For centuries, historians have debated what caused the Roman Empire to fall. Entire theories of political and cultural development have grown from speculation over why the world's strongest empire collapsed from within.
In the 1980s, attention turned to a new suspect: lead poisoning. The Roman elites flavored their food and wine with syrups cooked in iron pots. They were prone to gout, which is a sign of lead poisoning. Other symptoms of lead exposure include mental retardation, impaired judgment and erratic behavior. Among those afflicted were the erratic emperors Claudius, Caligula and Nero.
Now, for today's example: For decades, experts have debated the sharp rise of violent crime in America in the 1980s and the surprising fall in crime in the 1990s, a decline that continues today.
Theories abound: The rise in crime has been blamed on the permissive values of the 1960s, the persistence of poverty and racism, the spread of drugs, guns and violent video games. The fall in crime has been credited to the "broken windows" theory of community policing and to get-tough policies that have given the U.S. the highest incarceration rate in the world.
There's even a theory, espoused by the authors of "Freakonomics," that credits Roe v. Wade to the drop in violent crime. More abortions in the 1970s, they argue, meant fewer unwanted children to become delinquents in the 1990s.
Now, there's a new suspect: Lead poisoning.
If you didn't know violent crime is on the decline, you probably watch too much TV. The FBI's statistics are more reliable. They list 158 violent crimes per 100,000 population in 1961, more than doubling, to 396 in 1971. It nearly doubled again in the next decade, hitting 593 in 1981, then lurched even higher.
By 1991, the violent crime rate hit 758 crimes per 100,000 population. By then, politicians were shoveling money into more police forces, had been getting ever tougher on drugs, had been stiffening sentences and building prisons. Experts were predicting worse things to come, as juvenile "super-predators," raised in urban crack dens by single mothers, came of age.
Then crime started falling: By 1995, the violent crime rate hit 684; by 2002 it was 494. The 2011 rate, the latest available, was 386 per 100,000 population - the lowest rate since 1970.
What happened? Every theory has its advocates, and most have a special interest that benefits from its acceptance. But none of the theories matches up perfectly with the data.
The rapid fall in violent crimes, for instance, takes the wind out of the arguments blaming the rise in crime on drugs, single parenthood, violence in media and the decline in church-going. Those things didn't change, but crime came down anyway.
The lead poisoning thesis starts with a different curve. General Motors began encouraging the use of leaded gasoline in their vehicles back in the 1930s, to reduce engine knocking. Its use grew dramatically until 1970, when the Clean Air Act required new cars to have catalytic converters, which don't work with leaded gas. By 1990, leaded gasoline was prohibited entirely.
Lay the curves on top of each other, with a 20-year lag, and the similarity is stunning. As more cars spewed more lead into the air, more children absorbed it, especially in auto-congested cities. As they grew to adolescence and young adulthood, those kids committed more violent crimes.
In a compelling review of recent research in "Mother Jones," Kevin Drum shows this is no mere coincidence. In study after study, of different cities, different states, and different countries, the presence of lead poisoning in children correlates to higher rates of violent crime. It appears to be true even at the neighborhood level: a Tulane researcher found that New Orleans neighborhoods with high concentrations of lead matched neighborhoods with higher crime levels.
Drum also cites longitudinal studies. Researchers in Ohio followed a group of students in the 1980s, measuring lead levels in their blood every six months. "At age 7, kids with higher lead levels were doing worse in school," Drum writes. "At age 17, they were more heavily involved in juvenile delinquency. At age 27, they had higher arrest rates for violent crimes."
Other researchers, using new brain imaging technology, have learned more about the effects of lead poisoning. Among other things, even tiny levels of lead compromise functions in the frontal lobe, which is where impulse control, aggressiveness, concentration, emotional regulation and higher-level thinking are done.
As a result of these and other studies, experts have changed their definition of a "safe dose" of lead. Even in the smallest concentrations, lead reduces IQ, they've found. Now they say there is no safe dose.
This generations-long experiment in brain chemistry continues to surprise. We've always considered cities to be dangerous places, hotbeds of crime. But now that cars don't pour lead out their exhaust pipes, Drum writes, the difference in violent crime rates between cities and the country has all but disappeared.
Another surprising finding: The curve showing the rate of teen pregnancy in the U.S. matches the rise and fall of leaded gasoline as well.
Where do these conclusions take us? For starters, we need to look at the lead that still fills the streets and playgrounds of America's cities. Nobody is pumping leaded gas, but the old lead is still floating around. Lead paint in older homes is an even greater threat to children's development than we knew.
Then, if subsequent research confirms the lead poisoning thesis, we should do more. Tests for lead levels in blood should be expanded, especially in schools and jails. We could use more research on treatment and remediation. Let's also consider the 7 million people locked up in America's prisons, and whether the laws that put them there were misguided.
And maybe we should think about whether the Clean Air Act was the most important crime prevention measure in history, keeping the United States from going the way of the Roman Empire.
Rick Holmes, opinion editor for the Daily News, blogs at Holmes & Co. He can be reached at firstname.lastname@example.org.