The construction of public waterworks across the United States in the late 19th and early 20th centuries was one of the great infrastructure investments in American history. As David Cutler and Grant Miller have demonstrated, these waterworks accounted for “nearly half of the total mortality reduction in major cities, three-quarters of the infant mortality reduction, and nearly two-thirds of the child mortality reduction.” Within a generation, the scourge of waterborne infectious diseases – from cholera to typhoid fever – was largely eliminated. Moving to a city no longer took years off your life, a sociological trend that unleashed untold amounts of human innovation.

However, not all urban waterworks were created equal. Some systems were built with metal pipes containing large amounts of lead. (At the time, lead pipes were considered superior to iron pipes, as they were more durable and easier to bend.) Unfortunately, these pipes leached lead particulates into the water, exposing city dwellers to water that tasted clean but was actually a poison.

Over the last few decades, researchers have amassed an impressive body of evidence linking lead exposure in childhood to a tragic list of symptoms, including higher rates of violent crime and lower scores on the IQ test. (One study found that lead levels are four times higher among convicted juvenile offenders than among non-delinquent high school students.) In 2014, I wrote about a paper by Jesssica Wolpow Reyes that documented the association between leaded gasoline and violent crime:

Reyes concluded that “the phase-out of lead from gasoline was responsible for approximately a 56 percent decline in violent crime” in the 1990s. What’s more, Reyes predicted that the Clean Air Act would continue to generate massive societal benefits in the future, “up to a 70 percent drop in violent crime by the year 2020.” And so a law designed to get rid of smog ended up getting rid of crime. It’s not the prison-industrial complex that keeps us safe. It’s the EPA.

But these studies have their limitations. For one thing, the pace at which states reduced their use of leaded gas might be related to other social or political variables that influence the crime rate. It’s also possible that those neighborhoods with the highest risk of lead poisoning might suffer from additional maladies linked to crime, such as poverty and poor schools. To convincingly demonstrate that lead causes crime, researchers need to find a credible source of variation in lead exposure that is completely independent (aka exogenous) to the factors that might shape criminal behavior.

That is the goal of a new paper by James Feigenbaum and Christopher Muller. Their study mines the historical record, drawing from homicide data between 1921 and 1936 (when the first generation of children exposed to lead pipes were adults) and the materials used to construct each urban water system. If lead was responsible for higher crime rates, then those cities with higher lead content in their pipes (and also more acidic water, which leaches out the lead) should also experience larger spikes in crime decades later.

What makes this research strategy especially useful is that the decision to use lead pipes in a city’s water system was based in part on its proximity to a lead refinery. (Cities that were closer to a refinery were more likely to invest in lead pipes, as the lower transportation costs made the “superior” option more affordable.) In addition, Feigenbaum and Muller were able to look at how the lead content of pipes interacted with the acidity of a city’s water supply, thus allowing them to further isolate the causal role of lead.

The results were clear: cities that used lead pipes had homicide rates that were between 14 and 36 percent higher than cities that opted for cheaper iron pipes.

These violent crime increases are especially striking given that those cities using lead pipes tended to be wealthier, better educated and more “health conscious” than those that did not. All things being equal, one might expect these places to have lower rates of violent crime. But because of a little noticed engineering decision, the water of these cities contained a neurotoxin, which interfered with brain development and made it harder for their residents to reign in their emotions.

The brain is a plastic machine, molded by its environment. When we introduce a new technology – and it doesn’t matter if it’s an urban water system or the smartphone – it’s often impossible to predict the long-term consequences. Who would have guessed that the more expensive lead pipes would lead to spikes in crime decades later? Or that the heavy use of road salt in the winter would lead to a 21st century water crisis in Flint, as the chloride ions pull lead out of the old pipes?

One day, the scientists of the future will study our own blind spots, as we invest in technologies that mess with the mind in all sorts of subtle ways. History reminds us that these tradeoffs are often unexpected. After all, it took decades before we realized that, for some unlucky cities, even clean water came with a terrible cost.

James Feigenbaum and Christopher Muller. "Lead Exposure and Violent Crime in the Early Twentieth Century." Explorations in Economic History (2016)