Skip to main content

A novel new research projection suggests crime rates will decrease in the coming years if demographic and economic trends hold steady — even if people are imprisoned at lower rates. 

Crime Rates Unrelated to Prison Rates

The analysis from three top criminologists evaluated trends like inflation and teen birth rates, concluding that those factors “outweigh the impact of imprisonment rates on crime.”

Explaining the Past and Projecting Future Crime RatesDownload

Further, they found, “a state’s level of imprisonment has little to do with what happens to its rate of murder over time and nothing to do with its rate of robbery.” New York, for example, which decreased its prison population by about a third since the late 1990s, still saw crime decrease over 40% during the same time frame.

Reacting to the study, Alex Piquero, a widely cited criminologist and chair of the sociology department at the University of Miami, said it was an important contribution to the literature at a time when candidates in the current presidential race have made high-stakes claims about law and order. 

“People thought, for a long time, imprisonment is going to reduce crime. But it doesn’t have the effect anybody believes it does,” said Piquero, who was not involved in the study but has examined similar issues. “The report is useful; it gives people a barometer of where we are.”

The paper – authored by James Austin, Todd Clear and Richard Rosenfeld – was released last month by the Harry Frank Guggenheim Foundation, which funds research on “violence, aggression and dominance,” according to its website. In an interview, Austin said he was prompted to do the study to counteract a common perspective he heard from politicians in states and counties seeking to reduce their prison populations. 

“The first argument that always comes up is, if you do that, the crime rate is going to go up,” said Austin, a director with the JFA Institute, a criminal justice consulting firm for public agencies. “It’s something I needed to test out.”

Austin’s co-authors, professors Clear and Rosenfeld, are both past presidents of the American Society of Criminology. Clear works at Rutgers University and Rosenfeld at the University of Missouri-St. Louis. 

James Austin, Todd Clear, and Richard RosenfeldDownload

Scroll to Continue

Recommended Articles

For their recent study, the trio created two models measuring socioeconomic factors against crime data from the 1980s through 2016. Crime rose and fell dramatically in that time, consistent with steady price inflation, an aging population nationwide and a steep decline in the teen birth rate, among other demographic shifts. 

Current trends suggest violent crime will “remain flat through 2021,” the paper projects. Property crime would also continue declining, according to the study authors. And even under the worst-case scenarios – such as higher inflation or divorce rates – trends don’t suggest crime rates would come close to the peaks of the 1980s and 1990s.

Referring to household sizes and related factors, which tend to be slow-moving variables that have trended downward with crime, Austin said “it’s all kind of come together. We’re producing cohorts of boys to men who don’t have juvenile arrests.”

The authors also considered juvenile arrests as a predictor of future arrests as an adult. Even though the majority of individual juvenile offenders don’t become adult criminals, an overall decrease in juvenile arrests correlates with decreasing adult arrests.

The authors argue a steep drop in the number of juveniles arrested, from nearly 3 million in the early 1990s to 1 million five years ago, “bodes well for future crime rates.”

Still, the study warns that the “exogenous” effects – dramatic events like a pandemic, natural disaster, or severe recession – can upend even the most careful predictions.

The current coronavirus pandemic could qualify, but Austin said that barring permanent, deeper shifts in underlying trends like inflation or household size, he doesn’t expect the modest year-over-year increase in shootings in some parts of the country to continue.

Piquero at the University of Miami offered a similar assessment but also cautioned that these types of predictions become less reliable after a few years have passed, as variables like household size become harder to predict. 


“Projecting 20 years from now, that’s a fool’s errand,” he said. “But it’s OK to project the next three or four years. We’ll know the answer in the next few years, and this study is a useful baseline.”

Michael Fitzgerald
The Imprint

This story originally appeared in The Imprint, a daily news publication dedicated to rigorous, in-depth journalism focused on families and the systems that impact their lives.