The Most Dangerous Idea in Federal Reserve History ...

American Economic Review: Papers & Proceedings 2013, 103(3): 55–60
http://dx.doi.org/10.1257/aer.103.3.55

The Most Dangerous Idea in Federal Reserve History:
Monetary Policy Doesn’t Matter †

By Christina D. Romer and David H. Romer*

The hundredth anniversary of the founding of both the mid- and late 1970s, faced with high
the Federal Reserve is a natural time to reflect inflation, policymakers believed that monetary
on the record of US monetary policy. It is widely policy could not reduce inflation at any reason-
agreed that this record is far from perfect, and able cost. And there is evidence that in the past
that there have been some major failures of mon- few years, faced with high unemployment and a
etary policy over the past century. Our thesis is weak recovery, monetary policymakers believed
that overly pessimistic views about the power of that policy was relatively weak and potentially
monetary policy have been a critical source of costly. In each episode, these beliefs led to a
these failures. marked passivity in policymaking.

There is little doubt that the opposite prob- The next three sections discuss the link
lem—an overinflated belief in the power of mon- between pessimistic beliefs and policy inaction
etary policy—has also contributed to important in the 1930s, the 1970s, and the past few years,
policy errors. Most famously, policymakers in respectively. The final section concludes by
the mid-1960s believed that they faced a long- arguing that being a good central banker appears
run inflation-unemployment trade-off, and thus to require a balance of humility and hubris.
that monetary policy could move the economy
to a sustained path of very low unemployment I.  The 1930s
and low inflation. This belief led them to pursue The most significant error in the history of the
highly expansionary policy, starting the econ- Federal Reserve surely occurred in 1929–1933,
omy down the road to the inflation of the 1970s when the money stock fell 26 percent, the price
(for example, Romer and Romer 2002 and level declined 25 percent, and output decreased
Primiceri 2006). The record of such errors has 27 percent. There is vast evidence that an overly
led some to argue that perhaps the most impor- pessimistic assessment of the power of mone-
tant attribute of a successful central banker is tary policy to combat the downturn was a criti-
humility (for example, Booth 2012). cal source of this error (Friedman and Schwartz
1963; Meltzer 2003; and many others). Many
In this paper, we present evidence that an Federal Reserve officials believed that expan-
unduly pessimistic view of what monetary sionary policy would not be effective and that it
policy can accomplish has been a more impor- might involve substantial costs. The result was
tant source of policy errors and poor outcomes inaction in the face of the largest downturn in
over the history of the Federal Reserve. At vari- American history.
ous times in the 1930s, faced with the Great One early episode showing monetary policy­
Depression, Federal Reserve officials believed makers’ pessimism about what they could
that the power of monetary policy to combat the accomplish occurred in the summer of 1930,
downturn or stimulate recovery was minimal. In when the Federal Reserve Bank of New York
proposed expansionary actions. New York’s pro-
* C. Romer: Department of Economics, University of posal was opposed by most of the other Federal
California, Berkeley, CA 94720 (e-mail: cromer@econ. Reserve banks, and so little was done.
berkeley.edu); D.  Romer: Department of Economics, The opponents of expansion proffered two
University of California, Berkeley, CA 94720 (e-mail: main arguments that it would be ineffective.
[email protected]). We thank Donald Kohn for First, and crucially, the main indicators of
helpful comments. the stance of policy that they used—nominal

† To view additional materials, and author disclosure
statement(s),visit the article page at
http://dx.doi.org/10.1257/aer.103.3.55.

55

56 AEA PAPERS AND PROCEEDINGS MAY 2013

i­nterest rates, banks’ excess reserves, and bor- document the reasons that George Harrison
rowing from the Federal Reserve—implied (governor of the Federal Reserve Bank of New
that policy was already highly expansionary. York, and one of the architects of the program)
They therefore thought that monetary policy gave for the decision:
had done all it could. For example, one oppo-
nent argued, “With credit cheap and redundant When the figures of member bank reserves are
we do not believe that business recovery will be sufficiently high … , we shall probably have
accelerated by making credit cheaper and more done our part. If the commercial banks can’t or
redundant.”1 Another referred to “the fruitless- don’t use the credit which we provide, that is
ness and unwisdom of attempting to depress another problem.
still further the abnormally low interest rates It was thought best … not to use our ammu-
now prevailing.” Second, they believed that the nition until the chances of effective response
cause of the downturn was not monetary but lay from the banking and business community
in excesses in the 1920s, and thus that the down- would favor the success of our undertaking.
turn could not be solved by monetary policy.
One policymaker said, These ideas persisted into the recovery. For
example, the expression that at some point fur-
The consequences of … an economic debauch ther monetary easing is ineffective because “one
are inevitable. We are now suffering them. cannot push a string” appears to have originated
in Congressional testimony in 1935 by Marriner
Can they be corrected or removed by Eccles, the governor (that is, head) of the Federal
cheap money? We do not believe that they Reserve Board. Similarly, in 1937, the Federal
can. … [T]here is no short cut or panacea for Open Market Committee (FOMC) believed that
the rectification of existing conditions. “the existing volume of excess reserves and of
supplies of private capital is abundant at this
Policymakers also saw two costs to expan- time at low rates,” and therefore that “effective
sion, related to the two reasons they viewed action to meet and overcome the present busi-
expansion as unproductive. First, they believed ness recession should be taken outside the field
that an expansion that had little impact would of the System’s various monetary powers.”
damage their credibility, and so make later
expansion less effective. As one put it, In addition, the view that monetary expansion
could lead to inflation even when the economy
[With] an abundance of funds in the market, … it was far below capacity took on special impor-
should be the policy of the Federal Reserve tance in the mid-1930s. Policymakers were con-
System to maintain a position of strength, in cerned that expansion “might well add unwise
readiness to meet future demands, as and when stimulus to the inflation of prices” and that “a
they arise, rather than to put reserve funds into further increase in excess reserves of member
the market when not needed. banks might give added impetus to existing
Second, they feared that expansion could trigger inflationary tendencies.”
renewed speculation and inflation. For example,
one bank governor said, “Cheap money is a Consistent with these beliefs, the Federal
stimulant, … but a headache will follow if the Reserve was largely passive in the recovery, just
dose is large enough, and persisted in. It encour- as it had been during the downturn. The mone-
ages over-borrowing.” tary base rose rapidly during much of this period,
These beliefs prevented significant action but the increases were almost entirely the result
not just in 1930, but throughout the downturn. of gold inflows and the Treasury’s decision not
Consider, for example, the decision to end a brief to sterilize them, rather than of Federal Reserve
period of expansionary open-market operations actions.
in 1932. Hsieh and Romer (2006, pp. 169–72)
The Federal Reserve’s major policy initia-
1 The sources for all the quotations and data used in the tive in this period—the doubling of reserve
paper are given in the online Appendix.  requirements in 1936–1937 and working with
the Treasury to sterilize gold inflows at the
same time—was motivated by fear of inflation
in a still-depressed economy. Policymakers

VOL. 103 NO. 3 the most dangerous idea in federal reserve history 57

believed that banks’ excess reserves could “cre- conventional monetary policy to combat infla-
ate an injurious credit expansion,” and therefore tion. For example, in May 1971, the economist
“decided to lock up this part of the present vol- making the official staff presentation to the
ume of member bank reserves as a measure of FOMC said,
prevention.”2
The question is whether monetary policy could
II. The 1970s or should do anything to combat a persist-
ing residual rate of inflation … . The answer,
Another major failure of Federal Reserve I think, is negative. … It seems to me that we
policy occurred in the late 1960s and the 1970s, should regard continuing cost increases as
when inflation rose erratically from low levels to a structural problem not amenable to macro-
near 10 percent. In two parts of this era, Federal economic measures.
Reserve officials believed that inflation was very
unresponsive to economic slack, and thus that The belief that monetary policy would not be
monetary policy was an extremely ineffective effective in controlling inflation caused policy-
way to fight it.3 makers to advocate incomes policies, such as
wage and price controls, instead. For example,
The first part of the era when this view pre- in June 1971, Burns testified,
vailed was roughly from 1971 to 1973. After
inflation failed to fall in the mild recession of [A] substantial increase of unemployment has
1969–1970, Federal Reserve chairman Arthur failed to check the rapidity of wage advances
Burns and other policymakers concluded not or to moderate appreciably the rise of the gen-
that the natural rate was higher than they had eral price level.
previously believed, but that inflation was
almost impervious to high unemployment. With increasing conviction, I have there-
Federal Reserve documents record that in June fore come to believe that our Nation must
1971, Burns expressed the view that: supplement monetary and fiscal policy with
specific policies to moderate wage and price
[O]f late one found that at a time when unem- increases.
ployment was increasing prices continued to
advance at an undiminished pace and wages At the FOMC meeting the same month, Burns’s
rose at an increasing pace. … views were summarized as:

In his judgment a much higher rate of He thought the Administration had been much
unemployment produced by monetary policy too slow to recognize the need for an effec-
would not moderate [wage-cost] pressures tive incomes policy. He had urged that action
appreciably. be taken in that area and intended to continue
doing so.
In July, he testified that “even a long stretch of The second part of the 1970s when beliefs
high and rising unemployment may not suffice about the ineffectiveness of policy were preva-
to check the inflationary process.” lent occurred under the chairmanship of G.
William Miller in 1978 and 1979. Shortly after
As discussed by Romer and Romer (2002), becoming chairman, Miller testified,
these views led the Federal Reserve to not use Our attempts to restrain inflation by using con-
ventional stabilization techniques have been
2 Of course, the pessimistic views we have described less than satisfactory. Three years of high
were not the only source of the policy failures in the 1930s. unemployment and underutilized capital stock
Meltzer and Friedman and Schwartz show how views about have been costly in terms both of lost produc-
the proper role of monetary policy, including the importance tion and of the denial to many of the dignity
of defending the gold standard and of meeting credit demand that comes from holding a productive job.
rather than promoting macroeconomic stability, had impor- Yet, despite this period of substantial slack in
tant effects on policy. In addition, Friedman and Schwartz the economy, we still have a serious inflation
document how the fractured power structure of the Federal problem.
Reserve in the first part of the decade favored inaction over
action. 

3 Nelson (2005) documents similar beliefs in the United
Kingdom in this era. 

58 AEA PAPERS AND PROCEEDINGS MAY 2013

Other policymakers held similar views. For remained very high and has consistently been
example, in May 1979, Governor Henry Wallich, projected to remain so for years. And in contrast
generally regarded as one of the most antiinfla- to the 1970s and early 1980s—but similar to the
tionary FOMC members, said, “We also have 1930s—the high unemployment has occurred at
evidence that inflation in the American economy a time of low inflation, with core inflation and
is much less variable than it is in other countries the Federal Reserve’s inflation forecasts gener-
and is, therefore, much harder to bring down.” ally below its inflation target.
And at Miller’s final meeting in July 1979, the
staff presentation stated, “we expect that rising It is clearly too soon to reach firm conclu-
unemployment will do little to damp inflation,” sions about recent monetary policy. Much of
and that “[f]or monetary policy alone there seems the record of policymakers’ thinking is not yet
to be little in the way of policy options which available. More importantly, there has not been
would yield substantially improved results dur- enough time to confidently assess what monetary
ing the next year or two.” During the discussion, policy could and could not have accomplished.
the economist in charge of the presentation listed
several reasons that “we wouldn’t expect to get Nonetheless, it seems hard to assign pes-
the same price response from very weak mar- simism about the power of monetary policy a
kets” as had occurred just a few years before. large role in the crisis itself. Before the crisis,
monetary policymakers appear to have believed
This humility about their powers again caused that they would be able to largely counteract the
monetary policymakers to not pursue antiinfla- macroeconomic effects of a large fall in house
tionary policy, but instead to continue to stimu- prices. And during the crisis, they believed they
late the economy (Romer and Romer 2002). had the ability to prevent a collapse of the finan-
Miller testified in March 1979, “Real interest cial system and acted aggressively to do so.
rates … still appear to remain low by histori-
cal standards and thus continue to facilitate an There are, however, intriguing parallels
expansion of overall demands.” between policymakers’ beliefs in the period
from roughly the end of the recession to the
These views also led monetary policymakers latter half of 2012 and beliefs in the 1930s and
to again advocate nonmonetary steps to combat 1970s. Monetary policymakers in each period
inflation. At his first FOMC meeting in March have to some extent believed that their tools
1978, Miller argued that monetary policy was were not very effective and potentially costly.
not the best way to fight inflation, saying that
if the administration did not “take some more In the recent period, the strongest views of this
believable steps in fighting inflation … , inflation type have been among some of the presidents
is going to be left to the Federal Reserve and of the regional Federal Reserve banks. Indeed,
that’s going to be bad news.” The official sum- at times some have expressed views similar to
mary of the meeting said, “It was noted that an ones from the 1930s. For example, one argued
effective program to reduce the rate of inflation against additional action on the grounds that,
had to extend beyond monetary policy.” That “Why would the Fed provision to shovel bil-
same month, Miller testified that conventional lions in additional liquidity into the economy’s
policies “need to be complemented by programs boiler when so much is presently lying fallow?”
designed to enhance competition and to correct Another argued that “a zero-rate policy increases
structural problems.” the risk of misallocating real resources, creating
a new set of imbalances or possibly a new set
Thus, in both the 1930s and the 1970s, undue of bubbles.” A third argued that “the supply of
pessimism about what monetary policy could do bank reserves is already large enough to support
led to Federal Reserve inaction and highly unde- the economic recovery,” and that “further mon-
sirable economic outcomes. etary stimulus runs the risk of raising inflation
in a way that threatens the stability of inflation
III.  The Past Few Years expectations.”
The last several years have been another time
of dismal macroeconomic performance. The In addition, some bank presidents have attrib-
economy suffered its largest postwar recession uted high unemployment to structural prob-
in 2007–2009. Since then, unemployment has lems and have therefore doubted the ability of
monetary policy to reduce it without triggering
inflation. For example, one stated, “Most of the
existing unemployment represents mismatch

