The Lead-Crime Hypothesis: Your Children, Lead Exposure, and Crime Rates

The lead-crime hypothesis suggests that childhood lead exposure leads to higher crime rates roughly 20 years later.

Through the 1970s and 80s crime rates in New York City skyrocketed. Murders, robberies, and rapes were up 5-fold, 14-fold, and 4-fold respectively since the 1960s, leaving residents feeling like hostages in their own city. When Rudy Giuliani ran for mayor in 1993, he relied on his tough-on-crime image, as he was the former federal prosecutor who brought down the mob by successfully applying the RICO statute to its once untouchable leaders.

Giuliani won, crime rates fell sharply, and NYC experienced an unprecedented period of tranquility. Giuliani’s popularity soared, and his police chief, Bill Bratton, made the cover of Time. Their success was attributed to the broken windows policing philosophy, which explains that if a building has a broken window, the other windows are more likely to be broken. Using this logic, crimes are more likely to be committed if other crimes are being committed. Small crimes like loitering, jumping a turnstile, jaywalking, etc. made serious crimes like rape and murder more likely. Therefore, Giuliani and Bratton cracked down on minor crimes, and as expected, violent crime rates fell as well. For many at the time, and even today, broken windows was hailed as a success.

However, Kevin Drum in a 2013 article for Mother Jones disputed broken windows. He acknowledges that when Giuliani and Bratton took power in 1993, “rape rates had dropped 17 percent, assault 27 percent, robbery 42 percent, and murder an astonishing 49 percent,” but the decline actually began in 1990, before they took office. And the decline continued after they left office, when new mayors and new police chiefs used different policing philosophies. It seems as if the Giuliani and Bratton’s tenure was just a coincidence. Drum’s most damning argument, though, is that cities all over the US saw a similar pattern, in which crime peaked in the early 1990s and then dropped precipitously. As Drum put it, “Washington, DC, didn’t have either Giuliani or Bratton, but its violent crime rate has dropped 58 percent since its peak. Dallas’ has fallen 70 percent. Newark: 74 percent. Los Angeles: 78 percent.” Something else was happening. Other factors were at play.

Although the idea had already been circulating among researchers for decades, Kevin Drum made popular the lead-crime hypothesis. His logic was simple:

1. Academic literature has already shown that childhood exposure to lead, a well-known neurotoxin, causes a drop in IQ and self-control, and both of these make a person more likely to commit a crime.
2. Lead has been present in gasoline, paint, water pipes, and many other products for decades, and inner-city children are exposed at much higher rates than other children.
3. When the amount of lead in products dropped sharply in the late 1970s, children born around that time were exposed to lower concentrations.
4. When these children became adults roughly 20 years later and capable of committing serious crimes they were less likely to do so because they had higher IQs and more self-control.

For Drum, the clean streets of NYC and other major US cities was because the new generation of criminals was understaffed. Drum backed up his assertion with a handful of newly published studies. For example, Harvard researcher Jessica Wolpaw Reyes found the exact correlation Drum was looking for. She spent months cold calling state agencies to find lead emission data, and she saw that each state reduced their emissions through the 70s and 80s, thanks to the introduction of the catalytic converter and new EPA rules, though they did so at different rates. If Drum and Reyes were correct, then the crime rates 20 years later for each state should reflect these differences. By using FBI criminal data, Reyes found a nearly perfect match.

“Reduction in childhood lead exposure in the late 1970s and early 1980s is responsible for significant declines in violent crime in the 1990s, and may cause further declines into the future.”

Environmental Policy as Social Policy? The Impact of Childhood Lead Exposure on Crime, Jessica Wolpaw Reyes

But not everyone was convinced. After Drum brought Reyes’ and several other researchers’ work to the public, the skeptics—both professionals and amateurs—came out of the woodwork. Since Drum’s 2013 article, the debate has picked up steam, with new studies having been published supporting both sides.

The Case Against the Lead-Crime Hypothesis

First on the list of criticisms is that crime rates are strongly correlated with a lot of factors, such as education, nutrition, healthcare, etc., including their many sub-factors. The challenge for sociologists and economists has been to isolate these factors to get a clear picture of their effect on crime, which has historically proven difficult. Therefore, the effect Drum, Reyes, and others are attributing to lead exposure could be explained by something else. For example, Steven Levitt, the author of Freakanomics, and John Donohue explained the same crime reduction with the passing of Roe v. Wade in 1973. They used similar logic and found that “1) Legalized abortion leads to fewer ‘unwanted’ babies being born, and 2) unwanted babies are more likely to suffer abuse and neglect and are therefore at an increased risk for criminal involvement later in life.” Their graphs also matched the crime drop 20 years later, seemingly providing a plausible explanation.

Maybe both lead exposure and legalized abortion are at play. Maybe a 3rd factor is responsible, such as increased access to healthcare with the creation of Medicare and Medicaid under the Social Security Amendments of 1965. What about the sweeping education reforms of the 1970s? Theoretically, these too would cause the same crime drop 20 years later.

For many critics of the lead-crime hypothesis, the original research didn’t adequately control for these many possible outside factors. This is because a proper experiment is impossible. To get more accurate results on the connection between lead exposure and crime, participating children would have to be put into randomized groups, subjected to different levels of lead exposure, and observed for 20 years or more. By doing so, researchers would be able to control for other factors. But, of course, intentionally subjecting people, especially children, to lead would be a serious ethical violation. Instead, researchers are forced to look at data on non-randomized groups, meaning any number of factors could be at work.

