Wealth gradient in water quality : why the poor are disproportionately affected by the rainy season.

This paper examines the wealth gradient in having improved drinking water at point of use and compares that gradient between rainy and dry seasons. Pooling nationally representative survey data from 14 African countries, I show that households in the richest quintile are associated with a 21.2 percentage point increase in the probability of having improved drinking water at point of use compared to the poorest households in the dry season and I show a differential wealth gradient in the rainy season. However, there is substantial variation across countries; as a result, this estimate is close to zero and is statistically insignificant when including country fixed effects. I study a household’s endogenous decision to treat the water or choose a better source to examine the mechanisms driving the wealth gradient: I show the wealth gradient is driven by households in the fourth and fifth quintile being more likely to treat their water and by being more likely to use better water sources compared to households in the poorer wealth quintiles, consistent with and without the country fixed effects.