Employment, Hours, and Earnings in the Depression : an Analysis of Eightmanufacturing Industries

This paper employs monthly, industry-level data in a study of Depression-era labor markets. The underlying analytical framework is one in which, as in Lucas (1970), employers can vary total labor input not only by changing the number of workers but also by varying the length of the work-week. This framework appears to be particularly relevant to the 1930s, a period in which both employment and hours of work fluctuated sharply. With aggregate demand treated as exogenous, it is shown that an econometric model based on this framework, in conjunction with some additional elements (notably, the adjustment of workers' pay to permanent but not transitory variations in the cost of living, and the effects of New Deal legislation) can provide a good explanation of the behavior of the keytime series. In particular, the empirical model is able to explain the puzzle of increasing real wages during a period of high unemployment