accepted one of these criticisms and becomes averse to the criticized pattern
of behavior. It is possible to view this form of penitence as helping the Fed perfect its
approach to monetary policy. Particularly if one agrees with the critics, this penitence
would represent a form of learning: it leads the Fed not to repeatmistakes.
Without further evidence, however, it seems premature to view this form of
penitence as involving an accumulation of knowledge of the form one typically
associates with learning. To see this, it is sufficient to imagine a two-state system
that toggles from one state to the other whenever something bad happens outside
the system. Such a system responds to poor outcomes, but is essentially devoid of
historical information at all times.
The Fed has access to a rich menu of policy approaches, and one role of
outsiders is to help devise new ones. Still, there are two aspects of the Feds evolution that seem somewhat similar to the two-statesystem Ijust described, and which
raise concerns over the extent to which the Feds response to bad outcomes involves
the accretion of knowledge. First, many of the changes in Fed behavior that follow
such outcomes seem later to be reversed. In particular, the Fed both gained and lost
its aversion to stabilizing interest rates, as well as its aversion to inducing recessions
in response to inflation. Second, some knowledge seems to be lost when the Fed
develops a new aversion. Entire topics can practically vanish from the discussions
of the Federal Open Market Committee. As an example, the FOMC meeting of
January26, 1960, contained a remark by President of the Richmond Federal Reserve
Hugh Leach in which he based his assessment of the tightness of monetary policy
on the evolution of loans to build up inventories (Minutes, p.20). Information of
this sort stopped being incorporated into policy discussions when the Fed reduced
its attention to the asset composition of bank balance sheets.
Even if one believes that the changes in approach triggered by poor outcomes
have led to only limited accretions in Fed knowledge, the Fed may have accumulated a great deal of information at other times. During the Great Moderation, for
example, the Fed appears to have gradually learned to stabilize interest rates to an
ever-greater extent.
So how might the Feds knowledge and approach evolve in response to the
financial crisis of 2007? As was the case with previous bad outcomes, critics who
blame this crisis on Fed mistakes do not speak with a single voice. Fleckenstein
(2008) argues that the Fed started being prone to generate asset bubbles by having
low interest rates as far back as 1987, when it lowered rates in response to a stock
market crash. By contrast, Taylor (2012) applauds the Feds approach from 1987 to
2003, and singles out for criticism the post-2003 period in which the Fed set interest
rates below those implied by a backward-looking Taylor rule.
If such criticisms became accepted by the Fed to some extent, they could
lead to dramatic changes in the Feds approach by creating new aversions. The
Julio J. Rotemberg
83
Fed could, for example, seek to tamp down any potential increase in asset prices
that it regarded as a bubble, though it seems likely that such an approach would
quickly lead the Fed to be criticized for causing unnecessary losses in output.
Acceptance by the Fed that it had mistakenly kept interest rates too low starting
around 2003 could result in different aversions. If a consensus developed that the
Feds mistake was to abandon a Taylor rule based on past values for one based
on Fed projections, the Fed could become averse to using its forecasts in setting
policy, at least for a time.
Another move that could come to be seen as an error is the Feds policy of
announcing its expectations concerning future policy actions. At the December9,
2003, meeting of the Federal Open Market Committee, Governor Donald Kohn
said policy is quite easy, quite stimulative and nonetheless recommended that the
Fed continue to take [its] risks on the easy side of policy. At the same time, he
worried about the FOMCs flexibility to raise rates given that its August 2003 statement had said that policy accommodation can be maintained for a considerable
period (FOMC Transcript, December 9, 2003, p. 67). This raises the possibility
that Kohn felt trapped into keeping interest rates low to honor the Feds implicit
promise to do so. Thus, there is the possibility that the Feds use of forward guidance concerning its future policies could come to be seen as a mistake. Consistent
with penitence, the Fed might decide in the future to steer clear of communicating
in a way that seeks to affect expectations of future policy.
Papers in the volume Is Inflation Targeting Dead? (Reichlin and Baldwin 2013)
propose more gradual changes that would not require the development of an
aversion to past Fed practices. As discussed earlier, one possibility along these lines
would be to return partially to the pre-1963 view that monetary policy ought to
respond to the quality of assets held by institutions with monetary liabilities (Stein
2013). More gradual changes may prove less prone to reversals, and this would
constitute an advantage. To institute such gradual changes, more radical changes
may need to be held at bay. To successfully counter arguments for more radical
change it might help to understand how, in the past, critics often succeeded in
championing the abandonment of practices that, eventually, came to be seen as
beneficial once again.
This is a revised version of a paper prepared for a symposium on The First Hundred Years
of the Federal Reserve: The Policy Record, Lessons Learned, and Prospects for the Future,
held at the National Bureau of Economic Research on July10, 2013. I wish to thank ChangTai Hsieh, Robin Greenwood, Anil Kashyap, Ulrike Malmendier, Edward Nelson, Robert
Pindyck, Christina Romer, David Romer, and Timothy Taylor for comments and, especially,
Rawi Abdelal for several conversations and for his help in crystallizing the argument of
thispaper.
84
References
Axilrod, Stephen H., and David E. Lindsey.
1981. Federal Reserve System Implementation
of Monetary Policy: Analytical Foundations of the
New Approach. American Economic Review 71(2):
246 52.
Bernanke, Ben. 2002. On Milton Friedmans Ninetieth Birthday. Speech at the
Conference to Honor Milton Friedman,
University of Chicago, Chicago, IL, November 8.
http://www.federalreserve.gov/BOARDDOCS
/SPEECHES/2002/20021108/.
Bremner, Robert P. 2004. Chairman of the Fed:
William McChesney Martin and the Creation of the
American Financial System. New Haven: Yale University Press.
Bullard, James, and Stefano Eusepi. 2005. Did
the Great Inflation Occur Despite Policymaker
Commitment to a Taylor Rule? Review of Economic
Dynamics 8(2): 324 59.
Burns, Arthur F. 1969. The Business Cycle in a
Changing World. New York: NBER and Columbia
University Press.
Burns, Arthur F. 1979. The Anguish of Central
Banking. The 1979 Per Jacobson Lecture,
Belgrade, Yuogslavia, September 30.
Calomiris, Charles W., and David C. Wheelock.
1998. Was the Great Depression a Watershed for
Monetary Policy? In The Defining Moment: The
Great Depression and the American Economy in the
Twentieth Century, edited by Michael D. Bordo,
Claudia Goldin, and Eugene N. White, 23 65.
University of Chicago Press.
Carboni, Giacomo, and Martin Ellison. 2009.
The Great Inflation and the Greenbook. Journal
of Monetary Economics 56(6): 831 41.
Clarida, Richard, Jordi Gal, and Mark Gertler.
2000. Monetary Policy Rules and Macroeconomic
Stability: Evidence and Some Theory. Quarterly
Journal of Economics 115(1): 147 80.
DeLong, J. Bradford. 1997. Americas Peacetime Inflation: the 1970s. In Reducing Inflation:
Motivation and Strategy, edited by Christina D.
Romer and David H. Romer, 24776. University of
Chicago Press.
DiCecio, Riccardo, and Edward Nelson. 2013.
The Great Inflation in the United States and
the United Kingdom: Reconciling Policy Decisions and Data Outcomes. In The Great Inflation:
The Rebirth of Modern Central Banking, edited by
Michael D. Bordo and Athanasios Orphanides,
393 438. University of Chicago Press.
Eichengreen, Barry. 1992. Golden Fetters: The
Gold Standard and the Great Depression, 19191939.
New York: Oxford University Press.
Federal Reserve Goals and Tactics for Monetary Policy: A Role for Penitence?
85
86
Barry Eichengreen is the George C. Pardee and Helen N. Pardee Professor of Economics
and Political Science, University of California, Berkeley, California. His email address is
[email protected].
http://dx.doi.org/10.1257/jep.27.4.87
doi=10.1257/jep.27.4.87
88
of international factors in Fed decision making; and my personal favorite, William Adams Browns (1940)
The International Gold Standard Reinterpreted, which devotes successive chapters to the United States and
its centralbank.
Barry Eichengreen
89
90
Figure 1
Discount Rate, Federal Reserve Bank of New York
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
42
19
41
19
40
19
39
19
38
19
37
19
36
19
35
19
34
19
33
19
32
19
31
19
30
29
19
28
19
27
19
26
19
25
19
24
19
23
19
22
19
21
19
20
19
19
19
18
19
17
19
16
19
91
19
0.00 1
in Figure 1, with the objective of making the pound sterling a relatively more
attractive investment, inducing capital to flow from New York to London, and
helping the Bank of England push sterling up to the pre-World WarI exchange
rate against the dollar, where it could then be stabilized (Clarke 1967).2 To ensure
a wider impact on financial markets, the New York Fed also purchased US Treasury securities, in the course of so doing, helping to establish the efficacy of open
market operations as an instrument of monetary control. After importing gold for
51consecutive months from December 1920 to April 1925, the United States now
exported it instead. In January 1925, the Federal Reserve agreed to advance the
British Treasury an additional $200million in gold while encouraging a banking
syndicate led by J.P. Morgan to provide a $100million line of credit.
This policy was not an act of altruism: Strong firmly believed that helping Britain
back onto the gold standard, by paving the way for a broader stabilization of exchange
rates, would lend stimulus to US exports and economic growth. But that view was not
universally shared within the Federal Reserve System. Strongs initiative was criticized,
for example, by Adolph Miller, founding Governor of the Federal Reserve System
and previously professor at the University of California at Berkeley. Miller argued that
the resulting monetary policy was inappropriately loose for domestic circumstances.
Along with others like then-Secretary of Commerce Herbert Hoover, Miller warned
that low interest rates were fueling unsustainable real estate bubbles across Florida and
2
At this time, individual Federal Reserve Banks were free to change their own discount rates, subject to
the review and determination of the Federal Reserve Board.
Does the Federal Reserve Care about the Rest of the World?
91
from Detroit to Chicago. This was the first full-blown controversy within the Federal
Reserve over the relative importance of domestic and international objectives.
The second controversy arose in 1927, when Strong again proposed cutting
interest rates, this time in order to help Britain stay on the gold standard. Miller would
have objected, but he was on summer holiday in California. When he returned, he
mounted a strenuous attack on the policy as inappropriate for an economy already
recovering from a brief recession. Monetary historians have been similarly critical (as
described in Meltzer 2003), suggesting that a policy looser than appropriate from a
domestic standpoint helped to fuel the commercial real estate boom and Wall Street
run-up of the late 1920s, both of which came down with a crash. It would have been
better, they conclude, for the Fed to keep its eye on the domesticball.
The traditional constraint in which a central bank needs to hold interest rates
high to attract capital inflows and defend the exchange rate then reemerged with a
vengeance late in 1931. In a shock to financial markets, Britain departed from the
gold standard on September 21, 1931. The dollar weakened against the continental
European currencies, and gold losses mounted rapidly. In part, this reflected worries
about US competitiveness as it became clear that some two dozen other countries
were preparing to follow Britain in devaluing their currencies. Even more important was psychological contagionthe wake-up-call effect: if one reserve-currency
country could depreciate its currency, it was no longer inconceivable that the other,
the United States, might follow.
At this point the Fed made its priorities unambiguously clear. On October 8,
1931, the directors of the New York Fed voted to raise its discount rate by 100basis
points and then a week later by another 100 basis points (see Figure 1). Other
Reserve Banks followed. These higher interest rates were designed to attract financial capital from abroad, or at least discourage it from leaving, thereby supporting
the US dollar. The wisdom of the decision can be questioned, coming as it did in the
midst of the Great Depression when domestic conditions warranted lower interest
rates. But it clearly privileged exchange rate stability over price stability, financial
stability, and economic stability.
The final attack on the dollar came in FebruaryMarch 1933 in the interregnum
between the Hoover and Roosevelt administrations. Worries that the new president
might devaluesomething that only he, together with the Congress, and not the
Federal Reserve could decideencouraged capital flight from the United States
to France, Switzerland, and Belgium, countries seen as having stronger currencies
(Wigmore 1987). The decision in February 1933 to let Henry Fords Union Guardian
Trust Company go underthe equivalent for its time of the September 2008
bankruptcy of Lehman Brothersthen ignited a nationwide banking panic (Kennedy
1973; Awalt 1969). At this point there was essentially no choice but to embargo gold
exports, close the banks, and regroup. On his first day in office, President Franklin
Roosevelt invoked the Trading with the Enemy Act for the necessary authority.3
3
One cannot help but be reminded of UK Chancellor of the Exchequer Gordon Brown invoking the UK
Anti-Terrorism Act to prevent the repatriation of Icelandic assets in 2008.
92
At this point, the first era in which international considerations played a prominent role in US monetary policy drew to a close. President Roosevelt took the next
step in April 1933, making clear that there would be no early return to the gold
standard. Currency would no longer be exchanged for gold, and individuals were
required to turn in their gold and to receive currency in exchange. The Congress then
passed a joint resolution canceling all contracts, public and private, that called for
payment in gold. Starting in October, FDR used the authority of the Reconstruction
Finance Corporation, an emergency agency created in 1932, to purchase gold with
US dollars. Pushing up the dollar price of gold was equivalent to pushing down the
exchange rate of the dollar against the currencies of other countries still on the gold
standard. In effect, the Executive Branch had taken emergency control of monetary
policy. Finally, in January 1934 the president agreed to stabilize the price of gold at
$35 per ounce. This was not a formal return to the gold standard, since freedom
for individuals to hold gold and the reintroduction of gold clauses into private and
government contracts were not part of the bargain. Nonetheless, changes in the gold
stock again translated into changes in the monetary base (the money supply narrowly
defined)with some important exceptions detailed belowbecause the government
again bought gold for currency when it flowed in from abroad, to keep the US dollars
price from falling.
These steps inaugurated a new era in which international considerations played
little role in US monetary policy. There was no need for high interest rates to stem
capital outflows. The new higher dollar price of gold attracted the yellow metal toward
the United States; more generally, devaluation enhanced the countrys international
competitiveness. In addition, as the outlines of World War II became visible, foreign
capital fled in growing volumes to American shores. Foreign economic and financial
conditions mattered less for the US economy than in the 1920s, now that global trade
had collapsed due to the imposition of higher tariffs, and long-term international
investment had been discouraged by exchange controls and sovereign debtdefaults.
Barry Eichengreen
93
Confronted with a US Treasury that was seeking to carry out monetary policy in
this way, the Fed sought to regain the ability to influence money and credit markets.
Its campaign was unsuccessful: with the outbreak of World War II and the very
large government deficits of that time, the Fed was drafted into keeping federal
borrowing costs low by holding interest rates on Treasury bills at 0.375 percent
and Treasury bonds at 2.5 percent. The practice continued, despite growing Fed
resistance, for two years following World WarII. Current controversies over how a
change in Federal Reserve policy would affect federal government borrowing costs
in some ways echo this earlier situation.
This long period of fiscal dominance over monetary policy eventually came to
an end with the TreasuryFed Accord of 1951, which officially ended the expectation that the Fed would purchase US Treasury securities in whatever quantities
were necessary to keep interest rates at these low levels. Recent scholarship portrays
monetary policy over the balance of the 1950s in a relatively favorable light (for
example, Romer and Romer 2002a). More pertinent from the standpoint of this
paper, monetary policy at this time focused on keeping inflation low and, to a
lesser extent, on responding to temporary deviations from full employment. The
Minutes of the Federal Open Market Committee in this period (available online
at http://fraser.stlouisfed.org/publication/?pid=677) offer little emphasis or
even mention of the US dollar exchange rate, the US balance of payments, or the
effect of US monetary policy on the rest of the world. Of course, the discussions
do mention exports and imports, since these variables were seen by members
of the committee as containing information useful for forecasting the future
paths of inflation and what would later be called the output gap. But beyond
such comments, international factors do not appear to have impinged on the
committeesdeliberations.
There was a commitment after the Bretton Woods agreement of 1944
to continue stabilizing the price of gold at $35 an ounce and pay out gold on
demand to official foreign creditors, but no matter. This was the period of the
dollar shortage. The term refers to the difficulty experienced by other countries in acquiring, whether through exporting or foreign borrowing, the dollars
they needed in order to finance merchandise imports from the United States
(Kindleberger 1950). At this time, gold held by the US monetary authorities far
exceeded the foreign liabilities of the Federal Reserve, US commercial banks, and
the US government combined. In this sense, the orientation of monetary policy
was not constrained by internationalimplications.
This situation began to change around 1960, when US foreign monetary
liabilities first threatened to exceed US gold reserves. Investors worried that John
F. Kennedy, if elected president, might follow in Roosevelts footsteps and devalue
the dollar to get the economy moving again (to quote Kennedys campaign literature). Robert Triffin (1960) published the first of a series of books in which he
warned that if the dollar remained the only source of global liquidity (other than
gold, which still lurked behind the scenes at the agreed-upon price of $35 per
ounce), a crisis of confidence in the greenback would ultimately develop. Triffins
94
dilemma was based on the insight that the expansion of the global economy would
bring growing demands for international liquidity. If dollar-denominated claims,
and specifically US Treasury bonds, were the primary source of such liquidity (on
the margin), then US foreign liabilities would eventually come to exceed US gold
reserves. When this occurred, it would call into question the ability of the US
authorities to convert these liabilities into gold at the fixed price of $35 an ounce,
creating a crisis of confidence. Alternatively, if the authorities took steps to limit
US current account deficits and foreign lending, the rest of the world would be
starved of liquidity in an ongoing dollar shortage, and international transactions
generally would suffer.
The expectation that President Kennedy would devalue the US dollar proved
erroneous, but the other worries were not without foundation. Inflation accelerated in the later 1960s and grew more erratic. The goals of Federal Reserve
policy shifted from an overarching emphasis on inflation to greater attention to
unemployment and economic growth. Romer and Romer (2002b) argue that this
period saw a revolution in ideas in which policymakers forgot much of what they
had learned about the natural rate of unemployment. Instead, they overestimated
potential output and succumbed to the temptation to use monetary policy to target
real variables. Federal Reserve Chairman William McChesney Martin believed
that the Fed had an obligation to help keep federal debt service at manageable
levels, which constrained monetary policy in the direction of lower interest rates
as budget deficits grew.
Even if the Fed was concerned about its gold losses and other international
variables, it might nonetheless be hard to detect that concern amongst these
other changes. However, in Bordo and Eichengreen (2008), my coauthor and
Iargue that the Fed paid considerable attention to balance-of-payments considerations in the first half of the 1960s, tightening when it grew worried by the pace
of gold outflows. In addition to his concern with debt service, Fed Chairman
Martin was a firm believer in maintenance of the gold peg. Already in 1960,
the Fed abandoned its traditional bills-only policy (the policy of buying only
short-term Treasury debt) in order to let short-term interest rates rise, attract
capital flows, and strengthen the balance of payments (Solomon 1977, p. 36).
The Minutes of the Federal Open Market Committee (FOMC) regularly refer to
balance-of-payments considerations. Many of these statements, in an echo of the
1920s, came from the President of the Federal Reserve Bank of New York, Alfred
Hayes (Meltzer 2013). A count of references in the minutes and memoranda of
the FOMC, as available from 1950 through March 1976, normalized by pages,
shows mentions (and presumably concern) relating to the balance of payments
to be mounting in the first half of the 1960s and peaking around mid-decade, as
shown in Figure2.4
4
Where dots are missing in Figure2 (as in the first half of the 1950s), there were zero mentions. Normalizing by pages adjusts for the fact that the minutes and memoranda tended to grow longer over time,
although raw counts show basically the same picture. In principle, it should be possible to extend this
Does the Federal Reserve Care about the Rest of the World?
95
Figure 2
References to Balance of Payments and Related Terms in the Minutes
Recording Unit
0.8
Exchange rate
Balance of payments
Trade balance/balance of trade
0.6
0.4
0.2
0.0
1950
1955
1960
1965
1970
1975
Notes: Mentions for each term are taken from minutes and memoranda of discussion for all meetings and
telephone conferences of the Federal Open Market Committee from 1950 through March 1976. Data
are fit with a 2nddegree local polynomial LOESS regression with span parameter = 0.75 (indicating
that 75 of the data are used to estimate each local regression) and the bands represent 95 percent
confidenceintervals.
analysis beyond 1976 when transcripts of Federal Open Market Committee meetings become available.
However, the transcripts are sharply discontinuous with the minutes in terms of comprehensiveness; in
addition, pagination and font size are quite different, and the pagination and format of the transcripts
themselves are not constant over time. This makes trends in both raw and per-page counts more difficult
to interpret. I have resisted the temptation. The outlier in mid-1963 is from a meeting in a period of
heightened concern about dollar stability (Eichengreen 2000). The Federal Reserve system had recently
drawn its full $150million swap line with the Bundesbank, and dollar weakness had been a prominent
topic at the most recent monthly meeting of the Bank for International Settlements, where the Fed had
been represented by Charles Coombs of the Federal Reserve Bank of New York.
96
Romer and Romer (2002b, p. 57) similarly make the point that balance-of-payments considerations
prevented the Fed from being as expansionary as it would have otherwise wanted in the first half of
the1960s.
6
Indeed, the observation of how the federal funds interest rate changed with changes in inflation and
unemployment after 1983 is what led the eponymous Professor Taylor to develop his rule.
Barry Eichengreen
97
mattered principally insofar as they were relevant to the future evolution of inflation and the output gap.
A combination of factors explains why international factors were less influential in the late 1970s and early 1980s than in the 1920s, 1930s, and 1960s. After
the final collapse of the Bretton Woods agreement in 1973, there was no longer
an exchange rate peg or statutory gold price to defend. The 1977 legislation gave
the Fed a mandate to pursue price stability and full employment but said nothing
about the exchange rate, balance of payments, or international financial stability.
The experience of the 1970s had taught that a Federal Reserve that failed to achieve
price stability would lack the credibility to successfully pursue other economic and
financial goals. The US economy was large and closed enough that the Fed could
afford to act to a first approximation like the central bank of a closed economy.
The US share of world GDP peaked in 1985 at 33 percent, at a time when the
Soviet economy was in decline and Chinas growth spurt had just begun.7 The US
trade/GDP ratio was rising, but more slowly than in the subsequent quarter century.
The explosive growth of international capital flows and deepening of international
financial linkages was yet to come.
International considerations did in some circumstances play a role in monetary
policy during this time. Volckers inflation-control strategy itself had an international dimension; the fact that higher interest rates meant a stronger dollar made
for sharper disinflation through the channel of lower import prices (as argued by
Sachs 1985).8 The Feds decision to back away from a very tight monetary policy
in 1982 may have been influenced by the outbreak of the Latin American debt
crisis and the threat this posed to the solvency of major US banks (which is not
to deny that there was also an influence toward looser monetary policy from the
US recession in 1981 82). Central bank governors as well as finance ministers
were party to the Plaza Agreement of 1985 between the United States, the United
Kingdom, France, Germany, and Japan to attempt to stem the continued rise of
the dollar exchange rate. This led to coordinated foreign exchange market intervention and, in March 1986, coordinated interest rate reductions. Starting in 1986,
central bank governors of the so-called G -7 countries (the United States, United
Kingdom, France, Germany, Italy, Canada, and Japan) met regularly, together with
their finance ministry counterparts, on the sidelines of the spring and fall meetings of the IMF and World Bank. In addition, senior central bank officials met
bimonthly at the Bank of International Settlements. These meetings facilitated
information exchange. They also facilitated coordinated foreign-exchange-market
intervention, frequently before the mid-1990s and sporadically thereafter: in
7
One should be cautious about these comparisons, because they depend on the exchange rate used
to value transactions in dollars, and 1985 was when the dollar exchange rate was at a local peak. At
purchasing power parity, the US share of the global economy was more like 23 percent in 1985 and
reaches another local peak in 1999. Economists will of course debate which valuation method is more
relevant when considering the conduct of monetary policy.
8
Nelson (2005) argues that the Federal Open Market Committee had something similar in mind when
it tightened in 1978.
98
June1998 when the yen depreciated in the wake of the Asian crisis; in September
2000 when the euro weakened reflecting uncertainty about the policies of Europes
new central bank; and in March 2011 in response to the rise of the yen induced
by the Fukushima earthquake and the liquidation of foreign assets by Japanese
insurancecompanies.9
The global economic and financial crisis that unfurled in 2007 is another
reminder that the Fed cannot afford to neglect the impact of its policies on conditions abroad or the implications of conditions abroad for its policies. On October8,
2008, in the wake of the Lehman Brothers failure, the Federal Reserve coordinated a reduction in the federal funds rate with the lending rates of the European
Central Bank, Bank of England, Bank of Canada, Swiss National Bank, and Swedish
Riksbank. Irwin (2013), with a little exaggeration, calls this the first globally coordinated easing in history. Unusually, the Fed issued a joint statement together with
these other central banks announcing the action, which Iinterpret as an acknowledgment that coordinating policy with foreign central banks might produce better
outcomes under the circumstances.
In addition, the Fed arranged dollar swap lines with 14 foreign central
banks starting in December 2007, when the subprime crisis intensified. In these
arrangements, the Federal Reserve stands ready to swap US dollars with other
central banks for the currencies of that bank. The Fed renewed five of those
dollar swap lines, notably that with the European Central Bank, in May 2010.
These swap facilities were designed to alleviate financial problems abroad and
limit the blowback to US markets if foreign banks, unable to secure US dollar
funding, were forced to liquidate their holdings of US financial securities. These
swap lines were an acknowledgement that what happens abroad doesnt always
stay abroad and, while not modifying monetary policy to take this fact into
account, that the Fed must develop ancillary policy instruments to address strains
in foreign markets for US dollars. The Board of Governors (2013), in justifying
the practice to a skeptical Congress, noted that foreign currency swap lines might
also be helpful for addressing financial strains should US institutions experience
a shortage of foreign currency-denominated liquidity, although the most recent
swaps were not activated for this purpose.
To be clear, these intervention operations are decided in consultation by the US Treasury and the
Federal Reserve. Such operations are typically sterilized with the goal, sometimes questioned by
academics, of moving the dollar exchange rate without also moving the monetary base. On the effectiveness of sterilized intervention, a useful starting point is Rogoff (1984).
Does the Federal Reserve Care about the Rest of the World?
99
this into account. Iposit three trends that will heighten the impact of international
variables on the US economy.
First, Iassume that the United States will continue to grow more open to international trade and financial transactions. Admittedly, this assumption is contestable.
While technological progress works inexorably to reduce the costs of international
transactions, andI would arguealso reduces the cost of international relative
to domestic transactions, international openness depends not just on technology
but also on policy, which has been known to push in the other direction. But bear
withme.
Second, I assume that emerging and frontier markets will continue to grow
more rapidly than mature economies, so that the US economy will come to account
for a progressively declining share of the global economy. Again, continuing catchup and convergence are not inevitable; their progress will depend on policy choices.
But recent experience makes this assumption a reasonable starting point.
Third, I assume that the US dollar will lose its monopoly as a funding
currency for international banks and as the all but exclusive vehicle and currency
of denomination for international transactions (for discussion, see Shin 2011).
This is not to suggest, as does the film Looper,, that we will wake up tomorrow and
discover that all transactions are conducted in renminbi.10 Movement toward
other funding and vehicle currencies will be gradual, so that the dollar ends up
sharing its international role with other national units. But there is no fundamental reason why the United States should be the only country with deep and
liquid financial markets open to the rest of the world. Moreover, the US economy
alone will not be able to provide safe and liquid assets on the scale required by
an expanding global economy. It follows that US banks and firms will rely more
on foreign currency funding and liquidity in the future than the recent past
(Eichengreen 2011).
Taken together, these threetrends suggest that shocks to the exchange rate
and balance of payments will have a larger impact on the US economy in the
future, and that the implications may extend beyond those for inflation and
the output gap. For example, US dollar appreciation which creates competitiveness problems for the traded-goods sector will be more of a problem as a larger
share of US output and employment is exposed to international competition. If
dollar appreciation causes US firms to exit the market and they face fixed costs of
reentering (as in Baldwin and Krugman 1989), then transitory currency swings may
have permanent welfare-reducing effects. This is one explanation for why other
open economies adjust their policies in response to movements in the exchange
rate. It is an explanation for the aversion to freely floating exchange rates, known
as fear of floating, in emerging markets (Calvo and Reinhart 2002).
10
Looper, as film buffs know, is set in 2044. In the original script, the protagonist planned to move to
Paris in the future. When the director found filming in Paris prohibitively expensive, the future was
shifted to Shanghai, the Chinese distributor having offered to pay for a crew to film theresee http://
www.imdb.com/title/tt1276104/trivia.
100
A country on the receiving end of large capital inflows, in addition to experiencing a rising exchange rate, is likely to feel unwelcome pressure on housing
and financial markets. Capital inflows into the US economy in the period leading
up to the subprime crisis are an illustration of the problems that can arise. The
effect on local markets will be larger the smaller the economy is relative to the rest
of the world and the more open it is to global markets. In mid-2013, New Zealand
is a case in point of a country that is dealing with these kinds of housing and asset
market concerns due to exchange-rate and capital-flow pressures (Wheeler 2013).11
More generally, a variety of small open economies, and a number of middle-sized
countries like Brazil, have complained about the adverse impact of policies abroad
on their economies, operating through these channels, and have adjusted monetary
and other policies to address them: for example, Brazilian Finance Minister Guido
Mantega has registered strong complaints along these lines in a series of press interviews and speeches starting in September 2010.
Finally, as global markets grow relative to the US market, and as international
finance is provided in a wider range of currencies, US banks and firms will rely
more on foreign currency funding. As they accumulate liabilities denominated in
foreign currency, the Federal Reserve may then feel a growing reluctance to let
the dollar exchange rate move for fear of the destabilizing balance sheet effects
(specifically, that banks and firms with foreign-currency-denominated debts
will be unable to service them using their domestic-currency earnings). Those
adverse balance sheet effects are another explanation for why smaller, more
open economies where such currency mismatches are prevalent find it hard to
commit to regimes of flexible inflation targeting that imply benign neglect of the
exchangerate.
Assume, as a result of the changes posited above, that the effect of the
exchange rate and capital flows grows more important. Does it follow that the Fed
will have to modify the formulation and conduct of monetary policy to take them
intoaccount?
The answer, as with many economic questions, is yes and no. For example,
if a higher level of reliance on foreign currency funding causes exchange rate
movements to have more destabilizing balance sheet effects, then the first-best
response is not to use monetary policy to prevent those movements, but instead
to strengthen prudential supervision and regulation of banks and governance of
corporations to prevent excessive exposure to this form of balance sheet risk from
arising in the first place. Mishkin and Savastano (2001) is an early statement of the
tradeoff between, on the one hand, the strength of supervision and regulation of
balance-sheet mismatches, and on the other hand, policies of benign neglect
of the exchange rate.
Similarly, if capital inflows place worrisome upward pressure on housing and
other asset markets, then the first-best solution is to strengthen lending standards,
11
It is tempting to point to Ireland and Spain before 2009 as additional examples, but their cases are
special for obvious (euro-related) reasons.
Barry Eichengreen
101
raise margin requirements, and otherwise address problems in housing and asset
markets directly. The second-best set of responses in this case would be to address
the capital inflows that are the proximate source of the problem by applying taxes
to such inflows and, among other steps, tightening fiscal policy, which should put
downward pressure on interest rates and on the exchange rate. Trying to address
problems in asset markets caused by inflows of foreign capital by making adjustments in monetary policy would be no more than a third-best policychoice.
As yet another example, if the issue is the risk of permanent damage to tradedgoods sectors because temporary exchange rate movements have permanent effects,
then the first-best response is to eliminate the financial imperfections forcing
incumbents to exit and preventing them from reentering, or to use tax and other
policies (assuming that these can be enacted in a form compliant with World Trade
Organization rules) to address their problems directly.
Readers will detect echoes here of the lean versus clean debate (White
2009; Mishkin 2011). The question in that context was whether central banks
should lean against asset bubbles. Earlier thinking tended to favor letting other
agencies of governmentbank supervisors, regulators, those responsible for the
conduct of fiscal policyaddress problems of asset bubbles using instruments
better suited to the task than monetary policy, and for central banks to limit their
intervention to cleaning up the aftereffects. It would be presumptuous to assert
that recent events have permanently decided the question in favor of one view or
the other. But there is no question that those events have shifted the balance by
suggesting that central banks should think harder about the need to take preemptive action, both because other agencies of government may not be doing their
part and because cleaning up afterwards can be very costly.
The implication is that precisely the same issues will arise, with growing intensity over time, in connection with movements in exchange rates and capital flows.
There will be no return to the gold standard of the 1920s or the Bretton Woods
System of the 1950s and 1960s. But as the US economy grows even more open and
the rest of the world grows larger, international considerations, operating through
these channels, will impinge more directly on the central banks key objectives of
price and economic stability. The Fed will have no choice but to incorporate those
considerations more prominently into its policy decisions.
The first draft of this paper was prepared for the Symposium on the First 100 Years of
the Federal Reserve, held at the National Bureau of Economic Research in Cambridge,
Massachusetts, on July 10, 2013. I thank Christina Romer, David Romer, and the editors
of this journal for comments and Chris Krogslund for research assistance. The National
Bureau of Economic Research provided an honorarium in support of this paper. Other
than this the author has received no significant financial support from interested parties.
