LONGITUDINAL PATTERNS OF U.S. LABOR FORCE MIGRATION IN THE UNITED STATES, 1967-73: A REEVALUATION OF HUMAN INVESTMENT THEORY

The human investment approach to migration predicts that younger workers should respond more strongly to earnings incentives than older workers because they will have a longer period to realize the gains of investment in human capital. Premising that migration is an investment act, this thesis examines the comparative mobility behavior of younger versus older workers by means of a two-stage allocative shares migration framework. The data bases are migration tabulations for 52 large U.S. SMSAs derived from the Social Security Continuous Work History Sample one percent first-quarter file in the two time frames 1967 to 1970 and 1970 to 1973.;Destination choice behavior shows that the hypothesis that higher earnings attract migrants is supported among all race-sex groups with the exception of black males during 1970-73. However, contrary to expectations, responses of younger white males are not as strong as in the case of older age groups. The hypothesis that increasing distance acts as a barrier to migration is supported among all workers, regardless of age, race, or sex. An additional regression for white males 20-24 in 1970-73 including a migrant-stock variable shows that former area residents apparently strongly affected destination choices of later migrants.;Empirical analysis suggests two directions for policy. Since higher earnings do not significantly affect outmigration rates, efforts initiated by lagging areas to discourage outflows of workers may not have the desired outcome. A more appropriate direction for government to follow may be to pursue measures which offset the retarding effect of distance on destination choice.;Stage one of the analytical framework examines outmigration behavior by age, race and sex of worker. Stage two hypotheses concern the destination choice of migrants. A principal finding for both study periods is that initial earnings levels do not affect outmigration propensities significantly or even in the direction expected. Earnings growth rates also tend not to be significant. Among whites, the hypothesis that older workers should respond more than younger workers to climate conditions in the origin area tends to be substantiated. Coefficients on the size of work force variable have the hypothesized negative sign and are significant in the four aggregate demographic groups, again in both mobility periods.