CEA -- Chair's Remarks -- 03092009

Lessons from the Great Depression for Economic Recovery in 2009

Christina D. Romer
Council of Economic Advisers

As prepared for presentation at the Brookings Institution,
Washington, D.C., March 9, 2009

Click here to download printable PDF

In the last few months, I have found myself uttering the words "worst since the Great Depression" far too often: the worst twelve month job loss since the Great Depression; the worst financial crisis since the Great Depression; the worst rise in home foreclosures since the Great Depression. In my previous life, as an economic historian at Berkeley, one of the things I studied was the Great Depression. I thought it would be useful to reflect on that episode and what lessons it holds for policymakers today. In particular, what can we learn from the 1930s that will help us to end the worst recession since the Great Depression?

To start, let me point out that though the current recession is unquestionably severe, it pales in comparison with what our parents and grandparents experienced in the 1930s. Last Friday’s employment report showed that unemployment in the United States has reached 8.1%—a terrible number that signifies a devastating tragedy for millions of American families. But, at its worst, unemployment in the 1930s reached nearly 25%.1 And, that quarter of American workers had painfully few of the social safety nets that today help families maintain at least the essentials of life during unemployment. Likewise, following last month’s revision of the GDP statistics, we know that real GDP has declined almost 2% from its peak. But, between the peak in 1929 and the trough of the great Depression in 1933, real GDP fell over 25%.2

I don’t give these comparisons to minimize the pain the United States economy is experiencing today, but to provide some crucial perspective. Perhaps it is the historian and the daughter in me that finds it important to pay tribute to just what truly horrific conditions the previous generation of Americans endured and eventually triumphed over. And, it is the new policymaker in me that wants to be very clear that we are doing all that we can to make sure that the word "great" never applies to the current downturn.

While what we are experiencing is less severe than the Great Depression, there are parallels that make it a useful point of comparison and a source for learning about policy responses today. Most obviously, like the Great Depression, today’s downturn had its fundamental cause in the decline in asset prices and the failure or near-failure of financial institutions. In 1929, the collapse and extreme volatility of stock prices led consumers and firms to simply stop spending.3 In the recent episode, the collapse of housing prices and stock prices has reduced wealth and shaken confidence, and led to sharp rises in the saving rate as consumers have hunkered down in the face of greatly reduced and much more uncertain wealth.

In the 1930s, the collapse of production and wealth led to bankruptcies and the disappearance of nearly half of American financial institutions.4 This, in turn, had two devastating consequences: a collapse of the money supply, as stressed by Milton Friedman and Anna Schwartz, and a collapse in lending, as stressed by Ben Bernanke.5 In the current episode, modern innovations such as derivatives led to a direct relationship between asset prices and severe stress in financial institutions. Over the fall, we saw credit dry up and learned just how crucial lending is to the effective functioning of American businesses and households.

Another parallel is the worldwide nature of the decline. A key feature of the Great Depression was that virtually every industrial country experienced a severe contraction in production and a terrible rise in unemployment.6 This past year, there was hope that the current downturn might be mainly an American experience, and so world demand could remain high and perhaps help pull us through. However, during the past few months, we have realized that this hope was a false one. As statistics have poured in, we have learned that Europe, Asia, and many other areas are facing declines as large, if not larger, than our own. Indeed, rather than world demand helping to hold us up, the fall in U.S. demand has had a devastating impact on export economies such as Taiwan, China, and South Korea.

This similarity of causes between the Depression and today’s recession means that President Obama begins his presidency and his drive for recovery with many of the same challenges that Franklin Roosevelt faced in 1933. Our consumers and businesses are in no mood to spend or invest; our financial institutions are severely strained and hesitant to lend; short-term interest rates are effectively zero, leaving little room for conventional monetary policy; and world demand provides little hope for lifting the economy. Yet, the United States did recover from the Great Depression. What lessons can modern policymakers learn from that episode that could help them make the recovery faster and stronger today?

