The Payor of Last Resort · Episode 5 of 12 · 9 min
The Great Purchases
Three times public health bought the hazard instead of the person, and the arithmetic came back overwhelming: clean water, childhood vaccination, and unleaded gasoline. The episode walks each purchase with its dates, then puts the published correction beside every headline number, including a conceded data mistake in the famous clean-water return. What survives is a diagnostic of four shared conditions, with measles coverage now eroding.
Who's who
- Hiram F. MillsEngineer whose slow sand filter began treating Lawrence's water in 1893
- John L. LealPhysician who began chlorinating Jersey City's water in 1908
- George Warren FullerSanitary engineer who built the Boonton chlorination plant in ninety-nine days
- William J. MagieSpecial master who ruled chlorinated water lawful in 1910
- David M. CutlerEconomist co-author of the 2005 paper on public health improvements
- Grant MillerEconomist co-author of the 2005 paper and the 2022 reply
- D. Mark AndersonEconomist who re-analyzed the clean-water data in 2022
Where it happens
- Jersey City, New Jersey+1 more episodeFirst continuous municipal chlorination of a public supply, 1908
- Lawrence, MassachusettsSite of the 1893 slow sand filter
- Boonton Reservoir, New JerseySource of Jersey City's water and site of the chlorination plant
- Milwaukee, Wisconsin+1 more episodeSite of the 1993 Cryptosporidium outbreak
Timeline
- September 1893Lawrence, Massachusetts, begins operating a slow sand filter.
- September 26, 1908Jersey City's water is chlorinated, before a court has ruled it lawful.
- February 1985EPA publishes its regulatory analysis of the leaded-gasoline phasedown.
- October 1, 1994The Vaccines for Children program becomes operational.
- Spring 1993Roughly 403,000 people fall ill in Milwaukee's water outbreak.
- April 2022Cutler and Miller concede data mistakes and revise filtration's share of the mortality decline to 38 percent.
Still debated
The size of filtration's effect is disputed between the revised 38 percent and a re-analysis calling it fragile, and the 23-to-1 return has never been recomputed. Vaccination ratios rest on modeled counterfactuals and count productivity, and Milwaukee's death toll is a reported minimum.
