April 18, 2019 Reading Time: 5 minutes

There is no more heavily debated topic within macroeconomics than the causes of the Great Depression and why it came to an end. Even 80 years later, theories abound. And new theories keep coming.

A recent NBER working paper by Margaret M. Jacobson, Eric M. Leeper, and Bruce Preston puts forward a compelling new theory. They argue that FDR’s policies helped pull the United States out of the Great Depression, but not in the way that previous economists have argued. Instead of simply focusing on FDR’s spending and its effect on aggregate demand, the authors also bring in another major policy used to fight the Depression: the abandonment of the gold standard. It is only taken together that FDR’s fiscal and monetary policies led to a recovery.

And it truly was FDR conducting monetary policy, not the Federal Reserve. First, FDR abandoned the gold standard in April 1933. Then the Thomas Amendment in May 1933 granted the executive branch unprecedented monetary powers. It could issue greenbacks, fix the gold value of the dollar, and order the Fed to buy Treasury securities. This was the first step in making sure the Fed would not act to thwart the stimulative impact of fiscal expansion after its blunder in the Great Contraction of 1929–32.

Scholars across generations, from E. Cary Brown in the 1950s to Price Fishback in more recent years, have most often focused on the “Keynesian hydraulics” of FDR’s spending. This form of Keynesian fiscal theory says that increased government spending leads to increased aggregate demand, which ultimately leads to increased output. But, as many authors have pointed out, the size of the effects in the Depression do not line up with the Keynesian-hydraulics story. The increase in government spending was too small to generate the increase in output that followed. One man’s push cannot move a rocket ship.

Jacobson, Leeper, and Preston argue that the Keynesian-hydraulics story implicitly assumes FDR was operating under the gold standard. But, after 1933, he wasn’t. Under the gold standard, dollar-denominated government debt is real (gold denominated) debt. The credibility of the gold standard required that the government be ready to raise taxes in real terms to acquire gold if needed. When governments spend more, people expect higher future taxes and spend less today. In other words, higher expected taxes reduce the effectiveness of fiscal policy. Under a gold standard, monetary and fiscal policies cannot achieve just any desired price level.

Economists like Brown explicitly exclude government borrowing from their calculations. By doing that, Brown implicitly assumes that higher future taxes negate a substantial portion of the possible effects from the higher nominal debt that actually occurred in 1933. That rules out an effect on the price level. The assumption is reasonable for an economy on the gold standard.

But the link between spending and taxes is not so tight once the government abandons the gold standard. The price level is more flexible. Abandoning the gold standard allowed FDR to conduct “unbacked fiscal expansion” until a recovery came. As the authors explain, “Unbacked fiscal expansion increases government expenditures on purchases or transfers, issues nominal bonds to cover the deficit, and persuades people that surpluses will not rise to finance the bonds.” Since people are persuaded that taxes won’t be increased to create a government surplus, nominal bond prices must devalue. The ultimate effect may be inflation, but it need not be, especially if output increases drastically.

Even though FDR felt that the key to economic recovery was returning overall prices to their 1920s levels through inflation, increased spending after the gold standard did not lead to high-enough inflation to return to those levels. The authors compare different measures of output and the price level in Figure 2 in the paper (reproduced below). The yellow vertical line indicates when the U.S. abandoned the gold standard. While prices did rise slightly, they quickly stabilized at levels well below their 1921 highs. At the same time, output grew rapidly and returned to its 1929 level by 1937.

While both nominal and real debt levels increased under FDR, nominal debt grew about 30 percent more than real debt because prices were rising. What looks like unsustainable growth in nominal debt looks slightly better when thought of as real debt. More importantly, as Figure 8 in the paper shows, the debt “problem” goes away completely once we look at the debt as a percentage of GNP. That stabilized under FDR because of the recovery of 1933 (which increased the denominator).

The authors are further able to decompose any changes in the debt-to-GNP ratio in Figure 9 in the paper (shown below). That ratio goes in the opposite direction of GNP (Y) and the price level (P), and it goes in the same direction as real debt (B).

The authors argue that FDR was able to avoid runaway inflation because he was committed to increased spending but only until the recovery came: “Roosevelt never claimed to be aiming for what even he might have regarded as ‘irresponsible’ fiscal policy.” In fact, for non-emergency (ordinary) spending (the black line on the graph below), FDR could claim to have balanced the budget. The overall budget was not balanced, but the claim reinforced the idea that spending was only going to be high until a recovery came. After that, FDR would return to a sound fiscal path. We do not know whether FDR would have returned spending to normal if World War II had not come. But the stable inflation and interest rates following 1933 seem to indicate people at the time thought spending would return to normal.

Besides the narrative and graphical evidence that the authors provide, they run a more formal analysis using vector autoregressions and impulse responses. Their focus on unbacked fiscal expansion changes the formal analysis from what is usually done. Instead of looking at changes in the ratio of government surplus to output, which is what people have often looked at, the authors focus on the ratio of government surplus to debt. Fiscal policy’s ultimate impact on aggregate demand depends on total real (gold) backing relative to outstanding nominal debt.

To be clear: this amounts to asking what the government should have done in 1933 after the many policy mistakes from 1929-1932. But, given the Great Contraction, Jacobson, Leeper, and Preston offer a plausible account of the 1933 recovery. They do not provide the final word. The interplay of fiscal and monetary policy is complicated, as the last recession has reminded us. Simple fiscal Keynesian or monetarist stories will not suffice today. Nonetheless, they have put forth a serious alternative analysis that is worthy of deep consideration and subject to possible rebuttal.

Brian C. Albrecht


Brian Albrecht is Chief Economist of the International Center for Law & Economics (ICLE). Brian’s research focuses on price theory, information economics, competition and innovation, and political economy.

He has published in both academic journals, such as Contemporary Economic Policy, Public Choice, PLoS ONE, Journal of Macroeconomics, and the Journal of Economic Methodology, as well as popular media like the Boston Globe, Star Tribune, The Hill, and City Journal. Brian also writes the Economic Forces newsletter.

He earned his PhD in economics from the University of Minnesota in 2020. He previously earned his M.A. in economics, also from the University of Minnesota, and an M.Sc. in economics of public policy from the Barcelona Graduate School of Economics. He received his bachelor’s in physics and political science from St. Olaf College.

Get notified of new articles from Brian C. Albrecht and AIER.