Essays on emigration, remittances, and employment in the source country: Evidence from Mexico

This dissertation contributes to the emerging empirical literature on the developmental impact of international migration in the source developing economy. Given the policy priority placed on job creation and the emphasis on wage impacts in recent work, the focus here is on the consequences of emigration on the extent of employment and the nature of that employment of those left behind. In particular, our questions are (i) does international migration and remittances have an impact on unemployment rates in the source country? and (ii) do remittances encourage entrepreneurship, as reflected by the share of workers classified as self-employed? We investigate these questions using census data from Mexico, unlike much of previous work that has relied on household survey data. Our empirical strategy attempts to address the typical issues of self-selection and endogeneity that migration impact studies encounter. We classify workers into skill groups and employ the Borjas (2003) empirical strategy, carrying out our analysis at the national level. We also introduce the fractional logit estimator (Papke and Wooldridge (1996)), unused in this field before, to address the fact that the dependent variables of concern are both proportions, and thus avoid the flaws in typical studies that investigate the impact of selected variables on the conditional expectation of a proportion or vector of proportions. Our results show that (i) emigration and remittances decrease native unemployment rates, and (ii) remittances intensify self-employment activities among the receivers. These results are even stronger once we control for sample bias and endogeneity. In terms of economic policy, we therefore provide indirect evidence that migration and remittances could be growth enhancing through their effects on employment in the source country.