Printer Friendly

Order maintenance reconsidered: moving beyond strong causal reasoning.

A backlash has set in against order maintenance policing strategies, if not among policymakers and the public, then at least among criminologists. This backlash has several components, but the most prominent rests on empirical studies that have claimed to cast doubt on James Q. Wilson and George L. Kelling's broken windows theory--the theory that disorder, left unchecked, leads to crime by driving residents indoors and sending a message to would-be offenders that a neighborhood is out of control. (1) In this paper I argue that this backlash focuses too narrowly on the broken windows theory in its assessments of order maintenance policing, and I develop and apply alternative methods of analysis that focus more directly on the intrinsic merits of efforts to reduce disorder by using ethnographic research and normative analysis. In the process, I analyze the few grounded descriptions of order maintenance practice that have been presented in the literature to argue that at least some kinds of order maintenance policing are intrinsically valuable--regardless of the impact they have on serious crime--because they address important instances of accumulative harms and offenses. (2) Policing inappropriately ignores these problems when it only focuses on serious crime.

In making this argument, I draw on and extend recent ideas in policy analysis about the way scholarship can best inform public policy. In current debates, both opponents and proponents of order maintenance often presume that its benefits are best judged by its contribution to crime reduction--by its indirect effects on serious crime, rather than its direct effects on public order. (3) By tying the evaluation of order maintenance policing so closely to its indirect effects on crime, this literature offers an example of what Martin Rein and Christopher Winship have described as "the dangers of strong causal reasoning"--the dangers of policy analyses that rely on claims that an intervention will have large indirect effects on some important social problem (e.g., that incentives for marriage will improve the prospects for low-income children). (4) Rein and Winship argue that claims of this kind ask social science to do too much because it can rarely identify the tight causal relationships of the kind that would be necessary; in the meantime, the focus on indirect effects tends to crowd out questions about the intrinsic wisdom of policy interventions. Order maintenance policing is a case in point. Since the early empirical studies that called attention to the order maintenance function (5) and the scholarly debates that considered its intrinsic propriety, (6) criminologists have paid little attention to questions about whether order maintenance activities and the public order they hope to create are desirable in their own right, apart from their indirect contribution to crime prevention. These questions call for qualitative study and normative analysis of a kind that recent police literature has de-emphasized--in particular, for ethnographic study of order maintenance practice that illuminates what exactly this policing strategy involves, and for normative analysis that indicates how it implicates the ideals of liberal political theory. The literature about order maintenance policing is hardly the only area to neglect these forms of research: similar gaps appear elsewhere in criminology, and indeed in modern social science more generally. (7) In that respect, this paper's argument applies more widely than the order maintenance topic itself, which simply offers a timely example of how and why criminal justice scholarship should move beyond strong causal reasoning.

I make these arguments more fully in Part I, and in Part II, I suggest how scholarship about order maintenance policing could contribute more broadly to policy debate by drawing on recent ideas about ethical inquiry; in the process, I offer a tentative argument, based on what little relevant research exists, that some kinds of order maintenance policing are appropriate. Part III concludes.


Strong causal reasoning tries to identify policy interventions that, through a complex causal process, have major indirect effects on some unequivocally important social outcome. Rein and Winship's paradigmatic example is school desegregation. When desegregation reached the Supreme Court in Brown v. Board of Education, (8) social scientists came forward with arguments and evidence claiming that segregated schools had a deleterious effect on the personal development and educational achievement of black children, maintaining that these children would learn and develop better in integrated schools. The Court relied on these social scientific arguments in its opinion in Brown, but forty years later, the social science theories on which these arguments relied have come into question, as most research has been unable to identify large effects of integrated schools on student achievement. (9)

The conclusion that integration only produces small indirect benefits for black children's learning is, of course, relevant to the debate about school desegregation. But as in other policy debates where strong causal reasoning has played a major role, proponents of desegregation often claimed more than relevance for their arguments. They relied so extensively on social science theories about the indirect effects of integration that those effects often appeared to be the major justification for desegregation. Indeed, the objectivity of social science makes it very attractive in policy arguments because it holds out the hope of resolving intractable controversies through neutral methods of rational inquiry. But if we use causal analyses to bypass those controversies and the causal analyses come undone, we end up in a difficult position. The social scientific justification for the intervention has become illusory, and other issues that deserve attention have been forgotten or de-legitimized. A return to those issues appears quaint or hypocritical, the last gasp of a determined partisan who refuses to listen to evidence. Rein and Winship argue that "this is precisely the situation we are in today with respect to school desegregation." (10)
 [B]ecause of the perceived objectivity of the causal reasoning, the
 strong predictions made by social scientists, and the consequences
 of this for policy design, other arguments for integration were
 crowded out. Most importantly, after we change the rationale and
 terms of the justifications for the intervention we propose, it is
 difficult, if not impossible, to successfully advocate for a
 position which justifies integration as an important societal value
 in and of itself and not merely a means to promote the education of
 children. (11)


The debate about order maintenance policing has followed a similar course. Current interest in order maintenance tactics is generally credited to James Q. Wilson and George L. Kelling's 1982 article titled Broken Windows." The Police and Neighborhood Safety. (12) That article tried to support order maintenance strategies partly with what Rein and Winship would describe as a strong causal theory. Building on the work of psychologist Philip Zimbardo, (13) Wilson and Kelling's theory held that disorder, left unchecked, sends a message that a neighborhood is out of control and emboldens would-be criminals to commit serious crime; at the same time, it drives law-abiding residents indoors, reducing the informal surveillance and social control that can help to prevent serious crime in neighborhoods. In short, the theory justifies order maintenance activities by referring to their indirect effects on serious crime.

