«BEYOND SHOCKS: WHAT CAUSES BUSINESS CYCLES? AN OVERVIEW Jeffrey C. Fuhrer and Scott Schuh* In the summer of 1997, when the Federal Reserve Bank of ...»
Given the lack of private incentive to restrain the stimulative effects of this “oldest business cycle mechanism,” we come to the second factor that contributes to business cycle ﬂuctuations: government policy. Samuelson noted that he has often said, “When the next recession arrives, you will ﬁnd written on its bottom, ‘Made in Washington.’” This is not, as he points out, because the Fed is a sadistic organization. Rather, “if the central bank and ﬁscal authorities did not step on the brakes of an overexuberant economy now, they might well have to overdo that later.” When persistent macromarket inefﬁciencies threaten both employment and price stability and private incentives fail to encourage ﬁnancial markets back into line, only policymakers can take the systemic view necessary to guide the economy back into balance.
THE CAUSES AMERICAN BUSINESS CYCLES
Second, the Great Depression should be treated as a unique event. As Temin notes, output lost during this enormous downturn was almost one-half of the sum of output lost in all other downturns in the past century. The body of writing on the Great Depression is larger than that on all other business cycles combined. Consideration of the causes of the Great Depression provides useful lessons about the causes of the less prominent cycles of the past century. For example, it seems implausible that a single “shock” in 1929 pushed the U.S. economy into massive depression. Instead, Temin argues, the Great Depression was likely the result of a sequence of contractionary inﬂuences. Prominent among these were the fear that the hyperinﬂationary pressures in Eastern Europe following the First World War would spread to the United States, the adoption by industrialized countries of the relatively inﬂexible gold standard in response to these pressures, and the breakdown of banking and legal systems. The Great Depression was really a sequence of smaller
BEYOND SHOCKS: WHAT CAUSES BUSINESS CYCLES? AN OVERVIEW 7recessions large and persistent enough, given policy responses, to throw the world into depression.
Third, Temin cautions that his assignment of causes relies on the existing literature on the subject. The literature on recessions other than the Great Depression is quite sparse, with earlier recessions receiving considerably less attention than more recent ones. And within this limited set of sources, most authors focus on the transmission of cycles, rather than on the causes. Finally, most of the available sources do not highlight expectations and do not clearly distinguish anticipated from unanticipated changes.
Temin classiﬁes the reported causes of recessions as either domestic or foreign, and either real or monetary. Changes in the relative prices of assets, both real and ﬁnancial, are classiﬁed as real phenomena. Temin ﬁnds that the preponderance of cycles in the past century may be attributed to domestic causes, with the split between real and monetary causes roughly equal for the entire period. Monetary causes of recessions were more prevalent in the pre-World War I period than during the post-World War II period, however.
Temin focuses on the larger downturns. The cause of the Great Depression of 1931 is classiﬁed in Temin’s taxonomy as a foreign monetary phenomenon. The action of the Fed to maintain the gold value of the dollar by raising interest rates was to behave as a “traditional and responsible central banker” or, in other words, to follow a normal and expected endogenous policy course. Thus, the Fed’s behavior cannot be viewed as an exogenous cause of the Great Depression, in Temin’s view.
The search for causes then reverts to the question of what produced this monetary policy response. Temin suggests that U.S. monetary policy was responding to the external gold drain that arose from Britain’s departure from the gold standard, which threatened to weaken the dollar. The Fed’s reaction in increasing interest rates, and the bank panics and failures that followed, were endogenous responses to the gold drain.
In assessing the causes of the four largest downturns of the century— the Great Depression, and the recessions of 1920, 1929, and 1937—Temin concludes ﬁrst that no single cause explains all four downturns. Three of the four possible causes in Temin’s taxonomy appear as causes of the downturns. Second, three of the four recessions appear to be responses to domestic shocks. Most often, we cannot blame our downturns on foreign causes.
Taking all of the cycles studied into consideration, Temin offers the following conclusions: (1) “It is not possible to identify a single type of instability as the source of American business cycles.” Thus, Dornbusch’s statement, “None of the U.S. expansions of the past 40 years died in bed of old age; every one was murdered by the Federal Reserve,” is not supported by Temin’s analysis. (2) Domestic real shocks—ranging from inventory adjustments to changes in expectations—were the most freJeffrey C. Fuhrer and Scott Schuh quent source of ﬂuctuations. (3) Other than the two oil shocks of 1973 and 1979, foreign real shocks were not an important source of U.S. cycles. (4) Monetary shocks have decreased in importance over time. (5) When measured by the loss of output, domestic sources have loomed larger than foreign sources; real sources have caused about the same losses as monetary sources.
Christina Romer takes issue both with Temin’s classiﬁcation scheme and with his interpretation of the literature on the causes of recessions.
She suggests that an improved classiﬁcation scheme and a different reading of the literature would yield a more critical role for domestic monetary shocks, particularly in the inter- and postwar periods.
Romer suggests that Temin’s methodology is biased toward ﬁnding very few monetary causes of recessions. Whereas Temin classiﬁes most Fed behavior as a fairly typical response to prevailing conditions and therefore not the ultimate cause of the recession, Romer would prefer a more practical classiﬁcation of monetary policy actions. If the monetary policy action was the inevitable or highly likely result of a trigger, then we should consider the policy action endogenous and therefore not a cause. If, however, “a conscious choice was made” or if “alternative policies were... discussed at the time,” then the policy should be considered at least partly exogenous, and monetary policy should get some blame for the recession.
