The American Economic Review September 1988 [antikvár]

Forecasting the Depression: Harvard Versus YaleBy Kathryn M. Domínguez, Ray C. Fair, Matthew D. Shapiro*Was the Depression forecastable? After the Crash, how long should it have taken contemporary forecasters to realize how severe the downturn was going to be? Data assembled by the Harvard and Yale forecasters together with modern historical data are subjected to statistical analysis. Neither contemporary forecasters nor modern times-series analysts could have forecast the large declines in output following the Crash.James Tobin relates the following story, which was told to him and other members of the Harvard graduate economics club by Professor W. L. Crum. In the 1920s the Harvard Economic Service (HES) issued monthly reports on the current and expected future state of the economy. HES used three indexes, representing speculation, business, and money, to help predict the future. Crum claimed that in the sunraier of 1929 the statistical assistant at HES became alarmed when she noticed that the indexes indicated that a sharp downturn in economic activity was imminent. Crum did not see in the current business situation any cause for this adverse forecast. Moreover, he feared that a pessimistic forecast by the influential service could itself have an adverse effect on financial markets and economic activity. Therefore, he suppressed the pessimistic findings of the assistant;' the pubhshed report did not speak of a potential downturn.The data, in fact, provide only mild support for this account. The account does, however, raise the more general question of whether the Depression was forecastable. Was there anything in the data prior to theDomínguez: John F. Kennedy School of Government, Harvard University, Cambridge, MA 02138. Fair and Shapiro: Cowles Foundation for Research in Economics, Yale University, New Haven, CT 06520, and National Bureau of Economic Research. We are grateful for helpful comments from James Tobin, Paul A. Samuelson, and anonymous referees.And consequently he lost the opportunity to gain a reputation for Delphic wisdom.October 1929 stock market Crash that indicated the economy was about to enter a protracted slowdown? How should the news of the Crash have revised forecasts of economic growth? To address these questions, we study two sets of data assembled by contemporary business-conditions forecasters. The first consists of the three HES indexes. The second consists of commodity and stock price indexes compiled and analyzed by Irving Fisher. Fisher, who was at Yale from 1891 until his death in 1947, was a competitor of HES. He also released periodic reports on the state of the economy, and he was a critic of the HES indexes.We also use modem historical data on industrial production, producer prices, stock prices, time-loan rates, and the money stock in our statistical analysis. The use of these data allows us to address the broader question of the forecastabihty of the Depression without recourse to the services' idiosyncratic data. We can also ask whether the Depression-era data contain information not embodied in the modem data.The procedures, data, and pronouncements of the Harvard and Yale forecasting services are discussed in Section I, where the accuracy of Crum's account is also investigated. Both services failed to anticipate the Depression and remained optimistic about economic performance following the Crash. In the following section, we examine the data using time-series analysis. The analysis suggests that the services' optimism was not unwarranted.595