The Causal Impact of Trade on Migration: A Gravity Model Estimation

Studies on the causal impact of trade on migration are rare. Most previous studies have instead looked at the impact of migration on trade. The few empirical studies that have a causal interpretation have focused either on a single country, a single region, or within the members of a single trade agreement. This paper addresses the research question, does an increase in bilateral trade flows cause an increase in bilateral migration? We employ a novel instrumental variables strategy, using World Trade Organisation (WTO) affiliation and average tariff rates as instrumental variables within a gravity model framework. This approach mitigates against the endogeneity problem and allows us to extract the causal association between bilateral trade flows and bilateral migration flows. In the model, we employ data for 248 countries over the period 1990-2010. Our preferred estimator is the Poisson Pseudo-Maximum Likelihood Estimator, since it better handles the sparse nature of the data. Our findings suggest that trade is a statistically significant causal driver of migration. Based on our results, migration flows from country i to country j would increase by 11.3 percent if the corresponding trade flows increased by 10 percent.