Evidence from New York City and a Five City Social Experiment.”
In this paper, the authors reviewed the ongoing research being done on the “Broken Windows” theory of crime reduction in inner cities. The authors here look at a pair of studies designed to determine the efficacy of the Broken Windows theory and whether it has any validity at all. The reason the authors are considering this question is that considerable policing resources have been committed by major cities like New York, Chicago and Los Angeles toward implementing this theory. The following will review this paper, considering its evaluation of the theory involved.
This paper by Harcourt and Ludwig is essentially a review of two other studies conducted toward determining the effectiveness of the broken window theory. The broken window theory itself (as the authors here point out) was first propounded by James Wilson and George Kelling in 1982. The two presented their theory a highly influential and broadly discussed article published in the Atlantic monthly (Harcourt & Ludwig, 2006).
It was the theory of Wilson and Kelling that by reducing disorder in small ways in the urban environment, it would be possible to also reduce the likelihood and prevalence of crime in the inner city. This was particularly the case in poor neighborhoods where buildings were often run down, windows were broken and there was a general sense of abandonment and hopelessness. Homelessness and loitering were often issues as well (Wilson & Kelling, 1982). The principal means by which this change was to be accomplished was through the more stringent enforcement of crimes that are normally considered minor misdemeanors, such as vandalism or littering.
New York City was the first major city in the United States to try out Wilson and Kelling’s suggestions. From anecdotal observations, (not empirical research) there seem to be some evidence supporting Broken Window Theory. This research would be an example of inductive reasoning, in which a few anecdotal incidents are used to draw broader conclusions. But as Harcourt and Ludwig point out, this lack of clear experimental data left a question mark as to whether the resources being committed to this program would not be better spent on standard policing activities.
In their paper, Harcourt and Ludwig first turn to a report issued by the National Research Council, which came to the conclusion that (with one exception) current research provides no compelling evidence to support Wilson and Kelling's Broken Windows hypothesis (Harcourt & Ludwig, 2006). The exception referred to here was a 2001 study of crime trends in New York during the 1989-98 time period. This particular research was carried out by George Kelling and William Sousa. For the most part, the study carried out by Kelling and Sousa seem to confirm the original Broken Window Theory proposed by Wilson and Kelling.
In the Harcourt and Ludwig article, the authors reevaluate the aforementioned research by Kelling and Sousa and independently re-examine the same crime data used by Kelling and Sousa in their paper (Harcourt & Ludwig, 2006). In the process, the authors here came to very different conclusions than did Kelling and Sousa. In doing so, the authors were using deductive reasoning, beginning with the hypothesis and using observations to confirm or invalidate that hypothesis.
Harcourt and Ludwig also provided data from the MTO (Moving to Opportunity) experiment that was carried out in Boston, Baltimore, Los Angeles, Chicago and New York. The reason the authors included this particular research in their review was that the approach used in the MTO offered an alternative way to consider the broken window hypothesis that avoided some of the inherent problems with the supposedly empirical tests carried out by Kelling and Sousa. As described by the authors, the MTO involved giving over 4000 low income families that were currently residing in public housing located in high crime areas housing vouchers that would allow them to move into less chaotic and more privileged neighborhoods (Harcourt & Ludwig, 2006). The results from this experiment seemed refute the evidence from the Kelling and Sousa research.
This research is clearly quantitative in nature. It uses a series of statistical results to arrive at conclusions regarding the initial hypothesis. In both of the aforementioned studies, randomization was employed in an attempt to ensure the results were not biased in any way. For instance, in the MTO study, the participants were chosen randomly.
In conclusion, there are two problems here with the critique presented of the broken windows theory. The first is that the time window used in the first study by Kelling and Sousa was limited to 1989-1998. Both the original study by these researchers and the later critique by Harcourt and Ludwig might’ve carried more weight if the statistics were brought up to the present day. The second problem related to the second study considered in this article. The MOT study relied on the underlying premise that the broken windows theory would be invalidated if relocating is families to better neighborhoods did not result in lower incidence of crime among them. This premise fails to take into account the long-term effects that living in crime filled neighborhoods can have on individual psyches. A longer longitudinal study might have been more convincing.
References
Harcourt, B. E., & Ludwig, J. (2006). Broken windows: New evidence from new york city and a five-city social experiment. The University of Chicago Law Review, 73(1), 271-320.
Wilson, James Q.& Kelling, George L. (1982). Broken windows: the police and neighborhood safety, Atlantic Monthly 29, 38.