One crucial lesson from the 1930s is that a small fiscal expansion has only small effects. I wrote a paper in 1992 that said that fiscal policy was not the key engine of recovery in the Depression.7 From this, some have concluded that I do not believe fiscal policy can work today or could have worked in the 1930s. Nothing could be farther than the truth. My argument paralleled E. Cary Brown’s famous conclusion that in the Great Depression, fiscal policy failed to generate recovery "not because it does not work, but because it was not tried."8

The key fact is that while Roosevelt's fiscal actions were a bold break from the past, they were nevertheless small relative to the size of the problem. When Roosevelt took office in 1933, real GDP was more than 30% below its normal trend level. (For comparison, the U.S. economy is currently estimated to be between 5 and 10% below trend.)9 The emergency spending that Roosevelt did was precedent-breaking—balanced budgets had certainly been the norm up to that point. But, it was quite small. The deficit rose by about one and a half percent of GDP in 1934.10 One reason the rise wasn't larger was that a large tax increase had been passed at the end of the Hoover administration. Another key fact is that fiscal expansion was not sustained. The deficit declined in fiscal 1935 by roughly the same amount that it had risen in 1934. Roosevelt also experienced the same inherently procyclical behavior of state and local fiscal actions that President Obama is facing. Because of balanced budget requirements, state and local governments are forced to cut spending and raise tax rates when economic activity declines and state tax revenues fall. At the same time that Roosevelt was running unprecedented federal deficits, state and local governments were switching to running surpluses to get their fiscal houses in order.11 The result was that the total fiscal expansion in the 1930s was very small indeed. As a result, it could only have a modest direct impact on the state of the economy.

This is a lesson the Administration has taken to heart. The American Recovery and Reinvestment Act, passed less than thirty days after the Inauguration, is simply the biggest and boldest countercyclical fiscal action in history. The nearly $800 billion fiscal stimulus is roughly equally divided between tax cuts, direct government investment spending, and aid to the states and people directly hurt by the recession. The fiscal stimulus is close to 3% of GDP in each of the next two years. And, as I mentioned, a good chunk of this stimulus takes the form of fiscal relief to state governments, so that they do not have to balance their budgets only by such measures as raising taxes and cutting the employment of nurses, teachers, and first responders. We expect this fiscal expansion to be extremely important to countering the terrible job loss that last Friday's numbers show now totals 4.4 million since the recession began fourteen months ago.

While the direct effects of fiscal stimulus were small in the Great Depression, I think it is important to acknowledge that there may have been an indirect effect. Roosevelt's very act of doing something must have come as a great relief to a country that had been suffering depression for more than three years. To have a President step up to the challenge and say the country would attack the Depression with the same fervor and strength it would an invading army surely lessened uncertainty and calmed fears. Also, signature programs such as the WPA that directly hired millions of workers no doubt contributed to a sense of progress and control. In this way, Roosevelt’s actions may have been more beneficial than the usual estimates of fiscal policy suggest. If the actions President Obama is taking in the current downturn can generate the same kind of confidence effects, they may also be more effective than estimates based on conventional multipliers would lead one to believe.

A second key lesson from the 1930s is that monetary expansion can help to heal an economy even when interest rates are near zero. In the same paper where I said fiscal policy was not key in the recovery from the Great Depression, I argued that monetary expansion was very useful. But, the monetary expansion took a surprising form: it was essentially a policy of quantitative easing conducted by the U.S. Treasury.12

The United States was on a gold standard throughout the Depression. Part of the explanation for why the Federal Reserve did so little to counter the financial panics and economic decline was that it was fighting to defend the gold standard and maintain the prevailing fixed exchange rate.13 In April 1933, Roosevelt temporarily suspended the convertibility to gold and let the dollar depreciate substantially. When we went back on gold at the new higher price, large quantities of gold flowed into the U.S. Treasury from abroad. These gold inflows serendipitously continued throughout the mid-1930s, as political tensions mounted in Europe and investors sought the safety of U.S. assets.

