Reflecting on the subject: a critique of the social influence conception of deterrence, the broken windows theory, and order-maintenance policing New York style.

• Author: Harcourt, Bernard E.

INTRODUCTION

In 1993, New York City began implementing the quality-of-life initiative, an order-maintenance policing strategy targeting minor misdemeanor offenses like turnstile jumping, aggressive panhandling, and public drinking. The policing initiative is premised on the broken windows theory of deterrence, namely the hypothesis that minor physical and social disorder, if left unattended in a neighborhood, causes serious crime. New York City's new policing strategy has met with overwhelming support in the press and among public officials, policymakers, sociologists, criminologists and political scientists. The media describe the "famous"(1) Broken Windows essay(2) as "the bible of policing" and "the blueprint for community policing."(3) Order-maintenance policing has been called the "Holy Grail of the '90s."(4) "There is little dispute that the theory works," says the ABA Journal.(5) It has sparked "a revolution in American policing," according to the Christian Science Monitor, in an article captioned "One Man's Theory Is Cutting Crime in Urban Streets."(6) Even the recent U.S. News & World Report cover story on crime a cover story that debunks nearly every hypothesis for the national decline in crime -- makes a passing curtsy to the quality-of-life initiative: "Clearly, smarter policing was spectacularly decisive in some cities like New York."(7) Former Police Commissioner William Bratton, the principal architect of the quality-of-life initiative, credits the broken windows theory with falling crime rates in New York City. "These successes didn't just happen," Bratton contends. "They were achieved by embracing the concept of community policing."(8) Wesley Skogan, a political scientist at Northwestern University, has conducted an empirical study of the broken windows theory and concludes that "`[b]roken windows' do need to be repaired quickly."(9) George Kelling, co-author of Broken Windows and of a recent book entitled Fixing Broken Windows, contends that Skogan "established the causal links between disorder and serious crime -- empirically verifying the `Broken Windows' hypotheses."(10) In this euphoria of support, it is today practically impossible to find a single scholarly article that takes issue with the quality-of-life initiative.(11) It stands, in essence, uncontested -- even in the legal academy.

Dan Kahan, a leading social norm proponent in the area of criminal law, forcefully advocates order-maintenance policing and, in particular, New York City's quality-of-life initiative.(12) Kahan reports that order-maintenance policing "has been used with startlingly successful results in New York City."(13) He contends that the social influence conception of deterrence "makes it plausible to believe that order maintenance has in fact reduced crime in New York."(14) Kahan also suggests that "[t]he work of criminologist Wesley Skogan supplies empirical support for the `broken windows' hypothesis."(15) Other social norm proponents rely heavily on the broken windows theory and essentially endorse order-maintenance policing.(16)

In fact, order-maintenance policing is one of the leading recommendations along what Kahan calls "the new path of deterrence."(17) The new path is a loosely grouped set of initiatives in the area of crime and punishment, ranging from order-maintenance policing to curfews, gang-loitering laws, informal public-space zoning, reverse stings, and shaming penalties.(18) The new path seeks to revitalize the argument for deterrence by infusing it with cutting-edge social science. Social norm proponents locate the new path of deterrence between economics and sociology.(19) The new path represents, according to Kahan, "a third way, one that combines the virtues of both economics and sociology without succumbing to the vices of either."(20) From economics, the new path appropriates the idea that individuals are rational actors maximizing their utility. From sociology, the new path appropriates the idea that individuals are influenced, and their conduct is shaped, by social phenomena. The new path of deterrence is presented as an application of social norm theory to criminal law.(21)

