Identifying macro-determinants of remittance flows to a developing country: The case of Uganda

This paper employs minimum Lagrange Multiplier (LM) unit root tests for endogenous structural breaks combined with ARCH and GARCH models to investigate how key macrovariables impact diaspora remittances. Since remittances can reverse-cause exchange rate movements and domestic income, we use changes in the world price of oil denominated in U.S. dollars to proxy movements in the Uganda shilling nominal effective exchange rate. To control for endogenous bias between remittances and income, we use rainfall shocks as proxies for income shocks in a non-oil-producing developing economy dominated by agricultural sector and its related activities. In addition, large movements in oil price and rainfall shocks typically cause large supply shocks that can significantly impact size of remittance inflows. We control for interest rate differential, political business cycles and seasonality. Results indicate that accounting for structural change in intercepts (levels) and slopes (trends) of key macroeconomic determinants of remittances around their major structural break points significantly increases their explanatory power. In particular, positive (negative) innovations in income and depreciation (appreciation) in the currency of a recipient developing country are negatively (positively) correlated with remittance inflows. These results are robust across different model specifications.