He serves on the Centennial Advisory Council of the Board of Governors of the Federal
ReserveSystem.
102
References
Awalt, Francis Gloyd. 1969. Recollections of
the Banking Crisis in 1933. Business History Review
43(3): 34771.
Baldwin, Richard, and Paul Krugman. 1989.
Persistent Trade Effects of Large Exchange Rate
Shocks. Quarterly Journal of Economics 104(4):
635 54.
Board of Governors of the Federal Reserve
System. 2013. Central Bank Liquidity Swap
Lines. Washington, DC: Board of Governors of
the Federal Reserve System. http://www.federal
reserve.gov/newsevents/reform_swaplines.htm.
Bordo, Michael, and Barry Eichengreen. 2008.
Bretton Woods and the Great Inflation. NBER
Working Paper 14532, December.
Brown, William Adams, Jr. 1940. The International Gold Standard Reinterpreted, 19141934. New
York: National Bureau of Economic Research.
Broz, J. Lawrence. 1997. International Origins
of the Federal Reserve System. Ithaca, NY: Cornell
University Press.
Calvo, Guillermo A., and Carmen M. Reinhart.
2002. Fear of Floating. Quarterly Journal of
Economics 117(2): 379 408.
Clarke, Stephen V. O. 1967. Central Bank
Cooperation, 192531. Federal Reserve Bank of New
York.
Eichengreen, Barry. 2000. From Benign
Neglect to Malignant Preoccupation: U.S. Balance
of Payments Policy in the 1960s. In Economic
Events, Ideas and Policies: The 1960s and After,
edited by George Perry and James Tobin, 185 242.
Washington, DC: Brookings Institution.
Eichengreen, Barry. 2011. Exorbitant Privilege:
The Rise and Fall of the Dollar and the Future of the
International Monetary System. New York: Oxford
University Press.
Eichengreen, Barry, and Marc Flandreau.
2012. The Federal Reserve, the Bank of
England, and the Rise of the Dollar as an International Currency, 1914 1939. Open Economies
Review 23(1): 57 87.
Federal Open Market Committee. 19361994.
Minutes of the Federal Open Market Committee.
Available at Fraser Federal Reserve Archive. http://
fraser.stlouisfed.org/publication/?pid=677.
Friedman, Milton, and Anna Jacobson
Schwartz. 1963. A Monetary History of the United
States, 18671960. Princeton University Press for
the National Bureau of Economic Research.
Gal, Jordi, and Mark Gertler, eds. 2010. International Dimensions of Monetary Policy. University
of Chicago Press for the National Bureau of
Economic Research.
Does the Federal Reserve Care about the Rest of the World?
103
104
Martin Feldstein
Martin Feldstein is George F. Baker Professor of Economics, Harvard University, and President
doi=10.1257/jep.27.4.105
106
how much do you think that action contributed to the sharp rise in inflation in the
remainder of the decade?
VOLCKER: I certainly was a major proponent of suspending gold convertibility,
in fact the principal planner. Ihad come to the conclusion that we needed to negotiate a sizable exchange rate adjustment. At the time, we didnt have a choice, as
Isaw it, to suspending convertibility as a transition to a reformed system. In the end,
the Smithsonian Agreement1 was not a very reasonable outcome from my point of
view because Ididnt think the changes in exchange rates were big enough to instill
confidence. The United States didnt accept any responsibility itself by the way of
any kind of convertibility to support the new rates, so the foundation wasnt there
for a long-lasting solution in my view.
But my thought always was we suspend convertibility, get the necessary exchange
rate change, and then we would redesign the international monetary system. President
Nixon had no interest in redesigning the international monetary system, Ithink its
fair to say. Nor did Mr. Connally 2 have much interest, which led to, Ithink, a more
unsatisfactory situation internationally where it was easy to make the impression that
we were being irresponsible.3
FELDSTEIN: So given that you didnt get the second part of what you were
hoping would happen, does it still look like the right decision in retrospect? Or,
as you say, there was really no choice? You couldnt negotiate something different?
VOLCKER: Ithink it was the right decision. The question was what happened
afterwards. The whole international exchange rate situation got out of hand.
Federal Reserve policy, in my view, was not particularly credible at that point. Iwas
in the Treasury. Every time Iwent on a trip abroad, Iwould try to get to [Federal
Reserve Chairman] Arthur [Burns] to say, Dont ease up while Im abroad anyway.
The inflation took hold, as you know, and never was really brought under
control for some time. It was an unsatisfactory economic situation.
FELDSTEIN: Some people think that that was related to the fact that we had
gone off the gold convertibility. How much do you think that actually contributed
to the inflation that happened in the remainder of the decade?
VOLCKER: I didnt think we had any choice about the gold convertibility.
Isuppose, you know, we could have devalued, as we did eventually, and then try to
defend a new rate, but we didnt have enough credibility to do that. Ithink we didnt
accept any kind of constraint during that period, the early 1970s. The exchange
rate fell abruptly at one point or several points. We never had a reaction until late in
the decade. Ithink there was a lack of discipline that would have been useful at the
time, which was not recognized and wasnt acted upon.
At the Smithsonian in December 1971, representatives of ten countries Belgium, Canada, France,
Germany, Italy, Japan, the Netherlands, Sweden, the United Kingdom, and the United States established a new set of fixed exchange rates.
2
John Connally was US Secretary of the Treasury from 1971 to 1972.
3
Silber (2012), in his biography of Paul Volcker, covers this more fully.
Martin Feldstein
107
Paul Volcker
October 1979
FELDSTEIN: Let me move to a later time. When you moved from the New York
Fed in 1979 to become Chairman of the Board of Governors, consumer prices by
then were rising at more than 10percent a year. You decided to push short-term
rates up even higher than the rate of inflation.
VOLCKER: The market pushed those rates up.
FELDSTEIN: Yes. Absolutely. With a little facilitation from folks on Constitution Avenue.4 Anyway, that succeeded in bringing inflation down to 6 percent in
1982, and 3percent in 1983. Ive got several questions about all of that. First of all,
why do you think inflation had gotten so high at the end of the 1970s?
VOLCKER: Iguess Ihave to say the Federal Reserve policy contributed to it,
but there was the oil crisis in the early 70s and then repeated again in the later 70s.
It made a profound impression on me, if nobody else, that Arthur Burns titled his
valedictory speech The Anguish of Central Banking (Burns 1979). That was a long
lament about how, in the economic and political setting of the times, the Federal
Reserve, and by extension presumably any central bank, could not exercise enough
restraint to keep inflation under control. It was a pretty sad story. If you were going
4
108
to follow that line, you were going to give up, I guess. I didnt think you could
give up. If Iwas in that job, that was the challenge as the Chairman of the Federal
Reserve. You inherit a certain challenge.
FELDSTEIN: Some people thought at the time that it was just going to be too
costly in terms of lost GDP and higher unemployment to get from where inflation
was at that point down to low single digits. Ithink there was a recommendation, at
least among some academic economists, to stop further increases in inflation, but
not to try to bring inflation down because that would be so costly.
VOLCKER: The favorite word at the time, which was very popular within the
Federal Reserve, but Ithink popular in the academic community generally, was
gradualism. Idont quite remember them saying, Dont bring it down at all.
But instead, it was Take it easy. It will be a job of, I dont know, years, decades,
whatever, and you can do it without hurting the economy. I never thought that
wasrealistic.
The inflationary process itself brought so many dislocations, and stresses and
strains that you were going to have a recession sooner or later. The idea that this was
just going to go on indefinitely, and the inflation rate got up to 15percent, it was
going to be 20percent the next year.
One little story (Ithink of all these stories): Shortly after we began the disinflation, somebody, Ithink Arthur Levitt who was the head of the American Stock
Exchange, brought in some businessmenthey tend to be small businessmento
talk to me at the Federal Reserve. Ihad them for lunch, and Igave them my little
patter about, This is going to be tough, but were going to stick with it, and the
inflation rate is going to come down, and so forth. The first guy that responded
said, Thats all very fine, Mr. Volcker, but I just came from a labor negotiation
in which I agreed to a 13 percent wage increase for the next three years for my
employees. Im very happy with my settlement. Ialways wondered whether he was
very happy twoyears later on. But that was symbolic of the depths. He was happy at
a 13percent wage increase.
FELDSTEIN: Did you have a sense of what the cost would bethat is how
high was unemployment going to be? How long was this going to last?
VOLCKER: I had no sense that interest rates were going to be so high. In
October1979, we took the full panoply of restrictive measures and emphasized the
money supply, and so forth. Ithought this was all done to convince people we were
really serious. We had already raised the discount rate two or three times in the
space of two or threemonths. The last increase in the discount rate was by a 4 to 3
vote. Iwas a neophyte Federal Reserve Chairman. It was all right with me. Iknew
Ihad fourvotes. If we had to raise it again Id still have fourvotes. Thats not the
way the market interpreted it. They said, Ah, theyre at the end of their string. They
cant command the unified majority any more on the Board. So its the end of the
day for any Federal Reserve restraint.
Idecided we had to change the playbook a little bit, and we threw everything
we could into the October 1979 announcement. Ihad this naive hope. Iknew the
short-term rates would go up, but I thought, Ah, we will instill confidence and
109
long-term rates will not go up. Long-term rates went up, too, just about as much
as the short-term rates, which was a disappointment. But it showed how strong the
psychologywas.
FELDSTEIN: The inflation came down, I think, much faster than outsiders,
in any case, expected. Remember the tax cuts that were put in place in 1981 were
based on the idea that inflation would continue to increase tax revenue because of
bracket creep. We didnt start indexing income tax brackets for inflation until 1985.
Yet inflation, as Iquoted a minute ago, was down to 6percent in 1982 and half that
the next year. Were you surprised at how fast the process worked?
VOLCKER: Idont know that Iwas surprised, but Isure was relieved. Ididnt
know how long that could have gone on.
FELDSTEIN: Can you say a little more about what it took to persuade the
Federal Open Market Committee and the Board of Governors to go along with this?
VOLCKER: First of all, raising interest rates quite visibly and openly is not the
easiest thing in the world for central bankers or anybody. Its much easier to lower
interest rates than it is to raise interest rates, I think its fair to say, in almost any circumstances. That 4 to 3 vote that Ireferred to reflected something of thatreluctance.
Three or four years earlier, there was some pressure in the Open Market
Committee to adopt a much stricter money supply approach. So it wasnt entirely
unknown to the Federal Reserve. Now that approach had always been rejected when
it was raised. But there was a little bit of feeling, Iknew, among some of the Open
Market Committee membersin particular outside of the Board, the regional
Reserve Bank presidents had a certain amount of sympathy you had some kind of
instinctive support. Ithink people were upset and tired about the way things were
happening, they were looking for something different, something new, something
that had some hope. They realized, Ihate to overdramatize, this was a last chance.
Thats overdramatizing a bit.
People were willing to get together on the new policy. We knew pretty well
that it would have a sharp impact on short-term rates. It was meant to be highly
restrictive, no doubt about that. But the Board, the Open Market Committee, was
prettyunited.
FELDSTEIN: If I remember correctly, history has you warning President Carter
that something like this was going to happen, getting his at least implicit consent,
and then when it actually happened, he was a very unhappy man and acted up.
VOLCKER: I dont think it was quite that way.
FELDSTEIN: You were there, and I wasnt, so tell us.
VOLCKER: When I was appointed in August 1979, I had made clear to him
that Ithought Federal Reserve policy was too easy at the time. If Iwas going to be
chairman, I was going to be advocating a stronger policy. I remember I thought
Iwasnt going to be appointed after that conversation. But Iwas appointed anyway,
so that shows the difficulty of the time.
Then when it came to making the so-called October 1979 decision, I had
warned the Secretary of the Treasury and the Chairman of the Council of Economic
Advisers at the time, Bill Miller and Charlie Schultze.
110
Credit Controls
FELDSTEIN: But President Carter went on television and he triggered some
legal provisions which had some effect.6
VOLCKER: No, not then. This was a later catastrophe so far as Iwas concerned.
You may not remember this. This was later when interest rates got up to, Idont
know, 20percent or so maybe by the end of 1979.
Carter announced a budget in early 1980 which was very poorly received.
Nothing was happening. The Federal Reserve staff kept saying, Were having this
recession. The recession is beginning. There was no recession. In spite of all this,
the economy kept rising.
Carter was obviously under pressure, so he triggered a provision of law that
permitted the Federal Reserve to put on credit controls. He said, Iwant to put on
5
Silber (2012, p.166) reports an interview with Schultze that confirms this account.
For a transcript of President Carters televised address of March 14, 1980, in which he called upon the
Federal Reserve to impose credit controls, see Carter (1980).
Martin Feldstein
111
credit controls. Iwant to show Im on the team, sort of. Were going to go together,
and Ive got to change the budget. Im going to make the budget more restrictive.
Iwant to announce there will be credit controls. We were going to tighten policy
again. That was all part of the package: You wait to tighten policy. Well announce
creditcontrols.
We didnt like the idea because expansion of credit was not a problem at the
time, and we didnt want to get into all the mess of managing credit controls. We
thought it muddied the picture. I felt I didnt like it, but how can we rebuff the
President of the United States who is asking us to put on credit controls? Theoretically, we could have said no, but the Board reluctantly went along at my urging.
FELDSTEIN: He managed to ask you on television in March 1980.
VOLCKER: I guess maybe he did.
FELDSTEIN: It was not a secret to the American public that he had that view.
VOLCKER: No, it wasnt a secret. Anyway, this was a phenomenon engraved in
mymind.
We decided, All right. Well put on the most modest controls we can think of.
We will not put on any control over anything to do with housing, which is the biggest
source of credit. I think we exempted automobiles. Theres nothing the matter
withautomobiles.
So the only thing that was left, and in those days it was of little importance,
was installment credit that was not related to housing or cars, and credit cards,
which was not a very big area of the market. We said, Okay, youre going to have a
reserve requirement on credit cardsif credit cards exceed past peaks, you would
have a reserve requirement. We did that knowing, were now in March, the peak in
credit card use comes in November and December. We were way below it so there
was no possibility that this was going to become a factor for some time. This was all
announced at a big White House ceremony [laughs].
The economy at that point fell like a rock. People were cutting up credit cards,
sending in the pieces to the President as their patriotic duty. Mobile home and
automobile sales dropped within the space of a week or so. The money supply, we
didnt know why the money supply was dropping, but all of the sudden the money
supply was down 3percent in a week or something.
What happened was everybody was paying off their credit card debt by drawing
down their demand deposits and the money supply fell. We went into what the
National Bureau of Economic Research later determined was a recession, and
interest rates and the money supply dropped sharply.7 Well, it was a recession alright,
the economy went down, but it was an artificial recession. As soon as we took off the
credit controls in June, the economy began expanding again. Then things really
got tough. We reversed the easing during the recession and interest rates resumed
rising, including a discount rate increase a month or so before the election. That
was the only time President Carter publicly expressed concern.
7
For the June3, 1980, press release from the Business Cycle Dating Committee, see National Bureau of
Economic Research (1980).
112
High Real Interest Rates and the Debt Crisis in Latin America
FELDSTEIN: Let me look a little further down the road. Even after inflation
had dropped to 3percent, you kept, or the market kept, interest rates very high.
Treasury Bills were 8 and 9percent in 1983 and 1984, so we had real rates on Treasury Bills of 5 and 6percent. Why were they so high?
VOLCKER: Oh geez, I dont remember precisely. First of all when the economy
began expanding, it was expanding very fast. There was no problem with poor
economic activity at that point.8
Then, in the spring and early summer of 1984, you had the potential banking
crisis of Continental Illinois. A little bank in Oklahoma went bust. But it had sold a
lot of oil patch loans to Continental Illinois and other banks, and there was this sort
of a banking crisis.9 When Continental got in trouble, that had an effect on overnight bank lending. That kept, for a whilethat was during the election period,
actuallykept interest rates higher than we anticipated and really higher than
wewanted.
There was quite a debate at the Board of Governors at that time whether we
should react to what seemed an artificial increase in interest rates by easing our
policies or whether we would tough it out.
FELDSTEIN: You still had very high real rates, short real rates. Was it a concern
about inflation coming back?
VOLCKER: If Iput myself back in that position, Ithink we were totally satisfied
at what the economy was doing. You still had some inflation. Why would we be
easing up when the economy was expanding 6 or 7percent a year? Were getting
back to where we were. Everything was fine. Sort of fine.
FELDSTEIN: Another development at about this time was the debt crisis in
Latin America. What role did the high interest rates of the early 1980s play in causing
severe debt problems in emerging economies? What options did you consider to
deal with this?
VOLCKER: Thats an interesting question. This is something like, Isuppose,
the subprime mortgage thing. To understand what happened in the early 1980s, we
need to start earlier.
The 70s were characterized by a lot of liquidity growing out of the oil crisis
and the excess money that the Arabs had, and all the rest. That money was flowing
through the big banks to Latin America in a way that arguably looked constructive
for a while but was ultimately unsustainable.
Silber (2012, p.237) cites the 1984 Annual Report of the Council of Economic Advisors on the role of
the full employment federal deficit: [F]ederal borrowing to finance a budget deficit of fivepercent
of GNP . . . means the real rate of interest must rise.
9
The Oklahoma bank in question was Penn Square Bank. For an overview of the events surrounding the
Continental Illinois banking crisis that came to a head in 1984, see Federal Deposit Insurance Corporation (1997).
113
Arthur Burns, to his credit, was the Paul Revere on this thing. Hed go around
and make speeches: This cant continue. It shouldnt be continued. Weve got to
do something about it. 10
Iwas in the New York Fed then. We tried to do something about it, which was
totally ineffectual, Imust say. But, it went on and nobody was willing to say, If you did
something that was really effectualif thats a wordyou would have a crisis in Latin
America if you shut off the flow. Nobody much wanted to be all thataggressive.
By the early 1980s, interest rates had gotten very highIdont think it was the
interest ratesthe banks suddenly stopped lending to Mexico because they thought
they were overexposed, so you had a crisis. Once one stopped, they all stopped.
This had been building up. The figures were known. The president of Mexico
then, a left-wing guy, was being told by his own finance minister and central bank
governor, We ought to stop this or slow it down. He sent some people around to
talk to foreign banks and ask, Is this a problem?11 They all said no. This was in the
summer of 81. Now we come to 82. He got rid of the finance minister! He wasnt
going to get rid of the borrowing, but his finance minister instead. Its true.
The borrowing continued until the winter when a couple of banks stopped
lending. Mexico ran out of money. What do you do? You had a big crisis now. The
high interest rates were a burden for Mexico over time, but they didnt make the
crisis. They hadnt been in effect all that long. But theres no question that high
interest rates aggravated the problem.
What were you going to do? Were you going to conduct an easy-money policy
and go back on all the policy youd undertaken to try to save Mexico, which wouldnt
have saved Mexico anyway?
We did save Mexico, but by other means. It wasnt just Mexico. People forget.
This is a commentary on age. This was 82. How many years ago was that? Thirty-one
or -two years ago. Ihear all this talk about crisis. Nobody ever remembers the Latin
American debt crisis. Memories only go back to, somehow, the savings and loan
crisis in 1990 and dont make the leap back to 80.
The big US banks and some of the big foreign banks had more exposure to
Latin America than they had capital. It wasnt something you could just say, Okay,
knock off the loans by 50percent or something and everybody will be happy. They
all would have been bust. You look for other approaches, and it took nearly a decade
until Mr. Brady came along and settledthem.12
10
114
Martin Feldstein
115
was a phenomena of, what, threeyears maybe? From basically a standing start, in
threeyears it was a trillion dollars. Obviously that was kind of the focal point of the
crisis when it finally came. Look, this banking regulation stuff is very hard to deal
with. But I think there had been a relative lack of interest in it, which was unfortunate, to understate the matter.
The Volcker Rule, broadly understood, is that financial institutions that are eligible for deposit insurance and have access to the Federal Reserve and FDIC insurance should be limited in the risks they take
with their proprietary trading. For an early presentation of this argument, see Group of Thirty (2009).
15
For an overview of these events, in which a series of derivatives trades in spring 2012 cost JPMorgan
approximately $6 billion, see the hearings of the US Senate (2013), titled JPMorgan Chase Whale
Trades: A Case History of Derivatives Risks and Abuses.
116
survive $6billion. But what is the psychology that leads people to take that kind of risk?
Traders know that the rewards are hugeof a kind that have not been at all normal in
commercial banking now or in history. When youve got that kind of cleavage between
the culture on the investment banking side of the house and the traditional banking
side of the house, obviously the people in the commercial banking side say, Iwant
to make money, too. Maybe Ican make some big risks and Ill get some mortgages
together, and Ill package them up. Lets securitize them and stick them out. Well
make a commission on it. Its not a relationship matter. Were going to stick this out,
well stick somebody else with it. Its a different culture.
The guy that is most eloquent on this, it surprises me because he never used to
be friendly toward the Federal Reserve, is John Reed, head of the Citibank at the
time. He was a leader in commercial banks going in the investment business. They
bought Salomon Brothers investment bank back in 1998, you may recall. Salomon,
that subsidiary, went bust later. Not too much later, but it was part of Citibank, so
nothing happened. But he is very vocal: Mea culpa. We made a mistake. It destroyed
the culture of the institution.16 Thats my major worry about it.
FELDSTEIN: Another thing youve worried about: the size of the big banks. If
Iremember correctly, you were in favor of breaking up the big banks during this
crisis. Is thattrue?
VOLCKER: Inever took the view to break up the big banks. Iwanted to limit
their size, which Iguess is in the law someplacenot very effectively. Isort of lack
imagination. Idont see how you break them up without a lot of disturbance.
But even if you broke them up, you couldnt break them up into small enough
pieces so that they wouldnt be systemically significant, or whatever we call it now.
You break up JPMorgan in half, or Chase in half, theyre the same bank. Bank of
America, you slice them in half. Theyre not one two-trillion-dollar bank, theyre
twoone-trillion-dollar banks. Theyre still too big.
FELDSTEIN: That view, which Ive heard attributed to you, shouldnt have
been attributed to you about wanting to break them up?
VOLCKER: No. But I wouldnt mind if somebody else does it.
16
For example, see Reeds interview with Bloomberg as reported in Ivry (2009).
117
118
FELDSTEIN: Youve got a chance to do that now. That probably tells me something about what you think about the idea of unconventional policies. Maybe it
doesnt, so you tell us.
VOLCKER: I think this crisis required some unconventional policies, theres no
doubt about that. Extremely unconventional is the kindest word you can say about
it when you go back a few years ago.
Lets talk about this current version, this so-called QE3.17 Its a matter of judgment. I dont get alarmed about it, and I think they can manage their way out of it.
Chairman Bernanke has made that quite clear and I think hes right.18
It does have the dangers of speculative excesses. Its got pluses and minuses.
The pluses I dont think are very large. The minuses dont seem to be tremendous
right at the moment either.
Since I can say it, if I was conducting these policies, I dont really understand
why were paying interest on excess reserves when were worried about getting
interest rates as low as possible. The Federal Reserve pays more on their excess
reserves than the banks can get from lending to each other. So why pay them?
FELDSTEIN: You would stop paying interest on excess reserves?
VOLCKER: Yes, I would. It never dawned on me to pay interest on excess
reservesI guess a limitation of my own imagination. I was always in favor of
paying interest on required reserves.19 But the idea of paying interest on excess
reserves never occurred to me until the Federal Reserve began doing it. I can see,
in some circumstances, it may have some advantages. But I think youre going to
find it has some disadvantages, too. Some day you have to make that rate pretty
high if youre going to do it, I guess, when you want to tighten policy. Well see
how that goes.
FELDSTEIN: You mentioned the word deficits. Let me read another question
that came in: Can the US continue for long as the financial global hegemon with
the persistent large fiscal deficits?
VOLCKER: Whats a large fiscal deficit?
FELDSTEIN: A large fiscal deficit is what we have.
17
QE refers to the quantitative easing policies in which the Federal Reserve purchased financial
securities like US Treasury bonds and mortgage-backed securities. The first round of these policies,
QE1, started in November 2008; the second round, QE2, in August2010; and the third round, QE3, in
September2012.
18
For example, see Bernankes (2010) congressional testimony concerning The Federal Reserve
ExitStrategy.
19
Traditionally, the Federal Reserve did not pay interest on the reserves that it required banks to hold,
nor on any additional or excess reserves beyond the legal requirement that banks choose to hold. The
Financial Services Regulatory Relief Act of 2006 authorized the Federal Reserve to pay interest on reserves
beginning in October2011. That authority was accelerated to October2008 by the Emergency Economic
Stabilization Act of 2008, and the Federal Reserve began to pay interest on reserves on October 9,
2008. For additional explanation and interest rates that are paid, see the Federal Reserve website at
http://www.federalreserve.gov/monetarypolicy/reqresbalances.htm.
Martin Feldstein
119
VOLCKER: Yes, thats what we have. I think its hard, particularly the uncertainty over the decades ahead when Medicare, and Social Security, and so forth
seem to widen the deficit. The deficit, whats more up to the point to me, is the
balance of payments deficit. I think were back, in a way, in the Triffin dilemma.20
In the 1960s, we were in a position in the Bretton Woods system with the other
countries wanting to run surpluses and build their reserve positions, so the reserve
position of the United States inevitably weakenedweakened to the point where we
no longer could support the convertibility of currencies to gold. Now, how long can
we expect as a country or world to support how many trillions of dollars that the rest
of the world has? So far, so good. The rest of the world isnt in a very good shape, so
we look pretty good at the moment.
But suppose that situation changes and were running big deficits, and however
many trillion it is now, its another few trillions. At some point there is vulnerability
there, I think, for the system, not just for the United States. We ought to be conscious
of that and do something about it. Its not so easy to do something about it because
that comes back to the whole question of international monetary reform, which is
a favorite subject.
FELDSTEIN: The Chinese have been major buyers of US government debt.
Theyve been able to do that because they have had a large current account surplus.
That surplus has come down from 10percent a few years ago to less than 2percent
now. If they pursue the policies they say theyre going to pursue, it could easily
disappear in the next couple of years. How should we think about the implications
of that for the USeconomy?
VOLCKER: I think thats good news from the standpoint of what I just
mentioned. China is not the only other country in the world. But China was an
important accumulator of US dollars. If they stop accumulating, the kind of worry
Ijust expressed is somewhat alleviated. Its not gone, because our current account
deficitcontinues.
Our current account deficit is smaller, too, and hopefully can remain small.
Ihate to see our deficit go back to where it was. Id like to see it disappear, but its
hard to make it disappear. No question about that.
FELDSTEIN: On that semi-optimistic note, let me thank you again for taking
the time and giving us your views about all of this.
20
Robert Triffin (1960) testified before the Joint Economic Committee of Congress that the economy
with the worlds reserve currencythen and now the US dollarmust be willing to run ongoing
trade deficits so that the reserve currency will be available for the global economy, but this in turn
means that foreign governments hold large quantities of dollar assets, which at some point as events
change are likely to become a source of global financial instability. This situation became known as
the Triffindilemma.
120
References
Bernanke, Ben S. 2010. Federal Reserves
Exit Strategy. Testimony before the Committee
on Financial Services, US House of Representatives, Washington, DC, February 10. http://
www.federalreserve.gov/newsevents/testimony
/bernanke20100210a.htm.
Burns, Arthur F. 1977. The Need for Order in
International Finance. Address at the Columbia
University Graduate School of Business, New York,
NY, April 12. http://fraser.stlouisfed.org/docs
/historical/burns/Burns_19770412.pdf.
Burns, Arthur. 1979. The Anguish of Central
Banking: The 1979 Per Jacobsson Lecture.
http://www.perjacobsson.org/lectures/1979.pdf.
Carter, Jimmy. 1980. Anti-Inflation Program
Remarks Announcing the Administrations
Program. March 14. The American Presidency
Project. http://www.presidency.ucsb.edu/ws/?pid
=33142.
Federal Deposit Insurance Corporation. 1997.
Continental Illinois and Too Big to Fail. Chap.1
of History of the Eighties: Lessons for the Future,
Vol.1: An Examination of the Banking Crises of the
1980s and Early 1990s. Federal Deposit Insurance Corporation. http://www.fdic.gov/bank
/historical/history/.
Group of Thirty. 2009. Financial Reform: A
Framework for Financial Stability. http://www
.group30.org/images/PDF/Financial_Reform-A
_Framework_for_Financial_Stability.pdf.
Ivry, Bob. 2009. Reed Says Im Sorry for
Role in Creating Citigroup. Bloomberg.com,
November 6. http://www.bloomberg.com/apps
/news?pid=newsarchive&sid=a.z4KpD77s80&pos=6.
National Bureau of Economic Research, Business Cycle Dating Committee. 1980. Business
Cycle Peaked in January. June3. http://www.nber
.org/cycles/june1980.html.
Silber, William L. 2012. Volcker: The Triumph of
Persistence. New York: Bloomsbury Press.
Triffin, Robert. 1960. Statement of Robert
Triffin to the Joint Economic Committee of
Congress. December 8. http://babel.hathitrust
.org/cgi/pt?id=mdp.39015081224407;view=1up
;seq=234.
US Senate, Committee on Homeland Security
and Intergovernmental Affairs. 2013. JPMorgan
Chase Whale Trades: A Case History of Derivatives Risks and Abuses. March 15. http://www
.hsgac.senate.gov/subcommittees/investigations
/hearings/chase-whale-trades-a-case-history-of
-derivatives-risks-and-abuses.
Volcker, Paul A. 2005. An Economy On Thin
Ice. Washington Post, April 10, p. B7. http://www
.washingtonpost.com/wp-dyn/articles/A38725
-2005Apr8.html.
here are some things money cant buyfriendship, for example. If I want
more friends than I have, it clearly wouldnt work to buy some. A hired
friend is not the same as the real thing. Somehow, the money that would
buy the friendship dissolves the good I seek to acquire.
But most goods are not of this kind. Buying them does not ruin them. Consider
kidneys. Some people favor a market in human organs; others are opposed. But
those who oppose the buying and selling of kidneys cannot argue that a market in
kidneys would destroy the good being sought. A bought kidney will work, assuming
a good match. So if a market in human organs is objectionable, it must be for some
other reason. Money can buy kidneys (as the black market attests); the question is
whether it should be allowed to do so.
In my book What Money Cant Buy: The Moral Limits of Markets,, I try to show that
market values and market reasoning increasingly reach into spheres of life previously
governed by nonmarket norms (Sandel 2012). In procreation and childrearing,
health and education, sports and recreation, criminal justice, environmental protection, military service, political campaigns, public spaces, and civic life, money and
markets play a growing role. I argue that this tendency is troubling; putting a price
on every human activity erodes certain moral and civic goods worth caring about.
We therefore need a public debate about where markets serve the public good and
where they dont belong.
In this article, I would like to develop a related theme: When it comes to
deciding whether this or that good should be allocated by the market or by
Michael J. Sandel is the Anne T. and Robert M. Bass Professor of Government, Harvard
University, Cambridge, Massachusetts. His email is [email protected].
doi=10.1257/jep.27.4.121
122
nonmarket principles, economics is a poor guide. On the face of it, this may seem
puzzling. Explaining how markets work is a central subject of economics. So why has
economics failed to provide a convincing basis for deciding what should, and what
should not, be up for sale?
The reason lies in the conception of economics as a value-neutral science of
human behavior and social choice. As I will try to show, deciding which social practices
should be governed by market mechanisms requires a form of economic reasoning
that is bound up with moral reasoning. But mainstream economic thinking asserts its
independence from the contested terrain of moral and political philosophy. Economics textbooks emphasize the distinction between positive questions and normative
ones, between explaining and prescribing. The popular book Freakonomics states the
distinction plainly: Morality represents the way we would like the world to work and
economics represents how it actually does work. Economics simply doesnt traffic in
morality (Levitt and Dubner 2006, pp.11, 46, 190; see also Robbins 1932).
Moral Entanglements
Economists have not always understood their subject in this way. The classical
economists, going back to Adam Smith, conceived of economics as a branch of
moral and political philosophy. But the version of economics commonly taught
today presents itself as an autonomous discipline, one that does not pass judgment
on how income should be distributed or how this or that good should be valued.
The notion that economics is a value-free science has always been questionable. But
the more markets extend their reach into noneconomic aspects of life, the more
entangled they become with moral questions.
To be clear, I am not writing here about the standard textbook limitations
on markets. A considerable body of economic analysis is devoted to identifying
market failures, or situations in which unaided market forces are unlikely to
produce an efficient result, such as imperfectly competitive markets, negative and
positive externalities, public goods, imperfect information, and the like. Another
body of economic literature addresses questions of inequality. But this literature
tends to analyze the causes and consequences of inequality while claiming to be
agnostic on normative questions of fairness and distributive justice. Outsourcing
judgments about equity and fairness to philosophers seems to uphold the distinction between positive and normative inquiry.
