The existing literature on determinants of migration flows typically claims that income differences across countries should be a pushing factor for people's movement. We suggest that institutional quality is a better proxy for the factors that trigger migration. People may well want to stay in or move to relatively poor countries if institutions are good, partly because good institutions have an intrinsic value for people and partly because good institutions may be a sign of future economic growth. In contrast, low income and absolute poverty work both as push-factors and as credit constraints, so that people may want to leave, but few can afford to migrate when poverty is high.
We test our hypotheses using new data on bilateral migration flows from Abel and Sander (2014), the Worldwide Governance Indicators and the World Bank data on headcount poverty, using a migration gravity model with a spatial specification. Controlling for both source and destination income levels, we find that institutional quality matters significantly for migration. Poor institutions act as a push factor, while absolute poverty in a country of origin limits migration. We also find that omitting spatial factors biases the effect of institutions upwards.