Search About Newsletters Donate
Support independent, nonprofit journalism.

Become a member of The Marshall Project during our summer membership drive. Our journalism has tremendous power to drive change, but we can’t do it without your support.

Justice Lab

10 (Not Entirely Crazy) Theories Explaining the Great Crime Decline

From aging to gentrification to Prozac.

Over the course of the 1990s, crime rates dropped, on average, by more than one-third. It was a historic anomaly; one that scholar Frank Zimring dubbed “the great American crime decline.” No one was sure how long the trend would last. Then, in 2010, the Bureau of Justice Statistics announced that the homicide rate had reached a four-decade low. (Since then, overall crime rates have remained relatively flat.)

While everyone agrees this is fantastic news, no one, least of all researchers and experts, can agree on exactly why it happened. Below are 10 popular theories for the decline, from abortion to lead to technology to the broken windows theory, with unvarnished views from three leading researchers—Zimring; Richard Rosenfeld, chairman of a National Academy of Sciences roundtable on crime trends; and John Roman of The Urban Institute—on which are the most plausible.

The “abortion filter”

In 1999 and again in 2005, economists Steven Levitt and John Donohue triggered a sensation with their theory that the legalization of abortion was responsible for as much as half of the crime decline. The idea was that a drop in unwanted children led to better parenting and fewer delinquent young men. It’s easy to see why the media loved this hypothesis, which seemed to piss off everyone else: Conservatives hated the notion that abortion might have led to a major social good, while liberals pushed back against the assumption that bad parenting had been the cause of crime in the first place.

Today, the abortion hypothesis holds less water. “Generally speaking, it has been discredited,” Rosenfeld says. Violent crime continued to rise into the early years of the 1990s, when the first generation of boys who went through the “abortion filter” were already in their late teens to early 20s, older than the age—13—at which criminal behavior typically emerges. Evidence from other nations showed little relationship between the legalization of abortion and the crime rate. If abortion does account for a fraction of the American crime decline, it is likely a small one.

The happy pill thesis

Psychologists David Finkelhor and Lisa Jones suggest that crimes committed both by and against young people declined because of the ways in which antidepressants and anti-ADHD medications, like Prozac and Ritalin, improved human behavior and moods.

Rosenfeld considers this work fascinating but preliminary. “There are individual children who were highly aggressive, started their regimen of Adderall, and then calmed down. But it’s a far cry from that connection to suggest the population rate of crime has come down as a consequence of the spread of this new type of medication. Case not proved.”

The lead hypothesis

This is perhaps the trendiest crime decline hypothesis in 2014, and it was the subject of an entire session of the National Academy roundtable’s work. The effect of lead on children’s brains has been well documented: Exposure to the chemical causes aggressive behavior and cognitive delays. In the wake of the 1970 Clean Air Act, lead was removed from gasoline and, increasingly, from paint.

Economist Jessica Reyes estimates that phasing out lead was responsible for 56 percent of the reduction in violent crime (although she could find no relationship between lead and property crime). Experts still disagree about how plausible it is that lead alone could have been responsible for such a massive portion of the crime drop. “It’s fair to say that our assessment of this hypothesis is ‘case not proved’—or at least ‘not yet proved,’ ” Rosenfeld said.

Aging boomers

Historians agree that the crime wave of the 1960s and 1970s had a lot to do with the baby boom: There were more young men than ever before, and young men are the people who commit most crimes. As the boomers aged out of trouble in the early 1980s, crime fell. But from 1992 to today, the period of the crime decline, there was no significant decrease in the number of young men in the United States. This suggests the size of the youth population is no longer a major driver of crime.

A small caveat: Some experts believe the growth in the population over the age of 50 has contributed to better public safety, because there are more adults monitoring young people’s behavior. “My sense is that there is some connection there,” Rosenfeld says, but if so, it would account for only a small percentage of the crime drop.

The tech thesis

Air conditioning and television brought people from the streets into their homes, a safer environment. Cars are now much more difficult to steal, because of sophisticated ignition and locking systems. The introduction of debit cards, including for the distribution of welfare benefits, means people carry less cash, which makes them less likely to become victims.

“These forces change the dynamic of normal life,” Roman says. “How much do they explain the crime decline in the 1990s? A little bit. But they are dwarfed by the end of the crack epidemic and all the violence associated with it.”

Crack is wack

There is little doubt that decreased demand for crack and heroin drove some of the initial drop in crime in the early 1990s. Crack was “a single generation drug,” Rosenfeld says. “The younger generation saw the devastation crack was doing in the neighborhoods and to their older cousins, brothers, sisters, and parents. You think to the image of the crack head. It wasn’t a pretty sight. There was nothing cool about it.” Yet because the cooling of the crack market happened so long ago, it is not a likely cause for the continued decline over the past two decades.

The roaring ’90s (and Obama-mania)

Until recently, there was little doubt that the strong economy of the 1990s and 2000s played a major role. Unemployment was low, so young men were less likely to turn to the drug trade or other criminal activity for work. Consumers felt confident in their spending power, so they did not seek out cheap, stolen goods on the black market, which decreased the demand for property crime. And though this may be difficult to remember today, Americans generally felt optimistic about their public institutions during the Clinton years.