VOL. 103 NO. 3 the most dangerous idea in federal reserve history 59

that is not readily amenable to monetary policy.” entail “costs as well as …  benefits,” and went on
Another said: to detail the costs he perceived.

You can’t change the carpenter into a nurse Policymakers have been explicit that these
easily … . Eventually … [p]eople will be re- considerations have muted their policy response.
trained and they’ll find jobs in other indus- In October 2012, Bernanke said that “the Federal
tries. But monetary policy can’t retrain people. Reserve has generally employed a high hurdle
Monetary policy can’t fix those problems. for using” nontraditional tools. In April 2012,
Yellen said, “The FOMC’s unconventional pol-
Among the leading figures on the FOMC— icy actions … , in my judgment, have not entirely
chairman Ben Bernanke, vice-chair Janet compensated for the zero-bound constraint.”
Yellen, and president of the Federal Reserve These statements are consistent with the fact
Bank of New York William Dudley—the view that Federal Reserve policy in recent years has
that monetary policy tools are not very effective been less aggressive than some analysts have
and potentially costly has been milder and more urged (for example, Gagnon 2009).
nuanced, relative both to the views described
above and to those in the 1930s and 1970s. Another parallel with the earlier periods—
Nonetheless, there is evidence that it has been particularly the 1970s—is that concern about the
present. It appears to have had two key elements. effectiveness of their tools has led monetary pol-
icymakers to advocate nonmonetary measures.
One is that the power of the tools is limited. In September 2012, after saying that monetary
The language that these monetary policy­makers policy “is not a panacea” for addressing tight
have used to describe what their tools could financial conditions and high unemployment,
accomplish has consistently been measured. Bernanke said, “We’re looking for policymakers
In October 2012, for example, Bernanke said in other areas to do their part.” Using very similar
that “we expect our policies to provide mean- language in November 2011, Yellen elaborated
ingful help to the economy,” but that “mon- on her view that monetary policy alone could
etary policy is not a panacea” for “tackl[ing],” not solve an aggregate demand shortfall by say-
among other things, “the near-term shortfall ing that “it is essential for other policymakers
in aggregate demand.” Similarly, in November to also do their part.” And in January 2012, the
2011, after identifying “a dearth of aggregate Federal Reserve sent Congressional leaders an
demand,” Yellen also said that “monetary pol- unsolicited white paper discussing “current con-
icy is not a panacea.” The same month, Dudley ditions and policy considerations” concerning
said, “although a stimulative monetary policy is the housing market and housing policy.
essential for recovery, it may not be sufficient.”
Thus, concern about the power of policy
The second element has been the belief has limited the Federal Reserve’s response to
that the tools involve costs. Probably the most the very weak economy. Whether that concern
explicit statement of this view was made by has reflected unwarranted pessimism or a wise
Bernanke in August 2012. He listed four poten- assessment will not be known for many years,
tial costs to nontraditional policies: they “could if ever.
impair the functioning of securities markets,”
“reduce public confidence in the Fed’s ability Two pieces of evidence, however, are at least
to exit smoothly from its accommodative poli- suggestive of unwarranted pessimism. The first
cies,” create “risks to financial stability,” and is the analogy to the Depression. Then, as in the
cause “the possibility that the Federal Reserve past few years, nominal interest rates were very
could incur financial losses.”4 Similarly, Dudley low, and many attributed poor economic condi-
said in November 2011 that nontraditional tools tions to a speculative boom and bust rather than
to monetary causes. Yet the modern consensus is
4 Bernanke’s conclusion was that “the costs of nontra- that the beliefs that monetary expansion would
ditional tools, when considered carefully, appear manage- be ineffective and potentially costly were mis-
able”; and, as we discuss below, his speech came shortly taken. Second, the Federal Reserve’s decision
before a decision by the FOMC to use the tools more force- in September 2012 that it was appropriate to
fully. Nonetheless, the speech provides an unusually clear use its tools more aggressively, even though its
discussion of the costs that policymakers perceived.  economic outlook had improved since the previ-
ous meetings, suggests that policymakers may
now think they had been underestimating the

