Employment and Unemployment in the 1930s

This paper surveys recent research on employment and unemployment in the 1930s. Unlike earlier studies that tended to rely heavily on aggregate time series, the research discussed in this paper focuses on disaggregated data. This shift in focus stems from two factors. First, dissaggregated evidence provides many more degrees of freedom than the decade of annual observations associated with the depression and thus can prove helpful in discriminating between macroeconomic models. Second, and more importantly, disaggregation has revealed aspects of labor market behavior hidden in the time series that are essential to their proper interpretation and which are, in any case, important in their own right. In particular, the findings dispute the view that representative-agent models are useful for interpreting shifts in employment and unemployment over the course of the Depression.