Trends of improved water and sanitation coverage around the globe between 1990 and 2010: inequality among countries and performance of official development assistance

BACKGROUND As the Millennium Development Goals ended, and were replaced by the Sustainable Development Goals, efforts have been made to evaluate the achievements and performance of official development assistance (ODA) in the health sector. In this study, we explore trends in the expansion of water and sanitation coverage in developing countries and the performance of ODA. DESIGN We explored inequality across developing countries by income level, and investigated how ODA for water and sanitation was committed by country, region, and income level. Changes in inequality were tested via slope changes by investigating the interaction of year and income level with a likelihood ratio test. A random effects model was applied according to the results of the Hausman test. RESULTS The slope of the linear trend between economic level and sanitation coverage has declined over time. However, a random effects model suggested that the change in slope across years was not significant (e.g. for the slope change between 2000 and 2010: likelihood ratio χ2 = 2.49, probability > χ2 = 0.1146). A similar pro-rich pattern across developing countries and a non-significant change in the slope associated with different economic levels were demonstrated for water coverage. Our analysis shows that the inequality of water and sanitation coverage among countries across the world has not been addressed effectively during the past decade. Our findings demonstrate that the countries with the least coverage persistently received far less ODA per capita than did countries with much more extensive water and sanitation coverage, suggesting that ODA for water and sanitation is poorly targeted. CONCLUSION The most deprived countries should receive more attention for water and sanitation improvements from the world health community. A strong political commitment to ODA targeting the countries with the least coverage is needed at the global level.