But this intellectual division of labor is misleading, for two reasons. First, as
Atkinson (2009) has recently observed, economics is a moral science, despite
protestations to the contrary. Efficiency only matters insofar as it makes society
better off. But what counts as better off? The answer depends on some conception of the general welfare or the public good. Although welfare economics has
largely disappeared from mainstream economics in recent decades, Atkinson
writes, economists have not ceased to make welfare statements. Articles in journals of economics are replete with welfare statements and reach clear normative
Michael J. Sandel
123
124
125
relationships into transactions and treat all good things in life as if they were
commodities. The economic literature that acknowledges stigma and repugnance
makes implicit judgments about these questions; otherwise, it would be unable
to propose either market solutions or quasi-market alternatives. But it does not
articulate and defend the basis of these judgments. Doing so would carry economic
reasoning beyond the textbook distinction between positive and normative inquiry
and call into question the conception of economics as a value-neutral science of
social choice. I will try to show how this is so by considering arguments for and
against the use of market mechanisms in some contested contexts.1
A number of the sections of this paper draw upon Sandel (2012), especially from pp. 21133. For
those interested in following up specific discussions, here are the relevant page references to the 2012
book: Ticket Scalpers and Line Standers, pp.2123; Markets and Corruption, pp.3335; Refugee
Quotas, pp. 63 65; Fines vs. Fees, pp.6570; Tradeable Procreation Permits, pp.7072; Paying
to Shoot a Walrus, pp.82 84; Incentives and Moral Entanglements, pp.8891; The Case against
Gifts, pp.98103; Crowding out Non-market Norms, pp.113 120; The Commercialization Effect,
pp. 12022; Blood for Sale, pp. 122125; Two Tenets of Market Faith, pp. 125127; and Economizing Love, pp.127133.
126
And yet some people object. Senator Claire McCaskill, a Missouri Democrat,
has tried to ban paid Congressional line standing, without success. The notion
that special interest groups can buy seats at congressional hearings like they would
buy tickets to a concert or football game is offensive to me, she said (as quoted in
OConnor 2009; see also Hananel 2007).
But what exactly is objectionable about it? One objection is about fairness: It
is unfair that wealthy lobbyists can corner the market on Congressional hearings,
depriving ordinary citizens of the opportunity to attend. But unequal access is not
the only troubling aspect of this practice. Suppose lobbyists were taxed when they
hired line-standing companies, and the proceeds were used to make line-standing
services affordable for ordinary citizens. The subsidies might take the form, say,
of vouchers redeemable for discounted rates at line-standing companies. Such a
scheme might ease the unfairness of the present system. But a further objection
would remain: turning access to Congress into a product for sale demeans and
degradesit.
We can see this more clearly if we ask why Congress underprices admission to its deliberations in the first place. Suppose, striving mightily to reduce
the national debt, it decided to charge admission to its hearingssay, $1,000 for
a front row seat at the House Appropriations Committee. Many people would
object, not only on the grounds that the admission fee is unfair to those unable
to afford it, but also on the grounds that charging the public to attend a Congressional hearing is a kind ofcorruption.
We often associate corruption with ill-gotten gains. But corruption refers to more
than bribes and illicit payments. To corrupt a good or a social practice is to degrade
it, to treat it according to a lower mode of valuation than is appropriate to it (on
higher and lower modes of valuation, see Anderson 1993). Charging admission to
Congressional hearings is a form of corruption in this sense. It treats Congress as if it
were a business rather than an institution of representative government accessible to
allcitizens.
Cynics might reply that Congress is already a business, in that it routinely sells
influence and favors to special interests. So why not acknowledge this openly and
charge admission? The answer is that the influence peddling and self-dealing that
already afflict Congress are also instances of corruption. They represent the degradation of government in the public interest. Implicit in any charge of corruption is a
conception of the purposes and ends an institution (in this case, Congress) properly
pursues. The line-standing industry on Capitol Hill is corrupt in this sense. It is not
illegal, and the payments are made openly. But it degrades Congress by treating
access to public deliberations as a source of private gain rather than an expression
of equalcitizenship.
This does not necessarily mean that queuing is the best way to allocate access
to Congressional hearings or Supreme Court arguments. Another alternative,
arguably more consistent with the ideal of equal citizenship than either queuing
or paying, would be to distribute tickets by an online lottery, with the provision that
they be nontransferable.
Michael J. Sandel
127
128
129
and the effort of the community to accommodate them by setting aside certain
parkingspaces.
In practice, the distinction between a fine and a fee can be unstable. In China,
the fine for violating the governments one-child policy is increasingly regarded
by the affluent as a price for an extra child. The policy, put in place over three
decades ago to reduce Chinas population growth, limits most couples in urban
areas to one child. (Rural families are allowed a second child if the first one is
a girl.) The fine varies from region to region, but reaches 200,000 yuan (about
$31,000) in major cities a staggering figure for the average worker, but easily
affordable for wealthy entrepreneurs, sports stars, and celebrities (Moore 2009;
Bristow 2007; Coonan 2011; Mingai 2007).
Chinas family planning officials have sought to reassert the punitive aspect of
the sanction by increasing fines for affluent offenders, denouncing celebrities who
violate the policy and banning them from appearing on television, and preventing
business executives with extra kids from receiving government contracts. The fine
is a piece of cake for the rich, explained Zhai Zhenwu, a Renmin University sociology professor (Moore 2009). The government had to hit them harder where it
really hurt, at their fame, reputation, and standing in society (for discussion, see
also Xinhua News Agency 2008; Liu 2008).
The Chinese authorities regard the fine as a penalty and want to preserve
the stigma associated with it. They dont want it to devolve into a fee. This is not
mainly because theyre worried about affluent parents having too many children;
the number of wealthy offenders is relatively small. What is at stake is the norm
underlying the policy. If the fine were merely a price, the state would find itself
in the awkward business of selling a right to have extra children to those able and
willing to pay for them.
130
the rich would likely buy procreation licenses from the poor, the scheme would
have the further advantage of reducing inequality by giving the poor a new source
of income (de la Croix and Gosseries 2006).
Some people oppose restrictions on procreation, whether mandatory or marketbased. Others believe that reproductive rights can legitimately be restricted to avoid
overpopulation. Set aside for the moment that disagreement of principle and
imagine a society that was determined to implement mandatory population control.
Which policy would be less objectionable: a fixed quota that limits each couple to
one child and fines those who exceed the limit, or a market-based system that issues
each couple a tradable procreation voucher entitling the bearer to have onechild?
From the standpoint of economic reasoning, the second policy is clearly preferable. The freedom to choose whether to use the voucher or sell it makes some
people better off and no one worse off. Those who buy or sell vouchers gain (by
making mutually advantageous trades), and those who dont enter the market are
no worse off than they would be under the fixed quota system; they can still have
onechild.
And yet, there is something troubling about a system in which people buy and
sell the right to have kids. Part of what is troubling is the unfairness of such a system
under conditions of inequality. We hesitate to make children a luxury good, affordable by the rich but not the poor. Beyond the fairness objection is the potentially
corrosive effect on parental attitudes and norms. At the heart of the market transaction is a morally disquieting activity: parents who want an extra child must induce or
entice other prospective parents to sell off their right to have a child.
Some might argue that a market in procreation permits has the virtue of
efficiency; it allocates children to those who value them most highly, as measured
by the ability to pay. But trafficking in the right to procreate may promote a
mercenary attitude toward children and corrupt the norm of unconditional love
of parents for their children. For consider: Wouldnt the experience of loving
your children be tainted if you acquired some of them by bribing other couples
to remain childless? Might you be tempted, at least, to hide this fact from your
children? If so, there is reason to conclude that, whatever its advantages, a market
in procreation permits would corrupt parenthood in ways that a fixed quota,
however odious, would not.
In deciding whether to commodify a good, we must consider more than
efficiency and fairness. We must also ask whether market norms will crowd out
nonmarket norms, and if so, whether this represents a loss worth caring about.
Michael J. Sandel
131
hunting, with a small exception for aboriginal subsistence hunters whose way of life
had revolved around the walrus hunt for 4,500years.
In the 1990s, Inuit leaders approached the Canadian government with a
proposal. Why not allow the Inuit to sell the right to kill some of their walrus quota
to big-game hunters? The number of walruses killed would remain the same. The
Inuit would collect the hunting fees, serve as guides to the trophy hunters, supervise the kill, and keep the meat and skins as they had always done. The scheme
would improve the economic wellbeing of a poor community, without exceeding
the existing quota. The Canadian government agreed.
Today, rich trophy hunters from around the world make their way to the
Arctic for the chance to shoot a walrus. They pay $6,000 to $6,500 for the privilege.
They do not come for the thrill of the chase or the challenge of stalking an elusive
prey. Walruses are unthreatening creatures that move slowly and are no match for
hunters with guns. In a compelling account in the New York Times Magazine,, Chivers
(2002) compares walrus hunting under Inuit supervision to a long boat ride to
shoot a very large beanbag chair. The guides maneuver the boat to within 15yards
of the walrus and tell the hunter when to shoot. Chivers describes the scene as a
game hunter from Texas shot his prey: [The] bullet smacked the bull on the neck,
jerking its head and knocking the animal to its side. Blood spouted from the entry
point. The bull lay motionless. [The hunter] put down his rifle and picked up his
video camera. The Inuit crew then pull the dead walrus onto an ice floe and carve
up the carcass.
The appeal of the hunt is difficult to fathom. It involves no challenge, making it
less a sport than a kind of lethal tourism. The hunter cannot even display the remains
of his prey on his trophy wall back home. Walruses are protected in the United States,
and it is illegal to bring their body parts into the country.
So why shoot a walrus? Apparently, the main reason is to fulfill the goal of
killing one specimen of every creature on lists provided by hunting clubs for
example, the African Big Five (leopard, lion, elephant, rhino, and cape buffalo),
or the Arctic Grand Slam (caribou, musk ox, polar bear, and walrus).
It hardly seems an admirable goal; many find it repugnant. But from the
standpoint of market reasoning, there is much to be said for allowing the Inuit to
sell their right to shoot a certain number of walruses. The Inuit gain a new source
of income, and the list hunters gain the chance to complete their roster of
creatures killedall without exceeding the existing quota. In this respect, selling
the right to kill a walrus is like selling the right to procreate, or to pollute. Once
you have a quota, market logic dictates that allowing tradable permits improves
the general welfare. It makes some people better off without making anyone
worseoff.
And yet there is something morally disagreeable about the market in walrus
killing. Lets assume, for the sake of argument, that it is reasonable to permit the Inuit
to carry on with subsistence walrus hunting as theyve done for centuries. Allowing
them to sell the right to kill their walruses is nonetheless open to twomoral objections. First, it can be argued that this bizarre market caters to a perverse desire
132
that should carry no weight in any calculus of social utility. Whatever one thinks
of other forms of big-game hunting, the desire to kill a helpless mammal at close
range, without any challenge or chase, simply to complete a list, is not worthy of
being fulfilled. To the contrary, it should be discouraged. Second, for the Inuit to
sell outsiders the right to kill their allotted walruses arguably corrupts the meaning
and purpose of the exemption accorded their community in the first place. It is one
thing is to honor the Inuit way of life and to respect its long-standing reliance on
subsistence walrus hunting. It is quite another to convert that privilege into a cash
concession in killing on the side.
Of course, the moral judgments underlying these objections are contestable.
Some might defend the system of tradable walrus-hunting quotas on the grounds
that the desire to shoot a walrus is not perverse but morally legitimate, worthy of
consideration in determining the general welfare. It might also be argued that the
Inuit themselves, not outside observers, should determine what counts as respecting
their cultural traditions. My point is simply this: deciding whether or not to permit
the Inuit to sell their right to shoot walruses requires debating and resolving these
competing moral judgments.
133
use a cash incentive or a tradable quota requires that we evaluate, in each case, the
nonmarket values and norms such mechanisms may displace or transform.
Several other studies also demonstrate the crowding out effect:
Nuclear Waste Siting
When residents of a Swiss town were asked whether they would be willing to
approve a nuclear waste site in their community if the Parliament decided to build it
there, 51percent said yes. Then the respondents were offered a sweetener: Suppose
the Parliament proposed building the nuclear waste facility in your community
and offered to compensate each resident with an annual monetary payment.
(Frey, Oberholzer-Gee, Eichenberger 1996; Frey and Oberholzer-Gee 1997; see
also Frey 1997, pp. 6778). Adding the financial inducement did not increase
the rate of acceptance. In fact, it cut it in half from 51 percent to 25 percent.
Similar reactions to monetary offers have been found in other places where local
communities have resisted radioactive waste repositories (Frey, Oberholzer- Gee,
and Eichenberger 1996, pp. 1300, 1307; Frey and Oberholzer-Gee 1997, p. 750;
Kunreuther and Easterling 1996, pp.606608).
Why would more people accept nuclear waste for free than for pay? For many, the
willingness to accept the waste site apparently reflected public spirita recognition
that the country as a whole depended on nuclear energy, and that the waste had to
be stored somewhere. If their community was found to be the safest site, they were
willing to sacrifice for the sake of the common good. But they were not willing to sell
out their safety and put their families at risk for money. In fact, 83percent of those who
rejected the monetary proposal explained their opposition by saying they could not be
bribed (Frey, Oberholzer-Gee, and Eichenberger 1996, p.1306). The offer of a private
payoff had transformed a civic question into a pecuniary one. The introduction of
market norms crowded out their sense of civic duty (Kunreuther and Easterling 1996,
pp.61519; Frey, Oberholzer-Gee, and Eichenberger 1996, p.1301; for an argument
in favor of cash compensation, see OHare 1977).
Donation Day
Each year, on a designated day, Israeli high school students go door-to-door to
solicit donations for worthy causescancer research, aid to disabled children, and
so on. Gneezy and Rustichini (2000b) did an experiment to determine the effect
of financial incentives on the students motivations. They divided the students into
threegroups. One group of students was given a brief motivational speech about
the importance of the cause, and sent on its way. The second and thirdgroups were
given the same speech, but also offered a monetary reward based on the amount
they collected1percent and 10percent respectively. The rewards would not be
deducted from the charitable donations, but would come from a separate source.
Not surprisingly, the students who were offered 10percent collected more in
donations than those who were offered 1percent. But the unpaid students collected
more than either of the paid groups, including those who received the high commission. Gneezy and Rustichini (2000b, 802 807) conclude that, if youre going to use
134
financial incentives to motivate people, you should either pay enough or dont pay
at all. While it may be true that paying enough will get what you want, there is also
a lesson here about how money crowds out norms.
Why did both paid groups lag behind those doing it for free? Most likely, it was
because paying students to do a good deed changed the character of the activity.
Going door-to-door collecting funds for charity was now less about performing a
civic duty and more about earning a commission. The financial incentive transformed a public-spirited activity into a job for pay. As with the Swiss villagers, so
with the Israeli students: the introduction of market norms displaced, or at least
dampened, their moral and civic commitment.
Why worry about the tendency of markets to crowd out moral and civic ideals?
For two reasons one fiscal, the other ethical. From an economic point of view,
social norms such as civic virtue and public spiritedness are great bargains. They
motivate socially useful behavior that would otherwise cost a lot to buy. If you had to
rely on financial incentives to get communities to accept nuclear waste, you would
have to pay a lot more than if you could rely instead on the residents sense of civic
obligation. If you had to hire school children to collect charitable donations, you
would have to pay more than a 10percent commission to get the same result that
public spirit produces for free.
But to view moral and civic norms simply as cost-effective ways of motivating
people ignores the intrinsic value of the norms. Relying solely on cash payments
to induce residents to accept a nuclear waste facility is not only expensive; it is
corrupting. The reason it is corrupting is that it bypasses persuasion and the kind
of consent that arises from deliberating about the risks the facility poses and the
larger communitys need for it. In a similar way, paying students to collect charitable
contributions on donation day not only adds to the cost of fundraising; it dishonors
their public spirit and disfigures their moral and civic education.
Michael J. Sandel
135
process does not affect the product. Hirsch observed that this mistaken assumption
loomed large in the rising economic imperialism of the time, including attempts,
by Becker and others, to extend economic analysis into neighboring realms of social
and political life. The empirical cases weve just considered support Hirschs (1976)
insightthat the introduction of market incentives and mechanisms can change
peoples attitudes and crowd out nonmarket values.
A growing body of work in social psychology offers a possible explanation
for this commercialization effect. These studies highlight the difference between
intrinsic motivations (such as moral conviction or interest in the task at hand) and
external ones (such as money or other tangible rewards). When people are engaged
in an activity they consider intrinsically worthwhile, offering them money may
weaken their motivation by depreciating or crowding out their intrinsic interest or
commitment. (For an overview and analysis of 128studies on the effects of extrinsic
rewards on intrinsic motivations, see Deci, Koestner, and Ryan 1999).
Standard economic theory assumes that all motivations, whatever their character or source, are additive. But this misses the corrosive effect of money. The
crowding out phenomenon has far-reaching implications for economics. It calls
into question the use of market mechanisms and market reasoning in many aspects
of social life, including the use of financial incentives to motivate performance in
education, health care, the workplace, voluntary associations, civic life, and other
settings in which intrinsic motivations or moral commitments matter (Janssen and
Mendys -Kamphorst 2004).
Blood for Sale
Perhaps the best-known illustration of markets crowding out nonmarket norms
is a classic study of blood donation by the British sociologist Richard Titmuss. In
his book The Gift Relationship,, Titmuss (1971) compared the system of blood collection used in the United Kingdom, where all blood for transfusion was given by
unpaid, voluntary donors, and the system in the United States, where some blood
was donated and some bought by commercial blood banks from people, typically
the poor, who were willing to sell their blood as a way of making money. Titmuss
presented a wealth of data showing that, in economic and practical terms alone,
the UK blood collection system worked better than the American one. Despite the
supposed efficiency of markets, he argued, the American system led to chronic
shortages, wasted blood, higher costs, and a greater risk of blood contaminated by
hepatitis(pp.23132).
But Titmuss (1971) also leveled an ethical argument against the buying and
selling of blood. He argued that turning blood into a market commodity eroded
peoples sense of obligation to donate blood, diminished the spirit of altruism, and
undermined the gift relationship as an active feature of social life. Commercialization and profit in blood has been driving out the voluntary donor, he wrote.
Once people begin to view blood as a commodity that is routinely bought and sold,
Titmuss (pp.22324, 177) suggested, they are less likely to feel a moral responsibility to donate it.
136
Titmusss book prompted much debate. Among his critics was Kenneth Arrow
(1972). In taking issue with Titmuss, Arrow invoked twoassumptions about human
nature and moral life that economists often assert but rarely defend (for an insightful
contemporary reply to Arrow, see Singer 1973). The first is the assumption Ihave
examined above, that commercializing an activity doesnt change it. According to
this assumption, if a previously untraded good is made tradable, those who wish
to buy and sell it can do so, thereby increasing their utility, while those who regard
the good as priceless are free to desist from trafficking in it. This line of reasoning
leans heavily on the notion that creating a market in blood does not erode the value
or meaning of donating blood out of altruism. Titmuss attaches independent moral
value to the generosity that motivates the gift. But Arrow (1972, p. 351) doubts
that such generosity could be diminished or impaired by the introduction of a
market: Why should it be that the creation of a market for blood would decrease
the altruism embodied in giving blood?
The answer is that commercializing blood changes the meaning of donating
it. In a world where blood is routinely bought and sold, giving it away for free may
come to seem a kind of folly. Moreover, those who would donate a pint of blood at
their local Red Cross might wonder if doing so is an act of generosity or an unfair
labor practice that deprives a needy person of gainful employment selling his blood.
If you want to support a blood drive, would it be better to donate blood yourself, or
to donate $50 that can be used to buy an extra pint of blood from a homeless person
who needs the income?
The second assumption that figures in Arrows (1972) critique is that ethical
behavior is a commodity that needs to be economized. The idea is this: We should
not rely too heavily on altruism, generosity, solidarity, or civic duty, because these
moral sentiments are scarce resources that are depleted with use. Markets, which rely
on self-interest, spare us from using up the limited supply of virtue. So, for example,
if we rely on the generosity of the public for the supply of blood, there will be less
generosity left over for other social or charitable purposes. Like many economists,
Arrow (1972, pp.35455) writes, I do not want to rely too heavily on substituting
ethics for self-interest. I think it best on the whole that the requirement of ethical
behavior be confined to those circumstances where the price system breaks down
. . . We do not wish to use up recklessly the scarce resources of altruisticmotivation.
It is easy to see how this economistic conception of virtue, if true, provides yet
further grounds for extending markets into every sphere of life. If the supply of
altruism, generosity, and civic virtue is fixed, as if by nature, like the supply of fossil
fuels, then we should try to conserve it. The more we use, the less we have. On this
assumption, relying more on markets and less on morals is a way of preserving a
scarceresource.
Economizing Love
The classic statement of this idea was offered by Sir Dennis H. Robertson (1954),
a Cambridge University economist and former student of John Maynard Keynes,
in an address at the bicentennial of Columbia University. The title of Robertsons
137
lecture was a question: What does the economist economize? He sought to show
that, despite catering to what he called (p. 148) the aggressive and acquisitive
instincts of human beings, economists nonetheless serve a moral mission.
Robertson (1954) claimed that by promoting policies that rely, whenever
possible, on self-interest rather than altruism or moral considerations, the economist saves society from squandering its scarce supply of virtue. If we economists do
[our] business well, Robertson (p.154) concluded, we can, I believe, contribute
mightily to the economizing ... of that scarce resource Love, the most precious
thing in the world.
To those not steeped in economics, this way of thinking about the generous
virtues is strange, even far-fetched. It ignores the possibility that our capacity for
love and benevolence is not depleted with use but enlarged with practice. Think
of a loving couple. If, over a lifetime, they asked little of one another, in hopes of
hoarding their love, how well would they fare? Wouldnt their love deepen rather than
diminish the more they called upon it? Would they do better to treat one another in
more calculating fashion, to conserve their love for the times they really neededit?
Similar questions can be asked about social solidarity and civic virtue. Should
we try to conserve civic virtue by telling citizens to go shopping until their country
really needs them? Or do civic virtue and public spirit atrophy with disuse? Many
moralists have taken the second view. Aristotle ((Nicomachean Ethics,, BookII, chap.1,
pp. 1103a1103b) taught that virtue is something we cultivate with practice: We
become just by doing just acts, temperate by doing temperate acts, brave by doing
braveacts.
Rousseau (1762 [1973] Book III, chap. 15, pp. 239 40) held a similar view.
The more a country asks of its citizens, the greater their devotion to it. In a wellordered city every man flies to the assemblies. Under a bad government, no one
participates in public life because no one is interested in what happens there
and domestic cares are all-absorbing. Civic virtue is built up, not spent down, by
strenuous citizenship. Use it or lose it, Rousseau says, in effect. As soon as public
service ceases to be the chief business of the citizens, and they would rather serve
with their money than with their person, the state is not far from its fall.
The notion that love and generosity are scarce resources that are depleted with
use continues to exert a powerful hold on the moral imagination of economists,
even if they dont argue for it explicitly. It is not an official textbook principle, like
the law of supply and demand. No one has proven it empirically. It is more like an
adage, a piece of folk wisdom, to which many economists nonetheless subscribe.
Almost half a century after Robertsons lecture, Lawrence Summers, then the
president of Harvard University, was invited to offer the Morning Prayers address in
Harvards Memorial Church. He chose as his theme what economics can contribute
to thinking about moral questions. Economics, Summers (2003) stated, is too rarely
appreciated for its moral as well as practical significance. Summers observed that economists place great emphasis on respect for individualsand the needs, tastes, choices,
and judgments they make for themselves. He illustrated the moral implications of
economic thinking by challenging students who had advocated a boycott of goods
138
produced by sweatshop labor: We all deplore the conditions in which so many on this
planet work and the paltry compensation they receive. And yet there is surely some
moral force to the concern that as long as the workers are voluntarily employed, they
have chosen to work because they are working to their best alternative. Is narrowing an
individuals set of choices an act of respect, of charity, even of concern?
Summers (2003) concluded with a reply to those who criticize markets for
relying on selfishness and greed: We all have only so much altruism in us. Economists like me think of altruism as a valuable and rare good that needs conserving. Far
better to conserve it by designing a system in which peoples wants will be satisfied
by individuals being selfish, and saving that altruism for our families, our friends,
and the many social problems in this world that markets cannot solve.
Here was Robertsons (1954) adage reasserted. This economistic view of virtue
fuels the faith in markets and propels their reach into places they dont belong. But
the metaphor is questionable. Are altruism, generosity, solidarity, and civic spirit
like commodities that are depleted with use? Or are they more like muscles that
develop and grow stronger with exercise?
Michael J. Sandel
139
public good and where they dont belong, it should relinquish the claim to be a valueneutral science and reconnect with its origins in moral and politicalphilosophy.
I am grateful to the editors of this journal, David Autor, Timothy Taylor, and Ulrike
Malmendier, for their challenging comments and criticisms. Timothy Besleys recently
published essay (Besley 2013) on my book What Money Cant Buy helped me sharpen the
arguments of this paper, as did a valuable conversation with Peter Ganong. I would also
like to thank Robert Frank and the participants in New York Universitys Paduano seminar,
and my colleagues in Harvard Law Schools summer faculty workshop, for instructive and
penetrating discussions of an earlier version of this paper.
References
Anderson, Elizabeth. 1993. Value in Ethics and
Economics. Cambridge: Harvard University Press.
Aristotle. 1925. Nicomachean Ethics. Translated
by David Ross. New York: Oxford University Press.
Arrow, Kenneth J. 1972. Gifts and Exchanges.
Philosophy and Public Affairs 1(4): 343 62.
Associated Press. 2013. Tickets to Watch U.S.
Supreme Court Decisions are Costing People
Time, Money. San Francisco Examiner, March 25.
http://www.sfexaminer.com/sanfrancisco/tickets
-to-watch-us-supreme-court-decisions-on-gay
-marriage-are-costing-people-time-money/Content
?oid=2338811.
Atkinson, A. B. 2009. Economics as a Moral
Science. Economica 76(Issue Supplement S1):
791 804.
Besley, Timothy. 2013. Whats the Good of the
Market? An Essay on Michael Sandels What Money
Cant Buy. Journal of Economic Literature 51(2):
478 95.
Boulding, Kenneth E. 1964. The Meaning of the
Twentieth Century. New York: Harper & Row.
Bristow, Michael. 2007. Grey Areas
in Chinas One-Child Policy. BBC News,
September 21. http://news.bbc.co.uk/2/hi/asia
-pacific/7002201.stm.
Cain, Robyn Hagan. 2011. Need a Seat
at Supreme Court Oral Arguments? Hire a
Line Stander. FindLaw, September 2. http://
blogs.findlaw.com/supreme_court/2011/09
/need-a-seat-at-supreme-court-oral-arguments
-hire-a-line-stander.html.
Chivers, C. J. 2002. A Big Game. New York
Times Magazine, August 25. http://www.nytimes
.com/2002/08/25/magazine/a-big-game
.html?pagewanted=all&src=pm.
Coonan, Clifford. 2011. China Eases Rules on
One Child Policy. Independent (London), April1.
http://www.independent.co.uk/news/world
/asia/china-eases-rules-on-one-child-policy
-2258875.html.
Copeland, Libby. 2005. The Line Starts Here.
Washington Post, March 2, p. C1. http://www
.washingtonpost.com/wp-dyn/articles/A64231
-2005Mar1.html.
de la Croix, David, and Axel Gosseries. 2006. Procreation, Migration and Tradable Quotas. CORE
Discussion Paper No. 2006/98, November. Available
at SSRN: http://ssrn.com/abstract=970294.
Deci, Edward L., Richard Koestner, and Richard
M. Ryan. 1999. A Meta-Analytic Review of Experiments Examining the Effects of Extrinsic Rewards
on Intrinsic Motivation. Psychological Bulletin
125(6): 627 68.
Frey, Bruno S. 1997. Not Just for the Money: An
Economic Theory of Personal Motivation. Cheltenham,
UK: Edward Elgar Publishing.
Frey, Bruno S., and Reto Jegen. 2001. Motivation Crowding Theory. Journal of Economic Surveys
15(5): 589 611.
Frey, Bruno S., and Felix Oberholzer-Gee.
1997. The Cost of Price Incentives: An Empirical
Analysis of Motivation Crowding-Out. American
Economic Review 87(4): 746 55.
Frey, Bruno S., Felix Oberholzer-Gee, Reiner
Eichenberger. 1996. The Old Lady Visits Your
Backyard: A Tale of Morals and Markets. Journal of
Political Economy 104(6): 12971313.
140
conomists have made use of, and have contributed to the development of,
many branches of moral theory, including utilitarianism, social contract
theory, libertarianism, and maximin and capability theories of justice. In
contrast, virtue ethics the study of moral characterhas been an important
strand in moral philosophy for literally thousands of years, but has received little
attention from contemporary economists. That neglect has not been reciprocated.
A significant body of philosophical work in virtue ethics is associated with a radical
critique of the market economy and of economics. Expressed crudely, the charge
sheet is this: The market depends on instrumental rationality and extrinsic motivation; market interactions therefore fail to respect the internal value of human
practices and the intrinsic motivations of human actors; by using market exchange
as its central model, economics normalizes extrinsic motivation, not only in markets
but also (in its ventures into the territories of other social sciences) in social life
more generally; therefore economics is complicit in an assault on virtue and on
human flourishing. We will argue that this critique is flawed, both as a description of how markets actually work and as a representation of how classical and
neoclassical economists have understood the market. We will show how the market
and economics can be defended against the critique from virtue ethics.
Crucially, our response to that critique will be constructed using the language
and logic of virtue ethics.. In this respect, it is fundamentally different from a response
that many economists would find more naturalto point to the enormous benefits,
Luigino Bruni is Professor of Economics, LUMSA University, Rome, Italy. Robert Sugden
is Professor of Economics, University of East Anglia, Norwich, United Kingdom. Their emails
are [email protected] and [email protected].
http://dx.doi.org/10.1257/jep.27.4.141
doi=10.1257/jep.27.4.141
142
including income and leisure that can be devoted to intrinsically motivated activities, that we all enjoy as a result of the workings of markets, and to the essential role
of economics in explaining how markets work. Set against those benefits, it can be
argued, questions about whether market motivations are virtuous are second-order
concerns that economists can safely leave to moral philosophers. Thus, for example,
responding to the philosopher Michael Sandels objection to markets in carbon
dioxide emissions on the grounds that they express nonvirtuous attitudes to the
environment (Sandel 2012, pp.7276), Coyle (2012) writes, I would rather see an
effective scheme to reduce greenhouse gas emissions, but then Im an economist.
We are economists too, and have some sympathy with such sentiments. Nevertheless, the virtue-ethical critique of economics is gaining credence in public debate.
Many people see it as providing intellectual support for popular attitudes of opposition to capitalism and globalization, and of hostility to economics as a discipline.
Philosophically, the critique is grounded in an ancient and respected tradition of
ethical thought: it is not something that economics can or should simply brush
aside. Our premise is that economics needs a response to this critique that takes
virtue ethicsseriously.
Another possible reply, made for example by van Staveren (2009) and Besley
(2013), is that, in their critique of economics,, the virtue ethicists fail to recognize
the diversity of the discipline. Economics has never been unanimous or unconditional in advocating markets; indeed, it is possible to read the development
of normative economics in terms of a continually expanding catalog of market
failures and their remedies. In particular, a recent development in economics
has been the growth of a literature in which concepts of intrinsic motivation are
used to explain individual behavior. Although this work is not explicitly virtueethical in the normative sense, it allows economics to model a crowding-out
mechanism that is similar to the virtue ethicists account of the corrupting effects
of markets. However, pointing to the diversity of economics merely deflects the
virtue-ethical critique from economics in general to a particular but surely major
tradition of economic thoughtthat liberal tradition that understands the market
as a domain in which socially desirable consequences emerge from the pursuit of
private interests. In contrast, our response meets the critique head-on. We aim to
show that economists can teach about and defend the market without standing for
nonvirtue againstvirtue.1
The logic of our response requires that we use the modes of argument of virtue
ethics. We write as philosophically and historically inclined economists, hoping to
be read both by philosophers and by our fellow economists. For the benefit of the
economists and with apologies to the philosophers, we assume no prior knowledge
of virtue ethics on the part of the reader. Thus, we begin with a brief introduction
In this respect, our approach has more in common with McCloskeys (2006) account of the bourgeois
virtues. However, our analysis is more systematic and economics-specific than McCloskeys imaginative but discursive exploration of the seven virtues of traditional Christian thought and their role in
economiclife.
143
to virtue ethics. We then describe some prominent critiques of the market that are
grounded in virtue ethics and in the related economic and psychological literature
on intrinsic motivation.
Following this introduction, we use the methods of virtue ethics to develop a
conception of market virtue that is consistent with many classical and neoclassical
economists accounts of how markets work and of what purposes they serve. Our
central idea is that the public benefits of markets should be understood as the
aggregate of the mutual benefits gained by individuals as parties to voluntary transactions, and that the market virtues are dispositions that are directed at this kind
of mutual benefit. For a virtuous market participant, mutual benefit is not just a
fortunate by-product of the individual pursuit of self-interest: he or she intends that
transactions with others are mutually beneficial.
Using this idea, we identify some specific character traits that have the status
of virtues within the domain of the market. Our list of market virtues (which we do
not claim is complete) includes universality, enterprise and alertness, respect for the
tastes of ones trading partners, trust and trustworthiness, acceptance of competition,
self-help, non-rivalry, and stoicism about reward. We will argue that these market
virtues, grounded on ideas of reciprocity and mutual benefit, are closely associated
with virtues of civil society more generally. It is therefore a mistake to think that the
market is a virtue-free zone, or that the character traits that best equip individuals to
flourish in markets are necessarily corrosive of virtue in other domains oflife.
The idea that economic agents should understand their interactions as mutual
assistance is characteristic of a tradition of natural-law philosophy from which
mainstream economic thought turned away in the later eighteenth century. Nevertheless, as we will show, the idea that mutual benefit is in some sense the purpose of
the market is implicit in the writings of many major economists from the eighteenth
century to the present day. The specific market virtues that we present feature in
some canonical accounts of the desirable properties of markets. In this sense, our
paper can also be read as an attempt to reconstruct a submerged current of virtueethical thought in economics.
144
within any practice or domain of life, goodness is understood in relation to the telos
(literally, end or purpose) of that domainthat for whose sake everything is
done. For example, Aristotle (Book1, section1) treats medicine as a domain whose
telos is health and military strategy as a domain whose telos is victory. In relation to
a given domain, an acquired character trait is a virtue to the extent that the person
who possesses it is thereby better able to contribute to the telos of that domain. The
underlying idea is that human happiness or flourishing (eudaimonia)
(
) requires that
people are oriented towards their various activities in ways that respect the intrinsic
ends of the domains to which those activities belong.
How is the telos of a domain determined? Aristotle seems to think of the telos
as a natural fact that can be ascertained by intuition, but many modern virtue ethicists favor a communitarian approach. This approach, exemplified by the work of
MacIntyre (1984), understands the concept of flourishing as internal to specific
communities and cultural traditions. Thus, to identify the telos of a practice, one
must discover the meaning of that practice within the community of practitioners.
In this view, a claim about the telos of an practice is not just the expression of a
personal value judgement; it involves some (perhaps creative) interpretation of
what is already there (Sandel 2009, pp.184 192, 203 207; Anderson 1993, p.143).
As Sandel (p.98) puts it, we identify the norms appropriate to social practices by
trying to grasp the characteristic end, or purpose, of those practices.
There is much common ground between Aristotelian virtue ethics, with its
emphasis on the intrinsic value of practices, and those strands of modern positive
psychology that emphasise the importance of intrinsic motivation for human happiness, in particular the self-determination theory of Deci and Ryan (1985). In this theory,
the analog of flourishing is a concept of psychological health or well-being. The core
hypothesis is that individual autonomy is a source of psychological well-being, and
thus that human flourishing is linked with authenticity and self-realization. In Ryan
and Decis (2000) taxonomy of motivation, there is a continuum from amotivation, through increasingly autonomous forms of extrinsic motivation, to the full
autonomy of intrinsic motivation. A person who is extrinsically motivated performs
an activity in order to obtain some separable outcome. Extrinsic motivations can
become more internal (and thereby more autonomous) to the extent that the
individual has a sense of having chosen the objective on which he acts and endorsed
its value. But an intrinsically motivated person performs an activity for its inherent
satisfactions rather than for some separable consequence; such a person is moved to
act for the fun or challenge entailed rather than because of external prods, pressures,
or rewards (pp.5660). Thus, the analog of telos is the meaning that an individual
attaches to an activity when he sees the activity as an end initself.
145
back to Aristotle, who wrote (Book1, 5): The life of money-making is one undertaken under compulsion, and wealth is evidently not the good we are seeking;
for it is merely useful and for the sake of something else. This sentence makes
twoclaims that are echoed in critiques of economics made by modern virtue ethicists. The first claim is that when individuals participate in markets, they show a lack
of autonomythey act under compulsion.. The suggestion seems to be that a truly
autonomous person would not need to seek wealth (perhaps because he would
already have as much as he needed without having to seek for it).2 The second claim
is that the motivation for economic activity is extrinsic and thereby of an inferior
kindthe things that economic activity can achieve are merely useful and for the sake
of something else..
Here, we will focus on how threeprominent contemporary virtue ethicists apply
these themes in their writings about economics and the market. Of these criticisms
of the market, MacIntyres (1984) book After Virtue is the most radical. Taken literally,
MacIntyres elegant despair has no real point of contact with modern economics. But
precisely because it takes the critique of the instrumentality of markets to its logical
conclusion, it offers a useful point of reference. MacIntyre (p. 187) presents an
account of morality that is built on the concept of a practice.. A practice is a coherent
and complex form of socially established cooperative human activity which realizes
goods internal to that form of activity. A practice has intrinsic ends, and internal
standards of excellence that make sense in relation to those ends. Associated with
the practice are certain acquired character traits that assist in the achievement of
excellence, or in recognizing and internalizing communal understandings of the
meaning of the practice. The traits can be viewed as the virtues of the practice.
For MacIntyre (1984), a person who fails to treat an activity as a practice with
an internal end is failing to display virtue either because the activity falls within a
practice whose internal ends the person is failing to respect, or because the activity
is of such a morally impoverished and instrumental kind that it is not a practice at
allMacIntyres (p.187) questionable example of an activity that does not count as
a practice is bricklaying. This way of thinking immediately makes markets morally
suspect. The market motivation of creating goods for exchange conflicts with the idea
that activities, or the goods that they realize, are ends in themselves.. Thus, according
to MacIntyre, the exposure of a practice to market forces is liable to corrupt its excellences and virtues. MacIntyre does not quite claim that practices can never coexist
with market exchange. For example, he maintains that portrait painting from the
time of Giotto to that of Rembrandt was a practice with internal ends and standards
of excellence. He recognizes that many excellent painters were also able to achieve
(and presumably cared about) goods external to the practice of art, including the
income they were able to earn from the sale of their services (pp.189190). The
suggestion is that the corrupting tendencies of the market can be contained only
In a witty account of the history of Western intellectuals criticisms of capitalism, Alan Kahan (2010,
p.31) presents the Three Donts of anti-capitalism. The first is Dont make money (just have it).
146
to the extent that individuals are at least partially motivated by the internal ends of
practices (as, in MacIntyres account, the great painters were).
However, as MacIntyre (1984) recognizes, practices as he understands them
are not, and cannot be, characteristic of ordinary economic life in the world in
which we live. Treating the household as a paradigm case of communal life, he
argues that [o]ne of the key moments in the creation of modernity occurs when
production moves from the household to an impersonal domain of means-ends
relationships (p. 227). This thought reflects the presupposition that production
for exchange belongs to the domain of external goods. The implication is that an
economy of practices cannot make effective use of comparative advantage and the
division of labor. MacIntyres ultimate response to economic reality is a yearning for
an imagined and ill-defined economy of communal production somehow devoid of
the hierarchical power relationships found in real historical economies.
Similar themes, developed in somewhat less unworldly forms, are prominent
in the work of Anderson (1993) and Sandel (2009, 2012). These writers recognize,
at times reluctantly, that markets are a necessary part of social organization. But
they argue that the instrumental logic of markets is liable to corrupt virtues that are
proper to other domains of social life, and that it is therefore appropriate for the
state to impose limits on the scope of markets.
Thus, the first sentence of Andersons Value in Ethics and Economics (1993)
is: Why not put everything up for sale? This rhetorical question signals several
elements of her position: to allow all areas of social life to be governed by market
relationships would be morally objectionable; this truth ought to be obvious to
a morally aware reader; but some opinion-formers do want to put everything up
for sale, and their arguments need to be countered. More specifically, the people
against whom she is arguing fail to understand that there are ways we ought to value
people and things that cant be expressed through market norms (pp.xixiii).
Anderson (1993) proposes a pluralist theory of value in which different kinds
of goods ought to be valued in different ways (p.12). She tries to delimit the proper
scope of the market by identifying the norms that are characteristic of market relations, and the corresponding class of goods that are properly valued in terms of
those norms. For Anderson, the ideal economic good is a pure commodity. The
mode of valuation appropriate to pure commodities is use. She writes (p.144):
Use is a lower, impersonal, and exclusive mode of valuation. It is contrasted with
higher modes of valuation, such as respect. To merely use something is to subordinate it to ones own ends, without regard for its intrinsic value. This definition
immediately introduces the Aristotelian ranking of intrinsic value over instrumental
value. Anderson is presenting market norms as a kind of second-rate morality: the
markets mode of valuation is lower than that of other domains of social life; it is
merely use; it has no regard for intrinsic value. In this account, market norms are
impersonal and egoistic.. Impersonality is the idea that market transactions are viewed
instrumentally: each party to a transaction considers it only as a means to the satisfaction of his own ends. Egoism is the idea that those ends are defined in terms
ofself-interest.
147
In a more recent book, Sandel (2012) presents an argument about the moral limits of markets.
His paper in this issue takes up some of these arguments. As he acknowledges (p.208, note18), this
argument is similar to that of Anderson (1993). Sandel sees economics as complicit in the inappropriate
propagation of market values. Sandel is less precise than Anderson in explaining what those values are,
but it is clear that he sees them in opposition to the civic virtues of social solidarity, including shar[ing]
in a common life, and car[ing] for the common good (pp.128, 203).
148
bottled water in the aftermath of Hurricane Charley in Florida in 2004. At the time,
some economists argued that market-clearing prices promote efficiency in the use
of resources, and that this truth is not invalidated by hurricanes. Sandel sides with
the opinion that this kind of price gouging should be illegal. His reason is an
application of virtue ethics: the firms that charged scarcity prices were motivated
by greed; since greed is a vice, a bad way of being, the state should discourage it
(pp. 7 8). The second issue is the remuneration of senior corporate executives.
Sandel asks whether the chief executive officers of large American corporations
deserved the payments they received in the years leading up to 2008, when their
firms were generating large profits. We are invited to conclude that effort and talent
are qualities that are worthy of reward in business, but that when the market rewards
executives for profits that are not attributable to effort or talent, a principle of justice
is being violated (pp. 1218). The message from both examples, developed over
the course of the book, is that the market generates incomes that are not properly
aligned with the virtues of the people who receive them.
To an economically trained reader, these critiques of economics and the market
often seem divorced from the reality of everyday economic life. MacIntyre (1984)
and Anderson (1993) seem to find it hard to find moral significance in the ordinary
useful jobs by which most people earn their livings. Sandel (2009) seems to find
it hard to come to terms with the fact that market rewards depend on luck as well
as talent and effort. We will argue that virtue ethicists are failing to find virtue in
markets because they are not seeing the market as a practice in its own right.
149
reciprocity, or public spirit. If financial incentives are introduced into such a setting,
this prompts the thought that people who supply blood may be self-interested sellers
rather than altruistic donors. This can undermine the sense of would-be donors that
giving blood is a morally significant and socially valued act, and so lead to a reduction
in the supply of blood. A similar interpretation is now often given for the muchdiscussed finding that fines for lateness in collecting children from a day-care center
led to an increase in the incidence of lateness (Gneezy and Rustichini2000a).
The economic implications of the hypothesis of motivational crowding-out
were first explored by Frey (1994, 1997).4 Defining intrinsic motivation in essentially
the same way as Deci and Ryan do, Frey (1997, p.2) maintains that it is neither
possible nor desirable to build a society solely or even mainly on monetary incentives; intrinsic motivation has an essential role to play.
Within economics, there is growing interest in theorizing about how intrinsic
motivation can be shielded from market forces. One approach is summarized in
the slogan getting more by paying less. Suppose there is some occupation, say
nursing, in which workers are better able to provide the services that their employers
value if they are intrinsically motivated to pursue the internal ends of that occupationif, in Ryan and Decis (2000) terminology, they are attracted by its inherent
satisfactions and challenges. Viewed in the standard conceptual framework of
economics, a person with such a motivation for nursing has a lower reservation wage
for working as a nurse than for working in other occupations. So employers may be
able to separate the better workers from the worse by offering low wagesthey can
get more by paying less (Brennan 1996; Katz and Handy 1998; Heyes 2005). When a
person accepts the low wages of an employer who is looking for intrinsic motivation,
she signals to herself and to others that she is intrinsically motivated. So there need
be no crowding-outeffect.
We suspect that many readers will share our unease about this argument. Nelson
(2005) formulates this unease by raising two objections. First, because low wages
may screen out intrinsically motivated individuals who need to support themselves
and their families, access to intrinsically rewarding occupations may be restricted
to people with private incomes or well-off partners or parents. Second, when social
norms treat self-sacrifice as a characteristic virtue of caring occupations such as
nursing, they act as a cover for, and an incitement to, exploitation. These objections
draw attention to a questionable assumption of the getting more by paying less
argumentthat a person is virtuous or authentic to the extent to which that person
is willing to sacrifice material rewards in the pursuit of intrinsic ends. In a model in
which all motivations are represented as properties of individuals preferences, that
assumption is almost unavoidable, since an individuals preference for consuming
It is only very recently that economists have taken this hypothesis seriously. Titmusss (1970) work
was well-known to economists in the 1970s, but his crowding-out argument was viewed skeptically (for
example, Arrow 1972). Gneezy and Rustichini (2000a, 2000b) discussed motivational crowding-out as
a possible explanation of their findings, but favored a more conventional economic interpretation in
terms of incomplete contracts.
150
an intrinsic good is defined in terms of how much of other goods she is willing to
give up in exchange. However, it is not an essential part of a virtue-ethical approach
in which the exercise of virtue is associated with flourishing rather than sacrifice,
nor of a decision-theoretic approach in which intentions for mutual benefit are
represented as team reasoning (Bruni and Sugden 2008).
Folbre and Nelson (2000) suggest that the crowding-out problem can be countered by separating the payment of intrinsically motivated workers from the specific
services they provide, so that payment can be construed as an acknowledgement of
intrinsic motivation rather than as one side of a market exchange. The implication
seems to be that authentic caring is compromised if carers and cared see their relationship as that of seller and buyer. There is another echo here of the Aristotelian
idea that market relationships are instrumental and thereby nonvirtuous.
But how is the payment of service suppliers to be separated from exchange
relationships? One possibility is to use gift relationships. Consider the case of restaurant waiters who are paid less than the market wage, but with the expectation that
their earnings will be supplemented by tips from customers. Perhaps this practice
supports dispositions towards friendliness and efficiency that restaurant owners
value in their waiters and find costly to monitor, but one might think that it impairs
rather than supports the waiters sense of autonomy.
A different model (and probably the one that Folbre and Nelson 2000 have in
mind) is that of a salaried professional. Think of the role of the tenured academic in a
well-financed university, as that role used to be (and sometimes still is) understood.
The academic is awarded tenure in the expectation of a continuing intrinsic motivation to pursue excellence in teaching and research, but is subject to only the lightest
of monitoring. He is paid a good salary that has no direct relationship to the services
he provides, but is seen as expressing a social valuation of the excellence that is
expected. Actual excellence in teaching will be rewarded by the gratitude of students;
excellence in research, by the respect of peers. This kind of separation of payment
from services rendered can give professionals an enviable degree of autonomy; and it
can protect whatever intrinsic motivation they have from crowding-out effects. But
it also insulates them from pressures to respond to the interests of the people to
whom their services are being provided. Just as the waiter loses autonomy in having
to depend on the good will of the customer, so does the client in having to depend
on the professionals intrinsic motivation.
These examples illustrate the difficulty of shielding intrinsic motivation from
the supposedly corrosive effects of exchange relationships. These difficulties have
a common source: it is inherent in the concept of intrinsic motivation that an
individuals autonomy and authenticity are compromised whenever she enters into
exchange relationships, but such relationships are fundamental to the workings of
any economy that relies on comparative advantage and the division of labor. The
literature of intrinsic motivation invites us to aspire to the ideal of an economy in
which everyones actions and efforts are coordinated to realize gains from trade,
but in which no one is actually motivated to seek those gains. This ideal seems as
profoundly unrealistic as MacIntyres (1984) imaginary world of an economy built
151
on practices. If we are to reconcile the ideas of virtue and authenticity with real
economic life, we need a way of understanding market relationships that acknowledges that gains from trade are not realized by accident: they are realized because
individuals seek them out.
152
to set aside their skepticism for a moment, and to translate this question into
common-sense terms. What is the characteristic end or purpose or raison dtre
of the market? How would you describe, in the simplest and most general terms,
what markets do that is valuable? If you had to write a mission statement for the
market, what would it say?
Thoughtful economists have offered answers to such questions. For example,
Friedman (1962, p. 13) wrote that, in relation to the problem of coordinating
economic activity, the technique of the market place is voluntary cooperation of
individuals. Buchanan and Tullock (1962, p.103) wrote: The raison dtre of market
exchange is the expectation of mutual gains. We are not claiming here that Friedman,
Buchanan, and Tullock are virtue ethicists. All we are attributing to them is the idea
that markets have a point or purpose, and that that purpose is mutual benefit. Most
economists, faced with our questions, would probably invoke in one way or another
the idea of mutual benefit or gains from trade through voluntarytransactions.
If economists were asked to nominate one simple diagrammatic representation
of a market, the Edgeworth box would surely be one of the commonest choices,
and the point of that diagram is to understand markets as networks of mutually
beneficial voluntary transactions. Edgeworth (1881, pp.16 17) himself, in a famous
passage in which he declares that the first principle of economics is that every agent
is activated only by self-interest, distinguishes between war and contract, differentiated by whether the agent acts without, or with, the consent of others affected
by his actions; his analysis of competitive markets is presented as an analysis of
contract. If economists were asked to nominate a theorem to represent the market
in its best light, many would opt for the first fundamental theorem of welfare
economics, which is essentially equivalent to showing that in competitive equilibrium, no opportunities for mutually beneficial transactions, however complex,
remain unexploited. Another strong contender would be Ricardos (1817, Ch.7)
comparative advantage theorem, which shows that there are typically opportunities
for gains from trade between any pair of countries (and by extension, any pair of
individuals), whatever their respective endowments and productivity.
How else might one answer our question about the telos of the market? One
obvious alternative answer is that the telos of the market is wealth creation: after
all, the founding text of economics is called The Wealth of Nations.. But even for the
author of that text, the fundamental mechanism by which wealth is created is
the division of labor and the extension of the market, and the division of labor is the
consequence of the human propensity to truck, barter and exchange one thing
for another (Smith 1776 [1976], p.25). Other economists have emphasised how
the market creates wealth by exploiting comparative advantage (Ricardo 1817), the
division of knowledge (Hayek 1948), and increasing returns to scale (Marshall 1920,
pp. 222242; Arrow 1984, p. 188); but all of these mechanisms operate through
mutual gains from trade. Another possible answer is that the telos of the market is
economic freedom. The association between the market and freedom is a recurring
theme in economics; famous expositors of this idea include Mill (1848 [1910]),
Marshall (1920, p.8), Hayek (1948), and Friedman (1962). But economic freedom
153
is not the freedom of each person to get what he wants tout court;; it is his freedom to
use his own possessions and talents as he sees fit and to trade with whoever is willing
to trade withhim.
We suggest that the common core of these understandings of markets is that
markets facilitate mutually beneficial voluntary transactions. Such transactions can
be seen as valuable because individuals want to make them, because they satisfy
individuals preferences, because they create wealth, and because the opportunity
to make them is a form of freedom. We therefore propose to treat mutual benefit
as the telos of the market.
Market Virtues
On the supposition that the telos of the market is mutual benefit, a market
virtue in the sense of virtue ethics is an acquired character trait with twoproperties: possession of the trait makes an individual better able to play a part in the
creation of mutual benefit through market transactions; and the trait expresses an
intentional orientation towards and a respect for mutual benefit. In this section, we
present a catalog of traits with these properties, without claiming that our catalog
isexhaustive.
According to the logic of virtue ethics, such traits are properly or consistently
viewed as praiseworthy within the practice of the market, when that practice is understood as directed at mutual benefit. Thus, we should expect the traits in our catalog
to have been evaluated favorably in the tradition of liberal economic thought from
which we have distilled the telos of mutual benefit. We maintain that this is the case,
and will point to illustrative examples. Recall that virtue ethicists claim to uncover
the virtues of practices by philosophical reflection, and not simply by sociological
observation. It is in the spirit of such enquiry to look to thoughtful economists as
well as to market participants for insights into the nature of marketvirtues.
We will not claim that all market participants display the market virtues. (The
logic of virtue ethics does not require that kind of implausibility: virtue ethicists can,
for example, describe bravery as a military virtue without asserting that all soldiers
are brave.) But we do maintain that the market virtues are broadly descriptive of
traits that many people, including people who are successful in business, display
when they participate in markets. Readers who are accustomed to equating virtue
with self-sacrifice may suspect that this claim is overoptimistic, but we repeat that
such an equation is alien to virtue ethics. It is fundamental to the classical and
neoclassical understanding of markets that, under normal circumstances, each
party to a market transaction benefits from involvement in it. Thus, a disposition
to seek mutual benefit in markets will normally incline individuals towards the
kinds of individually beneficial behavior that economic theory has traditionally
described. Our account of market virtue is not a new theory of nonselfish behavior.
It is a description of a distinctive moral attitude to market relationshipsan attitude
characterized not by altruism but by reciprocity.
154
Universality
Our first market virtue is universalitythe disposition to make mutually beneficial transactions with others on terms of equality, whoever those others may be.
If the market is to be viewed as an institution that promotes the widest possible
network of mutually beneficial transactions, universality has to be seen as a virtue.
Its oppositesfavoritism, familialism, patronage, protectionismare all barriers to
the extension of the market.
It is intrinsic to the virtue of universality that market relations are not based on
personal ties of kinship, community, friendship, or gratitudethe kind of ties that
Anderson (1993) sees as characteristic of higher modes of valuation. As Smith
(1776 [1976], p. 27) makes clear in his account of how we get our dinners, it is
because the market is based on free horizontal relations between equals that it
allows us to satisfy our economic needs with independence and self-respect. This
independence can be compromised if economic transactions depend on relations
other than mutual benefit. However, this is not to say that market relations must
be impersonal in the sense that each party treats the other merely as a means to an
end. When trading partners intend their transactions to be mutually beneficial, it
is possible for their relations to have the characteristics of friendliness and goodwill
that we (Bruni and Sugden 2008) describe as fraternity.
Friedman (1962, pp.108 118) identifies another valuable aspect of universality
when he argues that market forces tend to counter racial and religious prejudice.
His leading example is the case of the Jews of medieval Europe, who (between
outbreaks of outright persecution) were able to survive in a hostile social environment by working on their own account and trading with non-Jews. For Friedman, it
must be said, universality is a desirable but unintended consequence of the pursuit
of self-interest, rather than a virtue in our sense; but nonetheless, the customer who
chooses where to shop on the basis of price and quality rather than the shopkeepers
religion can be thought of as exhibiting a market virtue.
Enterprise and Alertness
If the telos of the market is mutual benefit, enterprise in seeking out mutual
benefit must be a virtue. Discovering and anticipating what other people want
and are willing to pay for is a crucial component of entrepreneurship. (Think
of Freddie Lakers pioneering of no-frills aviation, Steve Jobss development of
graphical user interfaces, or Art Frys discovery of the commercial potential of the
Post-it.) Successful entrepreneurship requires empathy and imagination, as Jevons
(1871 [1970], pp.102103) recognized in one of the founding texts of neoclassical
economics: Every manufacturer knows and feels how closely he must anticipate the
tastes and needs of his customers: his whole success depends on it.
The virtue of alertness to mutual benefit applies to both sides of the market:
for mutual benefit to be created, the alertness of a seller has to engage with the
alertness of a buyer. Thus, the inclination to shop around, to compare prices, and
to experiment with new products and new suppliers must be a virtue for consumers.
Arguing that the law of one price has more application to wholesale than to retail
155
markets, Mill (1848 [1909], p. 441) wrote: Either from indolence, or carelessness, or because people think it fine to pay and ask no questions, three-fourths
of those who can afford it give much higher prices than necessary for the things
they consume. Notice how Mills empirical claim that well-off consumers are not
inclined to search for the lowest prices is linked with moral criticism.
Respect for the Tastes of Ones Trading Partners
One is more likely to succeed in making mutually beneficial transactions if one is
disposed to respect the preferences of potential trading partners. The spirit of this virtue
is encapsulated in the business maxim that the customer is always right. This virtue is
closely related to the idea that market transactions are made on terms of equality, and
opposed to the paternalistic idea that the relationship of supplier to customer is that
of guardian to ward. It is also opposed to the idea of virtues based on intrinsic motivation, or on professional and craft standards. It is perhaps true (asMacIntyre 1984 and
Anderson 1993 claim) that when professionals and craft workers sell their services,
they are liable to compromise the standards of excellence that are internal to their
respective practices, but that does not invalidate the proposition that producing what
customers do want to buy is an aspect of a practicethe practice of the marketwith
its own standards of excellence and its own forms of authenticity. From this perspective, it is unsurprising that Smith (1776 [1976], pp.758764) favored the payment of
university teachers by their students on a fee-for-service basisa practice that gives the
relationship between professional and client essentially the same status as that between
shopkeeper and customer.
In speaking of respect for the preferences of trading partners, we mean something more than the recognition that satisfying those preferences is a source of
profit. Consider a famous case in which this virtue is lacking. Gerald Ratner, the
chief executive of a (then) successful low-price British jewelery business, made
a speech in 1991 to the Institute of Directors in which he referred to his firms
products with the joke: People say, How can you sell this for such a low price?
Isay, because its total crap. When this was reported in the press, the business
lost 500 million in market value and eventually had to be relaunched with a new
nameand Ratner lost his job (Ratner 2007). Notice that Ratner was not saying,
as suppliers of lower-priced products often and quite properly do, that what he
was selling was cheap and cheerful and aimed at those consumers for whom value
for money was a priority. But nor, as we understand this story, was he confessing to
taking advantage of some lack of information on the part of his customers, and so
failing to return their trust: the objective properties of his products were transparent
enough. He was expressing contempt for the tastes to which his business catered,
and thereby for the idea that the relationship between supplier and customer is
one of mutual benefit.
Trust and Trustworthiness
Because the monitoring and enforcement of contracts is often difficult or
costly, dispositions of trust and trustworthiness (qualified by due caution against
156
157
This market virtue seems inescapable, given our approach, but there is no
denying that traders often find it hard to live by. For example, Adam Smith famously
claimed: People of the same trade seldom meet together, even for merriment and
diversion, but the conversation ends in a conspiracy against the public, or some
contrivance to raise prices (1776 [1976], p.45). Nevertheless, it is obvious from
the tone of these and similar remarksfor example about the wretched spirit of
monopoly (p.461)that Smith does not approve of this trait. The idea that cartel
agreements are unethicalunworthy of a virtuous traderis a recurring theme in
the writings of pro-market economists. Even Friedman (1962, pp. 131132), who
argues that market power is not a serious problem unless it is positively supported
by governments, approves the common law doctrine that combinations in restraint
of trade are unenforceable in the courts.
This is a convenient place to ask whether being concerned about externalities
resulting from ones activities should be included among the market virtues. One
way of posing this question is to ask whether the telos of the market is mutual benefit
among the parties to market transactions (considered severally), or mutual benefit among
everyone in a society.. We suggest the former. On this view, the existence of externalities can be a reason for governments to regulate markets, but self-regulation is not
part of the internal practice of the market.5
Self-Help
Within the practice of a market that is structured by mutual benefit, each individuals wants and aspirations are relevant to others only in so far as they can be
satisfied in mutually beneficial transactions. Thus, it is a market virtue to accept
without complaint that others will be motivated to satisfy your wants, or to provide
you with opportunities for self-realization, only if you offer something that they
are willing to accept in return. Smith (1776 [1976], p. 45) appeals to the virtue
of self-help or independence when, in relation to how we get our dinners, he
writes: Nobody but a beggar chuses to depend chiefly upon the benevolence of his
fellow-citizens. (The phrase chuses to is important here. Smith is not denigrating
dependence on others by people who have no other means of subsistence.)
A person who upholds the virtue of self-help will avoid asking others to reward
her for producing goods that those others do not value. Thus, for example, an artist
will not treat the intrinsic value of her work, as judged within the practice of art, as
a reason to be paid by people (whether as consumers or as taxpayers) who do not
recognize that work as beneficial to them. Nor will she treat the self-realization that
she achieves through that work as a reason to be paid. In this respect, the market
virtue of self-help conflicts with the positions taken by Anderson (1993) and Sandel
5
To this extent, we agree with Friedman (1962, pp.133 36) that social responsibility is not a proper
role of business. However, Friedman argues that the only responsibility of business is to use its resources
and engage in activities designed to increase its profits so long as it stays within the rules of the game,
which is to say, engages in open and free competition, without deception or fraud. Our idea that market
virtue involves intentions for mutual benefit is broader than this claim.
158
159
attitudes as virtuous. For example, Hume (1760 [1985], pp.32728) argues against
the narrow and malignant opinion that the relationship between commercial
economies is that of zero-sum rivalry: [T]he encrease of riches and commerce in
any one nation, instead of hurting, commonly promotes the riches and commerce
of all its neighbours.6 Writing almost a century later, Mill (1848 [1909], pp.581 82)
expresses the same sentiment: [C]ommerce first taught nations to see with good will
the wealth and prosperity of one another. Before, the patriot . . . wished all countries
weak, poor, and ill-governed, but his own: now he sees in their wealth and progress a
direct source of wealth and progress to his own country.
What about rivalry between firms, and in particular the case in which the
successful entry of one firm into an industry squeezes out another? Even in these
cases, the motivation of the entrant need not be positional. Indeed, even a selfinterested entrant would have no reason to want to displace an incumbent firm,
except as a means of making profit; and that profit can be earned only through
mutually beneficial transactions with customers. A virtuous entrant, one might say,
intends that the transactions he offers to make are mutually beneficial for the parties
that will be involved in them;; the entrant does not intend or take satisfaction in the
failure of competitors, even if that external effect is a predictable consequence of
successfulentry.
Stoicism about Reward
In a market structured by mutual benefit, each individual benefits according
to the value that other people place on their transactions with that individual. In
terms of any defensible concept of what people deserve, this form of economic
organization cannot consistently reward people according to their deserts. Desert
is a backward-looking concept: what people deserve can depend on how they
behaved in the past. But mutual benefit, in the sense that markets can be said
to facilitate its achievement, is defined in terms of peoples circumstances and
beliefs at the time at which they trade.. Because economic circumstances can change
unpredictably, efforts that were made with reasonable expectations of return may
turn out not to be rewarded by the market. Conversely, being in a position to gain
from mutually beneficial transactions with others at a particular time and place
can involve luck as well as foresight. Sandels (2009) example of being able to
benefit from possessing the human and physical capital of a hotelier or builder
in the aftermath of a hurricane is just an extreme case of this general feature of
market reward. If Sandels interpretation of the pay of senior corporate executives
in the pre-2008 period is that that those executives were benefiting from the good
luck of being able to exercise their trade in a bull market, that example illustrates
the same point.
That international trade promotes peace by making nations dependent on one another was argued
even earlier, by Montesquieu (1748 [1914], Book 20, Section 2). However, Hume is more explicit in
arguing that trade gives each country an interest in the prosperity of its trading partners.
160
To recognize this feature of markets is not to oppose all redistributive policies. Indeed, one might argue that a market economy is politically sustainable only
if everyone can expect to benefit in the long run from the wealth that markets
create, and that might require some collective commitment to redistribution. But if
the market is to function, rewards cannot be perfectly aligned with desert (Sugden
2004, 2012). To some critics, this disconnect between reward and desert comprises
a moral failure of the market. Sandel (2009) refers to a passage in which Milton
and Rose Friedman (1980, pp.136 137) argue that this aspect of the unfairness of
life is a price we have to pay for the freedom and opportunity that the market gives
us. Sandel (pp. 164 165) thinks this a surprising concession from advocates of
the market. His thought seems to be that material wealth is the currency of market
reward, and that individuals earnings from the market ought therefore to be in due
proportion to effort and talent.
Of course it is true that most people value material wealth, and that, in
this morally neutral sense, wealth is a currency of reward in the market, as it
is in other domains of life. But an adequate account of market virtue cannot
maintain that what a person earns from market transactions is a reward for the
exercise of virtue,, in the sense that a literary prize can be seen as a reward for
artistic excellence. Aperson can expect to benefit from market transactions only
to the extent that she provides benefits that trading partners value at the time
they choose to pay for them. To expect more is to create barriers to the achievement of mutual benefit. Thus, market virtue is associated with not expecting to
be rewarded according to ones deserts, not resenting other peoples undeserved
rewards, and (if one has been fortunate) recognizing that ones own rewards
may not have beendeserved.
This attitude of fortitude or stoicism towards the distribution of rewards in
a market economy is fundamental to Hayeks (1976) account of the moral status
of the market and the mirage of social justice. Hayek accepts that the market
often fails to reward desert, but writes: It is precisely because in the cosmos of
the market we all constantly receive benefits which we have not deserved in any
moral sense that we are under an obligation also to accept equally undeserved
diminutions of our incomes. Our only moral title to what the market gives us we
have earned by submitting to those rules which make the formation of the market
order possible(p. 94).