Under a gold standard, the Treasury could increase the money supply without going through the Federal Reserve. It was allowed to issue gold certificates, which were interchangeable with Federal Reserve notes, on the basis of the gold it held. When gold flowed in, the Treasury issued more notes. The result was that the money supply, defined narrowly as currency and reserves, grew by nearly 17% per year between 1933 and 1936.14

This monetary expansion couldn’t lower nominal interest rates because they were already near zero. What it could do was break expectations of deflation. Prices had fallen 25% between 1929 and 1933.15 People throughout the economy expected this deflation to continue. As a result, the real cost of borrowing and investing was exceedingly high. Consumers and businesses wanted to sit on any cash they had because they expected its real purchasing power to increase as prices fell. Devaluation followed by rapid monetary expansion broke this deflationary spiral. Expectations of rapid deflation were replaced by expectations of price stability or even some inflation. This change in expectations brought real interest rates down dramatically.16

The change in the real cost of borrowing and investing appears to have had a beneficial impact on consumer and firm behavior. The first thing that turned around was interest-sensitive spending. For example, car sales surged in the summer of 1933.17 One sign that lower real interest rates were crucial is that real fixed investment and consumer spending on durables both rose dramatically between 1933 and 1934, while consumer spending on services barely budged.18

In thinking about the lessons from the Great Depression for today, I want to tread very carefully. A key rule of my current job is that I do not comment on Federal Reserve policy. So, let me be very clear – I am not advocating going on a gold standard just so we can go off it again, or that Tim Geithner should start conducting rogue monetary policy. But the experience of the 1930s does suggest that monetary policy can continue to have an important role to play even when interest rates are low by affecting expectations, and in particular, by preventing expectations of deflation.

This discussion of fiscal and monetary policy in the 1930s leads me to a third lesson from the Great Depression: beware of cutting back on stimulus too soon.

As I have just described, monetary policy was very expansionary in the mid-1930s. Fiscal policy, though less expansionary, was also helpful. Indeed, in 1936 it was inadvertently stimulatory. Largely because of political pressures, Congress overrode Roosevelt’s veto and gave World War I veterans a large bonus. This caused another one-time rise in the deficit of more than 1½% of GDP.

And, the economy responded. Growth was very rapid in the mid-1930s. Real GDP increased 11% in 1934, 9% in 1935, and 13% in 1936. Because the economy was beginning at such a low level, even these growth rates were not enough to bring it all the way back to normal. Industrial production finally surpassed its July 1929 peak in December 1936, but was still well below the level predicted by the pre-Depression trend.19 Unemployment had fallen by close to 10 percentage points—but was still over 15%. The economy was on the road to recovery, but still precarious and not yet at a point where private demand was ready to carry the full load of generating growth.

In this fragile environment, fiscal policy turned sharply contractionary. The one-time veterans’ bonus ended, and Social Security taxes were collected for the first time in 1937. As a result, the deficit was reduced by roughly 2½% of GDP.

Monetary policy also turned inadvertently contractionary. The Federal Reserve was becoming increasingly concerned about inflation in 1936. It was also concerned that, because banks were holding such large quantities of excess reserves, open-market operations would merely cause banks to substitute government bonds for excess reserves and would have no impact on lending. In an effort to put themselves in a position where they could tighten if they needed to, the Federal Reserve doubled reserve requirements in three steps in 1936 and 1937. Unfortunately, banks, shaken by the bank runs of just a few years before, scrambled to build reserves above the new higher required levels. As a result, interest rates rose and lending plummeted.20

The results of the fiscal and monetary double whammy in the precarious environment were disastrous. GDP rose by only 5% in 1937 and then fell by 3% in 1938, and unemployment rose dramatically, reaching 19% in 1938. Policymakers soon reversed course and the strong recovery resumed, but taking the wrong turn in 1937 effectively added two years to the Depression.

The 1937 episode is an important cautionary tale for modern policymakers. At some point, recovery will take on a life of its own, as rising output generates rising investment and inventory demand through accelerator effects, and confidence and optimism replace caution and pessimism. But, we will need to monitor the economy closely to be sure that the private sector is back in the saddle before government takes away its crucial lifeline.21

The fourth lesson we can draw from the recovery of the 1930s is that financial recovery and real recovery go together. When Roosevelt took office, his immediate actions were largely focused on stabilizing a collapsing financial system. He declared a national Bank Holiday two days after his inauguration, effectively shutting every bank in the country for a week while the books were checked. This 1930s version of a "stress test" led to the permanent closure of more than 10% of the nation’s banks, but improved confidence in the ones that remained.22 As I discussed before, Roosevelt temporarily suspended the gold standard, before going back on gold at a lower value for the dollar, paving the way for increases in the money supply. In June 1933, Congress passed legislation establishing deposit insurance through the FDIC and helping homeowners through the Home Owners Loan Corporation.23 The actual rehabilitation of financial institutions, obviously took much longer. Indeed, much of the hard work of recapitalizing banks and dealing with distressed homeowners and farmers was spread out over 1934 and 1935.