The broken windows theory received early support from several studies designed to test it (14) and from re-interpretations of past studies. (15) Moreover, many academics, policymakers, and analysts argued that order maintenance policing was responsible for a major share of New York City's remarkable decline in serious crime. (16) Partly on the basis of these claims, hundreds of police departments around the country placed more emphasis on order maintenance strategies.

Over the past few years, however, social science has not been kind to the broken windows theory. A number of scholars reanalyzed the initial studies that appeared to support it, arguing in particular that Wesley Skogan's seminal study of the relationship between disorder and crime did not demonstrate the strong relationship that broken window proponents have claimed. (17) Others pressed forward with new, more sophisticated studies of the relationship between disorder and crime. (18) The most prominent among them concluded that the relationship between disorder and serious crime is modest, and even that relationship is largely an artifact of more fundamental social forces. (19) Still other social scientists have questioned the effect that New York City's police strategies had on that city's crime trends, arguing that factors like the decline of crack cocaine played larger roles than order maintenance, and that other cities that have not implemented order maintenance tactics have achieved comparable reductions in crime. (20)

These challenges to the broken windows theory have not yet discredited order maintenance policing with policymakers or the public. But among criminologists, order maintenance is clearly under siege. Thus Sampson and Raudenbush argue that "current fascination in policy circles ... on cleaning up disorder through law enforcement techniques appears simplistic and largely misplaced, at least in terms of directly fighting crime.... Attacking public disorder through tough police tactics may thus be a politically popular but perhaps analytically weak strategy to reduce crime...." (21) Their conclusions have been echoed (and often quoted) by other academic commentators on order maintenance policing. (22) These criticisms, like much of the sea change in opinion against order maintenance policing, rest on the empirical critique of the broken windows theory--on the recent evidence that claims to show that disorder and the order maintenance strategies designed to reduce it may not have large effects on serious crime.

In the process of making these critiques, scholars placed less emphasis than they once had on direct analysis of the merits of order maintenance as a use of police resources and authority--on questions about whether the maintenance of public order is a legitimate goal for policing and whether the strategies police use to pursue it are justifiable. Arguments claiming that order maintenance deserves greater emphasis in policing predate the broken windows theory, (23) and for a time in the 1970s and 1980s, policing scholars debated those arguments by focusing directly on the merits of "public order" as a goal of policing (rather than the contribution that public order made to crime reduction). (24) This debate, in turn, drew on a rich empirical, historical, and conceptual literature about the order maintenance function. (25) After the "Broken Windows" article appeared, however, scholarly attention largely turned to questions about the effect of order maintenance on crime, rather than the intrinsic merits of public order as a goal of policing. By 1986, Gary Sykes had already raised concerns about this trend in an article that foreshadowed Rein and Winship's arguments about the dangers of strong causal reasoning. "[T]he primary justification for the revival of peacekeeping strategies rests essentially on police effectiveness in crime control [i.e., via the broken windows theory]," Sykes noted. (26) "By salvaging the crime control function via order maintenance, the moral arguments in support of peacekeeping are not explored." (27)

As Rein and Winship acknowledge, (28) debates about the sort of "moral arguments" whose absence Sykes lamented are usually difficult to resolve, so the possibility that social science might offer a way to avoid them--that it might be able to replace contentious normative questions with causal questions resolvable using objective methods--is an attractive idea. For that reason, it is not surprising that policing scholars turned their attention to the more scientific enterprise of testing the broken windows hypothesis once it emerged as a possible rationale for order maintenance policies. If order maintenance has major effects on crime, the contentious and seemingly intractable question of whether "public order" is an intrinsically appropriate goal for policing could safely be set aside. Nevertheless, two difficulties confront this attempt to bypass Sykes's "moral arguments" about order maintenance policing.

First, it is not clear that the broken windows thesis could justify intervention against disorder on its own even if it were true. The broken windows theory claims to identify clear harms caused by disorder that would justify criminal justice intervention. As Bernard Harcourt puts it, "[w]hat the theory accomplished was to transform these quality-of-life offenses from mere nuisances or annoyances into seriously harmful conduct--conduct that in fact contributes to serious crimes," thereby identifying a clear-cut moral basis for police intervention. (29) But if disorder causes crime in the way the broken windows thesis indicates, it does so through the voluntary actions of third parties. For example, a panhandler's disorderly behavior sends a signal to a hardened criminal that a neighborhood is out-of-control and thus ripe for predation. But in cases of this kind--where A's actions unintentionally cause B to harm C--it is rarely appropriate to view A's actions as wrongful purely because of the harm to C. (30) The issues are not identical, but it is suggestive to compare the case of media violence, where a moviemaker's (A's) actions arguably induce criminals (B) to harm their victims (C). It is hard to imagine a case where it would be appropriate to hold a moviemaker criminally liable for producing a film that planted the idea for a crime in a criminal's head, and for that reason legal scholar Joel Feinberg has concluded that "legislatures ... should be leery of passing criminal legislation against the 'indirect causation' of harm through film broadcasts." (31) In a similar way, it is hard to imagine a case where otherwise-innocuous disorder would justify criminal justice intervention because it inspired serious crime by an independent party. (32) If disorder has large effects on crime, that may be a good reason to try to prevent it in other ways, and it may be a good reason for police to make disorder a larger priority if there are already other reasons to justify regulating it. But the effects of disorder on crime may not be a strong reason to criminalize otherwise-innocuous disorder on its own, so we cannot bypass Sykes's question so easily.