Conclusion
We have presented a view of the market as a domain of human life with a
distinctive constellation of virtues. We have argued that this view of the market is
compatible with, and to some extent implicit in, a long tradition of liberal economic
thought. The virtues we have discovered do not, as some moral critics of the market
might have expected, merely normalize egoism and instrumentality: they are
genuine virtues that can be upheld with authenticity.
161
We stress again that virtues are defined relative to practices. The traits that
make a person good as a participant in markets need not be evaluated positively in
all domains of human life. To acknowledge that there are market virtues is not to
claim that the market is the only morally relevant domain, nor that the market
virtues are the only virtues. We have argued (in agreement with some but not all
virtue ethicists) that the virtues of different domains can conflict with one another.
Thus, the market virtue of universality can conflict with loyalty to community and
tradition. Respect for ones trading partners tastes can conflict with upholding
standards of professional and craft excellence. The virtue of self-help, as viewed by
a potential philanthropist, can conflict with benevolence. Stoicism about market
reward can conflict with the pursuit of social justice. However, it should not be
thought that the market virtues apply only within the practice of the market. On
our account, the telos of the market is mutual benefit. Thus, market virtues will
apply in other domains of human life that are understood as cooperation among
equals for mutual benefit and that, as Mill (1861 [1976], pp. 29 30) argues,
thereby provide the environment in which the social feelings of mankind can
develop. As Mill and many later theorists of social capital recognize, market relations form one part of the network of cooperative relations of which civil society
is made up (for example, Putnam 1993). Thus, the market virtues are also virtues
of civil society in general.
We close with an expression of this idea by Antonio Genovesi (1765 67 [2005]),
an Italian contemporary of Adam Smith who, like Smith, tried to understand the
motivations driving the growth of commercial societies in his time and who made
an attempt to build a theory of commercial society based on the idea of mutual
assistance (Bruni and Sugden 2000). Significantly, the name that Genovesi tried to
give our discipline was not political economy but civil economy. We quote the final
words of his Lectures on Commerce, or on Civil Economy (Genovesi, 1765 67 [2005],
our translation), delivered at the University of Naples, where he was the worlds
first professor of economics. Having taught his students how a commercial society
works, he concludes: Here is the idea of the present work. If we fix our eyes at such
beautiful and useful truths, we will study [civil economy] . . . to go along with the
law of the moderator of the world, which commands us to do our best to be useful
to oneanother.
We are grateful for comments from participants at various conferences and workshops
at which earlier versions of this paper were presented, and from the editorial team at the
Journal of Economic Perspectives. Sugdens work was supported by the Economic and
Social Research Council through the Network for Integrated Behavioural Science (grant
referenceES/K002201/1).
162
References
Akerlof, George. 1982. Labor Contracts
as Partial Gift Exchange. Quarterly Journal of
Economics 97(4): 543 69.
Anderson, Elizabeth. 1993. Value in Ethics and
Economics. Cambridge, MA: Harvard University
Press.
Aristotle. c350 BCE [1980]. The Nicomachean
Ethics of Aristotle, translated by William David Ross.
Oxford University Press.
Arrow, Kenneth J. 1972. Gifts and Exchanges.
Philosophy and Public Affairs 1(4): 343 62.
Arrow, Kenneth J. 1984. Collected Papers of
Kenneth J. Arrow, Vol. 1: Social Choice and Justice.
Oxford: Basil Blackwell.
Besley, Timothy. 2013. Whats the Good of the
Market? An Essay on Michael Sandels What Money
Cant Buy. Journal of Economic Literature 51(2):
478 95.
Brennan, Geoffrey. 1996. Selection and the
Currency of Reward. In The Theory of Institutional
Design, edited by Robert E. Goodin, 256 75.
Cambridge University Press.
Bruni, Luigino, and Robert Sugden. 2000.
Moral Canals: Trust and Social Capital in the
Work of Hume, Smith and Genovesi. Economics
and Philosophy 16(1): 21 45.
Bruni, Luigino, and Robert Sugden. 2008.
Fraternity: Why the Market Need Not Be a Morally
Free Zone. Economics and Philosophy 24(1): 35 64.
Buchanan, James M., and Gordon Tullock.
1962. The Calculus of Consent. Ann Arbor: University of Michigan Press.
Coyle, Diane. 2012. Should You Pay to Jump
the QueueOr for a New Kidney? Its Hard to
Define Where Cash Has No Place. A Review of
What Money Cant Buy: The Moral Limits of Markets
by Michael Sandel. The Independent, May 5. http://
www.independent.co.uk/arts-entertainment
/books/reviews/what-money-cant-buy-the-moral
-limits-of-markets-by-michael-sandel-7711785.html.
Deci, Edward L. 1971. Effects of Externally
Mediated Rewards on Intrinsic Motivation. Journal
of Personality and Social Psychology 18(1): 105 115.
Deci, Edward L., and Richard M. Ryan. 1985.
Intrinsic Motivation and Self-Determination in Human
Behavior. New York: Plenum.
Edgeworth, Francis Ysidro. 1881. Mathematical
Psychics. London: Kegan Paul.
Folbre, Nancy, and Julie A. Nelson. 2000. For
Love or MoneyOr Both? Journal of Economic
Perspectives 14(4): 123 40.
Frey, Bruno S. 1994. How Intrinsic Motivation
Is Crowded In and Out. Rationality and Society
6(3): 334 52.
163
164
oday, per capita income differences around the globe are large, varying
by as much as a factor of 35 across countries (Hall and Jones 1999). These
differentials mostly reflect the Great Divergence (a term coined by
Huntington 1996)the fact that Western Europe and former European colonies
grew rapidly after 1800, while other countries grew much later or stagnated. What is
less well-known is that a First Divergence preceded the Great Divergence: Western
Europe surged ahead of the rest of the world long before technological growth
became rapid. Europe in 1500 already had incomes twice as high on a per capita
basis as Africa, and one-third greater than most of Asia (Maddison 2007). In this
essay, we explain how Europes tumultuous politics and deadly penchant for warfare
translated into a sustained advantage in per capita incomes.
Much of the European advantage in per capita incomes emerged after the Black
Death of 1350, which killed between one-third and one-half of the European population. In the three centuries after 1400, European per capita incomes grew rapidly,
while Africa and Asia stagnated (Maddison 2001). By 1700, Western Europeans
produced 2.5times more than Africans on a per capita basis, and 7085percent
more than Indians, Chinese, and Japanese.
http://dx.doi.org/10.1257/jep.27.4.165
doi=10.1257/jep.27.4.165
166
Table 1
The First Divergence Europe versus China
Urbanization rate
(percentage of population living in cities
with more than 10,000 inhabitants)
Year
China
762
1000
1120
1500
1650
1820
3%
3.1%
3.8%
4%
3.8%
Europe
Year
China
Europe
0%
1
960
1300
$450
$450
$600
$550
$422
$576
1700
1820
$600
$600
$924
$1,090
5.6%
8.3%
10%
Within Europe, there was also divergence: UK incomes were 75percent higher
in 1700 than they had been in 1500; Dutch incomes increased by 180percent. At the
opposite end of the spectrum were the laggards: Italy probably showed essentially
no increases in productive capacity over these two centuries; Spain grew by only
28percent. Urbanization rates tell a similar story. Where data on per capita income
is poor, urbanization rates make a good substitute (Wrigley 1985; Nunn and Qian
2011). This is because urbanization will reflect both the productivity of the urban
sector (creating goods that can be traded for food) and of agriculture (which needs
to generate a surplus above subsistence to feed cities).
Table1 shows comparative figures for Europe and China, for both urbanization rates and GDP per capita. (The urbanization rate is defined as the percentage
of the population living in cities with more than 10,000inhabitants.) Europe may
have been slightly ahead of China in terms of per capita incomes in Roman times;
by the High Middle Ages, it had declined both absolutely and in relative terms,
before showing rapid increases. The urbanization rate in China was already around
3 percent in the eighth century; Europe, in contrast, probably lagged behind
substantially.1 By 1500, European urbanization rates were already higher than in
China; by 1650, they were twice those in the Far East.2
1
Maddison (2007) estimates that the share of Europeans living in cities of more than 10,000inhabitants
was zero. Bosker, Buringh, and van Zanden (2008) present alternative figures, showing higher urban
shares in the Iberian peninsula, which was under Arab rule at the time.
2
The evidence in favor of Chinese underperformance has been questioned. Pomeranz (2001) points
out that comparing the most advanced countries of Europe such as England and the Netherlands with
all of China is unfair. The Yangtze area, Chinas leading agricultural producer, did much better than the
rest of the country. However, a decade of detailed research has now firmly established that early modern
European incomes were indeed much higher than Chinese ones. Broadberry and Gupta (2012) estimate
that Chinese and Indian wages already lagged European wages as early as 1550; by 1800, the gap was
huge. Allen, Bassino, Ma, Moll-Murata, and van Zanden (2011) similarly show that real urban wages in
China were much lower than in Europe. Allen (2009a) shows that even in the Yangtze area, per capita
incomes were on a downward path during the early modern period.
167
Figure 1
Scatterplot of per Capita Incomes in 1500, 1700, 1820, and 1998
25,000
20,000
15,000
10,000
5,000
25,000
20,000
15,000
10,000
5,000
25,000
20,000
15,000
10,000
5,000
0
500 1,000 1,500 2,000
2,000
2,000
1,500
1,500
1,000
1,000
500
400 600 800 1,000 1,200
500
500 1,000 1,500 2,000
Per capita income in 1700
2,000
1,500
1,000
500
400 600 800 1,000 1,200
Per capita income in 1500
Note: Data are from Maddison (2001), and the countries with data availability for 1500 1820 include
the European countries as well as Brazil, Mexico, China, India, Indonesia, Japan, Philippines, Iran, Iraq,
Turkey, Egypt, and Morocco.
The First Divergence matters not only for incomes at that time: The countries
that surged ahead also conquered vast parts of the globe in the 19th century, and
remain amongst the first rank of economic nations today. Countries that failed to
grow in the early modern period remained poor for centuries; only some caught up
morerecently.
For example, the same countries that surged ahead after 1500 were also the first
to undergo an Industrial Revolution (Comin, Easterly, and Gong 2010). Figure 1
illustrates the persistence of per capita income over the long term, plotting levels in
1500, 1700, 1820, and 1998 against each other. The correlation coefficient ranges
from 0.46 to 0.8, and is highly significant in every pairwise comparison. A nave
regression of income levels in 1998 on per capita income in 1500 can explain more
than 20 percent of total variance; incomes in 1820 predict 64 percent of crosssectional differences. If the relationship is already strong when looking at countries,
it is even stronger when adjusted for ancestral population movements (Putterman
and Weil 2010). One of the best predictors of an individuals income today is the
level of riches attained by that persons ancestors hundreds of years ago.
168
In this paper, we argue that Europes rise to riches during the First Divergence
was driven by the nature of its politics after 1350it was a highly fragmented continent characterized by constant warfare and major religious strife. Our explanation
emphasizes two crucial and inescapable consequences of political rivalry: war
and death. No other continent in recorded history fought so frequently, for such
long periods, killing such a high proportion of its population. When it comes to
destroying human life, the atomic bomb and machine guns may be highly efficient,
but nothing rivaled the impact of early modern Europes armies spreading hunger
anddisease.
In a Malthusian world, the amount of land per person was the prime determinant of per capita output. Wars were so common, and their impact was so severe,
that they raised average death rates in early modern Europe significantly.3 In turn,
this spelled higher land-labor ratios in agricultural production and thus higher per
capita income (Voigtlnder and Voth 2009, 2013). War therefore helped Europes
precocious rise to riches because the survivors had more land per head available
for cultivation. We argue that the feedback loop from higher incomes to more war
and higher land-labor ratios was set in motion by the Black Death in the middle
of the 14th century. As surplus incomes over and above subsistence increased, tax
revenues surged. These in turn financed near-constant wars on an unprecedented
scale. Wars raised mortality not primarily because of fighting itself; instead, armies
crossing the continent spread deadly diseases such as the plague, typhus, or small
pox. The massive, continued destruction of human life that followed led to reduced
population pressure. In our view, it was a prime determinant of Europes unusually
high per capita incomes before the Industrial Revolution.
A rapidly growing literature on persistence in economic performance has
sought explanations for the long arm of historythe puzzling extent to which past
economic performance continues to predict present economic outcomes. Focusing
on the British case, in Voigtlnder and Voth (2006), we model productivity advance
as an externality from capital use, and show how higher premodern incomes can
improve the chances of industrializing. Comin, Easterly, and Gong (2010) argue
that technological leadership is bequeathed from generation to generation, while
Spolaore and Wacziarg (2009) conclude that genetic distance to the technological
leadera proxy for how long ago twopopulations shared a common ancestoris a
key predictor of per capita incomes.
We begin by describing the economic logic of the Malthusian world,
explain why existing interpretations struggle to make sense of Europes early
and sustained lead in per capita income, and introduce the evidence for our
own interpretation in more detail. We also compare our results for Europe with
the Chinese case before discussing why alternative interpretations of the First
Divergence are ultimatelyunconvincing.
In England, for example, average life expectancy fell from 40years in 1580 to around 32years in 1700.
169
170
Figure 2
Unique Equilibrium in the Malthusian Model
Birth rate
b(w)
Death rate
d(w)
Wage (w)
Population (N )
w*
N *
w (N )
N*
w(N )
Technology
schedule
Wage (w)
Notes: In the Malthusian model, birth rates increase with wages, while death rates decline (upper panel).
Wages, in turn, depend negatively on population (lower panel) due to decreasing returns to labor in an
economy with fixed land supply. The intersection of birth and death rates yields zero population growth,
and thus a stable population N *. If mortality shocks move wages beyond w *, population grows. Rising
population exerts downward pressure on wages, and the economy returns to point C. If technology
improves, the w(N ) schedule shifts out, so that a higher population can be sustained at any given wage.
However, technological change does not affect the steady state wage w *.
171
High and Stable Incomes after the Black Death: The Effects of War
At its worst, early modern war from about 1400 to 1700 was more deadly
than World WarII in the most affected areas. During 1941 45, for example, the
Soviet Union lost an estimated 24million citizens, both combatants and civilians,
out of a population of 168million. This amounts to a loss of nearly 15percent.
German losses were somewhat smaller in proportion to the size of the population,
while Polish ones were greater17percent of Polands population died during
wartime after 1939. By comparison, the United States and the United Kingdom
172
during World War II registered losses of less than 1 percent. In contrast, the
twogreatest periods of conflict in the early modern periodthe Religious Wars in
late 16th-century France and the Thirty Years War in Germanyclaimed approximately 20 and 33percent of the population, respectively.4 While these estimates
have large margins of error, war in the age of the musket could clearly be more
devastating than in the age of tanks and aerial bombardment. How do we explain
this puzzlingfact?
The deadliness of war principally depends on two factorsthe lethal power of
weaponry, and the frequency with which noncombatants and soldiers succumb to
hunger and disease. The killing power of modern arms is many times greater than
it was in the past (Ferguson 2002, 2006), but death from hunger and disease has
become less frequent over time. Before the 19th century, the disease channel was
the most important driver of war-related mortality (Landers 2005). There are many
examples: When Europeans arrived in the Americas, even minor ailments like the
flu killed natives in large numbers. Major diseases like smallpox wiped out entire
populations (Diamond 1997). It has been estimated that European diseases caused
a collapse in Meso-American population size by 75percent or more.
Something similar, if milder, occurred when an army marched through isolated
villages in the European countryside. They brought the local population into
contact with new diseases, causing major spikes in mortality. Trade could also spread
diseases: the last plague outbreak in Western Europe, in Marseilles in 1720, was
caused by a ship from the Levant. But armies were more potent vectors of disease.
Typhus probably reached Europe via Spanish soldiers who contracted it on Cyprus;
syphilis might have been brought back from the Americas; and plague famously
spread throughout the Old World after a Mongol army infected the Genoese
defenders of a trading outpost on the Crimea in 1347 (McNeill 1976). The more
isolated populations were, the greater the mortality impact of a new disease. Exposure to new diseases was deadly for soldiers as well: during colonial wars in Africa,
annual death rates could reach one-fifth or more.5 As late as the early 19th century,
Russian troops occupying Swedish islands caused a major increase in death rates
without any fighting. Prussian troops contracted smallpox when campaigning in
France during the Franco-Prussian War, leading to an epidemic at home once they
returned (Landers 2003).
Total military and civilian deaths during World War II come from Clodfelter (2002) for the United
States, United Kingdom, Japan, and Poland; from Hubert (1998) for Germany; and from Ellman and
Maksudov (1994) for the USSR. Population estimates come from US Census Bureau (2000), Mitchell
(1988), Statistics Bureau of Japan (2011), Hubert (1998), Ellman and Maksudov (1994), and Piotrowski
(1997). For French Religious Wars, we use the death toll in Knecht (1996) and the population estimate
in Dupquier (1988), and we use Clodfelter (2002) for the German Thirty Years War.
5
During the two-month Logo expedition to what is now the country of Mali in 1874, the implied
annual death rate was 2,940/1,000, which means that the average soldier had a life expectancy of around
fourmonths (Curtin 1998).
173
Figure 3
Plague Outbreaks in Europe, 13501650
Total number of plague epidemics (per decade)
800
700
600
500
400
300
200
100
0
50
30
16
10
16
90
16
15
70
15
50
15
30
15
10
15
90
14
70
14
50
14
10
30
14
90
14
70
13
50
13
13
Decade
Source: Biraben (1975).
War did not just create temporary spikes in death rates; it raised average
death rates by up to one-third because it was so common.6 One way to understand
the effectiveness of war as a vector for disease is to look at the pattern of plague
outbreaks in Europe. Following the catastrophic outbreak of the plague in 1348, a
wave of epidemics followed. The detailed historical records demonstrate that many
of them were spread by marauding armies. Figure3 shows the number of epidemics
per year for the period 1350 1650. They increased gradually, from 150 per decade
in the second half of the 14th century to more than 400 on average during the first
half of the 17th century.
The disease channel was a particularly potent killer in Europe because of
geographical fragmentation. The continent is divided by large mountain ranges such
as the Alps and the Pyrenees, so the movement of armies brought populations into
contact with new germs. Political fragmentation also mattered: it went hand-in-hand
with frequent warfare. Since the fall of Rome, Europe has never been dominated
We explain the details of this calculation in Voigtlnder and Voth (2013), where we marry microevidence on changes in death rates in regions affected by war (traced from data on which areas saw
fighting or were traversed by armies) with estimates of the share of population exposed and the likely
mortalityimpact.
174
Table 2
Frequency of War
Century
16th
17th
18th
19th
20th
Number of wars
34
29
17
20
15
1.6
1.7
1.0
0.4
0.4
95%
94%
78%
40%
53%
175
conflict: at the height of early modern warfare, during the Thirty Years War, close to
half of the European population was affected by military conflict in a given year (for
a derivation of this estimate, see Voigtlnder and Voth2013).
Given these rising levels of war and disease, why were Europeans so much more
productive by 1700 than they had been in the Middle Agesand so much
more productive than the rest of the world? We argue that the particular type of
warfare that characterized Europe after the Middle Ages is an important part of the
answer. Before 1800, battlefield deaths and direct victims of armed force were few;
civilian and military deaths from disease were plentiful. The imbalance between
violent killing, and death from disease, also has important implications for the
economic impact of war. War in early modern Europe acted more like a neutron
bomb: because of disease, it destroyed human life quickly while not wreaking havoc
on infrastructure on a scale comparable to modern wars.7
Recovery could be quick in places where it was only wooden houses that
needed to be rebuilt. Since European agriculture did not rely on elaborate irrigation systems (as did the Middle East, for example), the direct effects of war were
limited to destroyed farm buildings, stores, and livestock. All of these typically could
recover or be rebuilt in short order. As Malthus (1798) observed: The fertile province of Flanders, which has been so often the seat of the most destructive wars, after
a respite of a few years, has appeared always as fruitful and as populous as ever.
Even the Palatinate lifted up its head again after the execrable ravages of Louis
the Fourteenth. Similarly, after the Turkish siege of Vienna in 1683, the speed
of recovery was astonishing. As one observer put it: the suburbs . . . as well as the
neighboring countryside ... have been completely rebuilt in a short space of time
(Tallett 1992). Land left fallow increased in fertility. Livestock had high rates of
reproduction, so herds could be rebuilt quickly.
War as practiced in this time therefore combined two characteristics that
mattered for economic outcomes: it was highly destructive of human life, and it was
largely ineffective in destroying infrastructure and capital stock. While the amount
of useful land and the size of the capital stock fell only a little as a direct result of
war, military conflict before 1800 was massively destructive of human life. In other
words, war was highly effective in increasing the ratio of land and capital relative
to the size of the population. In a Malthusian world, frequent war could act as a
powerful force raising per capita incomes for the survivors.
A key exception is the destruction of cities. When early modern cities were taken after a siege, they
sustained damage that is comparable or worse than that of aerial bombardment during World WarII.
For example, when Magdeburg was taken by Imperial forces in 1631, the entire city was burned to the
ground, more than 90percent of homes were destroyed, and an estimated 20,000 (out of 35,000) inhabitants lost their lives. The bombing of Dresden caused 20,000 25,000 casualties, out of a population of
650,000 (Cunningham 2000; Neutzner 2010). While the destruction of cities was not minor, there are few
countries where the urban share of total population exceeded 10percent before 1700.
176
Table 3
Tax Revenues in Europe
Year
1509
1559
1609
1659
1709
1759
1789
214
456
1,116
2,215
2,667
3,808
6,846
3.7
3.6
4.9
5.7
8.1
9.9
12.2
Source and Notes: Data are from Karaman and Pamuk (2010), who use countrylevel historical compilations of revenue statistics. The main database is the
European State Finance Database (ESFD), available at http://www.esfdb
.org. They use silver as the measure of fiscal revenue, because all national
currencies were convertible into it. The original source for the urban wage
series is Allen (2001).
177
render unto his prince the equivalent of less than four days pay; by 1789, on the eve
of the French Revolution, this had tripled to more than 12days wages.
Taxes after 1500 rose much faster than population size. One factor that
supported growing tax revenue was growth in per capita incomesthe very fact
at the root of the First Divergence. ORourke and Williamson (2002) estimate that
European surplus incomes that is, the amount above subsistenceon average
grew by 0.4 0.5 percent per year from a very low base, much faster than incomes
themselves. As incomes grew above subsistence levels, they could increasingly be
taxed by the belligerent princes of early modern Europe. Out of every unit increase
in surplus (above-subsistence) incomes, European states successfully appropriated about one-third (Voigtlnder and Voth 2013).
The vast majority of early modern tax revenues were spent on war. European
states at this time routinely spent 7080 percent of their income on armies and
navies. In wartime, spending exceeded revenue by a large margin; repaying the past
debts thus accounted for a good share of the remaining expenditure.
At this time, financial strength mattered a great deal for military success.
Warfare during the early middle ages had been a relatively cheap affair; armies
were small, and usually consisted mainly of vassals who were obliged to follow
their prince into battle (Landers 2003). In the early modern period, a military
revolution transformed the face of battle. Armies used firearms, which required
extensive training; standing armies became the norm. Huge new fortifications
were necessary to protect cities because medieval city walls had been rendered
ineffective by the invention of the cannon. These costs quickly bankrupted
fiscally weaker states, and financial prowess became a prime determinant of military success. As a Spanish 16th-century soldier and diplomat, Don Bernardino
de Mendoza, eloquently put it (as quoted in Parker 1977): [V]ictory will go to
whoever possesses the lastescudo.
War and Rising Riches
Did the continuous and near-universal warfare on the European continent lead
to higher incomes? We combine data on the incidence of early modern warfare
at the country level with two indicators for economic development: urbanization
and per capita GDP. The size of urban centers is from Bairoch, Batou, and Chvre
(1988), which we combine with population estimates from McEvedy and Jones
(1978) to obtain the percentage living in urban areas. As a result, we have countrylevel urbanization rates covering the early modern period between 1300 and 1700.
As a consistency check, we also use urbanization data from DeVries (1984), which
are available from 1500 onwards. In addition, we use per capita income data from
Maddison (2001), which is available in 100-year intervals from 1500 onwards.
To measure the extent of warfare, we employ data on the years of warfare from
Kohn(1999).
We analyze a cross-section of states in early modern Europe. To measure
economic progress we use the change in urbanization between 1300 and 1700 and
the change in per capita income between 1500 and 1700. Our explanatory variable
178
is war frequency over the time interval corresponding to each of the twooutcome
variables. Figure4, panelA, shows that a higher war frequency between 1300 and
1700 is associated with a larger increase in urbanization. Countries with aboveaverage frequencies of armed conflict, such as the Netherlands, France, and
England, gained urban population at a quick rate; those fighting fewer wars, such as
Ireland, Switzerland, and Norway, saw only limited progress. The correlation coefficient is 0.40. The same pattern is visible for per capita income (panelB), with a
correlation coefficient of 0.28. Here, we use data starting in 1500, as calculated by
Maddison(2001).
We look at the relationship between warfare and income growth in a variety
of ways. First, we split the countries into twogroups, those with below-average and
above-average war frequency. We find that both measures of urbanization as well as
per capita GDP grew significantly faster in countries with an above-average number
of wars. For example, urbanization rates grew by 7.4percentage points between 1300
and 1700 in countries with above-mean war frequency, versus 2.8 in the remainder.
Per capita income grew almost twice as fast over 1500 1700 in countries with aboveaverage warfare. These differences are statistically significant at the 5percent level.
In simple bivariate regressions, we also find a large and statistically significant
relationship between the number of wars and increases in urbanization. The baseline specification implies that in a country with one war per year on average between
1300 and 1700, urbanization rates rose faster, by 7.6 percentage points, over the
same period, as compared to a country without warfare. Two close examples for
these numbers are England, with 1.012wars per year between 1300 and 1700 and
a relatively high degree of urbanization, and Romania, with zero wars.8 The former
saw an increase in the urbanization rate by almost 13 percentage points, versus 2 in
thelatter.
179
Figure 4
Warfare and European Development
A: Wars and Urbanization, 13001700
.3
NL
.2
BE
EN
DK
.1
SER
RO AL
0
BG
HU
SWE
CH
NO
CZ
IRE
IT AT ES
PT
FR
PL
DE
RU
.1
.2
.4
.6
.8
Average number of wars per year 13001700
.8
EN
.6
.4
IRE
RO
BG
AL
SER
.2
CH
DK
NO CZ
SWE
BE
HU
AT
FR PT
ES
DE
PL
RU
IT
.5
Average number of wars per year 15001700
Sources: Change in urbanization from Bairoch, Batou, and Chvre (1988), in combination with
population estimates from McEvedy and Jones (1978), as explained in Voigtlnder and Voth (2013);
changes in per capita income from Maddison (2001). Average wars per year are derived from the dataset
used in Acemoglu, Johnson, and Robinson (2005), who construct years of warfare by time period based
on Kohn (1999).
180
That China had fallen behind significantly in per capita income by the early
19thcentury is not in doubt. Urbanization rates were low; incomes were a small fraction of their European equivalents. The country was also politically and militarily
weak, and was about to be humiliated at the hands of European powers.
Our interpretation attributes a good part of Chinas relative decline to its low
levels of military conflict. After the Yuan Dynasty (12711368) was deposed in a
series of revolts, comparative peace reigned. Under the Ming and Qing Dynasties
(1368 1644; 1644 1912), the country remained politically unified for a half a
millennium. Frequency of military conflict was dramatically lower in China: Europe
saw 443 wars during the period 1500 1800 (a frequency of 1.48 wars per year),
involving 1,071 major battles. The corresponding figures for China are 91 wars
between 1350 and 1800 (a frequency of 0.2 per year) and only 23major battles
most conflicts were peasant revolts (Tilly 1992; Jaques 2007). In other words, the
frequency of war per year was 85percent lower, and the number of major battles was
98percent lower, than in Europe.
Not only was war less frequent in China, it also caused fewer deaths. The
majority of the population tilled the soil within a few hundred kilometers of the
eastern seaboard. There were few natural obstacles to population movements and
trade. The epidemiological evidence, where it exists, suggests that disease pools
were largely integrated by the year 1000 (McNeill 1976). As a result, diseases spread
by troops did not have the same devastating impact in China as they did in Europe.
In this setting, China experienced considerable population pressure.
During the early modern period, Chinese population increased by an estimated
280percent; the corresponding figure for Europe is 140percent (Maddison 2007).
Europeans visiting China noticed the abundance and the cheapness of labor. As
Malthus (1798) observed: The country [China] is rather over peopled . . . and
labour is, therefore, so abundant, that no pains are taken to abridge it.
The view that China fell behind economically, and that its demography is partly to
blame, is controversial. In One Quarter of Humanity,, Lee and Wang (2001) challenged
the earlier consensus that Chinese fertility rates were much higher than European
ones. While marriage was universal, they argued, within-marriage fertility rates were
relatively low. The current consensus view is that there is some merit in the argument
but that total fertility rates in Europe were probably still markedly lower (especially
taking into account how much higher incomes were).9
If population pressure in China was much higher than in Europe but higher
fertility rates were only part of the answer, then lower mortality must be part of the
story. Notice that China being poorer should actually have produced relatively high
mortality rates: after all, many health risk factors and diseases before 1900 were
nutrition sensitive, and lower incomes probably resulted in higher death rates from
tuberculosis and the like. Thus, the absence of major war-induced mortality is a
9
The relevant variable in a Malthusian setting is the income-adjusted fertility rate, accounting for the
upward-sloping birth schedule in the Malthusian model. Total fertility rates measure the expected
number of children a woman would have over the course of her life.
181
plausible explanation for why Ming and Qing China experienced such a substantial
populationboom.
Alternative Interpretations
We are not the first to examine Europes relative rise to riches after 1500. Alternative interpretations have emphasized the role of technological innovations, of
institutional improvements, and of fertility restriction. Theoretically, it is possible
that a positive income shock driven by all or one of these factors gave rulers the
means to fight moreexplaining the positive correlation between war and income
growth without any causal connection. Such an alternative interpretation is unlikely,
for severalreasons.
Acemoglu, Johnson, and Robinson (2005) show that European outperformance
was largely driven by states bordering the Atlantic in combination with institutions
that fostered commerce. In England and the Dutch Republic, trading opportunities
strengthened the bourgeoisie, which in turn succeeded in constraining the powers
of rulers. On the Iberian Peninsula, in contrast, the discovery of the Americas gave
extra resources to powerful monarchs; as a result, institutional quality declined.
The implication of this argument is that North-Western Europe owed its precocious lead over the rest of the world to institutional improvements, most of which
occurred along the Western seaboard of the continent. Their interpretation is part
of a broader approach to Europeanand in particular, Britishoutperformance.
Another prominent interpretation emphasizes Europes growing ability to
innovate (Mokyr 1992) and contrasts it with technological decline elsewhere. While
medieval Europe had even forgotten some of the useful inventions of Romesuch
as concretetechnological creativity flourished after 1500. From the invention
of the printing press with movable letters and the barometer to vastly improved
sailing ships, steel ploughs, and hot air balloons, Europe excelled at producing new
and useful goods in the early modern period. In contrast, the famous fourgreat
inventions of Chinacompass, gunpowder, printing, and papermakingmarked
an even earlier period of technological advance that found no echo in the early
modern period. The underlying reasons for Europes technological advance at
this time are still a subject of research, but it seems plausible that the shortage
of labor helped to encourage a search for labor-saving devices and that ongoing
military conflict created pressure for innovations and a conduit for spreading ideas
(Allen2009b).
The principal problem with both the technological and the institutional interpretation is that they are not well-suited to explaining income divergence in a world
dominated by demographic forces. Ashraf and Galor (2011) demonstrate that there
are no significant gains in per capita incomes from productivity improvements
during the Malthusian era. The reason is that human populations typically grow
rapidly when faced with abundance. For productivity improvements to push up per
capita living standards, they would have to be faster than the rate of population
182
growth.10 In terms of orders of magnitude, this was highly improbable in the period
before 1950. As noted earlier, human populations can easily grow at more than
3percent per year, while technological change was probably less than 0.1percent
per year on average.
This leaves demographic interpretations. In any Malthusian system, incomes
are ultimately determined by mortality and fertility rates. Europes level of mortality
was uniquely high, and war was an important component of it. Other factors also
contributed to the fact that, at least for some part of this time period, mortality rates
could rise at the same time as incomes grew. As incomes rose, Europeans crowded
into more and larger urban centers. Cities in Europe before 1850 were veritable
death traps, with mortality rates much higher than fertility rates. Poor sanitation
and urban overcrowding were to blame. Therefore, not only did the development
of cities reflect rising per capita incomes; these cities also helped to sustain incomes,
much in the same way as war did, by keeping land-labor ratios high.