60 AEA PAPERS AND PROCEEDINGS MAY 2013

­effectiveness of the tools, or overestimating their One possible conclusion is that central bank-
costs.5 But both pieces of evidence are clearly ers should have a balance of humility and hubris.
far from definitive. They need a sound knowledge of both the limita-
tions and the powers of monetary policy. That is,
IV. Conclusion the most important characteristic to look for in
The view that hubris can cause central bank- central bankers is not their inherent optimism or
ers to do great harm clearly has an important ele- pessimism about the effectiveness of monetary
ment of truth. A belief that monetary policy can policy, but rather their understanding of how the
achieve something it cannot—such as stable low economy works and the possible contributions
inflation together with below-normal unemploy- of policy.
ment—can lead to the pursuit of reckless poli-
cies that do considerable damage. REFERENCES
But the hundred years of Federal Reserve
history show that humility can also cause large Booth, Philip. 2012. “The Bank of England Needs
harms. In the 1930s, excessive pessimism about a Humble Governor.” Institute of Economic
the power of monetary policy and about its Analysis, October 17.
potential costs caused monetary policymak-
ers to do little to combat the Great Depression Friedman, Milton, and Anna Jacobson Schwartz.
or promote recovery. In critical periods in the 1963. A Monetary History of the United States,
1970s, undue pessimism about the potential of 1867–1960. Princeton: Princeton University
contractionary monetary policy to reduce infla- Press.
tion led policymakers to do little to rein in the
Great Inflation. We have stressed that it is too Gagnon, Joseph E. 2009. “The World Needs Fur-
soon to reach conclusions about recent develop- ther Monetary Ease, Not an Early Exit.” Peter-
ments. But, faced with persistent high unem- son Institute for International Economics
ployment and below-target inflation, beliefs Policy Brief No. PB09-22.
that the benefits of expansion are small and the
costs potentially large appear to have led mon- Hsieh, Chang-Tai, and Christina D. Romer. 2006.
etary policymakers to eschew more aggressive “Was the Federal Reserve Constrained by the
expansionary policy in much of 2010 and 2011. Gold Standard During the Great Depression?
In hindsight, these beliefs may be judged too Evidence from the 1932 Open Market Pur-
pessimistic. chase Program.” Journal of Economic History
The approaches of two largely success- 66 (1): 140–76.
ful Federal Reserve chairmen—William
McChesney Martin and Paul Volcker—also sug- Meltzer, Allan H. 2003. A History of the Federal
gest that the value of humility in a central banker Reserve. Volume 1: 1913–1951. Chicago: Uni-
may be overstated. Both came into office believ- versity of Chicago Press.
ing that monetary policy could accomplish a
great deal, and both used policy aggressively. For Nelson, Edward. 2005. “The Great Inflation of the
example, Volcker undertook a highly successful Seventies: What Really Happened?” Advances
disinflation program because, as he stated at his in Macroeconomics 5 (1): Article 3.
confirmation hearings, “I don’t think we have
any substitute for seeking an answer to our prob- Primiceri, Giorgio E. 2006. “Why Inflation Rose
lems in the context of monetary discipline.” and Fell: Policy-Makers’ Beliefs and U.S.
Postwar Stabilization Policy.” Quarterly Jour-
nal of Economics 121 (3): 867–901.

Romer, Christina D., and David H. Romer. 2002.
“The Evolution of Economic Understanding
and Postwar Stabilization Policy.” In Rethink-
ing Stabilization Policy, 11–78. Kansas City:
Federal Reserve Bank of Kansas City.

5 For example, the Summary of Economic Projections in
September 2012, when the FOMC decided to make greater
use of the tools, involved a considerably lower level of
unemployment, and a similar rate of decline in unemploy-
ment and a similar path for inflation, than the projections
in November 2011, when it decided to take no substantial
new action.