“Something Unanticipated Happened”
Telling some “neo” stories about the Great Depressions of the 1930s
Published December 1, 2000 | December 2000 issue
In his epic novel of post-World War I Germany, A People Betrayed, Alfred Döblin describes a scene where a character pulls a number of crumpled hundred-mark bills out of his coat pocket, seemingly perplexed over how many he's found. He sends a girl out to buy cigars, at which his partner says:
"I can give you a cigar."
"I don't need a cigar. I just want to know what cigars cost today."
If an economist were writing this story, he might describe this as micro level uncertainty, or that it shows how inflation distorted people's behavior and the resultant suboptimality hindered productivity in Germany after World War I, or he might say that this is a story about 1m(2/L)(1/2I)G = 2x/K(2)r. All of which goes to show why economists aren't novelistsbut that's not the point.
The point, though, is that economists are, indeed, storytellers, andif a recent academic conference held at the Minneapolis Fed is any indicationtelling a story about complex economic events is a difficult endeavor. About 60 economists gathered at the Minneapolis Fed in October to present papers and discuss the Great Depressions of the 1930salong with investigations into more recent downturns in Japan and Latin America. A guiding premise of the conference was to apply neoclassical growth theory (discussed later) to events that occurred over 60 years ago, in the hopes of shedding light on one of the most vexing questions in economics.
"Economists and policymakers are still studying and debating what caused this catastrophic economic event," according to remarks in the Minneapolis Fed's Winter 1999 Quarterly Review, which included papers that took a new look at the U.S. Great Depression and partly inspired the conference.
Although many causes have been suggested for the Great Depression, economists have yet to agree on a uniform explanation. The standard approach of the profession since the 1940s has been to try to determine the causes of the depression by searching for relationships or correlations in the data. But since the Great Depression was so unique, there is no basis for comparison and, therefore, empirical analyses always come up short. Recently, some economists have begun applying mainstream economic theory, along with empirical analyses, to derive a new explanation. This attempt was initially met with much skepticism: How can normal theory explain such an abnormal period? But this new method has proved helpful in focusing new questions on these historical events, the answers to which may lead to new explanations.
Newand betterexplanations are important because they mean a better understanding of what went wrong and, ultimately, may help policymakers avoid similar situations in the future.
And better explanations mean better stories. It may strike some as odd to describe economists as storytellers, but it's a term they use when discussing themes and ideas. "My story today is ..." or "The story I want to tell is ..." are introductory words often used by presenters, and questioners frequently begin their queries similarly: "So, you're saying that your story is ..." or "I'm not sure that's the story you want to tell here, you should focus on ..."
While an economist's language may not be as fluid as a novelist'sjargon and equations rarely lend themselves to supple prosethere are still plot twists and surprises that keep the story moving along. As one economist said in the middle of his presentation: "And then, in 1933, something unanticipated happened." The task of those gathered in Minneapolis was to explain how those unanticipated events caused these economic depressions.
Modeling Smith and Ricardo
The study of economic growth is something of a revived phenomenon; the foundations of modern growth theory were laid in the 1950s, but only since the 1980s has the academy paid closer attention to questions about the relative poverty and prosperity of countries.
Of course, economists didn't just discover economic growth this century; Adam Smith's revolutionary 1776 work, after all, was titled "An Inquiry into the Nature and Causes of the Wealth of Nations." And many of Smith's economic growth factors are elements of modern growth theory: division of labor, accumulation of capital and technological progress. Add to that a working legal framework and an understanding of the importance of open trade, and a country is poised for growth, Smith maintained.
But along came David Ricardo to pour a little rain on the growth parade with his notion of diminishing returns. Additional investment in land, to use his example, tends to yield a lower return over time, implying that growth would eventually grind to a halt. These diminishing returns applied to all capital infusions. Trade could extend the benefits of investment, but only to a point; an economy's eventual slowdown was inexorable. As your mother might say: "There are limits."
These ideas of Smith and Ricardo, writing in the late 18th and early 19th centuries, provide the basis for modern growth theory.
Jump now to the 1950s and the work of Robert Solow and Trevor Swan, who developed a model incorporating perfect competition and growing output in response to increasing inputs of capital and labor. (A model is an abstraction of the economy, typically a mathematical description of economic interdependencies.) Solow and Swan's modeled economy also obeyed the law of diminishing returns, with each new unit of capital (given a fixed labor supply) bringing a slightly lower return than the preceding unit.
Thus was born neoclassical growth theory, of which there are two important foundations, one of which describes how labor and capital combine to create output, and the other which describes how households allocate their time between work and leisure, and consumption and savings. Macroeconomic performance, then, relies on (1) how output is allocated between consumption and investment, on (2) whether people spend their time working or playing (or, in economic parlance, on market or nonmarket activities) and on (3) productivity (or the amount of output per unit of input).
A key implication here is that to keep growing, an economy must benefit from continual infusions of new technology resulting in higher productivity. (Also, neoclassical growth theory implies that poorer countries should grow faster than rich ones because investment inputs in poor countries bring a higher return. Although an importantand disputedimplication, it is not necessarily germane to this discussion.) Neoclassical growth theory has been tweaked since its inception, especially to account for human capital (or learning); this learning explains how an economy can experience increases in productivity without diminishing returns.
A look through the neoclassical lens
What follows is a brief overview of the papers presented at the Great Depressions conference; discussants' comments and the sometimes lengthy (and technical) critiques following presentations are not included. It should also be clear that much of the work presented at the conference was in draft form and is not meant for complete disclosure. For those interested in a closer look at the work of the following economists and others at the conference.
When looking at the economy through the lens of the neoclassical growth model to determine causes for changes in economic growth, productivity shocks (or sudden changes in productivity) are the starting point; that is, the first question asked is how much can productivity shocks explain. In this case, the authors show, productivity shocks don't tell the story; rather, changes in institutional and market regulation are more likely culprits for the depth and duration of the French depression. Indeed, the size of those institutional changes may alone account for France's economic slump. This interpretation disputes the view that the French depression was largely dissimilar to the U.S. experience and suggests a search for common explanations.
"A Neoclassical Analysis of Britain's Interwar Depression": Harold L. Cole, Federal Reserve Bank of Minneapolis, and Lee E. Ohanian, University of California, Los Angeles. In this paper, neoclassical growth theory is applied to events in the United Kingdom, and the authors found that productivity shocks can't explain the persistence of the U.K. depression in the 1920s. (The economic slump in the United Kingdom, as described in this paper, began after World War I and lasted until 1936.) Here is a case where a positive productivity shocka rapid increase in productivity in the post-World War I United Kingdomdid not result in increased economic growth, as the theory would suggest. The standard explanation for the 1920s decline, that a deflationary monetary policy and an inordinately high fixed exchange rate were the prime factors, does not conform to the evidence either. A better explanation? According to the authors: "The most promising candidate shock is a substantial increase in unemployment benefits that began in the early 1920s." These benefits resulted in a greater incidence of nonmarket activities. As for the worsening of the depression in the 1930s United Kingdom, lower world income and productivity are likely answers.
At this stage of their research, the authors suggest that productivity changes may have been crucial in Germany, but they "do not have a story which can account for these changes." However, this paper evaluates the role of fiscal policy and real wages in the depression and recovery in Germany, and finds that fiscal policy had a limited effect on the economy, but that the behavior of real wages might have been an important factor in the economy.
"Understanding the Great Depression: What can we learn from the Italian experience?": Fabrizio Perri, New York University, and Vincenzo Quadrini, New York University and CEPR. Again, applying the lens of neoclassical growth theory and beginning with a look at productivity, the authors claim that the post-1929 slowdown in Italy cannot be explained solely by productivity shocks, but that other factorsnamely, trade restrictions and wage rigiditiesare to blame. A model is presented that predicts that trade restrictions account for about three-quarters of the observed slowdown, while wage rigidity (monetary shocks), account for the remaining fourth.
"The 1990s in Japan: A Lost Decade": Fumio Hayashi, Tokyo University, and Edward C. Prescott, University of Minnesota and Federal Reserve Bank of Minneapolis.
When observers account for the 1990s slump in Japan, the list of suspects usually includes these four: inadequate fiscal policy, the liquidity trap (or when even zero interest rates can't spur demand for money), low rate of investment and problems with financial intermediation (or the credit crunch brought on by Japan's reliance on bank-based lending and the now-infamous keiretsu conglomeration). Enter neoclassical growth theory and a new explanation: Japan was transitioning to a new and lower steady state growth path that was occurring because the growth rate of productivity was lower after 1991 than it was before. The neoclassical growth model, calibrated for the 1984-88 steady-state growth period, accounts for the important features of Japan's 1990s economy.
"A Decade Lost and Found: Mexico and Chile in the 1980s": Raphael Bergoeing and Raimundo Soto, ILADES-Georgetown University, Universidad Alberto Hurtado; Patrick J. Kehoe, Federal Reserve Bank of Minneapolis, and Timothy J. Kehoe, Federal Reserve Bank of Minneapolis and University of Minnesota.
In 1982 Mexico and Chile, countries with similar initial macroeconomic conditions, suffered economic crisesChile's more severe than Mexico'sand yet Chile recovered faster. Why? It wasn't because of different money growth rates, or the sharp depreciation of real exchange rates, or the problem of debt overhang deterring investment. Rather, the answer lies with Chile's implementation of market-based reforms (in banking, trade, privatization, fiscal policy and social security), some of which had begun in the 1970s and had allowed the country to better weather the shocks of 1982. Mexico, among other things, nationalized banks, levied tariffs and protected domestic industries from imports.
In the 1980s, Argentina's average annual per capita GDP declined by 1.5 percentsandwiched between growth rates of 2.3 percent and 1.8 percent in the decades preceding and following, respectively. What happened in the 1980s? The authors don't present definitive answers; rather, their intention is to present economic evidence within the framework of the neoclassical growth model to try to help sort out anomalies and better focus questions for future research. They note, though, that any investigative efforts into Argentina's economic performance will be hampered by sometimes-questionable data, an unfortunate holdover of "the heavy and long interventionist hand of successive Argentinian governments."
Neoclassical theory predicts that following a slump, the economy should recover strongly and with low real wagesprecisely opposite of what occurred from 1934-1939 in the United States, generally noted as the recovery years from the Great Depression, but cited by the authors as a continuation of the depression. The authors find that New Deal cartelization policies (National Industrial Recovery Act and its heirs) are an important factor in the post-1933 depression, and that the key depressing factor of New Deal policies was not collusion per se, but the link between paying high wages and collusion. Absent these New Deal policies, the authors suggest, the recovery from the Great Depression would have been stronger.
"A Dual Method of Empirically Evaluating Dynamic Competitive Equilibrium Models with Distortionary Taxes, Including Applications to the Great Depression and World War II": Casey B. Mulligan, University of Chicago and National Bureau of Economic Research.
As the title suggests, this paper was as much about the usefulness of the neoclassical model as it was about using this model to shed light on the Great Depression. Suffice to say that the former point garnered much discussion but would offer little insight to the general reader; on the latter point, the author challenges the neoclassical approach to explain why the marginal product of labor and the marginal value of leisure diverged so much and why that wedge persisted. In doing so, he argues that "explaining aggregate behavior during the period has been reduced to a public finance questionwere actual government policies distorting behavior in the same direction and magnitude as government policies in the model?"
"M" is for macro, and the Great Depression is a mystery story
The idea of finding the telltale "smoking gun" in the Great Depression story was often expressed during the conference, and so the first transparency that University of Chicago economics professor Nancy Stokey placed on the viewer was her artistic rendition of not one gun, but several. In other words, there are many smoking guns when it comes to solving the riddle of the Great Depression. Not only does more than one thing have to go wrong, Stokey said in her comments on the final paper, but these multiple causes are more than coincidentthey are contingent, even across countriesand they are often followed by misguided policies.
In his summary remarks at the close of the conference, Robert Lucas, University of Chicago economics professor and Nobel laureate, made a pitch for the continuing investigation of macro fundamentals. "Macro is ugly stuff," Lucas said, but answers lie in macro theory.
"We should continue to seek common factors," he said, and offered monetary instability as one area for further exploration. Big deflations are related to depressions, he said, and everywhere in the 1930s there was deflation. "If I was starting over again, I'd look there."
V.V. Chari, economics professor at the University of Minnesota and an adviser to the Minneapolis Fed, said that he detected an emerging consensus (generally speaking) among the group by the close of the two-day event. Essentially, it is a consensus that addresses the Great Depressions of the 1930s as, in effect, two closely linked events: the initial slump and the slower-than-anticipated recovery. The negative shocks of monetary instability and the attendant decrease in output set many economies slumping, and the reason they didn't bounce back as soon as expected is a compilation of bad policies that extended and worsened the Great Depression.
In the end, if the Great Depression is, indeed, a story, it has all the trappings of a mystery that is loaded with suspects and difficult to solve, even when we know the ending; the kind we read again and again, and each time come up with another explanation. At least for now.
[See Great Depressions of the Twentieth Century, published by the Federal Reserve Bank of Minneapolis.]