Broken windows theory's limited effectiveness in urban crime prevention

Broken Window Theory

The "broken windows" theory of crime prevention and control is perhaps one of the most widely discussed and least understood law enforcement paradigms, due to the relative simplicity of the theory and the ostensibly dramatic reductions in crime offered by the first studies of cities in which a "broken windows" policy was implemented. The policy was first proposed in the early 1980s, but it was not until the 1990s, when New York adopted a broken windows policy and saw a drop in crime rates, that the theory became widely popularized. However, subsequent analysis of these drops in crime as well as other detrimental effects of a broken windows policy helps to reveal that the gains initially promised by the results in New York and other cities is not indicative of a broken windows policy in general, and in fact, many of these reductions in crime may be attributed to other, less obvious factors.

Broken Windows

The broken windows policy of crime control was first introduced in 1982 by James Q. Wilson and George L. Kelling in an essay asserting "that crime and disorder are casually linked, and that policing 'little things is instrumental in preventing serious crime" (Geller, 2007). Thus, the theory takes its name from the idea that small forms of urban disorder, like broken windows (or graffiti, or in the famous case of New York, people hopping over subway turnstiles), contribute to a general sense of disorder, which actually increases crime by providing an environment in which it appears that disorder, and thus crime, will go unpunished. According to the theory, "social and physical disorder indirectly lead to serious crimes through a neighborhood breakdown of informal social controls," with broken windows functioning as a symbol of this social breakdown (Distler, 2011, p. 1). Put another way, the appearance of a broken window supposedly increases the likelihood of further damage, because potential criminals see the broken window as evidence of a general negligence on the part of police or residents.

Thus, for police hoping to reduce crime rates through the application of a broken windows policy of crime control, efforts are focused "on low-level, misdemeanor offenses" as well as stricter civil guidelines for urban upkeep and maintenance (Distler, 2011, p. 1). The appeal of the broken windows theory is attractive for obvious reasons; most crucially, it claims that all police need to do is focus on minor offenses and the rest will almost magically clear itself up, thus obviating the need for any larger structural or social changes. This has resulted in so-called "zero-tolerance" policies in a number of major cities, with any drop in crime rates over the last twenty years being almost exclusively attributed to these policies.

Causation vs. Correlation

Before investigating the evidence for and against the claims made by supporters of broken window policies, it will be helpful to briefly explicate two terms that have been the source of numerous misunderstandings and misapplications of policy. "Causation" and "correlation" are both terms used to describe the relationship between two factors in any given study, but the subtle difference between the two means the difference between a multimillion dollar law enforcement initiative being supported by robust evidence, or being a waste of time and money as a result of a misunderstanding of the evidence. Correlation means that the two factors under discussion have some kind of observable relationship, such that changes in the one are reflected by changes in the other. In the case of broken windows policies, there is clearly a correlation between lower levels of disorder and lower levels of violent crime, because correlation is generally easily observable, such as in the dramatic decrease in crime rates witnessed in New York during the 1990s.

Causation, on the other hand, requires more investigation and evidence, because it describes a relationship in which changes in one factor actually cause the changes in the other. In the case of correlation, there may be a third, unknown factor ultimately responsible for the changes in both known variables, such that one may observe correlation without causation. Broken windows theory relies on a perceived causal relationship between disorder and violent crime, because the claim is that a reduction in disorder causes a reduction in violent crime, but a closer look at the cases of broken windows policies which have been instituted over the last thirty years reveals that at best, one may only identify a correlation between the two, and not causation. This has led to an ongoing debate "over whether the relationship between disorder and crime is causal or correlational," because proponents of zero-tolerance policies can point towards the very real reductions in crime seen in major U.S. cities as justification for their position, but this represents an ultimately oversimplified understanding of statistics and the complex interplay of social factors (Yang, 2010, p. 140). In short, this is because those studies which seemed to show a causal relationship between disorder and violent crime failed to account for a number of other social, political, economic, and demographic changes occurring at the same time these policies were instituted, such that they failed to appreciate how these additional changes could actually precipitate the drops seen in both disorder and violent crime.

