Labor Force Migration, Unemployment and Job Turnover

IN recent years; economists have analyzed migration among geographic areas as a form of human investment whereby individuals are thought to incur present costs (both monetary and psychic) in the hope of receiving higher future earnings and other benefits.' The essence of human investment theory is the assignment of a primary causal role to present values of spatial differences in economic opportunity as a determinant of migration.2 An issue of considerable importance is the translation of these general human investment notions into concrete theoretical and empirical terms. The problem centers on the multi-period investment horizon and consequently on measures of job stability and turnover when there is unemployment and uncertainty. Virtually all empirical studies to date have taken as an (inverse) index of economic opportunity the unemployment rate for the area in question. Previous researchers who have used the unemployment rate have encountered serious empirical difficulties. Some have found higher migration rates into high unemployment areas,3 while others have found the unemployment rate to be statistically insignificant as an explanatory variable.4 In only a few studies has the unemployment rate been found to be an important deterrent to migration, and even then, only for certain population subgroups.5 Thus, Greenwood (1975, p. 411) concludes: "One of the most perplexing problems confronting migration scholars is the lack of significance of local unemployment rates in explaining migration." The empirical difficulty with the unemployment rate may well have a conceptual underpinning. A priori, we might suppose that unemployment rates are not very satisfactory measures of economic opportunity for potential migrants, who for reasons to be discussed below would presumably be more concerned with the probabilities of acquiring and retaining employment than with the average employment rate among all workers in the market. Thus, it would be expected that labor turnover variables (such as rates of new hires and layoffs) would play an important part in the explanation of migration. In this paper, we show how labor turnover considerations can be integrated into the human investment theory of migration and demonstrate that such a model provides a much better explanation for migration rates into major metropolitan areas than the conventionally-used unemployment rate. The method used here may be of interest as well to researchers working on other human investment problems that also have a multi-period dimension.