Nevertheless, the immediate actions to stabilize the financial system had dramatic short-run effects on financial markets. Real stock prices rose over 40% from March to May 1933, commodity prices soared, and interest-rate spreads shrank.24 And, the actions surely contributed to the economy’s rapid growth after 1933, as wealth rose, confidence improved, and bank failures and home foreclosures declined.

But, it was only after the real recovery was well established that the financial recovery took firm hold. Real stock prices in March 1935 were more than 10% lower than in May 1933; bank lending continued falling until mid-1935; and real house prices rose only 7% from 1933 to 1935.25 The strengthening real economy improved the health of the financial system. Bank profits moved from large and negative in 1933 to large and positive in 1935, and remained high through the end of the Depression, with the result that bank suspensions were minimal after 1933. Real stock prices rose robustly. Business failures and home foreclosures fell sharply and almost without interruption after 1932.26 And, this virtuous cycle continued as the financial recovery led to further narrowing of interest-rate spreads and increased willingness of banks to lend.27

This lesson is another one that has been prominent in the minds of policymakers today. The Administration has from the beginning sought to create a comprehensive recovery program. The Financial Stabilization Plan, which involves evaluating the capital needs of financial institutions, as well as crucial programs to directly increase lending, is central to putting the financial system back to work for American industry and households. Together with the Administration’s housing plan, these financial rescue measures should provide the lending and stability needed for economic growth. The fiscal stimulus package was designed to create jobs quickly. In doing so, it should lower defaults and improve balance sheets so that our financial system can continue to strengthen.

The fifth lesson from the Great Depression is that worldwide expansionary policy shares the burdens and the benefits of recovery. Research by Barry Eichengreen and Jeffrey Sachs shows that going off the gold standard and increasing the domestic money supply was a key factor in generating recovery and growth across a wide range of countries in the 1930s.28 Importantly, these actions worked to lower world interest rates and benefit other countries, rather than to just shift expansion from one country to another.

The implications for today are obvious. The more that countries throughout the world can move toward monetary and fiscal expansion, the better off we all will be. In this regard, the aggressive fiscal action in China and the reduction in interest rates in Europe and the U.K. announced last week were welcome news. They are paving the way for a worldwide end to this worldwide recession.

The final lesson that I want to draw from the 1930s is perhaps the most crucial. A key feature of the Great Depression is that it did eventually end. Despite the devastating loss of wealth, chaos in our financial markets, and a loss of confidence so great that it nearly destroyed Americans’ fundamental faith in capitalism, the economy came back. Indeed, the growth between 1933 and 1937 was the highest we have ever experienced outside of wartime. Had the U.S. not had the terrible policy-induced setback in 1937, we, like most other countries in the world, would probably have been fully recovered before the outbreak of World War II.

This fact should give Americans hope. We are starting from a position far stronger than our parents and grandparents were in in 1933. And, the policy response has been fast, bold, and well-conceived. If we continue to heed the lessons of the Great Depression, there is every reason to believe that we will weather this trial and come through to the other side even stronger than before.