Investigating the Evidence

The easiest way to demonstrate the problem with statistical evidence seemingly supporting broken windows theories is to address those applications of it which have been pointed to most frequently and vehemently; namely, New York and other major cities in the United States. New York, being the most populous U.S. city, is the most popular example to cite, as "crime rates in New York City have plummeted in recent years -- years that have coincided, in large part, with the implementation of the quality-of-life initiative" (Harcourt, 1998, p. 331). However, the assumption that this drop has occurred due to an increase in the enforcement of misdemeanor laws and other minor infractions is almost laughably simplistic, because there are a huge variety of other variables which may have contributed to the decline in both disorder and violent crime. According to Harcourt:

These include a significant increase in the New York City Police force, a general shift in drug use from crack cocaine to heroin, favorable economic conditions in the 1990s, new computerized tracking systems that speed up police response to crime, a dip in the number of eighteen-to twenty-four-year-old males, and increase in the number of hard-core offenders incarcerated in city jails and state prisons, the arrest of several big drug gangs in New York, as well as possible changes in adolescent behavior. (Harcourt, 1998, p. 332)

Thus, while "none [of these factors] may single-handedly account for the trend," each one likely contributed such the assumption that New York's drop in crime rates may be attributed to a broken windows policy is simply not sustainable (p. 332). Thus, even from the case of New York City alone, it seems that the apparent success of broken windows policies stems not from their inherent validity, but rather the fact that they were implemented alongside a variety of other developments such that broken windows policies serendipitously benefited from previously unconsidered variables.

One may look beyond New York for further evidence that broken windows policies are ultimately not responsible for the dramatic drops in violent crime rates seen in recent years. In particular, one may examine a program called Moving to Opportunity, in which families from socially disordered communities were moved to differential neighborhoods with varying levels of disorder. For nearly the last two decades this program "has been in operation […] in five cities: Baltimore, Boston, Chicago, Los Angeles, and New York" (Harcourt, 2006, p. 300). If the broken windows theory were to hold true, then the families moved to neighborhoods with higher levels of social disorder would be expected to be at a higher risk for participation in crime, and families moved to neighborhoods with less disorder would be expected to be at a lower risk for participation in crime. However, this was not the case, as examinations of juvenile delinquency and arrest records for families in the Moving to Opportunity program reveals that "while some subgroups do respond to moves to less disorderly neighborhoods by reducing their involvement in criminal behavior, most notably female youth, these effects are offset by increases in antisocial behavior among other subgroups," such that any changes are ultimately not convincing (Harcourt, 2006, p. 307). Furthermore, even if one did attempt to read these minor changes as some kind of evidence in support of broken windows theory, the fact remains that the theory necessarily does not make distinctions-based demographics, because disorder is supposed to affect all residents equally.

Thus, there are two different ways of demonstrating the inaccuracy of the assumptions underlying the broken window theory. First, any broken window policy was established alongside numerous other changes, such that one cannot easily distinguish between the drops in crime attributable to a broken window policy and a drop caused by any number of other variables. Furthermore, evidence from the Moving to Opportunity program demonstrates that disorder does not have a causal effect on participation in crime, such that the fundamental claim of the broken window theory is proved false.

Conclusion

It is crucial to investigate crime control policies such as the broken window theory closely, because millions of dollars are spent on sustaining these policies and modulating the public's behavior. In turn, if a policy is based on a faulty understanding of the facts, as in the case of the broken window theory, this misunderstanding can actually cause more harm than good. For example, when New York attempted to institute a harsh zero-tolerance policy, "the city sought to ban panhandling by passing a law prohibiting loitering for the purpose of begging," but it was struck down under the First amendment (Harcourt, 2000, p. 133). Following this, the transit authority actually instituted an advertising campaign "communicat[ing] to passengers that it was wrong to give money to panhandlers because it made them less likely to go seek help" (p. 134). The city essentially condemned charity out of a belief that it encouraged disorder, which meant a subsequent increase in crime. The perversity of this assumption is demonstrated if one follows the chain of logic, in which giving a panhandler money is interpreted to ultimately be responsible for violent crime.

Nonetheless, because the broken window theory offers such a seemingly simple explanation of the complex processes by which crime is encouraged and instigated, it was seized upon by policymakers and law enforcement personnel before a robust investigation of the evidence could occur. Since then, however, this investigation has progressed and demonstrated that the broken windows policy of crime control is predicated on an unnecessarily reductive and inaccurate interpretation of the relationship between disorder and crime.