'Broken Windows' Works

Crime, disorder and punishment.

Twenty-seven years ago, James Q. Wilson and I published “Broken Windows” in The Atlantic, proposing that untended disorder and minor offenses gave rise to serious crime and urban decay. We also hypothesized that government and community action to restore order might reduce crime.

Not surprisingly, responses to the article were mixed. The Justice Department’s research arm, the National Institute of Justice, prepared to fund a major experiment to study the links between disorder and serious crime, but senior officials nixed it as too controversial. Police were sympathetic to the Broken Windows theory but also wary, since they felt overwhelmed by 911 calls already and didn’t relish the prospect of still more work. And the article got little attention in the academy.

But after New York City’s astonishing crime drop in the ’90s--much of which Mayor Rudolph Giuliani and Police Commissioner William Bratton credited to the Broken Windows approach--a firestorm of academic criticism erupted, claiming that Broken Windows was racist, it harassed and criminalized the poor, it constituted cultural imperialism, it amounted to overzealous “zero tolerance” and so on. Moreover, the crime drop had nothing to do with Broken Windows (or any other police action); it was the result of changes in the economy or other broad social trends.

Some criminologists attacked Broken Windows to advance their careers, realizing that variations on the theme of “Broken Windows disproved” were an effective way to call attention to their own work. But for most, ideology was at stake. Not only did the effectiveness of Broken Windows undermine the decades-long assumption that only large-scale social and economic change could prevent crime; it also meant that breakthroughs in crime prevention could come from the Right--anathema to criminologists, most of whom occupied the far Left.

Still, critics of Broken Windows had one good point: New York provided, at most, anecdotal and correlational evidence of a relationship between disorder and crime. There were very few experimental studies--the most certain method of establishing causality--showing that the first caused the second.

But that changed last year, when University of Groningen researcher Kees Keizer and his colleagues published a paper in Science. In six experiments in the Netherlands, Keizer observed and compared the behavior of people under artificial conditions of order and disorder. Invariably, he found that disorderly conditions encouraged further and more serious levels of disorderly behavior. In one experiment, for example, Keizer placed an envelope conspicuously containing five euros in a mailbox. When the mailbox was clean, 13% of people who passed it stole the money; when it was covered with graffiti, 27% took it.

Also in 2008, Harvard University researcher Anthony A. Braga and his colleagues published the results of a complex set of field experiments in criminology. Researchers and police identified small neighborhoods in Lowell, Mass., and randomly assigned them to experimental and control conditions. In each of the experimental areas--where police were maintaining order, Broken Windows-style--crime dropped more sharply than in the control areas and, moreover, did not simply move to adjacent neighborhoods. The article also built on an earlier experiment, with the same results, that Braga had conducted in Jersey City a decade earlier.

While these studies do not settle, once and for all, the question of the relationship between disorder and serious crime, they do provide a substantial body of experimental evidence that fixing broken windows ought to be an integral part of any community’s response to crime. In fact, it’s hard to think of a policy option for fixing a major social problem that is as strongly supported--by both experience and solid research--as Broken Windows.

This piece originally appeared in Forbes

This piece originally appeared in Forbes