Does spending on refugees make a difference? A cross-sectional study of the association between refugee program spending and health outcomes in 70 sites in 17 countries

Timothy M Tan, Paul Spiegel, Christopher Haskew, P Gregg Greenough

Numerous simultaneous complex humanitarian emergencies strain the ability of local governments and the international community to respond, underscoring the importance of cost-effective use of limited resources. At the end of 2011, 42.5 million people were forcibly displaced, including 10.4 million refugees under the mandate of the United Nations High Commissioner for Refugees (UNHCR). UNHCR spent US$1.65 billion on refugee programs in 2011. We analyze the impact of aggregate-level UNHCR spending on mortality of refugee populations. Using 2011 budget data, we calculated purchasing power parity adjusted spending, disaggregated by population planning groups (PPGs) and UNHCR Results Framework objectives. Monthly mortality reported to UNHCR's Health Information System from 2011 to 2012 was used to calculate crude (CMR) and under-5 (U5MR) mortality rates, and expressed as ratios to country of asylum mortality. Log-linear regressions were performed to assess correlation between spending and mortality. Mortality data for 70 refugee sites representing 1.6 million refugees in 17 countries were matched to 20 PPGs. Median 2011 spending was $623.27 per person (constant 2011 US$). Median CMR was 2.4 deaths per 1,000 persons per year; median U5MR was 18.1 under-5 deaths per 1,000 live births per year. CMR was negatively correlated with total spending (p = 0.027), and spending for fair protection processes and documentation (p = 0.005), external relations (p = 0.034), logistics and operations support (p = 0.007), and for healthcare (p = 0.046). U5MR ratio was negatively correlated with total spending (p = 0.015), and spending for favorable protection environment (p = 0.024), fair protection processes and documentation (p = 0.003), basic needs and essential services (p = 0.027), and within basic needs, for healthcare services (p = 0.007). Increased UNHCR spending on refugee populations is correlated with lower mortality, likely reflecting unique refugee vulnerabilities and dependence on aid. Future analyses using more granular data can further elucidate the health impact of humanitarian sector spending, thereby guiding policy choices.