Based on an article titled Broken windowsby James Q. Wilson and George L. Kelling, which appeared in the March 1982 issue of The Atlantic Monthly. The title of the book comes from the following example:

“Consider a building with a broken window. If the window is not repaired, the vandals will tend to break a few more windows. Eventually, they may even break into the building, and if it's abandoned, it may be occupied by them or they may set fires inside.

Or consider a stool. Some garbage accumulates. Soon, more garbage is piling up. Eventually, people start leaving garbage bags from restaurants or breaking into cars.”

A good strategy for preventing vandalism, the book's authors say, is to fix problems while they're still small. Fix broken windows in a short period of time, say a day or a week, and the trend is that vandals will be less likely to break more windows or do more damage. Clean the sidewalks every day, and the trend will be that the garbage will not accumulate (or the garbage accumulated will be much less). Problems do not escalate and residents are prevented from fleeing the neighborhood.

So, the theory makes two hypotheses: that petty crimes and anti-social behavior will be decreased, and that first-degree crimes will, as a result, be prevented. Criticisms of the theory tend to focus solely on the second hypothesis.

The Broken Windows theory, developed by James Q. Wilson and George Kelling, is based on the premise that crime is the inevitable result of disorder. These criminologists found that crime, in any urban center, was higher in areas where carelessness, dirt, and mistreatment of public property prevailed. A broken window in a building, if not repaired soon, was the prelude to all the others soon being damaged.

This relationship had already been reported by Zimbardo, and is reported in more detail in general prevention. In 1969, Zimbardo conducted a very interesting experiment: he left two abandoned cars, of the same make, model, and color, one in Palo Alto, California, and the other in the Bronx, New York City. As expected, the first remained intact for a week, while the other was stolen and semi-destroyed. However, the fortunes for the Palo Alto car changed when Zimbardo himself broke a window.

The conclusion is clear: a car with a broken window that sits unattended is a car that no one cares about, and therefore can be looted.

If a community shows signs of deterioration and no one seems to care, it will show an increase in crime as a consequence. The most common manifestations of this deterioration are broken windows in abandoned buildings and graffiti. In fact, as in Zimbardo's experiment, they serve as inducers to cause epidemics of insecurity.

During the eighties, the New York City Subway became the archetype of New York insecurity. Tired of suffering violent assaults, intimidation, robberies, or traveling in deteriorated, graffiti-covered, slow cars, users began to abandon it, and as they did, the deterioration and insecurity of the facilities increased.

In the mid-1980s, Kelling was hired by the New York City Transit Authority as a consultant, and put the Broken Windows theory to work with Subway Director David Gunn. They were later joined by William Bratton as director of the Metro police. Its immediate objectives were two: to put an end to graffiti, and to persecute petty criminals, such as those who entered without paying, were drunk or did any kind of excess inside the facilities.

There was a reason for this: if a transgression, however small, is committed and left unprosecuted, there will always be imitators. If someone enters the Metro without paying and people see that they get away with it, they will think “and why not me”. So powerful is the engine of imitation encouraged by impunity.

Criticism was immediate from those who expected more radical and cumbersome solutions for major crimes. But Kelling argued that the only way to end insecurity was to prosecute petty crime.

Because those who commit small crimes are also involved in major ones. In the New York experience, the arrest of people who had not paid their entrance to the Metro or made improper use of its facilities, showed that 1 in 7 had an arrest warrant for a felony, and 1 in 20 illegally carried a weapon .

The relentless battle against graffiti also brought good results. As graffiti was emblematic of the decline of underground mass transit, its removal raised users' confidence that things were changing.

When Rudolph Giuliani became mayor of New York City in 1994, William Bratton was appointed head of the New York Police Department, and he applied similar but broader strategies: combating graffiti, targeting minor transgressions like urinating or littering in the public thoroughfare with the full weight of the law. Thus, with the prosecution of minor crimes that in turn allowed to attack the majors, and created cleaner, more cared for communities, which did not encourage the commission of crimes.

The success of Zero Tolerance and the application of the teachings of the Broken Windows theory broke with many prejudices that existed in the conception of crime as something due to genetic defects, bad education, lack of opportunities and other more or less well-founded hypotheses. It showed that the offender is not some kind of automaton, unable to stop committing crimes, but rather an individual highly sensitive to environmental changes in his immediate environment. Therefore, improvement in that environment is a better strategy, through not tolerating minor transgressions, than pursuing major crimes.

The results of such policies, summed up in general prevention, should at least give their critics something to think about. Unfortunately, most of them are dogmatic and act according to previously agreed agendas, group interests or prejudices, and only by exception with a legitimate – albeit mistaken – conviction of better alternatives to fight crime.