Essays on the Impact of Clean Water on Human Capital and Productivity

My dissertation investigates the effect of access to clean water on human capital. The first chapter examines the effect of municipal provisions of clean water---installation of water filtration plants---on school enrollment and child labor in American cities from 1880 to 1920. Numerous studies show that access to clean water reduces child mortality and morbidity, but little work has been done on the consequences for schooling and child labor. The effects are theoretically ambiguous because improved child health can raise schooling (e.g., better health makes education investments more productive) or lower it (e.g., the opportunity cost of attending school increases because there is a wage premium for healthier workers). Applying a difference-in-differences strategy which exploits variation of water filtration adoption across time and across cities, I find that municipal water filtration has a positive and statistically significant effect on school enrollment. Also, I find a negative effect on child labor, but it is not significant at conventional levels. These effects are most pronounced at ages 14 and 15, which map into the last years of elementary school and are beyond compulsory schooling age in some states. Additionally, I find that effects are larger for children who are exposed at an earlier age, can legally drop out of school, are from lower socioeconomic status families, or are female.

The second chapter uses a water services privatization program in Argentina during 1991 to 1999 to investigate the effect of early childhood exposure to clean water on educational attainment. By using a difference-in-differences strategy which exploits variation across regions and across cohorts, my results show that early childhood exposure to privatization has a zero effect on primary and compulsory school completion, and a small negative effect on secondary school completion. Furthermore, I find that the effect of this privatization program is heterogeneous across individuals who lived in nonpoor and poor municipalities. I find that, for primary and compulsory school completion, early childhood exposure to privatization has a zero effect in nonpoor municipalities and a negative effect in poor municipalities. A supplemental analysis adding information on years of primary school completed among individuals who did not complete primary schooling indicates that privatization induces individuals in poor municipalities to drop out of school at 5th, 6th, and 7th grade, which correspond to the final years of primary school.