Then came the Great Recession of 2008, from which we are still recovering. Unemployment spiked, consumer confidence plummeted, trust in political elites sank to record lows ... yet crime did not increase.

The Record

The best criminal justice reporting from around the web, organized by subject

Does this mean there is no connection between the economy, trust in public institutions, and crime? Maybe not. Criminologists went back to the drawing board to consider what they might have overlooked in their previous research. It turned out that one economic indicator—high inflation—has been correlated throughout history with high crime, Rosenfeld says. Yet inflation remained low throughout the recent recession. Does this mean crime will spike when and if inflation rises? Only time will tell. Zimring, however, is skeptical about the inflation explanation, noting that in Europe, changes in inflation have not been associated with changes in the crime rate. “There’s no consistent there there,” he says.

Rosenfeld also wonders if African-American optimism after Barack Obama’s presidential election as president may have helped keep crime rates down. “African-Americans, from 2008 to 2009, were more confident than whites were in the economy, even though on an objective level, they suffered more, economically. There is little question their spirits were buoyed by the Obama phenomenon. Can we link that in some rigorous fashion to the crime drop? Not really. But I think it’s an intriguing hypothesis.”

The prison boom

Because of increased drug-related arrests and mandatory minimum sentences, the American incarceration rate has more than quadrupled since 1970. Long prison sentences were supposed to incapacitate career criminals. Yet because criminals “age out” of law breaking whether or not they are in jail, incarceration itself could not have been the only driver of the crime decline.

The scholarly consensus is that mass incarceration accounted for about 10 to 20 percent of the overall crime drop since 1992. “If you did a thought experiment, let’s add a million people to the prison system, and let’s suppose 1 percent of them are really serious habitual offenders who commit 50 crimes per year, that’s a reduction of half a million crimes,” Roman says. “Whether the destruction of communities associated with mass incarceration is worth it? That’s a completely different question.”

Police on the beat

There are two theories of how policing prevented crime. The first is that Clinton administration funding for police departments allowed cities to put more cops on the beat, and that their increased presence prevented crime, regardless of what those officers were doing. The second is that particular policing tactics, like the broken windows approach or hot spot policing, were effective.

Hot spot, or place-based policing, in which departments use real-time crime data to flood dangerous micro-neighborhoods with officers, was effective in reducing crime, according to the work of David Weisburd, director of the Center for Evidence-Based Crime Policy at George Mason University. “In virtually every city in which it has been studied, the growth of hot spot policing tends to reduce crime without, fortunately, displacing it to other areas,” Rosenfeld explains.

Hot spot policing is often implemented in conjunction with the much heralded and critiqued broken windows strategy, in which police focus on making arrests for low-level misdemeanors—so-called “quality of life” crimes, such as public display of marijuana or selling single, untaxed cigarettes on the street. One of the best recent studies of policing tactics was conducted in Lowell, Mass., by Anthony Braga of Harvard and Brenda Bond of Suffolk University. It found that place-based strategies—such as increasing police presence around an abandoned building—were successful, but that broken windows-type arrests had the disadvantage of seeding distrust between communities and police.

Roman, however, is unsure if policing impacts crime rates as much as its proponents argue. He points out that three cities that experienced large crime declines—Washington, D.C., New York City, and San Diego—pursued vastly different policing strategies. In New York, the size of the police force quadrupled in the 1990s and the use of stop, question, and frisk created an “oppositional” relationship between neighborhoods and officers, Roman says. In San Diego, crime declined without a significant increase in the number of police on the beat. Meanwhile, in Washington, D.C., police chief Cathy Lanier has critiqued hot spot policing and stop, question, and frisk and has focused on community-oriented policing, even speaking about the need to hire officers who display “empathy.”

Immigration and Gentrification

Given that some American cities, regardless of policing tactics, experienced big crime declines while others remained dangerous, a few criminologists began to wonder if two other urban trends of the past 30 years had impacted crime: increased immigration and gentrification.

Roman notes that cities with more Central American immigration, such as New York, Dallas, and San Diego, experienced bigger crime declines than cities with fewer immigrants, such as Philadelphia or Baltimore. Why? After upwardly mobile immigrants flood a chronically high-poverty neighborhood once characterized by inter-generational poverty, middle-class artists and professionals begin to arrive, bringing “wealth, resources, and political capital that didn’t previously exist,” he says. “It creates a virtuous cycle. Places become safer and it expands throughout the city.” The powerful intervention, Roman writes, is not displacement of the poor (which only moves crime from one neighborhood to another), but rather the positive effects of integration, in which people of different races and classes live in close proximity to one another.

In the end, each of these 10 theories probably hold some limited explanatory power. What social scientists have learned over the past 30 years, according to Zimring, is that “serious crime is a more superficial behavior” than people assumed back in the era of so-called “superpredator” kids. Most criminals are responding to neighborhood, economic, and health conditions that can quickly shift and improve. “Crime rates can change when deep structures don’t change,” Zimring says.