Share this:

Nathan Link 80x108James Kelly 80x108Many attribute New York’s massive fall in crime that began in the early 1990s to the implementation of the “broken windows” policing policy that sought to reduce “incivilities” – the social and physical indicators of neighborhood deterioration. In new research, Nathan W. Link and James M. Kelly examine the common wisdom that urban incivility leads to fear of crime. They argue that some people can be more “tuned in” to incivilities in their environment than others, and that those who are more generally fearful perceive greater incivilities.

Broken Windows Theory—the famous progeny of the “incivilities thesis”—states that people are adversely affected by “incivilities”—social and physical indicators of neighborhood deterioration such as rowdy groups of teens in public, or trash strewn about the streets. These environmental cues lead residents and passersby to think that they are at high risk for crime victimization (the cognitive element of fear), which in turn leads them to become emotionally fearful. In contrast to this theoretical model, in new research we found that some folks are more “tuned in” to perceiving incivility in their environments than others. Earlier, higher levels of cognitive fear drive this perception. Consequently, they traverse their environment with “antennae up”—spotting and defining more aspects of their locale as concerning or problematic than the people with antennae down in the same environment. Although seemingly uncontroversial, this finding challenges the common wisdom of nearly forty years that it is perceiving incivility that leads to fear.

Implemented in 1994 by Police Chief Bill Bratton under Mayor Giuliani’s leadership, zero tolerance (also known as “order maintenance”) policing is thought  to be the logical policy extension of the theory that incivilities lead to heightened fear and mass withdrawal from sections of neighborhoods. This exodus of “non-disorderly types” decreases informal social control, concentrating crime-prone types in these now-unsupervised areas. Mayhem ensues. Upon overseeing its practical implementation, Giuliani reigned over a city that saw unprecedented declines in crime—including a 50 percent reduction in violence, as Figure 1 illustrates. 

Figure 1 – Violent crimes in New York per 1,000 1985 – 2013

Link Fig 1

Source: FBI.

But does the zero tolerance approach deserve credit for these precipitous declines? At least five reasons give social scientists pause. One, the beginning of the crime decline in New York City preceded the implementation of zero tolerance in 1994. In this sense, Giuliani’s timing was auspicious—he may have surfed down the crime wave he was credited with creating. Two, cities across the U.S. experienced similar contemporaneous crime declines, including places that did not employ the Broken Windows-based strategy. What’s more, historically crime and violence have been declining the world over. Three, the New York Police Department implemented several efforts other than zero tolerance at the same time, so how can one single it out as the effective strategy? Four, theorists like Robert Sampson and Stephen Raudenbush convincingly argue that Broken Windows conflates correlation with causation; incivilities and crime might both result from a neighborhood’s level of collective efficacy.

But perhaps the most compelling reason that the aggressive policing style is probably not responsible for much of the crime decline is that the theory on which it is based does not have strong empirical support. Some of the most meticulous work testing Broken Windows found that the impacts of incivilities on later crime pale in comparison to neighborhood structural factors, such as homeownership and demographics. Others, relying on cross-sectional data, have found strong statistical correlations between incivilities and fear and have surmised that this is support for the beginning sequence of the theory. However, one issue with this type of analysis is that we don’t know in which direction the causal arrow is pointing. Isn’t it conceivable that a fearful person might “see” more local problems than a less fearful person, which would explain the association?

Source: Author

Source: Authors

Indeed, “incivility”—the fundamental cause in Broken Windows theory—is amorphous. What exactly is an “incivility”? Is graffiti an incivility? Is panhandling? According to whom? These questions matter because words like “incivility” and “disorder” do not have specific intrinsic meaning. Rather, they are largely social constructions interpreted as such by whoever is defining the subject/object under question. What one considers to be unsightly and concern-inspiring graffiti, another interprets as art. The corner panhandler might bother one; another might donate with alacrity. This insight confronts the core assumption of the beginning sequence of Broken Windows theory. The interesting question that follows is if incivility is in fact a social construction, what accounts for the fact that individuals disagree over what constitutes it? We argue that it is theoretically plausible that the person who feels more localized fear is more likely to define things in his or her environment as incivility-ridden. In essence, this hypothesis turns Broken Windows theory on its head—agreeing that incivilities and fear may be linked, but arguing that the causal relationship between the two is opposite that posited by Broken Windows.

Our test of this reversed hypothesis relies on representative longitudinal data from Baltimore, Maryland. The advantage of using data measured at two points in time (one year apart) is that we can examine both pathways over time—the Broken Windows version and our reversed hypothesis. It also allows us to examine changes in incivilities and fear perceptions. Three hundred and five residents clustered within 50 streetblocks were asked about perceptions of crime, incivilities, and fear on their streetblock. In our study, we examined the cognitive element of fear—the perceived likelihood of personal victimization.

Source: Authors

Source: Authors

Accounting for a variety of demographic factors and streetblock differences known to affect incivilities and fear perceptions (we also accounted for all inter-streetblock variation by controlling for all 50 streetblocks in the model), we found more support for our reversed hypothesis. Those who reported feeling more fear at the first interview reported significant increases in perceptions of incivilities on their streetblock by the second interview. Turning to the Broken Windows model whereby perceiving incivilities leads to higher fear, we found no support for this pathway. Incivility—it appears—is a social construct informed, at least in part, by fear.

Interesting theoretical questions arise from these findings. Why do some people living in the same areas perceive more danger than others (why are their antennae up)? Is this because of personal factors such as anxiety? Is it driven by one’s prior experiences, possibly bad ones? Or is it something that gets mentally activated in some when they enter particular areas?

On a policy level, critics of zero tolerance claim that it is essentially a fight against the poor, that it balloons our jail and prison populations with low-level “offenders,” and that it disproportionately targets minorities. On a more fundamental level, the results of our work, along with that of several others, empirically challenges the theory upon which zero tolerance policing practices are premised. Taken together, especially in light of the fact that alternative policing approaches with empirical support exist, one might hope to see more careful dialogue on the (de)merits of this policing practice as part of the broader, ongoing criminal justice reforms in the U.S.

This article is based on the paper ‘Reversing Broken Windows: Evidence of Lagged, Multilevel Impacts of Risk Perceptions on Perceptions of Incivility’, in Crime & Delinquency.

Featured image credit: Mark Allover (Flickr, CC-BY-NC-SA-2.0)

Please read our comments policy before commenting.           

Note:  This article gives the views of the author, and not the position of USAPP– American Politics and Policy, nor of the London School of Economics.

Shortened URL for this post:


About the authors

Nathan Link 80x108Nathan W. LinkTemple University
Nathan Link is a Ph.D. student in Criminal Justice at Temple University. He researches issues in U.S. corrections and criminal justice policy, and recently he’s been thinking about financial debt emerging from justice processes. His work is published in Justice Quarterly, Crime & Delinquency, and Administration and Policy in Mental Health and Mental Health Services Research.

James Kelly 80x108James M. KellyTemple University
James M. Kelly is a Ph.D. student in economics at Temple University. His field training is in labor economics and econometrics with a specific interest in the economics of crime, health economics, and education. His recent work appears in Crime & Delinquency.