The second difficulty, more salient for the present discussion of the dangers of strong causal reasoning, is that scientific theories often prove less powerful than originally hoped, so policy interventions can rarely be justified substantially by their indirect effects. Rein and Winship explain:
 [S]ocial science has been able to provide only what we would call
 "weak" causal theories. In the vast majority of cases the effects
 that are found are of modest size and only a small amount of the
 variation in the dependent variable is explained.... What is
 problematic is that in arguing for particular policies we often
 argue as if social science's findings imply that there are strong
 determinative relations between particular causes and outcomes.
 Certainly, education affects earnings, but this does not mean that
 equalizing education will have much effect on earnings inequality.
 Similarly, economic poverty certainly affects child development,
 but this does not mean that reducing economic poverty will
 substantially improve child development. (33)

In the same way, disorder may affect crime, but that does not mean order maintenance will substantially reduce the level of violence in the most troubled cities and neighborhoods--a conclusion that seems to be borne out by recent disillusionment with the broken windows theory among many criminologists. Consequently, it is probably too optimistic to hope that the strong causal broken windows theory offers a scientific way to bypass difficult moral arguments about the propriety of "public order" as a goal of policing.

In cases like school desegregation and order maintenance policing, when dramatic claims about the indirect effects of an intervention fail to materialize, we should return, as Rein and Winship suggest, (34) to the direct effects and intrinsic features of the policy and ask whether those effects and intrinsic features are desirable. Is desegregation morally justified, whatever its indirect effects on black childrens' achievement? Is order maintenance a worthwhile and legitimate police activity, even if its indirect effects on serious crime are not large? Rein and Winship's analysis indicates clearly why these questions are important and indeed unavoidable, but it offers less guidance in suggesting how we should go about answering them. In the next section I develop one possibility by drawing on recent ideas about ethical inquiry.


Any claim that order maintenance is intrinsically appropriate, or that the production of "public order" is a legitimate goal for policing, may strike many social scientists as a mere assertion that cannot be subjected to scholarly analysis. In this section I will discuss how scholarly analysis can help to evaluate such claims. To evaluate claims about the intrinsic merits of a policy, it is necessary to have a clear view of both facts (about what exactly the policy entails in different contexts) and values (about which ideals the policy implicates in each of these contexts). To evaluate this particular claim about order maintenance policing, it is essential to develop a detailed, contextualized picture of what "order maintenance" activities involve--what exactly "disorder" means in practice and what steps police take to reduce it--and to identify and analyze the values that they implicate by drawing on and refining concepts from moral and political philosophy. I first explain the need for ethnographic research in these analyses, arguing that the lack of attention to this kind of research has hindered those few attempts that have been made to examine the intrinsic merits of order maintenance activities. I then explain the need for engagement with moral and political philosophy, arguing that familiar liberal theories about the scope and limits of state action offer a useful framework to assess order maintenance policing. In making these arguments, I will of necessity return several times with a critical, but hopefully constructive eye to the small number of recent studies that have considered the intrinsic merits of order maintenance policing. (35)


Questions about the intrinsic merits of an action are often viewed as purely normative matters, so my statement that empirical study plays a role in these assessments needs explanation. The essential point is that ethical analysis can rarely offer general judgments about a category of actions because whether or not some action is appropriate depends heavily on context. (36) For example, the most sophisticated practical ethicists do not ask whether the general category of acts called "lies" is right or wrong in the abstract; instead they identify and consider specific examples of lying in particular contexts in order to develop a taxonomy of cases that maps out examples of inappropriate and acceptable lies. (37) In part, this attention to context is necessary because general maxims (like "do not lie") are invariably ambiguous--particularly in marginal situations where key terms are ambiguous (as in cases where it is unclear whether a statement is a lie) or where one maxim conflicts with others (as in cases where the duty not to lie conflicts with the duty to minimize suffering). (38) In part, it is necessary because our normative judgments are often most certain when they refer to situations that are specific, concrete, and real; (39) we simply do not know whether abstract maxims like "do not lie" are valid ethical principles because we cannot adequately imagine the many situations to which they apply. Because context shapes ethical judgment in these ways, the first step in moral analysis is a kind of thick description of activities in their context. (40) As Joel Feinberg argued in his seminal treatise on the moral limits of the criminal law, "there is a limit to the power of abstract reasoning to settle questions of moral legitimacy;" when considering such questions, it is essential to consider the actions at issue in concrete detail. (41)

From this perspective, one of the most serious gaps in contemporary scholarship about order maintenance policing is the dearth of research that investigates what exactly order maintenance policing activities involve--what behaviors they target in which contexts, and what actions police take to control them. Since Kelling's brief descriptions of the order maintenance activities he observed while walking the beat with Newark foot patrol officers in the late 1970s, (42) I am aware of no ethnographic research in criminology that systematically describes what officers engaged in order maintenance policing do, or what kinds of "disorder" they target. In that sense, the thick description that practical ethics should begin with has been neglected in recent police scholarship, making it impossible to thoroughly analyze the intrinsic propriety of order maintenance activities. Criminologists have rushed over the question of what order maintenance policing is, relying instead on rudimentary proxies for it in order to analyze what it causes. As a result of this gap in empirical research, much of the literature about order maintenance policing has a stylized view of what it involves--both how police maintain order, and what they treat as disorderly. Consider each of these issues in turn.