In this Article, I critically examine the empirical evidence and the social influence explanation supporting New York City's experiment with order-maintenance policing. At the empirical level, I replicate the principal social scientific study that has attempted to establish the disorder-crime nexus, namely Wesley Skogan's Disorder and Decline: Crime and the Spiral of Decay in American Neighborhoods.(22) I conclude that Skogan's data do not support the claim that reducing disorder deters more serious crime. As a preliminary matter, the data are missing a large number of values (thirty to forty percent, on average, of the relevant dependent and independent variables) for such a small sample of neighborhoods (at most, forty neighborhood observations). But even setting aside that problem, my replication of Skogan's study establishes that (a) certain types of crime like rape, purse snatching, and pocket-picking are not significantly related to disorder; (b) other types of crime like physical assault and burglary are not significantly related to disorder when neighborhood poverty, stability, and race are held constant; and (c) although robbery remains significantly related to disorder, a cluster of five Newark neighborhoods exert excessive influence on the statistical findings. When those five Newark neighborhoods are set aside, the relationship between robbery victimization and disorder disappears. Accordingly, the data do not support the broken windows hypothesis.

Social norm proponents advance a second empirical argument in support of order-maintenance policing, namely the precipitous decline in crime rates in New York City.(23) The conventional explanations for the drop in crime, they argue, do not account for the magnitude of the drop in relation to other large cities. As we speak, however, there is a hotly contested debate raging among criminologists, legal scholars, policy-makers, journalists, and other experts over the causes of the decline in crime in New York City and nationally.(24) I review the various leading explanations and argue that it is far too simplistic to suggest that the quality-of-life initiative explains the extent of the decline of the crime rate in New York City.

The social influence conception of deterrence also does not withstand scrutiny at the theoretic level. The theory relies on a traditional sociological approach that does not sufficiently question the categories underlying the sociological analysis, or the relationship between its prescriptions and those categories. The theory's approach is similar to that of Emile Durkheim,(25) but ignores, I argue, some of the most insightful intellectual developments of the twentieth century. As a result, the set of policies emerging along the new path of deterrence are too limited. The policies do not sufficiently challenge our narrow way of conceptualizing crime.

Running through the social influence explanation and the broken windows theory is a recurrent and pervasive dichotomy between, what we could call in vulgar terms, honest people and the disorderly; between "committed law-abiders"(26) and "individuals who are otherwise inclined to engage in crime";(27) between "families who care for their homes, mind each other's children, and confidently frown on unwanted intruders"(28) and "disreputable or obstreperous or unpredictable people: panhandlers, drunks, addicts, rowdy teenagers, prostitutes, loiterers, the mentally disturbed."(29) Hand-in-hand with this set of categories is another ubiquitous dichotomy between order and disorder; between "norms of orderliness"(30) and "[p]ublic drunkenness, prostitution, aggressive panhandling and similar behavior";(31) between a "stable neighborhood"(32) and "an inhospitable and frightening jungle."(33)

The social influence conception of deterrence is grounded on these categories. The mechanisms of social influence assume these fixed identities because disorder operates on honest people and on the disorderly in different ways. Neighborhood disorder influences honest people to move out of the neighborhood or to lock themselves in their homes, but it influences the disorderly and especially criminals to move into the neighborhood and commit crimes.

These categories, however, do not have a pre-existent fixed reality, independent of the techniques of punishment implemented by the quality-of-life initiative. In other words, they do not pre-date the policing strategy. To the contrary, the category of the disorderly is itself a reality produced by the method of policing. It is a reality shaped by the policy of aggressive misdemeanor arrests. It is the product of a technique of punishment that combines several different historical modalities, including classical strategies of excessive force and modern disciplinary mechanisms like surveillance and spatial control. Michel Foucault's study, Discipline and Punish,(34) details these techniques of punishment, and it is there, I suggest, that we should turn to overcome the problems with the Durkheimian approach -- first, by rehearsing Foucault's analysis, but second, and more importantly, by refining his analysis.

The techniques of punishment that comprise the quality-of-life initiative create the disorderly person as an object of suspicion, surveillance, control, relocation, micromanagement, and arrest. According to the unwritten rules of a Newark police officer enforcing order, "[d]runks and addicts could sit on the stoops, but could not lie down. People could drink on side streets, but not at the main intersection. Bottles had to be in paper bags. Talking to, bothering, or begging from people waiting at the bus stop was strictly forbidden."(35) The fine art of policing creates the disorderly as a person with a full biography of habits, inclinations and...