Comment on 'Pubic Health Efforts and the Decline in Urban Mortality'

Abstract

In this comment, we address points raised by Mark Anderson, Kerwin Charles, and Daniel Rees’s August 2018 NBER working paper entitled “Pubic Health Efforts and the Decline in Urban Mortality.” The points generally fall into three categories: (1) assignment of differing clean water intervention dates, (2) construction of differing mortality rates using different population denominators, and (3) computation of correct standard errors. Ultimately, a large share of the discrepancies between our 2005 analysis and theirs is due to the construction of population denominators for mortality rates when such denominators are not known for certain. After carefully considering their points and correcting the unambiguous mistakes in our original data, our revised estimates suggest that municipal water disinfection (filtration) explains 38% of the mortality decline in our sample cities and study years – a result not dramatically different from the estimated 43% in the original paper.