Testing Self-Selection in Migration: Evidence from the Israeli Kibbutz

I use a longitudinal dataset of individuals entering and exiting Israeli kibbutzim, communities that engage in equal sharing of output, to test the mobility patterns induced by redistribution. I find evidence of negative selection in entry to kibbutzim and positive selection in exit. Entrants were negatively selected in their pre-entry earnings compared with non-entrants, especially among the more educated. Compared with stayers, individuals who left kibbutzim were positively selected in their observable skills such as education. Less educated kibbutz leavers were also positively selected on their ex ante-unobservable skills (measured by post-exit wage). Finally, the selection patterns into kibbutzim differed substantially from the selection patterns into other rural locations that did not engage in intensive redistribution. At the broader level, these findings also provide micro-level empirical support for Borjas' hypothesis that migrants' self-selection depends on the difference in returns to skills between the origin and the destination.