How Do Police Maintain Order? First, much of the order maintenance literature has not probed deeply into the question of what steps police might and do take to maintain order, so scholars have relied on a stylized view of order maintenance practice that does not offer an adequate basis for a valid assessment of this policing strategy. For example, Bernard Harcourt, a prominent critic of order maintenance policing, regularly equates order maintenance activity with arrests for minor offenses, describing it as a strategy that emphasizes "arresting misdemeanor and public order offenders--rather than issuing warnings or implementing alternative problem-solving techniques." (43) Similarly, other scholars equate order maintenance policing with "zero tolerance," i.e., a firm policy that even minor violations will result in arrest. (44) Although it is possible that such blunt approaches to order maintenance exist in some cities at some times, it is unlikely that this stylized description of order maintenance tactics fairly captures the range of possibilities that this policing strategy potentially offers. Consequently, it does not offer a sufficiently nuanced image of "order maintenance policing" to support a sophisticated assessment of that practice. Again, it is impossible to flesh out the nuanced picture of order maintenance practice that would be needed in the absence of more ethnographic study than recent literature has produced, but the few glimpses of practice that are available--primarily prescriptive statements about order maintenance policing from its practitioners--are suggestive. Consider the New Haven Police Department's "Order Maintenance Training Bulletin," which instructs officers to "use the least forceful means possible to achieve its purposes."
 The first level of intervention, whether by managers, supervisors,
 or by police officers, will be to educate the public about civility,
 the consequences of incivility, and the laws that oblige citizens to
 behave in particular ways. This can be done in neighborhood
 meetings, in schools, or in interactions with citizens. Some
 citizens do not fully understand their obligations, and if those
 obligations--for example, regarding a noisy car or public drinking
 in parks--are patiently explained, they will adhere to the law. The
 second level of intervention will be to remind citizens of their
 responsibilities if they are disorderly--that is, that they are
 breaking the law and subject to penalties if they persist. This too
 can be done in a variety of ways. It could be done by visiting a
 problem location and warning people that if their behavior continues
 they will be subject to penalties. Similarly, owners of locations
 that are chronic problems could be so warned by individual officers.
 The final level of intervention will be law enforcement--the use of
 citation and arrest. (45)

This account directly contradicts Harcourt's description of order maintenance practice, and other prescriptive statements about order maintenance policing have echoed it. For example, Kelling and Coles advised prospective order maintenance practitioners that "depending upon his or her assessment of the situation, an officer will have to decide to educate the offender, refer him or her to an agency for help, give a warning, order the offender to move on, cite him or her, make an arrest, or take some other action." (46) In the absence of ethnographic study, it remains an open question whether policing practice often realizes this nuanced ideal, but simply equating order maintenance policing with misdemeanor arrests is clearly too simple. The crucial point is that there exist possible approaches to order maintenance more subtle than "zero tolerance." An analysis of order maintenance policing that ignores them is akin to an analysis of lying that only examines, say, self-serving lies. (47)

Such simplifications make it impossible to offer a clear normative analysis of the intrinsic merits of order maintenance tactics. Consider Harcourt's analysis of order maintenance policing, which aims to shed light on order maintenance policing's intrinsic merits--or rather, its intrinsic dangers--by analyzing the role it plays in the carceral network to which Michel Foucault called our attention. (48) In the process, however, his equation of order maintenance with misdemeanor arrests leads him astray:
 To say ... that broken windows policing is part of the disciplinary
 project is to say too little; everything is today, since we live,
 according to Foucault, in a disciplinary society.... What we have to
 do today is compare the different genres of discipline. It is here
 that we can refine Foucault's analysis, for there are many things
 that order maintenance is not. It is not modeled on the
 rehabilitative ideal central to many disciplinary projects.... It
 does not feed into the psychotherapeutic. It does not coddle the
 disorderly. It does not aim to reform the disorderly so much as it
 does to punish them and to exclude them, in the sense of getting
 them off the street. (49)

If there is any descriptive truth to Kelling and Coles' prescriptive account of order maintenance policing, or to New Haven's order maintenance policy, then this interpretation is too quick. On Kelling and Coles' account, and under New Haven's policy, police do try to "reform" the disorderly--to help define and inculcate the norms of order that they must abide by. Thus it is impossible to identify the meaning of order maintenance policing accurately--to locate it within Foucault's analysis of the disciplinary project, or even to determine whether that analysis is relevant to it--without a more detailed account of what it involves. Foucault himself, the most prominent contemporary exponent of Nietzsche's genealogical method (which views historical and empirical study as a prerequisite for philosophical analysis), would insist on this empirical dimension to normative analysis. (50)

What Do Police Treat as Disorderly? In the same way that it relies on a stylized view of the techniques police use to control disorder, contemporary scholarship often relies on a stylized view about the activities that constitute "disorder," using broadly-defined behaviors like "loitering," "panhandling," "public drinking," and "public urination." To appreciate the limits of this approach, consider four specific examples of the way police departments and order maintenance practitioners have defined disorder. All of these examples come from Kelling's writings because he has been one of the few scholars to offer anything that approaches the detailed account of order maintenance practice that is needed.

"Obstructing" in the New York Subway. When the New York City Metropolitan Transit Authority developed its order maintenance initiative, it spent substantial time refining the regulations that transit police would enforce. Kelling and Coles offer one example:
 One rule, against "obstructing" was especially troublesome. It was
 sufficiently vague both to invite legal challenges, and to trouble
 line officers charged with enforcing it. Was a person lying halfway
 across a walkway that was otherwise unimpeded obstructing? How about
 a person who stopped at the bottom of stairs, just to get oriented?
 What about someone who set luggage on the platform, inadvertently
 blocking free movement? Was the person obstructing and therefore
 deserving police attention? Police were nervous that too much
 discretion was involved in answering these and a myriad of other
 questions about obstructing. As the group took photographs and
 videos of actual subway conditions and played out enforcement
 scenarios, members realized that the problem was not really
 obstructing, but lying down. (51)

Thus the broad concept of "obstructing" came to be replaced by the more concrete behavior of "lying down."

