Differences-in-Differences: Handwashing and Waterborne Transmission of Cholera

 From: Carlos Chavez, On Causality

Difference-in-differences has undergone a methodological renaissance. The method’s intellectual roots reach back further than most economists realize - to the 1840s and 1850s.

• In 1847, Ignaz Semmelweis used comparative clinic analysis to show that handwashing prevented childbed fever, comparing mortality between Vienna’s physician-staffed ward (where medical students performed autopsies) and its midwife-staffed ward, then tracking outcomes after implementing chlorine disinfection.

•  John Snow’s 1854 investigation of cholera in London is credited as the first recorded application of difference-in-differences proper. Snow compared mortality rates across households served by two water companies: the Lambeth Company, which had moved its intake upstream of the city’s sewage outflow in 1852, and the Southwark & Vauxhall Company, which continued drawing contaminated water. By comparing changes in cholera deaths before and after Lambeth’s relocation - while Southwark & Vauxhall served as the control - Snow produced a cleaner diff-in-diff design, with parallel pre-treatment trends giving way to dramatic divergence. His “Grand Experiment” found 315 deaths per 10,000 households among Southwark & Vauxhall customers versus only 37 per 10,000 among Lambeth customers, providing powerful evidence for the waterborne transmission of cholera decades before germ theory was established.