An intense debate has broken out regarding whether the recent rise in violent crime in a subset of American cities indicates a return to the horrific bloodshed of a generation ago. The argument emerges in part from the fact that criminologists have never fully agreed on why America had such an astonishing rise in violent crime from the mid-1960s to early 1990s nor why it dissipated so rapidly afterward.

The most popular class of explanations focus on the character of different generations. One theory holds that large birth cohorts (e.g., the baby boomers) are more prone to deviant behavior because they overwhelm the capacity of relatively smaller, older generations to supervise and socialize them. Another account maintains that the introduction of leaded gasoline and paint exposed several generations to a neurotoxin that made them more prone to violence, and that when lead was phased out by regulators, subsequent generations became more peaceable. A third hypothesis is that America’s drop in violent crime was a result of the arrival of widely available legalized abortion in the 1970s, which removed from the population millions of children who were at higher-than-average risk of growing up to become criminals.

Yet these generation-based explanations run into an inconvenient reality highlighted by researchers Phil Cook and John Laub using violent-crime arrest data from the U.S. Bureau of Justice Statistics. Focusing on the peak violence year of 1994, Cook and Laub showed that multiple generations contributed both to the waxing and the waning of violent crime. A concrete example of how the data in the chart do not fit generational theories is illuminating: If the extraordinarily high violent crime rate of 20-year-olds in 1994 had been due to the character of that generation (i.e., people born in 1974), it’s very hard to explain how in 2004, when they were 30 years old, this “violent generation” was no more violent than were the 30-year-olds of 1984 (e.g., the generation born in 1954).

(Chart generated by Keith Humphreys and Phil Cook using data from the U.S. Bureau of Justice Statistics)

Violent crime spikes thus cannot be understood simply as generational quirks; rather they are driven by the spirit of the times. The 1980s and 1990s, for example, were characterized by thriving, violent markets in cocaine and firearms, which swept up many adolescents and adults. When the “crack wars” subsided, violent crime rates dropped sharply among teenagers and people in their 20s and 30s. The violent crime epidemic was the result of changes in context, not the bad character of a particular generation.

Does this mean that generation-specific explanations of crime are wrong? Cook, a professor of public policy at Duke University, says: "Not necessarily — but explanations that focus on how young people, rather than the social environment, are changing should be treated with great caution. At the height of the violence epidemic, the most prominent explanation was a growing prevalence of evil youths — ‘superpredators.’  Shortly thereafter, youth violence rates plunged."

Generations of Americans differ through the life span in their propensity to commit crime. For example, even in late life, the baby boomers are somewhat more criminally active than their predecessors. But these differences between generations seem smaller than those between different historical periods, suggesting that policymakers striving to reduce crime should focus less on "kids these days" and more on malleable environmental forces that can push multiple generations of Americans simultaneously into or out of crime.

Keith Humphreys is a professor of psychiatry at Stanford University.