| Abstract: | This paper tests the dynamic implications of cumulative causation and network theory on the self‐sustaining nature of the migration process. A sequential migration model is derived and estimated with a panel of Mexican household heads for the years 1980–89. Consistently with cumulative causation and network theory, the empirical results show that, after controlling for unobserved heterogeneity, previous migration experience and migration‐related variables are the strongest predictors of current migration decisions. This persistence in migration patterns implies that immigration policies inducing small changes in expected costs and benefits of US work fail to prevent entry into, or encourage exit from, the US labor market by experienced migrants. However, large temporary disruptions such as the 1994 Mexican peso devaluation may permanently increase migratory flows by inducing new migrants to enter the US labor market. |
