Migration and Wages of Young Men

We specify a two-period nested logit migration model with selectivity. We estimate the model using a sample of young male workers and study their choices among the nine U.S. census divisions during the first two years after leaving fulltime schooling. We find evidence of the existence of unobserved similarities in the divisions within each census region. We find that locational choices in one period affect locational choices of individuals in subsequent periods. We test for selectivity in division-specific wage equations and find little evidence of selectivity in the wage equations of most divisions and evidence of negative selection in the wage equations of two divisions.