"Youth Parties" in Dayton. In Dayton, Ohio, police repeatedly received complaints about student parties, which created tensions between students and nearby residents. The department did not define all parties and late-night activity as "disorder" that would not be tolerated. Instead it consulted with students and residents to develop guidelines that regulated such activity and revisited them for reconsideration each year. For example, the guidelines discussed the traffic problems student parties created as follows:
 The residents recognize that large numbers of students in attendance
 at parties may require the temporary use of the streets. It is
 important, however, that vehicles be permitted to pass through these
 crowds so that residents can gain access to their property and so
 emergency vehicles can pass. Police crews are expected to ensure
 vehicles are not being blocked unless barricades have been erected
 in accordance with our block party procedures. (52)

"Public Urination" in New Haven. When the New Haven Police Department considered the issue of public urination in its order maintenance training bulletin, it did not define all cases of "public urination" as disorderly. Instead it delimited the contexts where police should target public urination by informing officers that "we would be less concerned about a person who urinated publicly if the person attempted to find a solitary location and maintain a sense of modesty than someone who flagrantly exposed him or herself in a highly visible location." (53)

Panhandling. The most controversial aspect of order maintenance policing has been its focus on panhandling. Kelling has advised police departments to oppose outright bans on panhandling, writing "because some beggars are no threat to anyone, it seems appropriate to limit legislation and police action to aggressive panhandling." Elaborating on that distinction, he explains that it requires detailed attention to context:
 Identical behavior can be passive or aggressive, depending upon its
 context. At least five contextual elements give meaning to behavior:
 time, location, number or aggregation of events, condition of the
 victim/observer relative to the perpetrator/actor, and the previous
 behavior/reputation of the perpetrator/actor. A person standing at
 the top of subway stairs, holding a paper cup, and begging during
 the 8 a.m. rush hour is not much of a threat. At 10 p.m., the same
 panhandler, behaving in the same manner as observed during the rush
 hour, is a genuine threat to a solitary elderly woman returning home
 from work. The time and the location and condition of the woman
 combine to turn the beggar's behavior into a threat that anyone in
 the woman's position would feel. (54)

Whether any of these specific cases or categories of actions justify police attention may remain controversial. The point is just that our judgments about these specific cases and categories may be different from--and ultimately more reliable and fine-grained than--our judgments about the general categories commonly used in the literature. For that reason, attempts to determine whether "public order" is a legitimate goal of policing are premature without greater empirical detail. Definitions of "disorder" that rely on broad categories like "obstructing," "youth parties," "public urination," and "panhandling" need to be replaced with a more context-specific definition of disorder--a definition that relies on more specific categories like "lying down on public steps," "flagrant public urination in a highly visible location," and "parties that persistently block vehicular access to streets," as well as particular cases like the beggar who approaches a solitary elderly woman in the dark to panhandle. Ethnographic observation is the most effective way to develop this kind of a nuanced picture of disorder.

Once again, inattention to these empirical subtleties undermines those few efforts that have been made to assess the intrinsic merits of "public order" as a goal for policing. Consider the following passage from Harcourt:
 What exactly does disorder mean? What does it communicate?
 Proponents of the broken windows theory assume that it means a
 neighborhood has lost control and doesn't care about crime. But
 surely there are other plausible meanings. It could signal artistic
 ferment, a youth hangout, rebellion, or an alternative lifestyle. It
 could communicate frenzied commercial activity, a frantic pace of
 life, or a bullish stock market. (55)

In this passage, Harcourt implies that "disorder" may be interpreted in a more positive way than the proponents of order maintenance policing have interpreted it--that disorder signifies positive social values like "artistic ferment" and healthy "rebellion," so efforts to suppress it are often misguided. (56) In particular, he argues that we only view disorder as threatening because the broken windows theory has taught us to do so; there is no intrinsic reason to view disorder as undesirable. (57) Is this alternative interpretation of "disorder" plausible? To the extent that police define "disorder" the way the departments summarized in the previous paragraph intended to define it, it is not. It is very hard to argue that lying down on a public stairway generally signifies any healthy form of political rebellion or artistic ferment, that young partygoers who refuse to make room for vehicles to pass a public street signify a desirable or even tolerable kind of youth hangout, and that flagrant public urination in highly visible areas signifies an alternative lifestyle in any positive sense. On the other hand, in cities or neighborhoods where police do indiscriminately prohibit all forms of "panhandling," "youth parties," or even "public urination," critiques like Harcourt's are more plausible. In those cases, order maintenance activity may unjustly suppress innocuous and even valuable activity.

Efforts to defend (rather than criticize) order maintenance policing have sometimes suffered from the same inattention to the empirical complexity of disorder. In the 1986 article mentioned above, Gary Sykes mounted a general moral defense of order maintenance policing on the grounds that disorder (which he defines roughly as "violations of community norms") undermines community life. (58) But as Carl Klockars noted at the time, any general defense of this kind will inevitably overreach. Sykes's defense, Klockars argued:
 suffers from being so abstract and general that it is not very
 helpful in distinguishing types of order maintenance policing which
 are good and should be encouraged from those which are bad and
 should not. Sykes leaves us with a thesis that says all order
 maintenance policing is a good thing; and that, I believe Sykes
 would agree, is just not true. (59)

It is just as problematic to imply that police should be empowered prohibit all instances of "panhandling," "loitering," and even "public urination" as it is to imply that they should have no power to regulate behavior that falls into these categories. Any viable attempt to defend order maintenance activity needs to draw careful and complex distinctions among kinds of public behavior.