While the mechanism of disease-ridden cities adding to mortality is the same
as for war, it is quantitatively much smaller. Even as late as 1800, only 10percent of
Europeans lived in cities with more than 10,000inhabitants (DeVries 1984). Even
if these city-dwellers suffered markedly higher mortality, they could not influence
aggregate death rates by much.11
A similar argument applies to the effect of trade. Trade typically increases
with incomes, and it can act as a potent vector for diseases. While little is known
quantitatively about the volume of trade before 1800, it is reasonable to assume
that increasing contact between distant population centers led to the exchange
of germs and higher mortality. In related work, we estimate the size of the plausible effects, and find that they are quantitatively small, adding no more than
0.25percentage points to annual European death rates of approximately 3percent
(Voigtlnder and Voth2013).
Another factor that could have helped hold down Europes population growth
and thus improve its performance in a Malthusian world is fertility restriction.
Europe evolved a unique form of fertility limitation. Europeans of this time typically
married latein their mid-20s, not much earlier than they do today. A significant
share of women also remained unmarried (Hajnal 1965). The reasons for this
phenomenon (which only existed west of a line from St. Petersburg in Russia to
Trieste in Italy) are complex. Most interpretations emphasize economic factors,
such as the culturally determined need for newlyweds to set up a new household
(neo-locality), the access of women to urban labor markets, inheritance rules,
and the increasing use of females in pastoral agriculture (De Moor and van Zanden
2010; Voigtlnder and Voth forthcoming).
10
In Voigtlnder and Voth (2013), we show that in a calibrated Malthusian model, even a sudden jump
from technological stagnation to ongoing technological growth at the early modern rate of 0.1percent
per year would not have had a substantial impact on per capita income or urbanization.
11
There are some exceptions: In the Netherlands, for example, the urbanization rate was so high that
urban mortality on its own may have increased overall death rates by 0.5percentage points, relative to a
baseline of 3percent (Voigtlnder and Voth 2013).
183
Conclusion
The First DivergenceEurope pulling ahead long before the Industrial
Revolutionhas long posed a puzzle for growth theorists and economic historians.
In a world with strong Malthusian forces, incomes should not have had much scope
to rise and then stay elevated over long periods. And yet, this is what happened in
early modern Europe.
In this paper, we argue that a good part of Europes precocious rise to riches
reflected gifts from Marspermanently high per capita incomes for the survivors
were an indirect consequence of near-constant, and deadly, warfare. We first show
that despite small army size and relatively primitive weapons technology, war in
the centuries before 1800 was a potent destroyer of human lives. The main cause
of death was not armed force itself, but the spreading of disease: A single army of
6,000 8,000men, dispatched from La Rochelle by Cardinal Richelieu to fight in the
Mantuan war in 1628 may have killed up to a million people by spreading the plague
on its march from France to northern Italy (Landers 2003; Biraben 1975).
In a Malthusian world, population losses generally created higher incomes
for the survivorsthere was more land available per capita. These effects should
have been transitory: as population recovered, gains in per capita output ought
to have been reversed. After the Black Death hit in 1348 50, population losses were
massive and so were gains in per capita income, but one would expect these gains
to fade over time. However, with income gains much greater than what could be
eroded by population growth in a generation or two, rulers found ample surplus
income (over and above subsistence) to tax. As they appropriated this surplus to
a growing extent after 1350, war frequency surged. Rulers effectively treated war
as a luxury good, and as money became available, fought ever more of them. The
high frequency of war in turn made it easier to sustain the gains in living standards
for those whosurvived.
The war channel for greater riches was particularly potent in Europe because
of political fragmentation. Plagues also hit Justinian Rome, China, and the Middle
East (McNeill 1976), without similar consequences. The Black Death in 1348 50
only acted as a catalyst for a simultaneous rise in the frequency of warfare and of
per capita incomes because there were so many European states and statelets that
could fight each other. And fight they did: war became a near-constant feature of
early modern Europe.
In Dan Browns (2013) bestseller, Inferno, the chief villain is a geneticist about
to unleash a diabolical virus. He points to the experience of the Black Death to
suggest that population losses can be beneficial and that good economic times
will follow. Our research suggests that this is misguidedsometimes, history
offers little guidance for policy implications today. Most parts of the world have
clearly escaped from Malthusian constraints; land-labor ratios no longer determine per capita income except in the poorest countries. Instead, human capital,
institutions, and technology are key. This also means that the synergistic link
between war, population losses, and higher incomes that we described is unlikely
184
References
Acemoglu, Daron, Simon Johnson, and James
Robinson. 2005. The Rise of Europe: Atlantic
Trade, Institutional Change, and Economic
Growth. American Economic Review 95(3): 546 79.
Allen, Robert C. 2009a. Agricultural Productivity and Rural Incomes in England and the
Yangtze Delta, c.1620c.1820. Economic History
Review 62(3): 525 50.
Allen, Robert C. 2009b. The British Industrial
Revolution in Global Perspective. Cambridge University Press.
Allen, Robert C., Jean-Pascal Bassino, Debin
Ma, Christine Moll-Murata, and Jan Luiten van
Zanden. 2011. Wages, Prices, and Living Standards in China, 17381925: In Comparison with
Europe, Japan, and India. Economic History Review
64(Issue Supplement S1): 8 38.
Anderson, Michael, and Ronald Lee. 2002.
Malthus in State Space: Macro EconomicDemographic Relations in English History, 1540
to 1870. Journal of Population Economics 15(2):
195 220.
Ashraf, Quamrul, and Oded Galor. 2011.
Dynamics and Stagnation in the Malthusian
Epoch. American Economic Review 101(5): 2003 41.
Bairoch, Paul, Jean Batou, and Pierre Chvre.
1988. La population des villes europennes, 800-1850:
Banque de donnes et analyse sommaire des rsultats
[The population of European cities, 8001850:
Data bank and short summary of results.] Geneva,
Switzerland: Droz.
Biraben, Jean Nol. 1975. Les hommes et la peste
en France et dans les pays Europens et Mditerranens.
Paris: Mouton.
Bosker, Maarten, Eltjo Buringh, and Jan Luiten
van Zanden. 2008. From Baghdad to London:
185
Population, Production, and Power in the Pre-Industrial West. Oxford University Press.
Landers, John. 2005. The Destructiveness of
Pre-Industrial Warfare: Political and Technological
Determinants. Journal of Peace Research 42(4):
455 70.
Landes, David S. 1999. The Wealth and Poverty of
Nations: Why Some Are So Rich and Some So Poor. New
York: W. W. Norton.
Lee, James Z., and Feng Wang. 2001. One
Quarter of Humanity: Malthusian Mythology and
Chinese Realities, 17002000. Cambridge, MA:
Harvard University Press.
Maddison, Angus. 2001. The World Economy,
Vol.1: A Millennial Perspective. Paris: OECD.
Maddison, Angus. 2007. Contours of the World
Economy 12030 AD: Essays in Macro-Economic
History. Oxford University Press.
Malthus, Thomas Robert. 1798. An Essay on the
Principle of Population. London.
McEvedy, Colin, and Richard Jones. 1978. Atlas
of World Population History. Puffin.
McNeill, William H. 1976. Plagues and Peoples.
New York: Knopf.
Mitchell, Brian R. 1988. British Historical Statistics. Cambridge University Press.
Mokyr, Joel. 1992. The Lever of Riches: Technological Creativity and Economic Progress. New York:
Oxford University Press.
Neutzner, Matthias. 2010. Abschlussbericht
der Historikerkommission zu den Luftangriffen auf
Dresden zwischen dem 13. und 15. Februar 1945. City
of Dresden.
Nicolini, Esteban A. 2007. Was Malthus Right?
A VAR Analysis of Economic and Demographic
Interactions in Pre-industrial England. European
Review of Economic History 11(1): 99 121.
Nunn, Nathan, and Nancy Qian. 2011. The
Potatos Contribution to Population and Urbanization: Evidence from a Historical Experiment.
Quarterly Journal of Economics 126(2): 593 650.
ORourke, Kevin H., and Jeffrey G. Williamson.
2002. After Columbus: Explaining Europes Overseas Trade Boom, 15001800. Journal of Economic
History 62(2): 417 56.
Parker, Geoffrey. 1977. The Dutch Revolt.
Penguin.
Piotrowski, Tadeusz. 1997. Polands Holocaust:
Ethnic Strife, Collaboration with Occupying Forces
and Genocide in the Second Republic, 19181947.
McFarland & Company.
Pomeranz, Kenneth. 2001. The Great Divergence:
China, Europe, and the Making of the Modern World
Economy. Revised. Princeton, NJ: Princeton University Press.
Putterman, Louis, and David N. Weil. 2010.
Post-1500 Population Flows and the Long-Run
186
Structural Change and Physical Capital Accumulation during the Industrial Revolution. Journal of
Economic Growth 11(4): 319 61.
Voigtlnder, Nico, and Hans-Joachim Voth.
2009. Malthusian Dynamism and the Rise of
Europe: Make War, Not Love. American Economic
Review 99(2): 248 54.
Voigtlnder, Nico, and Hans-Joachim Voth.
2013. The Three Horsemen of Riches: Plague,
War, and Urbanization in Early Modern Europe.
Review of Economic Studies 80(2): 774 811.
Voigtlnder, Nico, and Hans-Joachim Voth.
Forthcoming. How the West Invented Fertility
Restriction. American Economic Review.
Wells, H. G. (Herbert George). 1905. A Modern
Utopia. London: Scribners.
Wrigley, E. A. 1985. Urban Growth and Agricultural Change: England and the Continent in the
Early Modern Period. Journal of Interdisciplinary
History 15(4): 683 728.
Wrigley, E. A., and R. S. Schofield. 1989.
The Population History of England 15411871.
Cambridge University Press.
rban populations have skyrocketed globally and today represent more than
half of the worlds population. In some parts of the developing world, this
growth has more-than-proportionately involved rural migration to informal
settlements in and around cities, known more commonly as slums densely
populated urban areas characterized by poor-quality housing, a lack of adequate living
space and public services, and accommodating large numbers of informal residents
with generally insecure tenure.1 Worldwide, at least 860million people are now living
in slums, and the number of slum dwellers grew by sixmillion each year from 2000
to 2010 (UN-Habitat 2012a). In sub-Saharan Africa, slum populations are growing at
4.5percent per annum, a rate at which populations double every 15years.
The global expansion of urban slums poses questions for economic research,
as well as problems for policymakers. Some economists (Frankenhoff 1967; Turner
1969; World Bank 2009; Glaeser 2011) have suggested a modernization theory of
1
Perhaps not surprisingly, the identification of slum inhabitants suffers from the lack of a consistent
terminologyfor example, slums and squatter settlements are used almost interchangeably,
although tenure and ownership institutions vary greatly across informal settlements. UN-Habitat (2006)
applies the notion of slum household to any household lacking access to improved water, improved
sanitation, sufficient living area, durable housing, and secure tenure. Slum areas are generally thought
of as geographic areas accommodating informal residents that combine several of these characteristics.
doi=10.1257/jep.27.4.187
188
189
Table 1
Two Lists of the Developing Worlds Largest Slums
UN-Habitat (2003)
Name of slum
Dharavi
Orangi
Kibera
Villa el Salvador
Ashaiman
Davis (2006)
City,
Country
Population
estimate
Name of slum
Mumbai,
India
Karachi,
Pakistan
Nairobi,
Kenya
Lima,
Peru
Tema,
Ghana
Over 500,000
Neza/Chalco/Izta
Over 500,000
Liberatador
400,000
El Sur/Ciudad
Bolivar
San Juan de
Lurigancho
Cono Sur
300,000
150,000
Ajegunle
Sadr City
Soweto
Gaza
Orangi
City,
Country
Mexico City,
Mexico
Caracas,
Venezuela
Bogota,
Columbia
Lima,
Peru
Lima,
Peru
Lagos,
Nigera
Baghdad,
Iraq
Gautung,
South Africa
Palestine
Karachi,
Pakistan
Population
estimate
4 million
2.2 million
2.0 million
1.5 million
1.5 million
1.5 million
1.5 million
1.5 million
1.3 million
1.2 million
190
can occur. The left-hand column lists five of the worlds largest slums, with rough
population estimates, as provided by UN-Habitat (2003). An unofficial ranking
provided in Davis (2006) lists 30slums with over 500,000inhabitants, including ten
in sub-Saharan Africa and nine in Latin America. The tenslums with largest population by this measure are shown in the right-hand columns of Table1. Dharavi and
Kibera, the first and thirdslums in the UN-Habitat (2003) list both appear further
down in the Davis (2006) ranking with an estimated population of 800,000each.
Of course, the fact that a lot of the global slum expansion takes place in poor
economies does not invalidate a modernization or transitory view of slums. Since
rural migration continues unabated even in those countries, urban productivity
must be rising relative to rural productivity, either because capital accumulation
and technological progress are concentrated in cities or because rural productivity
is declining. The location decision of slum dwellers also indicates that standards of
living in slums are somewhat preferable to those in the rural hinterlands. Glaeser
(2011) provided evidence that the urban poor worldwide are on average richer and
happier than their rural counterparts, and Chowdhury, Mobarak, and Bryan (2009)
showed that seasonal urban migration in Bangladesh can generate welfare improvements for families ofmigrants.
However, these facts say little about the nature of poverty in slumsand
whether it can be escaped via the competitive market forces offered by the city.
There has been, in fact, a relative lack of empirical research work conducted in
slums of the developing world to test this argument.2 One reason is that data collection in slums is problematic due to a variety of factors, including safety issues for
research fieldworkers, the high mobility and turnover rates of respondents, and the
fact that target households are regularly absent from their dwellings. As a result, few
studies have tried to document the degree of intergenerational social mobility of
slum dwellers in the developing world. Here, we start by illustrating what life is like
in slums and what characteristics seem to be common across slums, using recent
surveys collected in slums of fourcountries: Bangladesh, India, Kenya, and Sierra
Leone. We rely most heavily on the case of the Kibera slum in Nairobi, Kenya, where
we have conducted extensive fieldwork over the past two years. Table 2 provides
specific sources of data for these slum areas.
A nonexhaustive list of recent empirical research conducted in slums includes Banerjee, Duflo,
Glennerster, and Kinnan (2010) in Hyderabad; Banerjee, Pande, Vaidya, Walton, and Weaver (2011)
in Delhi; El-Zanaty and Way (2004) in Cairo; Field (2007) in Peru; Galiani et al. (2013) in slums of
three Latin American countries; Gulyani, Bassett, and Talukdar (2012) in Dakar and Nairobi; and
Perlman (2010) in Rio de Janeiro. Gupta, Arnold, and Lhungdim (2009) analyze the NFSH-3 dataset
on which we alsorely.
191
Table 2
Description of Datasets
Country
Locations
Bangladesh
Tongi, Jessore
Bangladesh
Tongi, Jessore
India
India
Kenya
Kibera (Nairobi)
Kenya
Kibera (Nairobi)
Kenya
Kibera (Nairobi)
Sierra Leone
Source
Year
Number of slum
households
surveyed
2000
26,830
2000
1,120
2005
2006
8,669
2005
2,800
Banerjee, Duflo,
Glennerster, and Kinnan
(2010)
Kenya National Bureau of
Statistics (national census)
Marx, Stoker, and Suri
(2013)
Marx, Stoker, and Suri
(2013)
UNICEF Sierra Leone
SMART Survey
1999,
2009
2012
64,588
(approx.)
31,765
2012
1,093
2010
789
Sources: The Bangladesh data is available from http://dvn.iq.harvard.edu/dvn/ and was collected in 2000
as part of the SHAHAR census and baseline surveys. The India data was collected with approximately
8,669 slum households in eight Indian cities as part of the 2005 2006 National Family Health Survey
(NFSH-3). It is available from http://www.measuredhs.com/ (India Standard DHS, 2005 06). The
Hyderabad dataset (Banerjee, Duflo, Glennerster, and Kinnan 2010) was collected as part of a randomized
evaluation on the impact of microfinance. It is publicly available from the data repository of the Jameel
Poverty Action Lab ( J-PAL) at http://thedata.harvard.edu/dvn/dv/jpal. Our Kenya data come from
twodifferent sources. First, we collected onecensus and onehousehold survey in Kibera, Nairobis largest
slum area. Second, we obtained from the Kenya National Bureau of Statistics (KNBS) twowaves of complete
micro census data collected in the same area in 1999 and 2009. The Sierra Leone data was collected as part
of the 2010 UNICEF SMART Survey and is proprietary from UNICEF. We use these sources throughout
unless otherwise indicated.
There is a wide literature on poverty traps, including a theoretical literature highlighting the specific mechanisms leading to poverty traps (for excellent definitions
and reviews, useful starting points include Basu 2003, Matsuyama 2005, and Bowles,
Durlauf, and Hoff 2006). The literature has also described spatial poverty traps,
but mostly in rural settings ( Jalan and Ravallion 2002; Golgher 2012). We argue
that urban slums present a different challenge to communities and governments
administering them, and that the very nature of life in the slums makes it difficult
to achieve improvements in standards of living through marginal investments in
housing, health, or infrastructure alone. We now discuss some of the mechanisms
relevant to slum contexts that may lead to poverty traps.
192
Human Capital
Despite tremendous variations across slums, issues common to all slum settings
are a lack of adequate living space, insufficient public goods provision, and the poor
quality of basic amenities, all of which lead to extremely poor health and low levels
of human capital.3 In this section, we focus on the health aspects of human capital.
Education, albeit another important component of human capital, is a less relevant
metric for our argument given that universal free primary education laws have
reduced disparities in access to education between rural and urban settings,4 and that
ruralurban differences in the quality of education are extremely difficult tomeasure.
In the Kibera area of Kenya, informal households reported in 2009 an average
dwelling size of 1.17 habitable rooms (with average household size of 3.15), as
opposed to 1.95 for urban households and 2.97 for rural households. For perspective, according to UN-Habitat (2006), a dwelling provides sufficient living space
if each room is shared by no more than threeindividuals. In the Zimbabwe slum
in Abidjan, population density was reported to be as high as 34,000inhabitants per
square kilometer (UN-Habitat 2003). As a point of comparison, Manhattans population density was 26,924 per square kilometer in 2010 (US Census Bureau 2013).
Across slum settings, the adverse health effects of overcrowding are aggravated
by poor access to water and sanitation facilities. Table3 shows that the majority of
slum dwellers across our datasets have no private latrine, and many use inferior-type
latrines (such as an open space or traditional pit that is not connected to a sewage
network), no source of private water, and no garbage collection (meaning that
garbage is either left in a roadside ditch or burnt next to the household dwelling).
These data are corroborated by a range of studies documenting the poor water
access and overall hygiene of slum neighborhoods. For example, in Mumbais Shiva
Shakti Nagar slum, community taps are reportedly shared by 100people on average
(World Bank 2009). In a survey of slum dwellers in Delhi, Banerjee et al. (2011)
found that the environment of 83percent of toilet sites was infected with fecal or
other waste matter.
Absent or deficient water and sewage systems translate into a broad range of
health and sanitation issues, whether through direct exposure to bacterial agents,
contaminated drinking water, or other channels. Duflo, Galiani, and Mobarak (2012)
described the disease burden arising from the unsanitary living conditions in slums.
In the slums of Tongi and Jessore in Bangladesh, 82percent of respondents report
any household member being sick in the past 30days. In Kibera, 16percent of our
respondent households have at least one member chronically ill in the previous
threemonths. In Sierra Leone, a country whose slums routinely experience cholera
3
The World Bank (2009) argued that urban areas fare consistently better than rural areas worldwide
along a variety of health indicators. In this section, we compare rural areas with slum areas specifically.
4
Lopez (2007) shows evidence of this trend for Latin America. Hannum, Wang, and Adams (2008) study
the case of China, where some urbanrural disparities remain, but the primary enrollment of 716year
olds in rural areas is nearly universal. Worldwide, school attendance rates have increased in the majority
of low-income countries since 1990, and rural areas were the primary beneficiaries of this trend (World
Bank2009).
193
Table 3
Public Goods and Basic Amenities across Slums
Tongi (Dhaka)
Hyderabad
Kibera
Kolkata
Mumbai
All Indian slums NFSH-3
No private
latrine
Inferior
latrine type
No private
water source
No garbage
collection
70%
46%
NA
75%
78%
68%
34%
43%
63%
46%
8%
49%
81%
61%
92%
57%
12%
25%
64%
NA
73%
NA
NA
NA
Source: Authors using data from the SHAKAR Project for Bangladesh; Marx, Stoker, and Suri (2013) for
Kenya; and the 2005 2006 National Family Heath Survey (NFSH-3) for India.
Notes: In the firstcolumn, we report the fraction of slum households who share their latrine with other
households. In the secondcolumn, we report the fraction of households using an open space or traditional
pit as latrine. To compute this number, we combine open air and septic tank/pit toilet in Hyderabad;
traditional pit latrine, bucket, and bush or river or stream in Kibera; flush to septic tank, flush to
pit latrine, flush to somewhere else, flush, dont know where, ventilated improved pit latrine, pit
latrine with slab, pit latrine without slab/open pit, no facility/bush/field, composting toilet, and
dry toilet in the 2005 2006 National Family Heath Survey (NFSH-3) data (Kolkata, Mumbai and All
Indian slums). In the thirdcolumn, we report the fraction of households who share their main drinking
water source with other households. This combines share restricted and shared unrestricted in the
Bangladesh data, public tap and public water tank in Kibera, public tap/standpipe and tube well
or boreholes outside the dwelling in the NFSH-3 data. In the fourthcolumn, we report the fraction of
households whose garbage is not collected by a public or a private company. NA means not available.
outbreaks, slum households exhibit poorer health outcomes than their rural counterparts. The prevalence of underweight, stunting, and wasting (acute malnutrition) is
in fact greater in the slum outskirts of the capital Freetown than in rural areas nationwide, as children under five living in slums have significantly lower weight-for-age and
weight-for-height indexes than children under five in rural areas.5 Across cities in
the developing world, there is some evidence that life expectancy is lower, and infant
mortality higher among the urban poor than among comparable groups in rural and
formal urban areas (Bradley, Stephens, Harpham, and Cairncross 1992).
Health and sanitation issues are rendered more problematic by the lack of
provision of a social safety net in slums. Slum living involves a wide range of risks:
in our Kibera data, 10percent of households have experienced being evicted from
their dwelling, and 4percent report at least one death in the household in the past
sixmonths. In Bangladesh, 56percent of respondents say they do not meet their
basic needs of food, water, shelter, and healthcare; 48percent do not feel safe in their
house during bad weather; and of the households reporting onemember being ill
in the previous 30days, 26percent say that they could not afford to seek medical
attention. In Hyderabad slums, where 70percent of households classify themselves
as poor, only 12percent report receiving any assistance from the government.
Calculated by the authors using the UNICEF SMART Survey data for Sierra Leone.
194
195
Table 4
Ownership Type of Main Dwelling
Tongi (Dhaka)
Hyderabad
Kibera
Kolkata
Mumbai
Own
Rent
Occupy
Other arrangements
52%
68%
7%
37%
72%
41%
28%
92%
56%
26%
7%
3%
1%
NA
NA
0%
1%
1%
8%
2%
Source: Authors using data from the SHAKAR Project for Bangladesh; Marx, Stoker, and Suri (2013) for
Kenya; and the 2005 2006 National Family Heath Survey (NFSH-3) for India.
Notes: Occupy refers to situations where the respondent lives in a dwelling without paying for rent
and without holding a title on the land. Other arrangements include land given by the government in
Hyderabad, land owned by a friend or relative in Kibera, and land obtained as part of an employment
or any other arrangement in Mumbai and Kolkata. The Other arrangements category may include
occupiers in Mumbai and Kolkata, although this is not clear from the survey questionnaire. NA means
notavailable.
stands in stark contrast to some of the problems that characterize rural poverty,
where relatively cheap technologies can often lead to substantial improvements in
income and welfare (for example, the results of green revolution technologies
in agriculture). Not only are private investments in housing infrastructure low in
slums, but Duflo, Galiani, and Mobarak (2012) have documented a low willingnessto-pay for improved public goods in poor urban areas. A big push approach would
then seem necessary to address the lack of investment in slums, and to generate
aggregate demand spillovers in the area of public goods and basic services the
seminal model described Murphy, Shleifer, and Vishny (1989) would justify this type
ofintervention.
A third, less well-known cause for low investment levels in slums could be the
high rent premiums that dwellers must pay to live in close proximity to the city,
and which reduce opportunities for savings accumulation. While slum dwellers are
typically thought of as squatters occupying vacant public land, available evidence
suggests that a large number of dwellers across slums are in fact rent-paying tenants,
as shown in Table4.
In a survey of city officials in Kibera (Marx, Stoker, and Suri 2013), we found
that the two most common causes of landlordtenant disputes in the slum were the
tenants inability to pay their rent and rent increases asked by the landlord. Looking
at amounts paid in rent across different income brackets in Kibera, households in
the poorest quantiles do not pay lower rents per square meter occupied (as shown
in Table5). This finding undermines the notion that rural migrants can pay a modest
premium to live in close proximity to the city. Although rents for urban poor look
rather low in amount (about 12percent of consumption), most rural households (for
example, 90percent in Kenya) do not pay any rent. In 2008, the average monthly rent
for rural households was 266Kenyan shillings (about $3US), or 1percent of household consumption, as opposed to 3,303 Kenyan shillings ($39 US), or 10 percent
196
Table 5
Rent Prices across Consumption Quintiles in Kibera
(rent per month)
1st quintile
Area rented per capita
(square meters)
Rent per capita
(Kenyan shillings)
Rent per square meter
(Kenyan shillings)
2nd quintile
3rd quintile
4th quintile
5th quintile
5.3
7.9
10.9
14.1
15.9
310.9
435.2
488.6
666.0
1121.4
117.6
107.5
96.8
109.5
127.0
of consumption, for urban households ( Jack and Suri forthcoming). In the Kibera
slum, where food expenditure represents 61percent of consumption, housing rents
represent in fact almost a third of nonfood expenditure.6
A fourthset of factors that can cause low investment involve the extreme coordination failures and governance gap intrinsic to slum life. Widespread governance
failures work against the prospects for the urban poor to find creative solutions to
upgrade the quality of their neighborhoods (as envisioned in Turner and Fichter
1972). A large amount of anecdotal evidence suggests that allocation mechanisms
in slums are inefficient and that private actors or bureaucratic entrepreneurs fill
the governance space, as opposed to legitimate local governments or community
representatives. For example, land and housing markets are often controlled by
a handful of powerful or well- connected individuals: landlords, local bureaucrats,
or gang members. Davis (2006) reported on the example of Mumbai, where it is
claimed that 91individuals control all vacant land.7 The slums of Nairobi, where
land permits on vacant land have been illegally awarded by the local administration since the 1970s, are a poster case. The Nairobi slumlords can often rely on
the support of the local administration to settle rent disputes ( Joireman 2011),
and they may collude with local chiefs to discourage improvements in the housing
infrastructure that could lead to more entrenched tenancy rights. In areas where
chiefs are not able to enforce their authority, gangs sometimes fill the governance
Across the developing world, poor households tend to spend a large share of their income on food, as
discussed in this journal by Banerjee and Duflo (2007).
7
Owning and renting out structures in slums can be immensely profitable to landowners: in Nairobi,
Amis (1984) reported that annual returns on the housing capital stock in slums could be as high as
131percent.
197
space to enforce rules of their own, levy taxes, and control expenditure and investments in their neighborhoods (Marx, Stoker, and Suri 2013). In other areas, the
formal governance system is entirely absent and has been replaced by these other
interestsan example is the role of drug cartels in the favelas of Rio de Janeiro
(Ferraz and Ottoni 2013).8
A fifth potential contributor to low investment traps in slums comes from the
well-known Todaro paradox (1976): slum living standards cannot be improved
without generating an additional influx of rural migrants, which in turn depresses
public and private investments in the existing settlements. This may give little
incentive for the public sector to invest in infrastructure and public goods in slums.
However, there is little rigorous evidence on the link between these pull factors and
slum growth in the developing world, while push factors (such as overcrowding on
the fertile lands and an overall decline in agricultural output) have been better documented. For instance, Lipton (1977) and Bates (1981) argued that the urban bias
of policy and tax-based income transfers between peasants and city dwellers until the
1990s was a chief cause of ruralurban migration. The seminal model on the issue
of ruralurban migration was that of Harris and Todaro (1970), who modeled the
ruralurban wage gap as a driving force behind migration decisions. However, this
work had little to say about locations decisions of migrants within cities, and we are
not aware of any more recent theoretical attempt to model those location choices.9
The Policy Trap
In addition to being trapped in low-human-capital and low-investment equilibriums, slum areas are generally places of extreme policy neglect, well beyond the
lack of public goods provision discussed above. This policy trap stems from political economy factors that are different from the market failures we just discussed.
First, the informal nature of slum neighborhoods implies that these areas are
usually considered not eligible for urban planning or public upgrading projects.
This may be for purely administrative reasons or because public investments could
amount to more entrenched occupancy rights for the slum residents an outcome
that governments generally do not favor (Fox 2013). Over the past twodecades, the
few countries that made significant advances in the struggle against slum growth
were those where political support was widespread for reducing the prevalence of
slums and where a genuine political commitment was expressed for curbing slum
expansionfor example, in Egypt and Mexico (UN-Habitat 2006).
Second, enumeration problems and the fact that slum populations are often
(deliberately or mistakenly) undercounted distorts the weight assigned to slum areas
8
Ferraz and Ottoni (2013) study the effects of a pacification program in the Rio favelas where military
interventions were necessary to re-establish a police presence in the slums.
9
For US cities, one canonical modeling approach to looking at location decisions in urban areas is to
think of a city with a central business district and how housing, commerce, and schools are organized
relative to that city center, along the lines of Alonso (1964), Muth (1969), and Mills (1972). But these
sorts of models do not seem an apt description of the location decisions and outcomes in the cities of
low-income countries.
198
in the political process. For example, Sabry (2010) showed how the undercounting
of ashwaiyyat (informal settlement) populations of Greater Cairo led to an undersampling of slum households in household surveys. In India, slum populations were
comprehensively enumerated for the first time in 2001, but discrepancies in the
state-level definitions of slums and the refusal of some states to validate the slum
statistics resulted in gross under-estimation/under-coverage of slum populations in
the country (Government of India 2011). Lack of representation can have dramatic
consequences when issues of eviction are at stake. For example, the Makoko neighborhood in Lagos, Nigeria, one of the oldest slums in the world until its partial
demolition in 2012, had not been covered by the countrys last national census in
2007 (Babalola 2009). Not having accurate census data on slums is problematic for
many reasons. The true population of the slum is unknownfor example, there has
been controversy over the population of the Kibera slum, with estimates ranging
from about 170,000 to over onemillionand policy interventions are impossible to
plan without accurate population numbers.10
Third, catering to the interests of the silent majority of slum dwellers might not
even be in the best interest of the people in charge in the slum. As discussed above,
planning or regulatory powers in slums often do not belong to legitimate governance
bodies, but are usually split between a galaxy of private actors, landlords, chiefs
and bureaucrats, and gangs. Conflicting interests between these actors, and policy
conflicts between central government and municipal authorities could explain why
status quo interests have often prevailed in slums (Fox 2013). In other words,
maintaining high transaction costs and opaque governance mechanisms can be very
beneficial to a minority of bureaucratic entrepreneurs willing to garner support in
slum patronage politics or to extract rents from their informal control over the land
(Amis 1984; Marx, Stoker, and Suri 2013). In line with this, Fox (2013) documents
how members of Tanzanias ruling Chama Cha Mapinduzi (CMM) Party are involved
in transactions on slum land markets. Syagga, Mitullah, and Karirah-Gitau (2002)
found that 57percent of landlords in Nairobi slums were public employees.
10
Aside from the policy constraints that lack of data poses, the poor census implies that the sampling
frame for any survey (national or specifically of the slum) cannot draw a representative random sample.
It is therefore much harder to accurately collect socioeconomic indicators of conditions in a slum and
understand how these conditions evolve over time.
199
Figure 1
Patterns in Urban and Slum Growth, 1990 2007
120
100
80
60
40
20
0
ig
ki
Pa
ia
er
sta
sh
es
in
ia
de
la
pp
ili
ng
Ba
Ph
a
in
Ch
il
az
Br
es
on
d
In
ia
d
In
co
yp
40
i
ex
Eg
20
60
80
% increase in slum population
Source: Figure compiled by the authors using data from UN-Habitats Global Urban Indicators online
database (UN-Habitat 2013). The data on GDP per capita come from the World Banks World
Development Indicators online database.
Notes: Figure 1 compares slum growth and urban growth in the ten countries that had more than
10million slum households in 1990. The countries are ranked by the percentage growth in their slum
population from 1990 to 2007, shown by the bars. The dotted line shows for each country in the sequence
the percentage increase in GDP per capita from 1990 2007 divided by 10 (to fit on the same scale).
growth, and the prevalence of slums? Is there evidence that standards of living are
improving within slums, and/or across generations of slum dwellers?
