One quick example, claiming cities in the 1920s that switched to lead pipes saw ~25% increase in homicides vs those that didn't. http://scholar.harvard.edu/files/jfeigenbaum/files/feigenbau...