Furthermore, some researchers believe the connection between lead exposure and crime is exaggerated due to publication bias. This is a phenomenon in which positive, more interesting results are more likely to be published, making a conclusion seem more accurate in the literature than it should be. Negative, less interesting studies tend to get weeded out. A meta-analysis from the University of Glasgow looked at 24 studies and found strong evidence of publication bias, meaning that “the effect of lead is overstated in the literature.” The researchers recalculated the effect of lead exposure on crime using the many studies that hadn’t been selected for publication. They found that lead exposure contributed to 0-5% of crime, instead of the 3-36% found by the more popular studies. However, the researchers acknowledge that their results have not closed the case, as they need to analyze more studies and the studies analyzed may not be comparable.

The Manhattan Institute reached out to Reyes for a reaction on this meta-analysis. She said it was well done, but she remains “confident that there is a strong causal relationship in which childhood lead exposure increases the likelihood of committing violent crimes.”

The Case for the Lead-Crime Hypothesis

While the original studies on the lead-crime hypothesis left many experts unconvinced, more creative studies have been published that can provide a clearer picture.

For example, in this paper published in the journal Explorations in Economic History, authors Feigenbaum and Muller set their sights on lead water pipes. They hypothesized that cities with both lead water pipes and water acidic enough to leach lead into the environment would have higher crime rates roughly 20 years after. Most lead pipes were installed around the turn of the century, so the researchers focused on homicide rates between 1921 and 1936. They found a nearly identical pattern to the one found by Drum, Reyes, etc.

“Lead water pipes exposed entire city populations to much higher doses of lead than have previously been studied in relation to crime. Our estimates suggest that cities’ use of lead service pipes considerably increased city-level homicide rates.”

Lead exposure and violent crime in the early twentieth century, Feigenbaum and Muller

In a different paper from the National Bureau of Economic Research, researchers Aizer and Currie focused on 120,000 children born in Rhode Island between 1990 and 2004. These children were from the same community, meaning the researchers could eliminate factors like education, access to healthcare, income, etc. because they were more or less the same. Instead, Aizer and Currie looked at proximity to busy roads, as the soil near these roads had higher levels of lead contamination due to the long-time use of leaded gasoline. As expected, the closer children lived to a busy road the higher BLL (blood lead level) they had. By comparing rates of deviant behavior in school, they found that “A one-unit increase in lead increased the probability of suspension from school by 6.4-9.3 percent and the probability of detention by 27-74 percent.”

Furthermore, this paper analyzed the behavioral changes caused by CDC intervention on children with a high BLL. When a child’s blood test twice shows that he or she has toxic level of lead in the blood, the CDC gives the family lead abatement services in their surroundings and nutritional counseling. Authors Billings and Schnepel focused on children born in Charlotte-Mecklenburg County, North Carolina between 1990 and 1997 who received CDC help. They compared the children’s BLLs and their records for antisocial behaviors, such as suspension from school, violent crimes, etc. As expected, they found “that the negative outcomes previously associated with early-life exposure can largely be reversed by intervention.”

Public Policy and the Future

The original studies on the lead-crime hypothesis may have been unconvincing for everyone, but the field continues to grow, making the idea seem more and more plausible. Assuming it’s true, what do we do about it?

For starters, the Environmental Protection Agency needs to beef up its regulations. The first attempt at reducing the amount of lead the public was exposed to dates back to 1925, when the US government issued a non-enforceable limit of 0.1 mg/L. Even back then some experts had reason for concern. Although some states had passed enforceable limits since then, it was not until 1986 that federal government passed the Safe Drinking Water Act, which mandated lead-free plumbing. And it wasn’t until 1996 that leaded gasoline and 1978 that lead-based paint were fully banned. The EPA warns that homes, playgrounds, products, etc. created before these dates have a much higher likelihood of being contaminated with lead.

Today, politicians are focused on strengthening the Lead and Copper Rule. First enacted in 1991, this requires water providers to perform rigorous testing, control pipe corrosion, and replace pipes at risk of contaminating users. After the water crisis in Flint, Michigan, in early 2021, the Trump administration pushed the EPA to strengthen and expand this legislation, although these changes have been held up by the Biden administration for review.

Furthermore, federal and state governments need to provide funding for replacing lead-based infrastructure and mitigating the damage already done. The Biden administration pushed for $45 billion to replace old water pipes, although this was chopped down to $15 billion during Senate negotiations over the new infrastructure bill. Biden also pushed for $12 billion in general drinking water funds that can be used by the states to clean industrial sites, perform testing, etc. Also, as already discussed above, the CDC plays a vital role in helping those suffering from lead toxicity. Billings and Schnepel estimated that for every $1 invested in intervention, society saw a return of $1.80, meaning an increase in funding for intervention could be easily justified.

On an international level, 800 million children (about 1 in 3) have lead toxicity levels high enough to warrant action, according to UNICEF. It’s impossible to measure, but it’s worth wondering how this affects global politics, war, terrorism, etc. If childhood lead exposure increases antisocial behavior at the city, state, and national level, it’s not crazy to assume it does the same at the international level, right? Therefore, it’s in the best interest of national governments to help combat this problem in other countries. While doing so would certainly be expensive the costs would be recouped, as UNICEF said that “Childhood lead exposure is estimated to cost lower- and middle-income countries almost USD $1 trillion due to lost economic potential of these children over their lifetime.”

So evidence in favor of the lead-crime hypothesis seems to be strengthening, and if it’s true, governments around the world need to drastically step up efforts to remove lead from water systems, paint, products, etc. and to mitigate the damage that lead has already done. By doing so, crime is likely to continue to fall, making the world more peaceful and productive.

Support The Happy Neuron by clicking the links below:

• Twitter
• Facebook
• LinkedIn
• RSS Feed
• Mail