Over the past 20years, countries that experienced fast economic growth are
also the ones that achieved the most significant reductions in the proportion of
urban households living in slums. In a cross-country regression framework, Arimah
(2010) found that the prevalence of slums in any given country was significantly
correlated with a variety of aggregate economic indicators, including GDP per
capita (negatively), the debt stock and debt service, and inequality measured by the
Gini coefficient (positively). However, cross-country correlations overlook widely
heterogeneous situations, as rapid urbanization rates in developing countries are
often not associated with fast economic growth. In fact, a number of the least developed countries have experienced a rapid growth of their urban population without
experiencing much economic growth at all. Extreme rural poverty, natural disasters,
and civil wars have been the main drivers of this urbanization without growth. An
example is given by the Democratic Republic of the Congo, where the population
of the countrys capital Kinshasa more than tripled in size between the beginning of
the Mobutu regime (1965) and the end of the Second Congo War (2002).
Figure1 compares slum growth and urban growth in the ten countries that
had more than 10 million slum households in 1990. The countries are ranked
200
by the percentage growth in their slum population from 1990 to 2007, shown by
the bars. These bars can be compared to overall urban population growth, shown
by the dark solid line. A few countries experienced explosive urban growth
with limited or no slum expansion (for instance, India, Indonesia, and Brazil),
whereas in others (Pakistan and Nigeria), slum growth accounts for most of
urban growth. The countries where urban population growth outstripped slum
population growth have a declining share of urban population living in slums:
between 1990 and 2009, this proportion decreased from 55 to 29 percent in
India, from 44 to 29percent in China, and from 51 to 23percent in Indonesia
(UN-Habitat2012a).
The dotted line in Figure 1 shows for each country in the sequence, the
percentage increase in GDP per capita over 19902007 divided by 10 (to fit on
the same scale). Very roughly, per capita growth in GDP was similar between the set
of countries where slum populations fell (Egypt, Mexico, and Indonesia), and the
set of countries where most of urban growth was slum populations (the Philippines,
Pakistan, and Nigeria). In fact, it appears that the connection between economic
growth and slum growth across countries is quite diverse, without a uniform
pattern. Below, we discuss how different policy choices may have contributed to
these heterogeneous experiences.
An Intergenerational Perspective
In many slums in low-income countries, living standards do not seem to be
improving over time. In Kibera, Kenya, census data suggest that living conditions
have either deteriorated or at best stagnated over the 19992009 period, a period
during which the economy as a whole grew at 3.5percent per annum. The share
of household heads with a primary education fell from 47 to 40percent over this
time; the number of rooms per capita held essentially the same at 0.68 in 1999, and
0.67 in 2009; and the share of those using a pit latrine (rather than the main sewer
system) fell only slightly, from 82percent in 1999 to 77percent by 2009.
Of course, this simple analysis of living standards over a given period overlooks a fundamental selection problem: households that improved their condition
over the period may no longer live in the slum, while other poor households may
have migrated into the slum. The available evidence, overall, does not provide
prima facie evidence of rapid changes in the composition of slums. In our Kibera
data, respondent households have lived in the slum for 16 years on average,
and incomes do not increase (nor decrease) with the duration of residency, as
illustrated in Figure2 (panelA). In the Bangladeshi settlements studied, income
per capita correlates negatively and significantly with the total number of years
that that household has spent in the slum and with the number of years since the
households first left the countryside (Figure 2, panel B). Similarly, UN-Habitat
(2003) reported that 41percent of Kolkata slum dwellers had lived in slums for
over 30years, and more than 70percent had lived in slums for over 15years. In a
case study of Bangkok slums, 60percent of individuals were reportedly born in the
same slum (UN-Habitat 2003).
201
Figure 2
Living Standards and Duration of Residency in Slums of Bangladesh and Kenya
A: Kenya
Monthly consumption
per capita in US$ (2013)
300
200
100
0
0
20
40
Years of residency in the Kibera slum
60
80
20
40
60
Number of years since the household left rural area
80
B: Bangladesh
Monthly income
per capita in US$ (2000)
25
20
15
10
5
0
Notes: The first scatterplot shows the result of a single variable regression of monthly consumption per
capita on the number of years the household has lived in the slum from our Kibera survey data. The
estimated slope is 0.017 (SE 0.15). In this dataset the average monthly consumption is USD 64.
The second scatterplot presents the result of a regression of monthly income per capita (defined as
the total household income in US$ (non PPP) divided by the household size) on the number of years
since the household first left rural areas for Tongi and Jessore slums in Bangladesh, controlling for
the age and education of the household head, the number of adults in the household, and district
fixed effects. The average monthly consumption is US$15 (US$21 in 2012 equivalent). The slope of
the fitted line is the coefficient of interest in that regression, and the estimated value of this coefficient
is 0.055 (SE: 0.019). The negative sign of the slope is robust to removing these controls and to the
inclusion of more controls.
202
To our knowledge, Perlman (2010) is one of the few studies to have attempted
to address selection issues in surveys of slums by tracing slum respondents and their
descendants over an extended period of time: 19692001, in the favelas of Rio de
Janeiro. Her findings were somewhat ambiguous. She reported that a majority of
individuals interviewed in 1969 and found again in 2001 were no longer living in
the favelas (63percent of the original interviewees, as well as 64percent of their
children and 68 percent of their grandchildren), and that the fraction that had
remained in the favelas had better access to public services (including education)
as well as improved household amenities. However, only a nonrandom 41percent
of the original sample (307 out of 750) could be accurately relocated, of which
22percent were still alive and 19percent were deceased. In addition, from a new
random sample of 425individuals in the same survey areas, she found considerably
higher unemployment rates (51percent, up from 32percent in 1969) and a higher
fraction of respondents with no income (23percent, up from 17percent in 1969).
One question that remains is why households that initially migrated to slums
but did not experience welfare gains do not return to their province of origin. A
dataset on twoslums in Nairobi (APHRC 2013)11 that collects information on where
outmigrants go once they leave the slum provides some answers to that question.
Between 2003 and 2007, 15percent of slum residents moved locations. Of these,
26 percent moved to another slum, 4 percent moved to another location in the
same slums, 22percent moved to a nonslum area in Nairobi, and 41percent moved
to rural Kenya. The notion of slums being a poverty trap implies that households
do not necessarily move there planning to stay, but instead get caught in a low-level
equilibrium. The APHRC data does speak to thisvery few slum residents moved,
only about 15percent, and one-third of these stayed in slum areas.
Empirically, little is known about the outside options that slum households have
or contemplate at any given time. Although there is a small probability of success
(for example, of finding consistent employment), perhaps the expected gains are
large enough to make this decision individually rational. A related literature is
that on entrepreneurshipmany studies find that entrepreneurship does not pay
(Hamilton 2000; Moskowitz and Vissing-Jrgensen 2002), but some of this may be
due to nonpecuniary benefits of being self-employed (Hamilton 2000) or perhaps
to the overconfidence that entrepreneurs have in their own skills (Bernardo and
Welch 2001). In the literature on education, researchers also found that information provided to parents on the returns to education can have effects on enrollment
rates of their children ( Jensen 2010; Nguyen 2008). Issues of overconfidence
and misinformation about individual success probabilities may also be at stake in
slumsthere is room for research on expectations and their role in migration decisions, especially in the context of slum populations.
11
This data covers two slums, Korogocho and Viwandani, which form a Demographic Surveillance Site
in Nairobi. The data is collected by the African Population and Health Research Center. A subset of the
data is available at http://www.aphrc.org/.
203
204
12
Slum upgrading programs until now rarely provided a natural experiment framework that could
isolate causal effects (Field and Kremer 2006). A recent exception is the work of Galiani et al. (2013),
who provide causal estimates of the impact of housing upgrading projects in slums of three Latin
Americancountries.
205
In the past 20years, national governments and the World Bank implemented urban
land titling projects in more than 18African, Asian, and Latin American countries
(Durand-Lasserve, Fernandes, Payne, and Rakodi 2007).
The positive relationship between tenure security and investments in land has
been well documented in the empirical literature for rural settings (for example,
Banerjee, Gertler, and Ghatak 2002; Besley 1995; Goldstein and Udry 2008;
Hornbeck 2010). However, few academic studies have looked at the impact of urban
titling programs on the investment decisions made by households in slums. For
urban households in Peru, Field (2005) found that land titling increases the rate
of housing renovations (by about two-thirds), and Field (2007) showed that titling
increased household labor supply (by 10 to 15percent) by freeing resources that
were previously used to protect household assets. Galiani and Schargrodsky (2010)
showed that the allocation of formal land titles in Buenos Aires led to increased
investments and education amongst households who benefited from the titling. On
the policy side, however, by 2005 concerns began to emerge about how urban land
titling programs were being widely promoted without much if any evidence on their
effectiveness (Durand-Lasserve, Fernandes, Payne, and Rakodi 2007).
The major limitation of the land titling argument is that it assumes that a lack
of formal titles implies weak or nonexistent property rights. However, there is no
systematic evidence that land rights are always weakly enforced in slums. As one
counterexample, Lanjouw and Levy (2002) have argued that informal rights can
effectively substitute for formal titles in slum settlements in Ecuador. In Kibera,
where all vacant land was formally reclaimed by the Kenyan government in 1969,
the members of one ethnic group still claim land rights based on tenancy permits
allocated by the British colonial authorities in the early twentieth century. The
fact that most individuals recognized as landlords live outside the slum (Syagga,
Mitullah, and Karirah-Gitau 2002) also implies that their informal rights over the
renters are strongly enforced.
In a similar vein, analyzing the impact of two titling projects in Senegal and
South Africa, Payne, Durand-Lasserve, and Rakodi (2008) found that residents
in most informal settlements in both [Senegal and South Africa] already enjoy
de facto tenure security. Hence the impact of titling projects on individuals who
already enjoy tenure security is, from the start, uncertain: Palmer (1998) pointed
out how the effectiveness of land titling programs actually depended on whether
they could achieve an increase in the total security that poor households enjoy.
Durand-Lasserve, Fernandes, Payne, and Rakodi (2007) reviewed evidence from
land titling projects that may have actually decreased tenure security (because they
allowed for lawfully enforced evictions) in Afghanistan, Cambodia, Egypt, India,
and Rwanda. Titling programs alone cannot be expected to lift households out of
poverty and to overhaul existing social and economic dynamics within the slum,
because existing systems of ownership act to preserve these dynamics. In fact, land
titling is more likely to benefit the slumlords (whose informal ownership rights
are often well-recognized locally) and hurt, at the bottom of the pyramid, the slum
renters, either in the form of outright evictions or increased rents in the titled area.
206
These features have been documented in Senegal, for instance (Payne, DurandLasserve, and Rakodi2008).
Ultimately, individual approaches such as upgrading and formal titling have
largely failed to improve livelihoods in slums. Recognizing that the effect of
titling projects and conventional land administration systems had been limited,
UN-Habitat (2012c) recently advocated a continuum of land rights relying on
participatory land enumeration and record-keeping to improve tenure security
for the urban poor. This would ideally be included in a more holistic approach of
slum policy. Countries that managed to curb the growth of slums, such as Brazil
or Egypt, indeed appear to be those where slum policy relied on a combination of
instruments including efforts to increase the transparency and efficiency of land
markets, to improve local governance, to increase public investments massively, and
to increase the supply of cheap housing (UN-Habitat 2010a).
Conclusion
Slums represent a major policy challenge for developing economies in the
twenty-first century. Adding migrants to the existing slum populations, UN-Habitat
(2012b) estimated that 450million new housing units would be needed in the next
20years just to accommodate households in urgent need of housing. Yet the challenge of slums is not simply one of housing policy: a holistic approach is needed to
address housing needs for rural migrants, health and sanitation issues, local governance, private savings and investments, and land market institutions. Both formal
and informal systems of property rights may be necessary to curb the rapid growth
of slum areas worldwide. In the absence of strong policy agendas similar to those
adopted in Singapore or, more recently, in Brazil, it seems unlikely that slums will
disappear in the foreseeable future, as implicitly assumed by a modernization view
of the issue.
Overall, there has been very little theoretical and empirical economic research
about how the public policy challenges posed by slums in low-income economies
should be addressed. A research agenda on slums could focus on three distinct
sets of methodological and policy questions. First, the methodological problems
that hamper field research in slums should be addressed. In particular, efforts to
enumerate slum populations and to track panel respondents over several generations of slum dwellers could be stepped up, and empirical methods used to deal with
survey attrition in other contexts (Fitzgerald, Gottschalk, and Moffitt 1998) should
be more consistently applied. This would allow for a better understanding of the
most pressing issues faced by slum dwellers and a better integration of these dwellers
in national political processes. Relatedly, there has been little effort to systematically
study the movements into and out of slums, or to collect data to track the individuals
who do move, even if those exiting are few and far between. Similarly, understanding
the intergenerational correlation in incomes and other socioeconomic outcomes
would be an important contribution.
207
The authors would like to thank David Autor, Chang-Tai Hsieh, Ulrike Malmendier, and
Timothy Taylor for comments that greatly improved the paper. We thank the MIT Sloan School
of Management for funding the Kibera surveys used in this paper. In addition, the authors
acknowledge use of publicly available data for Bangladesh from the Supporting Household
Activities for Health, Assets and Revenue Surveys; for India, from the National Family
Health Survey (NFSH-3); and for Hyderabad, from the data repository of the Abdul Latif
Jameel Poverty Action Lab ( J-PAL). The authors are extremely grateful to the Kenya National
Bureau of Statistics for access to their 1999 and 2009 Population Censuses and UNICEF for
access to the 2010 UNICEF SMART Survey for Sierra Leone.
References
Aggarwal, Smita. 2009. Slum-free India Deadline
Now 7Years. The Indian Express, August28. http://
www.indianexpress.com/news/slumfree-india
-deadline-now-7-yrs/508138/.
Alonso, William. 1964. Location and Land Use:
Toward a General Theory of Land Rent. Cambridge,
MA: Harvard University Press.
Amis, Philip. 1984. Squatters or Tenants: The
Commercialization of Unauthorized Housing in
Nairobi. World Development 12(1): 87 96.
APHRC. 2013. NUHDSS (Nairobi Urban Health
208
209
210
Timothy Taylor
This section will list readings that may be especially useful to teachers of undergraduate economics, as well as other articles that are of broader cultural interest. In
general, with occasional exceptions, the articles chosen will be expository or integrative
and not focus on original research. If you write or read an appropriate article, please
send a copy of the article (and possibly a few sentences describing it) to Timothy Taylor,
preferably by email at [email protected], or c/o Journal of Economic Perspectives,,
Macalester College, 1600 Grand Ave., Saint Paul, Minnesota, 55105.
International Trade
The World Investment Report 2013 from UNCTAD (the UN Conference on Trade
and Development) has the theme Global Value Chains: Investment and Trade for
Development. About 60per cent of global trade, which today amounts to more than
$20trillion, consists of trade in intermediate goods and services that are incorporated
at various stages in the production process of goods and services for final consumption. For instance, even the relatively simple GVC [global value chain] of Starbucks
(United States), based on oneservice (the sale of coffee), requires the management
of a value chain that spans all continents; directly employs 150,000people; sources
coffee from thousands of traders, agents and contract farmers across the developing
world; manufactures coffee in over 30plants, mostly in alliance with partner firms,
doi=10.1257/jep.27.4.211
212
usually close to final market; distributes the coffee to retail outlets through over
50 major central and regional warehouses and distribution centres; and operates
some 17,000retail stores in over 50countriesacross the globe. This GVC has to be
efficient and profitable, while following strict product/service standards for quality.
It is supported by a large array of services, including those connected to supply chain
management and human resources management/development, both within the firm
itself and in relation to suppliers and other partners. The trade flows involved are
immense, including the movement of agricultural goods, manufactured produce, and
technical and managerial services. At http://unctad.org/en/PublicationsLibrary
/wir2013_en.pdf.
The World Trade Report 2013 from the World Trade Organization has the theme
Factors Shaping the Future of World Trade. Measured in gross terms, the dollar
value of world merchandise trade increased by more than 7per cent per year on
average between 1980 and 2011, reaching a peak of US$ 18trillion at the end of that
period. Trade in commercial services grew even faster, at roughly 8per cent per year
on average, amounting to some US$ 4trillion in 2011. . . . Developing economies
only accounted for 34 per cent of world exports in 1980 but by 2011 their share
had risen to 47 per cent, or nearly half of the total. . . . Improvements in transport, telecommunications and information technology, together with increased
economic integration and greater trade openness, have resulted in higher levels
of technological diffusion and increased mobility and accumulation of productive
factors over time. As a result, countries have become less specialized in the export
of particular products, and therefore more similar in terms of their export composition. Comparative advantage, or international differences in relative efficiencies
among products, has become weaker over time in many countries. At http://www
.wto.org/english/res_e/publications_e/wtr13_e.htm.
Arvind Subramanian and Martin Kessler discuss The Hyperglobalization of
Trade and Its Future. This paper describes seven salient features of trade integration in the 21st century: Trade integration has been more rapid than ever
(hyperglobalization); it is dematerialized, with the growing importance of services
trade; it is democratic, because openness has been embraced widely; it is crisscrossing because similar goods and investment flows now go from South to North
as well as the reverse; it has witnessed the emergence of a mega-trader (China),
the first since Imperial Britain; it has involved the proliferation of regional and
preferential trade agreements and is on the cusp of mega-regionalism as the worlds
largest traders pursue such agreements with each other; and it is impeded by the
continued existence of high barriers to trade in services. Peterson Institute for
International Economics, July 2013, Working Paper 13-6, http://www.iie.com
/publications/wp/wp13-6.pdf.
The Bank of International Settlements has published the Triennial Central
Bank Survey, subtitled Foreign Exchange Turnover in April 2013: Preliminary
Global Results. Trading in foreign exchange markets averaged $5.3 trillion per
day in April 2013. This is up from $4.0trillion in April 2010 and $3.3trillion in April
2007. . . . The US dollar remained the dominant vehicle currency; it was on one side
Timothy Taylor
213
of 87% of all trades in April 2013. The euro was the second most traded currency,
but its share fell to 33% in April 2013 from 39% in April 2010. The turnover of the
Japanese yen increased significantly between the 2010 and 2013 surveys. So too did
that of several emerging market currencies, and the Mexican peso and Chinese
renminbi entered the list of the top 10most traded currencies. September 2013,
http://www.bis.org/publ/rpfx13fx.pdf.
Lecture Notes
Alan Krueger spoke at the Rock and Roll Hall of Fame on Rock and Roll,
Economics, and Rebuilding the Middle Class. He used the music industry as a
microcosm for technological trends that have led to greater income inequality in
recent decades. I want to highlight fourfactors that are important in generating
a superstar economy. These are technology, scale, luck and an erosion of social
norms that compress prices and incomes. All of these factors are affecting the
music industry. . . . Technological changes through the centuries have long made
the music industry a super star industry. Advances over time including amplification, radio, records, 8-tracks, music videos, CDs, iPods, etc., have made it possible
for the best performers to reach an ever wider audience with high fidelity. And
the increasing globalization of the world economy has vastly increased the reach
and notoriety of the most popular performers. They literally can be heard on a
worldwide stage. But advances in technology have also had an unexpected effect.
Recorded music has become cheap to replicate and distribute, and it is difficult to
police unauthorized reproductions. This has cut into the revenue stream of the best
performers, and caused them to raise their prices for live performances. My research
suggests that this is the primary reason why concert prices have risen so much since
the late 1990s. In this spirit, David Bowie once predicted that music itself is going
to become like running water or electricity, and, that as a result, artists should
be prepared for doing a lot of touring because thats really the only unique situation thats going to be left. While concerts used to be a loss leader to sell albums,
today concerts are a profit center. June12, 2013, at http://www.whitehouse.gov
/blog/2013/06/12/rock-and-roll-economics-and-rebuilding-middle-class/.
Raghuram Rajan delivered the Andrew Crockett Memorial Lecture at the Bank
of International Settlements, titled A Step in the Dark: Unconventional Monetary
Policy after the Crisis. Twocompeting narratives of the sources of the crisis, and
attendant remedies, are emerging. The first, and the better known diagnosis, is
that demand has collapsed because of the high debt build up prior to the crisis. . . .
But there is another narrative. And that is that the fundamental growth capacity in
industrial countries has been shifting down for decades now, masked for a while
by debt-fuelled demand. More such demand, or asking for reckless spending from
emerging markets, will not put us back on a sustainable path to growth. Instead,
industrial democracies need to improve the environment for growth. The first
narrative is the standard Keynesian one, modified for a debt crisis. It is the one
214
most government officials and central bankers, as well as Wall Street economists,
subscribe to, and needs little elaboration. The second narrative, in my view, offers a
deeper and more persuasive view of the blight that afflicts our times. Rajan argues
that central banks took the right actions during the financial crisis, but that the
wisdom of the ultra-low interest rate policies in the aftermath of the crisis are not
yet clear. Churchill could well have said on the subject of unconventional monetary
policy, Never in the field of economic policy has so much been spent, with so little
evidence, by so few. Unconventional monetary policy has truly been a step in the
dark. June23, 2013, at http://www.bis.org/events/agm2013/sp130623.htm.
Potpourri
An IMF staff team led by Bernardin Akitoby offers some thoughts on Reassessing the Role and Modalities of Fiscal Policy in Advanced Economies. The
prevailing consensus before the crisis was that discretionary fiscal policy had a limited
role to play in fighting recessions. The focus of fiscal policy in advanced economies
was often on the achievement of medium- to long-run goals such as raising national
saving, external rebalancing, and maintaining long-run fiscal and debt sustainability
given looming demographic spending pressures. For the management of business
cycle fluctuations, monetary policy was seen as the central macroeconomic policy
tool. . . . While debate continues, the evidence seems stronger than before the crisis
that fiscal policy can, under todays special circumstances, have powerful effects
on the economy in the short run. In particular, there is even stronger evidence
than before that fiscal multipliers are larger when monetary policy is constrained
by the zero lower bound (ZLB) on nominal interest rates, the financial sector is
weak, or the economy is in a slump. Earlier research often assumed that the impact
of fiscal policy was similar across different states of the economy, but a number of
recent empirical studies suggest that fiscal multipliers may be larger during periods
of slack. September 2013, IMF Policy Paper at http://www.imf.org/external/np
/pp/eng/2013/072113.pdf.
The Congressional Budget Office discusses Rising Demand for Long-Term
Services and Supports for Elderly People. By 2050, one-fifth of the total U.S.
population will be elderly (that is, 65 or older), up from 12percent in 2000 and
8percent in 1950. The number of people age 85 or older will grow the fastest over
the next few decades, constituting 4percent of the population by 2050, or 10times
its share in 1950. That growth in the elderly population will bring a corresponding
surge in the number of elderly people with functional and cognitive limitations. . . .
On average, about one-third of people age65 or older report functional limitations
of one kind or another; among people age 85 or older, about two-thirds report
functional limitations. . . . The total value of long-term services and supports for
elderly people, including the estimated economic value of informal (or donated)
care, exceeded $400 billion in 2011 . . . June 26, 2013, at http://www.cbo.gov
/publication/44363.
215
Frank Levy and Richard J. Murnane consider the interaction between workers
and machinery in Dancing with Robots: Human Skills for Computerized Work.
On March22, 1964, President Lyndon Johnson received a short, alarming memorandum from the Ad Hoc Committee on the Triple Revolution. The memo warned
the president of threats to the nation beginning with the likelihood that computers
would soon create mass unemployment: A new era of production has begun. Its
principles of organization are as different from those of the industrial era as those of
the industrial era were different from the agricultural. The cybernation revolution
has been brought about by the combination of the computer and the automated
self-regulating machine. This results in a system of almost unlimited productive
capacity which requires progressively less human labor. Cybernation is already
reorganizing the economic and social system to meet its own needs. The memo
was signed by luminaries including Nobel Prize winning chemist Linus Pauling,
Scientific American publisher Gerard Piel, and economist Gunnar Myrdal (afuture
Nobel Prize winner). Nonetheless, its warning was only half right. There was no
mass unemploymentsince 1964 the economy has added 74 million jobs. But
computers have changed the jobs that are available, the skills those jobs require, and
the wages the jobs pay. For the foreseeable future, the challenge of cybernation
is not mass unemployment but the need to educate many more young people for
the jobs computers cannot do. Third Way, 2013, http://content.thirdway.org
/publications/714/Dancing-With-Robots.pdf.
C. Robert Taylor and Diana L. Moss have written The Fertilizer Oligopoly:
The Case for Antitrust Enforcement, as a monograph for the American Antitrust
Institute. Collusive agreements between fertilizer producers on prices and market
shares pepper the history of the global commercial fertilizer industry dating back
to the 1880s. The underlying structure of the current global industry remains
conducive to anticompetitive coordinationa landscape that undoubtedly
prompted Wall Street Journal commentators to observe that fertilizer markets are
so manipulated, they might make a Saudi prince blush, and the global price sets
a benchmark so American farmers pay essentially what the cartels dictate. Indeed,
the global industry is dominated by two government-sanctioned export associations
in the U.S. (PhosChem)
(
) and Canada (Canpotex);
(
); a privately traded monopoly sanctioned and likely controlled by the Moroccan government (Office Chrifien des
Phosphates (OCP)); and a cabal of three potash companies in the former Soviet
Union (Belaruskali, Silvinit, and Uralkali, operating through their marketing
cartel, Belarusian Potash Company (BPC)). . . . Damages from supra-competitive
pricing of fertilizer likely amount to tens of billions of dollars annually, the direct
effects of which are felt by farmers and ranchers. But consumers all over the world
suffer indirectly from cartelization of the fertilizer industry through higher food
prices, particularly low-income and subsistence demographics. American Antitrust Institute, 2013, at http://www.antitrustinstitute.org/~antitrust/sites/default
/files/FertilizerMonograph.pdf.
Cato Unbound offers four essays on The Political Economy of Recycling.
In the lead essay, Michael Munger asks: Recycling: Can It Be Wrong, When It
216
Timothy Taylor
217
Discussion Starters
Richard V. Reeves, Isabel Sawhill, and Kimberley Howard discuss The
Parenting Gap. Gaps in cognitive ability by income background open up early in
life, according to research by Tamara Halle and her colleagues at Child Trends, a
non-profit research center focused on children and youth. Children in families with
incomes lower than 200percent of the federal poverty line score, on average, onefifth of a standard deviation below higher-income children on the standard Bayley
Cognitive Assessment at nine monthsbut more than half a standard deviation
below higher-income peers at twoyears. This is the social science equivalent of the
difference between a gully and a valley. These early months are critical for developing skills in language and reasoningand, of course, months in which parents
play the most important role. Closing ability gaps in the first two years of life
pre-pre-K, if you likemeans, by definition, closing the parenting gap. . . . Parents
influence their childs fortunes right from their first breath, while pre-K is aimed at
4-year-olds. In child-development terms, fouryears is an eon. By the time pre-K kicks
in, big differentials in test scores are already apparent. . . . In the last fiveyears, the
federal government has allocated $37.5billion to Head Start25times as much as
promised to home-visiting programs over the next five. This may not be the optimum
ratio in terms of promoting greater mobility and opportunity. . . . Democracy,, Fall
2013, pp.4050. http://www.democracyjournal.org/30/the-parenting-gap.php.
Jacqueline Deslauriers tells of the great maple syrup heist in Liquid Gold.
The Federation of Quebec Maple Syrup Producers was set up in 1966 to represent
and advocate for producersmost of them dairy farmers who supplemented their
income by tapping maple trees. By the 1990s, maple syrup output had grown rapidly,
and by 2000 the industry was producing a surplus of between 1.3 and 2 million
gallons a year. . . . Any output that cannot be sold must be transferred to the federations reserve. Producers do not receive payment for this excess production until
the federation sells it. . . . Maple syrup is sold from the reserve when current production does not meet the demand from authorized buyers. In 2009, after fourdismal
years of production, the global maple syrup reserve ran dry. Since then production
has bounced back and the reserve is overflowing. . . . The $18million theft was from
one of three warehouses the federation uses to stash excess production and was
discovered in mid2012 during an audit of the warehouse contents. The warehouse,
about 60miles southwest of provincial capital Quebec City, was lightly guardedin
retrospect, perhaps, too lightly guarded. The thieves set up shop nearby, and over
the course of a year, according to police, made off with roughly 10,000barrels of
maple syrupabout 323,000gallons, or about 10percent of the reserve. Finance
& Development,, June 2013, pp. 4851. http://www.imf.org/external/pubs/ft
/fandd/2013/06/deslauriers.htm.
John C. Williams explains that Cash Is Dead! Long Live Cash! [S]ince the
start of the recession in December 2007 and throughout the recovery, the value of
U.S. currency in circulation has risen dramatically. It is now fully 42% higher than
it was fiveyears ago. . . . Over the past fiveyears, cash holdings increased on average
218
about 7% annually, more than threetimes faster than the economys growth rate
over this period. At the end of 2012, currency in circulation stood at over $1.1trillion, representing a staggering $3,500 for every man, woman, and child in the
nation. . . . As fears about the safety of the banking system spread in late 2008, many
people became terrified of losing their savings. Instead, they put their trust in cold,
hard cash. Not surprisingly, as depositors socked away money to protect themselves
against a financial collapse, they often sought $100bills. Such a large denomination
is easier to conceal or store in bulk than smaller bills. Indeed, in the sixmonths
following the fall of the investment bank Lehman Brothers in 2008, holdings of
$100bills soared by $58billion, a 10% jump. 2012 Annual Report,, Federal Reserve
Bank of San Francisco, http://www.frbsf.org/files/2012_Annual_Report.pdf.
Michael Manville and Jonathan Williams advocate ending free parking placards for the disabled in Parking without Paying. The government isnt going to
hand out free gasoline anytime soon, but at least 24states and many local governments do distribute free parking passes, in the form of disabled placards. These
placards not only grant access to spaces reserved for people with disabilities, but
also let their holders park free, often for unlimited time, at any metered space.
Nor are placards difficult to get. In California, for example, doctors, nurses, nurse
practitioners, optometrists and chiropractors can all certify people for placards, for
everything from serious permanent impairments to temporary conditions like a
sprained ankle. We recommend that cities and states limit or eliminate free parking
for disabled placards. We believe the payment exemption has high costs and few
benefits. It harms both the transportation system and the environment, and offers
little help to most people with disabilities. Access,, Spring 2013, pp.1013, http://
www.uctc.net/access/42/access42_parkingwoutpaying.shtml.
Ben A. Minteer and Leah R. Gerber propose Buying Whales to Save Them.
Under this plan, quotas for hunting of whales would be traded in global markets.
But again, and unlike most catch share programs in fisheries, the whale conservation market would not restrict participation in the market; both pro- and antiwhaling
interests could own and trade quotas. . . . Conservation groups, for example, could
choose to buy whale shares in order to protect populations that are currently threatened; they could also buy shares to protect populations that are not presently at
risk but that conservationists fear might become threatened in the future. Despite
the widely acknowledged failure of the IWC [International Whaling Commission]
moratorium to curtail unsustainable whaling, the whale conservation market idea
has proved to be wildly controversial within conservation and antiwhaling circles.
. . . Many critics of the idea are also plainly not comfortable with the ethics of putting
a price on such iconic speciesthat is, with using contingent market methods for
what they believe should be a categorical ethical obligation to preserve whales.
On the other hand . . . the vulnerable status of many whale populations and the
failure of the traditional regulatory response to halt unsustainable harvests call for a
more innovative and experimental approach to whale policy, including considering
unconventional proposals, such as the whale conservation market. Issues in Science
and Technology,, Spring 2013, http://www.issues.org/29.3/minteer.html.
We are grateful for the research assistance of Ben Sprung-Keyser, whose work on a related project
helped us to uncover this error.
http://dx.doi.org/10.1257/jep.27.4.219
doi=10.1257/jep.27.4.219
220
Table 1
Balance Sheets for Households Aged 6569 in 2008
Percent of households with
positive balance
Mean holding
(dollars)
All households
Net worth
Social Security
Defined benefit pension
Non-annuitized wealth
Financial assets
Personal retirement accounts
Housing and other real estate
99.2
88.2
42.1
90.8
86.7
52.2
81.3
870,908
204,264
99,147
567,496
132,484
121,137
271,605
100.0
23.5
11.4
65.2
15.2
13.9
31.2
Single-person households
Net worth
Social Security
Defined benefit pension
Non-annuitized wealth
Financial assets
Personal retirement accounts
Housing and other real estate
98.8
86.6
38.0
84.4
82.3
36.4
67.8
530,556
134,006
62,555
333,996
83,082
47,074
188,813
100.0
25.3
11.8
63.0
15.7
8.9
35.6
Married couples
Net worth
Social Security
Defined benefit pension
Non-annuitized wealth
Financial assets
Personal retirement accounts
Housing and other real estate
99.5
89.6
45.5
96.0
90.3
65.1
92.3
1,148,873
261,645
129,033
758,196
172,830
181,625
339,222
100.0
22.8
11.2
66.0
15.0
15.8
29.5
Asset category
Source: Authors tabulations using Health and Retirement Study, Wave 9, 2008. Two components of
net worth, business assets (mean value $45,966 for all households) and debt ($3,697) are included
in net worth and non-annuitized wealth, but are not in any of the subcategories (financial assets,
personal retirement accounts, or housing and other real estate). The sum of these three subcategories
therefore does not equal non-annuitized wealth.
