NYCuriosity Data ← Croton Aqueduct

Death Rate

Visuals in this post: Population Water Use Cost Death Rate Build Times
Infrastructure November 2026

Clean Water Was the Start of Beating Disease

Croton water arrived in 1842, yet the city's worst cholera years came after it: about 46 deaths per 1,000 residents in 1849 and 1854. The death rate averaged 33 per 1,000 from 1842 to 1865, against 26 before 1832, and fell for good only from the 1890s, after the Metropolitan Board of Health (1866), sewers and, later, filtration.

New York's death rate rose after Croton opened and fell for good only in the 1890s

Crude death rate, deaths per 1,000 residents, New York City, 1804–1920. Dashed lines mark the arrival of Croton water (1842) and the creation of the Metropolitan Board of Health (1866).

Source: Michael Haines, The Urban Mortality Transition in the United States, 1800–1940, NBER Historical Paper 134 (2001), Figure 1 Note: Values read from the published figure's plotted points and are accurate to about half a death per 1,000. The figure has no point for 1906.

The deadliest years came after the aqueduct. Cholera returned in 1849, 1854 and 1866, when Irish and German immigration was crowding lower Manhattan and the city still had few sewers. Haines finds that baseline mortality was rising across American cities before the Civil War, even as their economies grew.

Croton was the precondition. Abundant water made sewers, street cleaning and household plumbing possible, while the Metropolitan Board of Health, created by the State in 1866, and later filtration and chlorination turned that water into lower death rates. Nationally, clean water accounts for nearly half of the drop in big-city mortality from 1900 to 1936 (Cutler and Miller, 2004).

Source: Michael Haines, The Urban Mortality Transition in the United States, 1800–1940, NBER Historical Paper 134 (2001), Figure 1. Haines compiles New York City vital registration from the City Inspector's and Health Department reports and Rosenwaike's Population History of New York City (1972), and notes "serious mortality peaks, notably from the cholera epidemics of 1832, 1849, 1854, and 1866."

Method: the paper prints the series only as a chart. Each yearly value here was read from the coordinates of the chart's plotted points in the PDF and mapped to its axis (10 to 50 deaths per 1,000), so values are approximate. The period averages in the text are the author's arithmetic on those values.