For these reasons, to evaluate whether order maintenance efforts are appropriate, and to distinguish "types of order maintenance policing which are good and should be encouraged from those which are bad and should not," we need a detailed picture of what police might potentially treat as disorderly in practice to serve as a basis for considered judgments as to which of those situations justify police intervention. This picture must be constructed through ethnographic observation of the conditions of social life that police encounter in troubled neighborhoods. Just as we cannot say in the abstract whether "lying" is right or wrong, we cannot say in the abstract whether "disorder" should be prohibited. What we can say with greater confidence is whether particular kinds of behavior, described in the context where they occur, are a legitimate focus for police attention. We can confidently pass judgment on police who arrest a single panhandler who politely asks passersby in a busy commercial district for spare change, and we can confidently make the opposite judgment about police who ignore a man who flagrantly urinates in public on a neighborhood street traversed by families on their evening walk home. With more difficulty and less certainty, we can consider the many kinds of complicated cases that police encounter between these two extremes, asking whether and how they resemble the paradigm cases just described--on the one hand, cases of behavior that clearly should be permitted, and on the other hand, cases of behavior that clearly call for police attention. (60) The effort to describe and analyze such cases would begin with ethnographic accounts of how police carry out their peacekeeping function in practice, focusing on the kinds of situations police might potentially treat as disorderly and the steps they take when they do intervene. That effort would help to draw the line, one case at a time, between behavior and conditions that are innocuous and those that are intolerable--to draw a line around the normative concept of "disorder" that defines a particular kind of legitimate focus for police attention. In Albert Jonsen and Stephen Touhnin's terms, this effort would generate a moral taxonomy of order and disorder. (61)


Practical judgment about particular cases can often take us a long way, since judgments about concrete situations are often relatively firm. But particularly when controversial cases are at issue (such as those between the two extreme cases in the previous paragraph), it is useful to check these concrete judgments against more general moral principles. (62) To paraphrase Nelson Goodman, (63) moral principles and concrete judgments are justified by being brought into agreement with each other. General principles are amended if they yield a concrete judgment that we are unwilling to accept (e.g., the general principle that "disorder is a sign of artistic ferment" may need to be modified if we are unwilling to accept the inference that "flagrant public urination on a neighborhood street is a sign of artistic ferment"). But concrete judgments may also be rejected if they violate general principles we are unwilling to amend (e.g., the initial judgment that police should cite someone who urinates in public even if he does so "modestly" and in a side alley may need to be rejected by those who accept the analysis of homelessness and freedom presented below). That, at least, is the basic insight behind John Rawls's influential account of ethical reasoning as the pursuit of "reflective equilibrium," which drew explicitly on Goodman's account. (64)

In keeping with this view about ethical reasoning, analyses of order maintenance policing's intrinsic merits should draw from the abstractions provided by normative theory as well as the concrete images provided by ethnographic research. Where the latter offers concrete cases about which we can make tentative judgments, the former offers general principles with which those judgments can be compared. The pursuit of reflective equilibrium involves the encounter between these two features of our ethical commitments, in which we ask whether we have reasons to modify any of our existing beliefs. We use the concrete judgments to ask whether general principles are adequate, and we use the general principles to ask whether our concrete judgments should be overturned.

In the study of order maintenance policing, scholars have proposed various normative theories as starting points for this kind of analysis. In this section, I will briefly review two of those theories, and I will then argue that conventional liberal ideals (those that define individual rights and limit the goals that criminal justice may legitimately pursue with reference to conventional principles of philosophical liberalism) offer a better framework for analyzing the moral basis of order maintenance.

Foucauldian Political Theory. Among critics of order maintenance policing, the most sustained attempt to use normative theory to assess the intrinsic merits of order maintenance policing again comes from Harcourt, who offers an argument against order maintenance policing that draws on the philosophy of Michel Foucault. I have already argued that Harcourt has not offered an adequate empirical background to support that analysis. Nevertheless, it is possible that more thorough empirical study would not contradict his overall argument (indeed, I am inclined to think such study would actually reveal more of Foucault's core techniques of disciplinary power), so it is worth examining it in more detail.

Harcourt's argument is complex, but one important strand draws on Foucault's idea of "subject creation" to argue that order maintenance policing introduces an arbitrary dichotomy between "the orderly" and "the disorderly." In making this argument, Harcourt suggests that our current understanding of orderliness (and our preference for order over disorder) is socially constructed, and for that reason we should question its wisdom. (65) This analysis does not, however, have the normative force that Harcourt implies for it. The possibility that our ideas about "disorder" might be otherwise is not yet an argument that they should be. (66) The fact that current conceptions of order and disorder "may be constructed, may change, may be changed" (67) does not tell us whether those conceptions are justifiable, and the observation that these socially-constructed conceptions make efforts to police disorder "seem natural, necessary, preordained" (68) does not tell us whether those efforts are appropriate. Ronald Dworkin's words in a different context are an apt response to Harcourt's challenge: "We can do no more ... in the face of that caution, than to think again with that possibility in mind. The prospect of future change ... should no doubt make us more tentative and thoughtful. But it cannot count, in itself, as any argument against our present opinions." (69)