Note: Numbers in bold have changed.
the share of wealth in primary homes, secondary homes, and other real estate is
31.1percent rather than 25.9percent as previouslyreported.
The corrected calculations suggest that one household in five has defined
benefit pension wealth of at least $169,400 (in contrast to $238,500 in our earlier
analysis), and that a household at the 30th percentile of the Social Security wealth
distribution has $126,800 (not $214,500) of such wealth, while a household at
the 90th percentile has $384,800 (not $643,100). The median net worth values
at the 70th and 90th percentiles of the net worth distribution are $911,900 (not
$1.1 million) and $1,826,400 (not $2.1million), respectively.
221
Table 2
Distribution of Wealth Components for Households Aged 6569 in 2008
(in 1,000s)
Net
worth
Social
Security
Denfined
benefit
pension
Nonannuitized
wealth
Financial
assets
All households
10
127.3
30
289.3
50
548.2
70
911.9
90
1,826.4
0.0
126.8
187.4
227.8
384.8
0.0
0.0
0.0
83.0
329.6
0.1
71.8
221.7
518.0
1,274.0
0.0
2.0
15.0
70.0
358.0
0.0
0.0
5.0
75.0
347.0
0.0
42.0
120.0
229.5
585.0
Single-person households
10
97.4
30
176.2
50
295.8
70
544.1
90
1,094.1
0.0
98.2
136.7
177.2
230.1
0.0
0.0
0.0
51.2
206.2
0.0
14.0
100.0
272.0
892.0
0.0
0.4
5.0
34.0
240.0
0.0
0.0
0.0
10.1
124.0
0.0
0.0
60.0
150.0
392.0
Married households
10
240.9
30
509.2
50
769.1
70
1,234.1
90
2,224.2
0.0
195.3
284.0
342.6
425.5
0.0
0.0
0.0
116.1
440.4
24.7
158.0
357.0
755.7
1,677.8
0.0
6.0
27.8
107.0
459.2
0.0
0.0
35.0
137.0
464.0
12.0
90.0
170.0
300.0
725.0
Percentile
Personal
Housing &
retirement
other real
account assets
estate
Source: Authors tabulations using 2008 (Wave 9) Health and Retirement Study; see Table 1 and text for
further description.
Note: Numbers in bold have changed.
References
Poterba, James, Steven Venti, and David Wise. 2011. The Composition and Drawdown of Wealth in
Retirement. Journal of Economic Perspectives 25(4): 95 118.
222
Correspondence
224
References
Bessen, James, and Eric Maskin. 2009. Sequential
Innovation, Patents, and Imitation. RAND Journal of
Economics 40(4): 61135.
Correspondence
225
References
Anton, James J., and Dennis A. Yao. 1994. Expropriation and Inventions: Appropriable Rents in the
Absence of Property Rights. American Economic
Review 84(1): 190 209.
Bessen, James, and Allesandro Nuvolari. 2013.
Diffusing New Technology without Dissipating
Rents: Some Historical Case Studies of Knowledge
Sharing. Unpublished paper, Boston University
School of Law.
Boldrin, Michele, and David K. Levine. 2004.
Rent Seeking and Innovation. Journal of Monetary
Economics 51(1): 127 60.
Boldrin, Michele, and David K. Levine. 2010.
Against Intellectual Monopoly, Cambridge University
Press.
Dunn, Peter M. 1999. The Chamberlen Family
(15601728) and Obstetric Forceps. Archives of
Disease in ChildhoodFetal and Neonatal Edition
81(3): F232F234.
Henry, Emeric, and Carlos Ponce. 2009. Waiting
to Imitate: On the Dynamic Pricing of Knowledge.
CEPR Discussion Paper DP7511. Available at SSRN:
http://ssrn.com/abstract=1507507.
Kahn, B. Zorina, and Kenneth L. Sokoloff. 2001.
History Lessons: The Early Development of Intellectual Property Institutions in the United States.
Journal of Economic Perspectives 15(3): 233 46.
Loudon, Irvine. 1989. Review of The Obstetric
Forceps: A Short History and Descriptive Catalogue of
the Forceps in the Museum of the Royal College
of Obstetricians and Gynaecologists, by Bryan M.
Hibbard. Medical History 33(1): 14142. http://www
.ncbi.nlm.nih.gov/pmc/articles/PMC1035787/.
Nuvolari, Allesandro. 2004. Collective Invention
during the British Industrial Revolution: The Case
of the Cornish Pumping Engine. Cambridge Journal
of Economics 28(3): 34763.
Ponce, Carlos J. 2011. Knowledge Disclosure as
Intellectual Property Rights Protection. Journal of
Economic Behavior & Organization 80(3): 418 34.
To be considered for publication in the Correspondence section, letters should be relatively shortgenerally less
than 1,000 wordsand should be sent to the journal offices at [email protected]. The editors will choose which letters
will be published. All published letters will be subject to editing for style and length.
226
Notes
For additional announcements, check out the continuously updated JEP online Bulletin Board (http://www
.aeaweb.org/bulletinboard.php). Calls for papers, notices of professional meetings, and other announcements of interest
to economists should be submitted to Ann Norman, at [email protected], in one or twoparagraphs containing the
relevant information. These will be posted at the JEP online Bulletin Board. Given sufficient lead time (at least one
month before an issue goes online), we will also print a shorter, one-paragraph version of your notice in the Notes
section of the Journal of Economic Perspectives. We reserve the right to edit material received.
Some of the papers presented at the annual meeting are published in the May American Economic
Review (the Papers & Proceedings).
). The Presidentelect includes at least three contributed sessions
(12papers) from among those submitted in response
to this Call for Sessions and Papers.
The 2014 Annual Meeting will take place in
Philadelphia, PA, on January 3 5, 2014. The headquarters hotel will be the Philadelphia Marriott
Downtown. Registration is now open, please go to
https://conf.aeaweb.org/.
AEA Continuing Education Program is held immediately after the Annual Meeting in January. The
program aims to help mid-career economists and
others maintain the value of their human capital. It
is tailored primarily to faculty at liberal arts colleges
and teaching-oriented state universities who may
have fewer research opportunities than colleagues
at universities with PhD programs. The lecturers are
leading scholars who also are excellent expositors.
The focus is on content to help improve teaching
and research.
The three topics for January 2014 in San Diego
are: Cross Section Econometrics (Alberto Abadie,
Harvard University, and Joshua Angrist, MIT),
Education and the Economy (Susan Dynarski, University of Michigan, and Brian Jacob, University of
Michigan), and Economic Growth (Oded Galor,
Brown University, and David Weil, Brown University).
For more information go to http://www.aeaweb.org
/cont_education/.
Election Results. The American Economic Association is pleased to announce the election results
for the 2013 Election for Officers of the American
228
229
l of E
The Jou
ectives
ic Persp
rnal of
ectives
ic Persp
conom
Econom
tion
the
Associa
nal of
A jour an Economic
Americ
Spring
2011
e
Volum
mber
25, Nu
2
230
231
The proposal provides the editors and authors an opportunity to preview the
substance and flow of the article. For proposals thatappear promising, the editors
provide feedback on the substance, focus, and style of the proposed article. After
the editors and author(s) have reached agreement on the shape of the article
(which may take one or more iterations), the author(s) are given several months
to submit a completed first draft by an agreed date. This draft will receive detailed
comments from the editors as well as a full set of suggested edits from JEP s
Managing Editor. Articles may undergo more than one round of comment and
revision prior to publication.
Readers are also welcome to send e-mails suggesting topics for JEP articles and
symposia and to propose authors for these topics. If the proposed topic is a good
fit for JEP, the JEP editors will work to solicit paper(s) and author(s).
Correspondence regarding possible future articles for JEP may be sent
(electronically please) to the assistant editor, Ann Norman, at <anorman@
JEPjournal.org>. Papers and paper proposals should be sent as Word or pdf
e-mail attachments.
New
Principles of Macroeconomics
Activist vs. Austerity Policies
Seventh Edition
Howard J. Sherman and Michael A. Meeropol
Jerry W. Markham
New
Economics of Crime and Enforcement
Anthony M. Yezer
New
Understanding Central Banking
The New Era of Activism
David M. Jones
AD1309H
M.E. Sharpe
Cambridge Surveys of
Economic Literature
Productivity Convergence
Anarchy Unbound
Edward N. Wolff
Peter T. Leeson
$90.00: Hb: 978-1-107-02580-6: 250 pp.
Sweatshops
Improving Lives and
Economic Growth
Benjamin Powell
Economic Politics
in the United States
2nd
Edition
2nd
Edition
William R. Keech
Coming Soon!
Behavioral Economics
A History
Floris Heukelom
Historical Perspectives on
Modern Economics
Economics of Agglomeration
Cities, Industrial Location,
and Globalization
Coming Soon!
Eminent Economists II
Questioning
Credible Commitment
Perspectives on the
Rise of Financial Capitalism
Democracy
and its Elected Enemies
The Analytics of
Uncertainty and
Information
Coming Soon!
Shalendra Sharma
Schumpeterian Analysis of
Economic Catch-up
Knowledge, Path-Creation,
and the Middle-Income Trap
Keun Lee
$55.00: Hb: 978-1-107-04268-1: 256 pp.
www.cambridge.org
@CambUP_Econ
THE NEWEST
IN ECONOMICS
TITLES FROM
OXFORD
NEW IN PAPERBACK
paper
$55.00
BEN RAMALINGAM
2013 cloth
2013
$185.00
cloth
$40.00
WRONG
Nine Economic Policy Disasters and What We Can Learn from Them
RICHARD S. GROSSMAN
STEPHEN H. AXILROD
$74.00 / paper
$16.95
2013
Exodus
cloth
$27.95
cloth
$27.95
cloth
$65.00
To place your order, contact customer service at 1-800-451-7556 or visit us online at oup.com/us.
Prices are subject to change and apply only in the US.
ONOMI
C
EC
URNAL
JO
ERICA
N
AM
ONOMI
C
EC
GABRIEL CHODOROW-REICH
LAURA FEIVESON
ZACHARY LISCOW
WILLIAM WOOLSTON
Does State Fiscal Relief During
Recessions Increase Employment?
Evidence from the American Recovery
and Reinvestment Act
Volume 4 (3), (August 2012, 118145)
AEJ: Macroeconomics
ONOMI
C
EC
URNAL
JO
ERICA
N
AM
URNAL
JO
ERICA
N
AM
URNAL
JO
ERICA
N
AM
JENNIFER HUNT
MARJOLAINE GAUTHIER-LOISELLE
How Much Does Immigration Boost
Innovation?
Volume 2 (2), (April 2010, 3156)
AEJ: Microeconomics
MICHAEL OSTROVSKY
MICHAEL SCHWARZ
Information Disclosure and Unraveling
in Matching Markets
Volume 2 (2), (May 2010, 3463)
RAJ CHETTY
Harvard University
Distinguished Fellows
The recognition of Distinguished Fellows began in 1965. Past presidents of the Association are Distinguished Fellows. Additional Distinguished Fellows may be elected,
but not more than four in any one calendar year from economists of high distinction
in the United States and Canada. The following economists are Distinguished
Fellows for 2013:
HAROLD DEMSETZ
University of California, Los Angeles
STANLEY FISCHER
Bank of Israel
JERRY HAUSMAN
Massachusetts Institute of Technology
PAUL JOSKOW
Massachusetts Institute of Technology
AEA
rCareer Services)VOESFETPG
SFDSVJUFSTVTFPVSi+0&u +PCTGPS
&DPOPNJTUT
QSPHSBNUPBEEZPVOH
UBMFOUFENFNCFSTUPUIFJSSPTUFST
rCollaboration6UJMJ[FNFFUJOHT
DPNNJUUFFQBSUJDJQBUJPO
BOEDPOUJOVJOHFEVDBUJPOQSPHSBNTUPGPTUFS
NFOUPSTIJQ
POHPJOHMFBSOJOHBOEQFFSDPOOFDUJPOT0OMZ"&"NFNCFSTDBO
TVCNJUUIFJSQBQFSTBU"44"
rPeer Recognition"XBSETQSPHSBNTBDLOPXMFEHFUIFDPOUSJCVUJPOTPGUPQ
FDPOPNJTUT3FDJQJFOUTPGUFODJUFUIF"&"BTBDSJUJDBMQBSUOFSJOUIFJS
TVDDFTT
rLearning Resources(FUFYDMVTJWFDPOUFOUBUUIF"&"XFCTJUFJODMVEJOH
HPWFSONFOUEBUB
HSBEVBUFQSPHSBNT
CMPHT
OFXTMFUUFST
JOGPSNBUJPOGPS
TUVEFOUT
SFGFSFODFNBUFSJBMT
+&-$PEFHVJEF
BOENPSF
rSpecial Member SavingsPOBSUJDMFTVCNJTTJPOGFFT
DPOUJOVJOHFEVDBUJPO
DPVSTFT
"&"BSDIJWFTPO+4503
JOTVSBODF
BOEKPVSOBMQSJOUBOE$%PQUJPOT
Economics Research
Starts Here
EconLit
Authoritative Content. Easy to Use. All in One Place.
EconLit includes the most sought-after economics publications from around
the worldpeer-reviewed journal articles, working papers from leading
universities, PhD dissertations, books, collective volume articles,
conference proceedings, and book reviewsall expertly indexed, classified,
and linked to library holdings.
Students and professionals at all levels rely on EconLits broad scope and
authoritative, up-to-date coverage.
With over 1.2 million records and coverage as far back as 1886.
www.econlit.org
EconLit
Four Journals . . .
Five Years
American Economic Journal
Economic Journal
MOHAMAD MAHMOUD AL
AL-ISSISS
ISSISS AND NOLAN MILLER
What Does Health Reform Mean for the Health Care Industry?
Evidence from the Massachusetts Special Senate Election
HUNT ALLCOTT
The Welfare Effects of Misperceived Product Costs: Data and
Calibrations from the Automobile Market
E LOECKER
ing Learning by Exporting
MARC FLEURBAEY
AND ERIK SCHOKKAERT
Behavioral Welfare Economics and
Redistribution
ZLEM BEDRE-DEFOLIE
AND EMILIO CALVANO
Pricing Payment Cards
ADAM FREMETH,
BRIAN KELLEHER RICHTER,
AND BRANDON SCHAUFELE
Campaign Contributions over CEOs
Careers
FRANOIS GOURIO
Credit Risk and Disaster Risk
EMI NAKAMURA, JN STEINSSON, ROBERT BARRO
AND JOS URSA
Crises and Recoveries in an Empirical Model of Consump
Disasters
GLENN ELLISON
How Does the Market Use Citation Data?
The Hirsch Index in Economics
JUAN J. CRUCES
AND CHRISTOPH TREBESCH
Sovereign Defaults: The Price of
SHING-YI WANG
Marriage Networks, Nepotism, and Labor
Market Outcomes in China
MATTHIEU BUSSIRE,
GIOVANNI CALLEGARI,
FABIO GHIRONI,
GIULIA SESTIERI,
AND NORIHIKO YAMA
Estimating Trade Elastici
Demand Composition an
Trade Collapse of 2008
B. KELSEY JACK
Private Information and the Allocation
of Land Use Subsidies in Malawi
DAVID A. MATSA
AND AMALIA R. MILLER
A Female Style in Corporate
Leadership? Evidence from Quotas
ADAM FREMETH,
BRIAN KELLEHER RICHTER,
AND BRANDON SCHAUFELE
Campaign Contributions over CEOs
Careers
Macroeconomics
Applied Economics
Microeconomics
MOHAMAD MAHMOUD AL
AL-ISSISS
ISSISS AND NOLAN MILLER
What Does Health Reform Mean for the Health Care Industry?
Evidence from the Massachusetts Special Senate Election
FRANOIS GOURIO
Credit Risk and Disaster Risk
EMI NAKAMURA, JN STEINSSON, ROBERT BARRO,
AND JOS URSA
Crises and Recoveries in an Empirical Model of Consumption
Disasters
DEN,
DAVID COOK
AND MICHAEL B. DEVEREUX
Sharing the Burden: Monetary and
Fiscal Responses to a World
Liquidity Trap
SHING-YI WANG
Marriage Networks, Nepotism, and Labor
Market Outcomes in China
MATTHIEU BUSSIRE,
GIOVANNI CALLEGARI,
FABIO GHIRONI,
GIULIA SESTIERI,
AND NORIHIKO YAMANO
Estimating Trade Elasticities:
Demand Composition and the
Trade Collapse of 20082009
B. KELSEY JACK
Private Information and the Allocation
of Land Use Subsidies in Malawi
DAVID A. MATSA
AND AMALIA R. MILLER
A Female Style in Corporate
Leadership? Evidence from Quotas
ADAM FREMETH,
BRIAN KELLEHER RICHTER,
AND BRANDON SCHAUFELE
Campaign Contributions over CEOs
Careers
BEN LOCKWOOD
AND FRANCESCO PORCELLI
Incentive Schemes for Local
Government: Theory and Evidence
from Comprehensive Performance
Assessment in England
ZLEM BEDRE-DEFOLIE
AND EMILIO CALVANO
Pricing Payment Cards
ROBERTA PIERMARTINI
AND LINDA ROUSOV
The Sky Is Not Flat: How Discriminatory Is the
Access to International Air Services?
s on
GLENN ELLISON
How Does the Market Use Citation Data?
The Hirsch Index in Economics
DANIEL M. HUNGERMAN
Substitution and Stigma: Evidence
on Religious Markets from the
Catholic Sex Abuse Scandal
MARC FLEURBAEY
AND ERIK SCHOKKAERT
Behavioral Welfare Economics and
Redistribution
TER,
ELE
CEOs
HUC,
FRANOIS GOURIO
Credit Risk and Disaster Risk
IRENA GROSFELD,
ALEXANDER RODNYANSKY,
AND EKATERINA ZHURAVSKAYA
Persistent Antimarket Culture:
A Legacy of the Pale of Settlement
after the Holocaust
ARTURS KALNINS
AND FRANCINE LAFONTAINE
Too Far Away? The Effect of Distance
to Headquarters on Business
Establishment Performance
Macroeconomics
MARK R. JACOBSEN
Fuel Economy and Safety: The Influences of Vehicle Class
and Driver Behavior
LIBERTAD GONZLEZ
The Effect of a Universal Child Benefit on
Conceptions, Abortions, and Early Maternal
Labor Supply
ANJA SAUTMANN
Contracts for Agents with Biased Beliefs:
Some Theory and an Experiment
MATTHIEU BUSSIRE,
GIOVANNI CALLEGARI,
FABIO GHIRONI,
GIULIA SESTIERI,
AND NORIHIKO YAMANO
Estimating Trade Elasticities:
Demand Composition and the
Trade Collapse of 20082009
JUAN J. CRUCES
AND CHRISTOPH TREBESCH
Sovereign Defaults: The Price of Haircuts
R
e
Quotas
location
awi
HUNT ALLCOTT
The Welfare Effects of Misperceived Product Costs: Data and
Calibrations from the Automobile Market
JAN DE LOECKER
Detecting Learning by Exporting
HANNA HAABURDA AND YARON YEHEZKEL
Platform Competition under Asymmetric Information
Labor
MARC FLEURBAEY
AND ERIK SCHOKKAERT
Behavioral Welfare Economics and
Redistribution
Macroeconomic
MARK R. JACOBSEN
Fuel Economy and Safety: The Influences of Vehicle Class
and Driver Behavior
DANA BURDE AND LEIGH L. LINDEN
Bringing Education to Afghan Girls:
A Randomized Controlled Trial of Village-Based Schools
ARTURS KALNINS
AND FRANCINE LAFONTAINE
Too Far Away? The Effect of Distance
to Headquarters on Business
Establishment Performance
ZLEM BEDRE-DEFOLIE
AND EMILIO CALVANO
Pricing Payment Cards
ed Economics
ANJA SAUTMANN
Contracts for Agents with Biased Beliefs:
Some Theory and an Experiment
MATTHIEU BUSSIRE,
GIOVANNI CALLEGARI,
FABIO GHIRONI,
GIULIA SESTIERI,
AND NORIHIKO YAMANO
Estimating Trade Elasticities:
Demand Composition and the
Trade Collapse of 20082009
DAVID A. MATSA
AND AMALIA R. MILLER
A Female Style in Corporate
Leadership? Evidence from Quotas
ROBERTA PIERMARTINI
AND LINDA ROUSOV
The Sky Is Not Flat: How Discriminatory Is the
Access to International Air Services?
JUAN J. CRUCES
AND CHRISTOPH TREBESCH
Sovereign Defaults: The Price of Haircuts
B. KELSEY JACK
Private Information and the Allocation
of Land Use Subsidies in Malawi
BEN LOCKWOOD
AND FRANCESCO PORCELLI
Incentive Schemes for Local
Government: Theory and Evidence
from Comprehensive Performance
Assessment in England
JAN DE LOECKER
Detecting Learning by Exporting
SHING-YI WANG
Marriage Networks, Nepotism, and Labor
Market Outcomes in China
DANIEL M. HUNGERMAN
Substitution and Stigma: Evidence
on Religious Markets from the
Catholic Sex Abuse Scandal
Microeconomics
GLENN ELLISON
How Does the Market Use Citation Data?
The Hirsch Index in Economics
IRENA GROSFELD,
ALEXANDER RODNYANSKY,
AND EKATERINA ZHURAVSKAYA
Persistent Antimarket Culture:
A Legacy of the Pale of Settlement
after the Holocaust
ARTURS KALNINS
AND FRANCINE LAFONTAINE
Too Far Away? The Effect of Distance
to Headquarters on Business
Establishment Performance
FRANOIS GOURIO
Credit Risk and Disaster Risk
LIBERTAD GONZLEZ
The Effect of a Universal Child Benefit on
Conceptions, Abortions, and Early Maternal
Labor Supply
ANJA SAUTMANN
Contracts for Agents with Biased Beliefs:
Some Theory and an Experiment
Macroeconomics
MARK R. JACOBSEN
Fuel Economy and Safety: The Influences of Vehicle Class
and Driver Behavior
cs
Applied Economics
economics
Applied Econom
s
MOHAMAD MAHMOUD AL-ISSISS
AL ISSISS AND NOLAN MILLER
What Does Health Reform Mean for the Health Care Industry?
Evidence from the Massachusetts Special Senate Election
MARK R. JACOBSEN
Fuel Economy and Safety: The Influences of V
and Driver Behavior
HUNT ALLCOTT
The Welfare Effects of Misperceived Product Costs: Data and
Calibrations from the Automobile Market
GLENN ELLISON
How Does the Market Use Citatio
The Hirsch Index in Economic
LIBERTAD GONZLEZ
The Effect of a Universal Child Benefit on
Conceptions, Abortions, and Early Maternal
Labor Supply
SHING-YI WANG
Marriage Networks, Ne
Market Outcomes in C
IRENA GROSFELD,
ALEXANDER RODNYANSKY,
AND EKATERINA ZHURAVSKAYA
Persistent Antimarket Culture:
A Legacy of the Pale of Settlement
after the Holocaust
B. KELSEY JACK
Private Informatio
of Land Use Sub
DAVID A. MAT
AND AMALIA
A Female Style
Leadership? Ev
DANIEL M. HUNGERMAN
Substitution and Stigma: Evidence
on Religious Markets from the
Catholic Sex Abuse Scandal
ADAM FREME
BRIAN KELLE
AND BRANDO
Campaign Contri
Careers
BEN LOCKWOOD
AND FRANCESCO PORCELLI
Incentive Schemes for Local
Government: Theory and Evidence
from Comprehensive Performance
Assessment in England
YANN ALGAN, PI
AND ANDREI SHLE
Teaching Practices and S
ROBERTA PIERMARTINI
AND LINDA ROUSOV
The Sky Is Not Flat: How Discriminatory Is the
Access to International Air Services?
Macroeconomics
Applied Economics
Microeconomics
FRANOIS GOURIO
Credit Risk and Disaster Risk
EMI NAKAMURA, JN STEINSSON, ROBERT BARRO,
AND JOS URSA
Crises and Recoveries in an Empirical Model of Consumption
Disasters
JAN DE LOECKER
Detecting Learning by Exporting
HANNA HAABURDA AND YARON YEHEZKEL
Platform Competition under Asymmetric Information
al
MATTHIEU BUSSIRE,
GIOVANNI CALLEGARI,
FABIO GHIRONI,
GIULIA SESTIERI,
AND NORIHIKO YAMANO
Estimating Trade Elasticities:
Demand Composition and the
Trade Collapse of 20082009
Applied Economics
MOHAMAD MAHMOUD AL
AL-ISSISS
ISSISS AND NOLAN MILLER
What Does Health Reform Mean for the Health Care Industry?
Evidence from the Massachusetts Special Senate Election
JAN DE LOECKER
Detecting Learning by Exporting
Microeconomics
FRANOIS GOURIO
Credit Risk and Disaster Risk
EMI NAKAMURA, JN STEINSSON, ROBERT BARRO,
AND JOS URSA
Crises and Recoveries in an Empirical Model of Consumption
Disasters
JUAN J. CRUCES
AND CHRISTOPH TREBESCH
Sovereign Defaults: The Price of Haircuts
ADAM FREMETH,
BRIAN KELLEHER RICHTER,
AND BRANDON SCHAUFELE
Campaign Contributions over CEOs
Careers
BEN LOCKWOOD
AND FRANCESCO PORCELLI
Incentive Schemes for Local
Government: Theory and Evidence
from Comprehensive Performance
Assessment in England
ROBERTA PIERMARTINI
AND LINDA ROUSOV
The Sky Is Not Flat: How Discriminatory Is the
Access to International Air Services?
ALAN J. AUERBACH AND YURIY GORODNICHENKO
Corrigendum: Measuring the Output Responses to Fiscal Policy
Macroeconomics
DAVID A. MATSA
AND AMALIA R. MILLER
A Female Style in Corporate
Leadership? Evidence from Quotas
MARC FLEURBAEY
AND ERIK SCHOKKAERT
Behavioral Welfare Economics and
Redistribution
ZLEM BEDRE-DEFOLIE
AND EMILIO CALVANO
Pricing Payment Cards
America
Micr
B. KELSEY JACK
Private Information and the Allocation
of Land Use Subsidies in Malawi
ARTURS KALNINS
AND FRANCINE LAFONTAINE
Too Far Away? The Effect of Distance
to Headquarters on Business
Establishment Performance
FRANOIS GOURIO
Credit Risk and Disaster Risk
SHING-YI WANG
Marriage Networks, Nepotism, and Labor
Market Outcomes in China
ANJA SAUTMANN
Contracts for Agents with Biased Beliefs:
Some Theory and an Experiment
MATTHIEU BUSSIRE,
GIOVANNI CALLEGARI,
FABIO GHIRONI,
GIULIA SESTIERI,
AND NORIHIKO YAMANO
Estimating Trade Elasticities:
Demand Composition and the
Trade Collapse of 20082009
DAVID COOK
AND MICHAEL B. DEVEREUX
Sharing the Burden: Monetary and
Fiscal Responses to a World
Liquidity Trap
Macroeconomics
GLENN ELLISON
How Does the Market Use Citation Data?
The Hirsch Index in Economics
JUAN J. CRUCES
AND CHRISTOPH TREBESCH
Sovereign Defaults: The Price of Haircuts
MARK R. JACOBSEN
Fuel Economy and Safety: The Influences of Vehicle Class
and Driver Behavior
HUNT ALLCOTT
The Welfare Effects of Misperceived Product Costs: Data and
Calibrations from the Automobile Market
CHRISTOPH BASTEN AND FRANK BETZ
Beyond Work Ethic: Religion, Individual, and Political Preferences
SAURABH BHARGAVA AND VIKRAM S. PATHANIA
Driving Under the (Cellular) Influence
ELIZABETH CASCIO, NORA E. GORDON,
AND SARAH J. REBER
Local Responses to Federal Grants: Evidence from the
Introduction of Title I in the South
LIBERTAD GONZLEZ
The Effect of a Universal Child Benefit on
Conceptions, Abortions, and Early Maternal
Labor Supply
MARK R. JACOBSEN
Fuel Economy and Safety: The Influences of Vehicle Class
and Driver Behavior
DANA BURDE AND LEIGH L. LINDEN
Bringing Education to Afghan Girls:
A Randomized Controlled Trial of Village-Based Schools
HAROUNAN KAZIANGA, DAN LEVY, LEIGH L. LINDEN,
AND MATT SLOAN
The Effects of Girl-Friendly Schools: Evidence from the
BRIGHT School Construction Program in Burkina Faso
GLENN ELLISON
How Does the Market Use Citation Data?
The Hirsch Index in Economics
IRENA GROSFELD,
ALEXANDER RODNYANSKY,
AND EKATERINA ZHURAVSKAYA
Persistent Antimarket Culture:
A Legacy of the Pale of Settlement
after the Holocaust
www.vanderbilt.edu/AEA/
NG 014
I
M 2
CO ARY
NU
A
FASTER
J
ACCESS TO
THE LATEST RESEARCH
The American
Economic Review
ARTICLES
DAVID H. AUTOR AND DAVID DORN
The Growth of Low-Skill Service Jobs and the Polarization of the US Labor Market
YING FAN
Ownership Consolidation and Product Characteristics: A Study of the US Daily
Newspaper Market
DAVID ATKIN
Trade, Tastes, and Nutrition in India
MARK AGUIAR, ERIK HURST, AND LOUKAS KARABARBOUNIS
Time Use During the Great Recession
VASCO CARVALHO AND XAVIER GABAIX
The Great Diversification and its Undoing
SULEYMAN BASAK AND ANNA PAVLOVA
Asset Prices and Institutional Investors
FERNANDA BROLLO, TOMMASO NANNICINI, ROBERTO PEROTTI, AND GUIDO TABELLINI
The Political Resource Curse
NICOLE MAESTAS, KATHLEEN J. MULLEN, AND ALEXANDER STRAND
Does Disability Insurance Receipt Discourage Work?
Using Examiner Assignment to Estimate Causal Effects of SSDI Receipt
JAMES W. ROBERTS AND ANDREW SWEETING
When Should Sellers Use Auctions?
PRASHANT BHARADWAJ, KATRINE VELLESEN LKEN, AND CHRISTOPHER NEILSON
Early Life Health Interventions and Academic Achievement
HENRIK JACOBSEN KLEVEN, CAMILLE LANDAIS, AND EMMANUEL SAEZ
Taxation and International Migration of Superstars: Evidence from the European
Football Market
12
www.vanderbilt.edu/AEA/
Founded in 1885
EXECUTIVE COMMITTEE
Elected Officers and Members
President
CLAUDIA GOLDIN, Harvard University
President-elect
WILLIAM D. NORDHAUS, Yale University
Vice Presidents
RAQUEL FERNNDEZ, New York University
PAUL R. MILGROM, Stanford University
Members
MONIKA PIAZZESI, Stanford University
MICHAEL WOODFORD, Columbia University
ANIL K KASHYAP, University of Chicago
ROSA L. MATZKIN, University of California at
Los Angeles
AMY N. FINKELSTEIN, Massachusetts Institute of
Technology
JONATHAN LEVIN, Stanford University
Ex Officio Members
CHRISTOPHER A. SIMS, Princeton University
ORLEY C. ASHENFELTER, Princeton University
Appointed Members
Editor, The American Economic Review
PINELOPI KOUJIANOU GOLDBERG, Yale University
Editor, The Journal of Economic Literature
STEVEN N. DURLAUF, University of Wisconsin,
Madison
Editor, The Journal of Economic Perspectives
DAVID H. AUTOR, Massachusetts Institute of
Technology
Editor, American Economic Journal: Applied Economics
ESTHER DUFLO, Massachusetts Institute of Technology
Editor, American Economic Journal: Economic Policy
ALAN J. AUERBACH, University of California at
Berkeley
Editor, American Economic Journal: Macroeconomics
JOHN LEAHY, New York University
Editor, American Economic Journal: Microeconomics
ANDREW POSTLEWAITE, University of Pennsylvania
Secretary-Treasurer
PETER L. ROUSSEAU, Vanderbilt University
OTHER OFFICERS
Editor, Resources for Economists
WILLIAM GOFFE, State University of New York
at Oswego
Director of AEA Publication Services
JANE EMILY VOROS, Pittsburgh
Managing Director of EconLit Product Design and Content
STEVEN L. HUSTED, University of Pittsburgh
Assistant Secretary-Treasurer
JOHN J. SIEGFRIED, Vanderbilt University
and University of Adelaide
Counsel
TERRY CALVANI, Freshfields Bruckhaus Deringer LLP
Washington, DC
ADMINISTRATORS
Administrative Director
REGINA H. MONTGOMERY
Convention Manager
MARLENE HIGHT
Economic Perspectives
Fall 2013, Volume 27, Number 4
Articles
Economic Perspectives
A journal of the
American Economic Association
The Journal of
Fall 2013
Perspectives
Symposia
The First 100 Years of the Federal Reserve
The Journal of
Fall 2013