James Q. Wilson and George L. Kelling recently revisited their broken windows theory of policing in a brief essay that appears in the November issue of The Atlantic. In a 1982 issue of that same magazine, they advanced their now well-known hypothesis that a decrease in visible signs of public disorder would lead to a reduction in crime rates. To be fair, the format of the essay—which appeared as one of many commenting on the “The Future of the American Idea”—did not lend itself to nuanced reflection. But because broken windows helped make community policing commonplace, sparked proposals for dramatic changes in criminal procedure doctrine, and is a key element in the biography of Rudolph Giuliani, Wilson and Kelling’s readers can be forgiven for hoping for a bit more.
In their new essay, Wilson and Kelling write, “Virtually all of the evidence we have from studies of police suggests that restoring order is associated with a drop in crime. This is reassuring, but it may not be conclusive. The idea has never been fully tested.” This does not satisfactorily answer scholars who have questioned whether a reduction in serious crime actually follows an increase in public order. Most importantly, it does not address whether a drop in serious crime—if it occurs—stems from the reduction in public disorder or from the increased surveillance that aggressive misdemeanor arrests make possible.
Wilson and Kelling also write, “Decency in public places may be only a small part of the American idea, but especially for those people living in dangerous, gang-ridden neighborhoods, it is an important one.” This is a statement with which it is difficult to argue. But it says nothing about whether constraining disorder is the best use of limited police resources, or how the police choose their targets in a public order campaign, or whether addressing disorder can ever mean more than moving it to a less visible place. We would all benefit from knowing more about how the original proponents of broken windows would answer these difficult questions.