In the end, the question is not whether these natural-seeming ideas are socially constructed but whether we have good reasons to revise them. For example, for those who hold them, is there any reason to question the beliefs that police should intervene when someone flagrantly urinates in public, when student partiers block traffic and refuse to move, or when a panhandler approaches an elderly woman exiting the subway at night? Though Harcourt's analysis of the Foucauldian concept of subject creation is creative and thorough, I cannot see how it identifies justifiable normative principles that would call our intuitive judgments about these cases into question. Moreover, this gap does not result from the way Harcourt uses Foucault's philosophy; it is an inherent weakness of that philosophy. Foucault's analysis, although empirically rich in its descriptions of the techniques of modern power, lacks normative principles that can help to assess those techniques in cases where our intuitions about them are conflicted. Nancy Fraser expressed this view succinctly in an influential article:
 Because he has no basis for distinguishing, for example, forms of
 power which involve domination from those which do not, Foucault
 appears to endorse a one-sided, wholesale rejection of modernity as
 such. Furthermore, he appears to do so without any conception of
 what is to replace it.... Clearly what Foucault needs and needs
 desperately are normative criteria for distinguishing acceptable
 from unacceptable forms of power. (70)

In sum, Foucauldian genealogy can be an important aspect of ethical analysis, but it only offers the empirical component discussed in the last section, not the normative principles needed to complement it.

Communitarianism. Among order maintenance proponents, many scholars have drawn on communitarianism to justify police efforts to eliminate disorder. On this view, police derive the authority to maintain order from "community traditions" (71) or "community norms," (72) which establish the standards of orderly conduct that police enforce. Police can legitimately enforce these norms, even at some cost to individual freedom as that notion is conventionally understood because a vibrant and orderly community life is a prerequisite of true individual freedom and flourishing--because "no one is truly autonomous, [because] liberty can only exist in an environment of reasonable order, and [because] personal development requires familial and neighborhood support," as James Q. Wilson put it. (73) It is unreasonable to criticize order maintenance for infringing on personal liberty because liberty loses its value without a stable, orderly, supportive community environment in which to exercise it. Thus, Gary Sykes argued that:
 [I]n responding to the mandate for order maintenance, the police
 create a sense of community that makes social life possible. Where
 police are unwilling or unable to play this moral leadership role or
 define the community boundaries of right conduct, the quality of
 life declines and the existence of every other cherished value may
 be jeopardized. (74)

It has always been objected that this perspective grants police too much power to regulate public behavior, since some community norms should not be supported. (75) Even many communitarians have recognized that "neighborhoods empowered to impose their values upon individual behavior and expression can be both coercive and cruel," (76) so a principle holding that police may intervene in any situation that the community defines as disorderly is unacceptable because it grants police too much power. The hard work lies in defining the limits of that principle the limits beyond which the will of the community cannot justify police intervention.

Whether communitarianism can define those limits on its own is a matter of considerable controversy, (77) and I do not intend to engage in that dispute here because I think it is premature to do so. Scholars who rely on communitarian ideals to justify order maintenance policing often do so because they assume that liberalism cannot accomplish this task on its own. Thus, James Q. Wilson believes that liberalism's "rights-oriented legal tradition does not easily deal with" the problem of public disorder, (78) and this belief apparently leads him to turn to communitarian ideals to justify order maintenance. To the extent that order maintenance is consistent with existing liberal ideals, there is no need to turn to the controversial tenets of communitarianism to defend it.

Liberalism. Skepticism about liberalism's ability to justify criminal justice intervention against disorder seems to rest on a misunderstanding of the harm principle that underlies the liberal theory of criminal justice--the principle that no action can be prohibited by criminal law unless it causes harm to others. (79) The skeptics worry that any individual incident of disorder will not be viewed as "harmful," or that its harm will be viewed as trivial (which is just as bad, since the harm principle comes with the corollary that the law does not concern itself with trifles). (80) For example, in concluding that liberalism's rights-oriented legal tradition "does not easily deal with" the problem of disorder, Wilson emphasizes the way it focuses narrowly on individuals:
 The court will, typically, hear a case brought by (or on behalf of)
 an individual beggar, sleeper, or solicitor. Such an individual
 rarely constitutes much of a threat to anyone, and so the claims of
 communal order often seem, in the particular case, to be suspect or
 overdrawn. But the effects on a community of many individuals taking
 advantage of the rights granted to an individual ... are often
 qualitatively different from the effects of a single person.... The
 judge finds it hard to believe that one broken window is all that
 important.... (81)

Wilson is clearly right to conclude that this situation is more challenging to liberal ideals than more conventional criminal offenses like burglary or assault. But liberal theory does not utterly lack conceptual resources to deal with this kind of situation because it does not restrict the reach of criminal justice only to those acts that cause significant harms on their own. First, an act that does not cause dramatic harm in isolation may still justify criminal justice intervention if it qualifies as an accumulative harm. (82) Second, a harmless act may still justify criminal justice intervention if it qualifies as an offense to others. (83) With the aid of these concepts, liberalism has less difficulty than Wilson suggests in justifying police intervention against disorderly conduct that intuitively seems to call for it, and it offers useful principles to aid intuition in more complicated cases.