NOTES

1. Unemployment data for the 1930s are from Historical Statistics of the United States: Colonial Times to 1970 (Washington, D.C.: Government Printing Office, 1975), Part 1, p. 135, series D86.
2. Real GDP data are from the Bureau of Economic Analysis, http://www.bea.gov/national/nipaweb/SelectTable.asp?Selected=Y, Table 1.1.3.
3. Christina D. Romer, "The Great Crash and the Onset of the Great Depression," Quarterly Journal of Economics 105(August 1990): 597-624.
4. Data on the number of banks is from Banking and Monetary Statistics (Washington, D.C.: Board of Governors of the Federal reserve System, 1943), Table 1.
5. Milton Friedman and Anna Jacobson Schwartz, A Monetary History of the United States: 1867-1960 (Princeton: Princeton University Press for NBER, 1963), and Ben S. Bernanke, "Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression," American Economic Review 73 (June 1983): 257-276.
6. Christina D. Romer, "The Nation in Depression," Journal of Economic Perspectives 7 (Spring 1993): 19-39.
7. Christina D. Romer, "What Ended the Great Depression?" Journal of Economic History 52 (December 1992): 757-784.
8. E. Cary Brown, "Fiscal Policy in the ‘Thirties: A Reappraisal," American Economic Review 46 (December): 857-879.
9. The 2009 figure is an extrapolation to the current quarter based on estimates of potential output by the Congressional Budget Office, http://www.cbo.gov/ftpdocs/99xx/doc9957/Background_Table2-2_090107.xls. For 1933, the estimate is based on the facts that the economy does not appear to have been substantially above trend in 1929 and that real GDP fell 25% from 1929 to 1933. Normal growth would have added at least 10% to GDP over this period.
10. The deficit figures are from Historical Statistics of the United States: Colonial Times to 1970, Part 2, p. 1194, series Y337. Nominal GDP data are from the Bureau of Economic Analysis, http://www.bea.gov/national/nipaweb/SelectTable.asp?Selected=Y, Table 1.1.5. I average calendar year figures to estimate fiscal year values.
11. The data on state and local fiscal stance are from the Bureau of Economic Analysis, http://www.bea.gov/national/nipaweb/SelectTable.asp?Selected=Y, Table 3.3.
12. Romer, "What Ended the Great Depression?"
13. For a comprehensive history of the role of the gold standard in the Great Depression, see Barry Eichengreen, Golden Fetters: The Gold Standard and the Great Depression, 1919-1939 (New York: Oxford University Press, 1992).
14. Friedman and Schwartz, A Monetary History of the United States, Table A-1, column 1 and Table A-2, column 3. The growth rate refers to the period December 1933 to December 1936.
15. The GDP price index data are from the Bureau of Economic Analysis, http://www.bea.gov/national/nipaweb/SelectTable.asp?Selected=Y, Table 1.1.4.
16. Romer, "What Ended the Great Depression?"
17. Peter Temin and Barrie A. Wigmore, "The End of One Big Deflation," Explorations in Economic History 27 (October 1990): 483-502.
18. Data on the components of spending are from the Bureau of Economic Analysis, http://www.bea.gov/national/nipaweb/SelectTable.asp?Selected=Y, Table 1.1.3.
19. Industrial production data are from the Board of Governors of the Federal Reserve, http://www.federalreserve.gov/releases/g17/iphist/iphist_sa.txt.
20. The data on interest rates are from Banking and Monetary Statistics, Table 120; the data on lending are from the same source, Table 2.
21. Of course, every episode is different, and the Federal Reserve will come to its own independent management of monetary policy.
22. See Friedman and Schwartz, A Monetary History of the United States, pp. 328, 421-428, for more information on the 1933 Bank Holiday.
23. For a good description of the various financial stabilization measures Roosevelt took, see Lester V. Chandler, America’s Greatest Depression, 1929-1941 (New York: Harper & Row, 1970), Chapter 9.
24. Stock price data are from Robert Shiller, http://www.econ.yale.edu/~shiller/data.htm. Temin and Wigmore, "The End of One Big Deflation," describe the behavior of commodity prices and interest-rate spreads in 1933.
25. The data on bank lending are from Banking and Monetary Statistics (Washington, D.C.: Board of Governors of the Federal reserve System, 1943), Table 2; the data on house prices are from Robert Shiller, http://www.econ.yale.edu/~shiller/data.htm.
26. The data on bank suspensions and profits are from Banking and Monetary Statistics (Section 7 and Table 56, Column 5, respectively). The data on business failures are from Historical Statistics of the United States: Colonial Times to 1970, Part 2, p. 912, series V27. The data on home foreclosures are from Historical Statistics of the United States: Colonial Times to 1970, Part 2, p. 651, series N301.
27. Data on U.S. and corporate bond yields are available in Banking and Monetary Statistics, Table 128.
28. Barry Eichengreen and Jeffrey Sachs, "Exchange Rates and Economic Recovery in the 1930s," Journal of Economic History 45 (December 1985): 925-946.