In search of the remittances resource curse across developing countries

This paper analyses the impact of remittances on corruption in developing countries and how the impact varies with the level of corruption itself. Using a novel instrumental variable for remittances and two indicators of corruption, we estimate mean regression effects and unconditional quantile treatment effects with an endogenous treatment variable in a sample of 135 developing countries over the period 1996–2020. We contribute new findings that, on average, remittances do not impact corruption. This finding holds when we consider the influence of democracy and rule of law as moderating factors in the remittances-corruption relationship. In quantile regressions, when they do have an impact however, remittances appear to exacerbate corruption in countries that have better control of corruption, while they also appear to reduce corruption for countries that have more difficulty with corruption control. Our results suggest that the notion of a remittances resource curse may not be a phenomenon for most countries. Importantly, developing countries with challenges in controlling corruption and who are dependent on remittances may not experience a remittances resource curse that hinders long-term development.