Revisiting the Migration- Development Nexus

This paper presents empirical estimates of a gravity model of bilateral migration that properly accounts for non-linearities and tackles causality issues through an instrumental variables approach. In contrast to the existing literature, which is limited to OECD data, we have estimated our model using a matrix of bilateral migration stocks for 127 countries. We find that the inverted-U relationship between income at origin and migration found by other authors survives the more demanding bilateral specification but does not survive both instrumentation and introduction of controls for the geographical and cultural proximity between country pairs. We also evaluate the effect of migration on origin and destination country income using the geographically determined component of migration as a source of exogenous variation and fail to find a significant effect of migration on origin or destination income.