No abstract

Income uncertainty contributes substantially to explaining the fall in consumption that marks the onset of the Great Depression. Consistent estimates of the variance of income measure income uncertainty from 1921‐30 and are produced using a linear moment model. This series provides a statistical link between the large erratic swings in income uncertainty after September 1929 and the Great Crash in the stock market. Comparison of the behavior of income uncertainty in the 1920s to the pre‐World War I and post‐World War II eras suggests that the experience after the Great Crash was historically unique.

Building upon Beaudry and Koop's (1993) analysis, we consider a "current depth of the recession" (CDR) variable in modeling the time-series behavior of the postwar quarterly U.S. unemployment rate. The CDR approach is consistent with the state-dependent behavior in the unemployment rate documented in the business-cycle asymmetry literature. We show that while the CDR effect is significant in-sample, no statistically significant out-of-sample forecast improvement is obtained relative to the linear alternative. Augmenting an AR(2) model by inclusion of the CDR term, however, does not significantly worsen the out-of-sample forecast performance. Acknowledgment. We wish to thank an anonymous referee and Norm Swanson for helpful comments.Since the publication of Neftci's (1984) seminal paper, many researchers have documented significant nonlinearity in the movement of the postwar quarterly U.S. unemployment rate over the course of the business cycle. Such nonlinear dynamic behavior is said to be asymmetric in the sense that the unemployment rate increases quickly in recessions, but declines relatively slowly during expansions. Cyclical asymmetry of this type is inconsistent with the conventional linear time-series tools used in the analysis of macroeconomic data.The importance of this issue for time-series modeling of the unemployment rate is (at least) twofold. First, the standard second-order approach for producing in-sample diagnostics may be misleading. For example, Potter (1995) shows how a state-dependent impulse response function differs rather strongly from one based on an estimated autoregressive moving average (ARMA) model. Second, it may be possible to reduce out-of-sample mean squared prediction error (MSPE) with a nonlinear forecast relative to a conventional ARMA forecast. Rothman (1998), for example, conducts an extensive out-of-sample forecasting competition between a linear model and many nonlinear alternatives for the postwar U.S. unemployment rate. In several cases, the nonlinear forecasts produce statistically significant MSPE reductions.Our primary interest in this paper is in out-of-sample forecasting for the U.S. unemployment rate. In particular, we focus on the out-of-sample forecasting performance of a class of models introduced by Beaudry and Koop (1993). These authors augment an autoregressive model of U.S. GNP growth rates by adding a term they label CDR, a variable that measures the "current depth of the recession" in the following sense: for each

No abstract

We demonstrate three facts consistent with the debt deflation/credit view explanation of the Great Depression. First, private medium‐ and long‐term nominal debt during the 1920 s exhibited a combination of a high initial value relative to income and a rapid growth rate that is unparalleled in a consistent data set covering more than half a century. Second, the debt issued during the 1920 s occurred in a stable price regime. Third, near the onset of the Depression, the price process switched to one of deflation. Taken together, the evidence suggests that debt deflation was operative during the Depression.