Romer shows that, using this criterion, many more of the twentiethcentury recessions have an important monetary policy aspect. Monetary factors would likely be given an important causal role in the 1931 recession, for example, as “reasonable men at the time were urging the Fed to intervene” in the face of ﬁnancial panics. Thus, the choice not to intervene but to raise the discount rate was not inevitable or even most likely. Romer also questions the extent of the constraint imposed by the gold standard, as U.S. gold reserves in 1931 were probably adequate to have allowed the Fed to pursue expansionary open market operations while maintaining the gold value of the dollar, as in fact it did in 1932.
Turning to the 1973 recession, for which Temin ascribes no monetary role, Romer argues that the central bank was not simply acting as “a respectable central bank [that] resists inﬂation,” and therefore responding only as expected. Romer points out that the decision to tighten in 1974 was not a foregone conclusion but rather a conscious choice, as “the economy was already in a downturn and many were calling for loosening.” Thus, “monetary policy and the oil shock share responsibility for the 1973 recession.” Romer also challenges Temin’s attribution of the 1957 and 1969 recessions to declines in government spending. She points out that the high-employment budget surplus actually falls throughout the late 1950s, suggesting a net stimulative impulse from the federal government for the
BEYOND SHOCKS: WHAT CAUSES BUSINESS CYCLES? AN OVERVIEW 91957 recession. For both recessions, Romer asserts that the Federal Reserve made a conscious decision to tighten in order to reduce inﬂation.
As Romer sees it, “the key change has not been from monetary to real shocks or vice versa, but from random shocks from various sources to governmental shocks.” Since the Second World War, the government has been more effective at counteracting most shocks, accounting for the diminished frequency of cycles. However, the combination of a tendency toward overexpansion and a few large supply shocks caused inﬂation to get out of hand. In sum, Romer would agree with the thrust of Dornbusch’s statement, which is that monetary policy has played a vital role in postwar recessions. She might re-cast the role of the Fed, however, as “more like a doctor imposing a painful cure on a patient with an illness than a murderer.”
THE INTERNATIONAL EXPERIENCEMichael Bergman, Michael Bordo, and Lars Jonung examine the broad cyclical properties of GDP, using a newly compiled data set of annual observations for a sample of “advanced” countries. Their data set spans the years 1873 to 1995. The authors show that the duration of business cycles (the calendar time from peak to peak or trough to trough) has been fairly similar across countries and fairly stable over time. The average duration rose from about four years in the pre-World War I period to about ﬁve and one-half years during the interwar period, falling back to just under ﬁve years in the period following World War II. The most severe recessions appear to have occurred prior to 1946, and the magnitude of all ﬂuctuations in GDP seems to have decreased in the postwar period.
Formal statistical tests of diminished cyclical ﬂuctuations in the postwar period generally conﬁrm the visual evidence. This observation has often been interpreted as evidence that countercyclical policy has been more effective in the postwar period. However, an alternative explanation is that the increased integration of the world economy serves to mitigate the negative inﬂuence of any one country’s disruptions on other countries.
Conventional wisdom holds that downswings are sharper and “steeper,” whereas upswings are more gradual. Bergman, Bordo, and Jonung test this proposition and ﬁnd that, for the United States, upswings are indeed more gradual than downswings. The evidence for other countries is more mixed, however, with most exhibiting this asymmetry prior to World War II but only a minority displaying asymmetry in the postwar period.
The authors then attempt to determine the extent to which different components of GDP—including consumption, investment, government expenditures and revenues, exports, and imports—account for its cyclical 10 Jeffrey C. Fuhrer and Scott Schuh volatility. For virtually all countries and time periods, all components of GDP except consumption generally are more volatile than GDP. This ﬁnding is consistent with the presence of a consumption-smoothing motive, that is, the desire of consumers to maintain a relatively smooth stream of consumption over time in the face of volatility in their income and wealth.
The authors ﬁnd that larger countries experience deeper recessions;
the average decline in GDP below trend is larger for large countries than for small, open European countries. For most countries, the downturn in GDP during a recession is accounted for by declines in consumption, investment, and net exports.
Finally, Bergman, Bordo, and Jonung consider the patterns of international co-movement of output and prices in their data. They ﬁnd that the correlations among real output in the 13 countries have increased over time, suggesting a more integrated world economy and possibly a stronger coherence of the business cycle across countries. During the gold standard, real GDP for most countries exhibited little or no correlation with real GDP in other countries. During the interwar period, U.S. GDP was signiﬁcantly correlated with seven other countries, but corresponding correlations between other countries were not evident. The authors suggest that this correlation arises from the role of the United States as the “epicenter” of the Great Depression. Output linkages among European countries strengthened considerably in the postwar period, perhaps the result in part of the establishment of the European common market and in part of the common inﬂuence of the oil shocks in the 1970s.
Price levels appear to be much more consistently correlated across countries. Like output, price levels have become increasingly correlated over time, perhaps consistent with “increased global integration of goods markets,” the authors suggest.
Richard Cooper offers a different perspective on Bergman, Bordo, and Jonung’s conclusion that “the cyclical pattern... appears to remain surprisingly stable across time, regimes, and countries” and on the broad question of the international origin and transmission of the business cycle. He examines years in which the raw data for real GDP declined, for a set of nine countries during the periods 1873 to 1913 and 1957 to 1994.
Cooper prefers this approach, as the authors’ results may depend on the ﬁltering and detrending methods that they used in constructing their data.
The conclusions that he draws for the earlier period are as follows: