How Do Remittances Affect Child Mortality at Different Levels of Mortality? Estimating Unconditional Quantile Treatment Effects on Three Leading Causes of Child Mortality

There is an emerging view that remittances improve health outcomes in developing countries, though less is known about the conditions under which they may be effective. This study investigates whether and how the impact of remittances on child mortality depends on the level of mortality itself. Using a sample of 134 developing countries over the period 1990 to 2018, we estimate unconditional quantile treatment effects with an endogenous treatment variable, on aggregate child mortality and the three leading cause-specific child mortality rates: mortality from neonatal disorders, lower respiratory infections, and diarrheal diseases. We find that the impact of remittances differs systematically across the mortality distribution and across mortality indicators. Remittances appear less effective at reducing child mortality at the lowest and highest mortality rates than at the average. Remittances, money sent by migrant workers to their home countries, amounted to US$605 billion for developing countries in 2021. Recent research finds that remittances, on average, improve overall child health. However, it is theoretically unclear whether such benefits accrue to developing countries that cannot be considered average. In particular, do countries with non-average levels of child health also benefit? The purpose of this paper is to analyze whether and to what extent remittances influence child health at all levels of child health, and not just the average. We do so using novel quantile modeling techniques that enable us to more accurately quantify these differences. Secondly, we investigate whether remittances improve child health from three leading causes of child mortality. We find that remittances reduce child mortality in countries with low and high levels of mortality. However, remittances seem to have the most benefit for child health in countries with an average level of child mortality. Remittances also reduce the rate of the three leading causes of child mortality, but they seem to be more effective in mortality arising from infectious diseases. Our findings imply that countries at different levels of mortality cannot rely on remittances in the same way to improve public health outcomes. Coordinated planning at a national budget level and among donors is indeed required to improve healthcare alongside remittance inflows, even in countries experiencing high remittance inflows. One limitation is that we do not consider the impact of non-monetary migrant transfers which anecdotally may be relatively large.