Consider accumulative harms first, since they speak directly to the worry Wilson expressed above. Accumulative harms are actions that do not cause significant harm in isolation but do harm some public interest (such as the viability of a neighborhood or a street corner) once a threshold has been crossed. Familiar examples include littering (dropping a single candy wrapper on the sidewalk will not spoil the streetscape but dropping many will) and air pollution (one car's emissions will have a negligible effect on air quality but the emissions of many cars may cause unsafe concentrations of pollutants). It is true that each instance of an accumulative harm is trivial in isolation, but as the instances accumulate, they eventually pass the threshold from inconvenience to harm. For reasons of fairness and practicality, it rarely makes sense to criminalize only the act that crosses that threshold--to prosecute the last straw, but not those that set the stage for it. The alternatives are to prohibit all instances of the accumulative harm (which is how we deal with littering) or to set up a system of regulation that controls the amount of harm that will accumulate, backing up that regulatory system with criminal sanctions (which is how we deal with air pollution). The regulatory approach is especially appropriate when the relevant behavior has great value to society or those who perform it (as in the case of driving a ear but not the case of dropping a candy wrapper on the sidewalk), since regulation makes it possible to allow some room for valuable activities while keeping their cumulative effects in check. (84)

Many instances of disorder can be viewed in this way, and they present neither more nor less difficult challenges for liberal theories of criminal justice than littering and pollution. Consider the case of "lying down" in the subway station from the previous section. It may be true that a single person lying down on a station staircase or platform is only a minor inconvenience, not a significant harm that justifies prohibition. But many people doing the same thing could easily shut the station down, and there is no fair and practical way to regulate the level of lying down in the station to avoid that outcome short of outright prohibition. In more complex cases, a regulatory response may be more appropriate, as proponents of order maintenance policing like Kelling have recognized. "Skillful order maintenance activities," Kelling writes, "acknowledge that [activities like] squeegeeing, ... prostitution, panhandling, and drug dealing, will never be eliminated, but good policing can determine the conditions under which and how such activities can take place...." (85) As Feinberg argues, this enterprise is most legitimate when it aims to maintain the overall level of harm below the crucial threshold where inconvenience fades into harm while interfering as little as possible with socially or personally valuable activities (e.g., by reserving alternative times and places for them). (86)

Some examples of disorder that intuitively seem to justify police intervention still cannot be accommodated within an expanded view of the harm principle that encompasses accumulative harms. Public lewdness, for example, is not usually harmful to anyone (at least not so significantly that the harm principle offers a strong reason to regulate it), but it seems inappropriate for police to ignore it entirely. (87) In such cases it may be appropriate to turn to the offense principle--the principle that "when public conduct causes offense to someone, the fact of that offense is relevant to the permissibility of the conduct." (88) To interpret this principle properly, it is important to note that Feinberg uses "offense" as a term of art: "Offended states" include a wide variety of harmless but unpleasant reactions including anxiety, disgust, fear, and shame. (89) The offense principle holds that the criminal justice system may regulate some behaviors that cause others to experience offended states. Feinberg has argued that an adequate liberal theory should not restrict itself to the harm principle but should also accept some version of the offense principle. Offenses "are nuisances, making it difficult to enjoy one's work or leisure in a locality which one cannot reasonably be expected to leave in the circumstances," (90) and for that reason they may justify criminal justice attention.

At the same time, Feinberg recognizes the danger the offense principle holds, in that it may justify an overly broad range of criminal justice intervention, so he advises caution in its application:
 People take offense--perfectly genuine offense--at many socially
 useful or even necessary activities, from commercial advertisement
 to inane chatter. Moreover, bigoted prejudices of a very widespread
 kind (e.g., against interracial couples strolling hand in hand down
 the main street of a town in the deep South) can lead onlookers to
 be disgusted and shocked, even "morally" repelled, by perfectly
 innocent activities, and we should be loath to permit their
 groundless repugnance to outweigh the innocence of the offending
 conduct. For these and similar reasons, the offense principle must
 be formulated in a very precise way, and supplemented by appropriate
 standards or mediating maxims, so as not to open the door to
 wholesale and intuitively unwarranted legal interference. (91)

In response to this worry, Feinberg nominates a number of factors that must be considered in applying the offense principle: (92)
 The extent of the offense, which consists of "the intensity and
 durability of the repugnance produced, and the extent to which
 repugnance could be anticipated to be the general reaction of
 strangers to the conduct displayed or represented (conduct offensive
 only to persons with an abnormal susceptibility to offense would not
 count as very offensive);"

 The avoidability of the offense, that is, "the case with which
 unwilling witnesses can avoid the offensive displays;" at one
 extreme, it is rarely appropriate to prohibit offense to those who
 consent to be offended, while at the other, captive observers
 deserve special consideration;

 The value of the offending activity, as comprised by its "personal
 importance to the actors themselves and its social value generally;"

 "[T]he availability of alternative times and places where the
 conduct in question would cause less offense;" and

 The motives that drive the offending activity--particularly "the
 extent, if any, to which the offense is caused with spiteful
 motives." (93)

It is particularly important to analyze the extent of the offense, the value of the offending activity, and the availability of alternative times and places to perform it.

Analyzing the extent of the offense can be particularly difficult, but one important piece of evidence has to do with how widespread the offended reaction is--whether a behavior offends most people or only those with abnormal susceptibility. As Feinberg puts it, "the more fragile our sensitive sufferer's psyche, the less protection he can expect from the criminal law.... If a mere sneeze causes a glass window to break, we should blame the weakness or brittleness of the glass and not the sneeze." (94) Thus, "the more widespread the offense ... the stronger the case for prohibition of the conduct that produces it, that is, the weightier must be the considerations on the other side to counterbalance it." (95) In the case of disorder, it is primarily those potentially disorderly behaviors that offend a wide cross-section of people rather than the skittish few that may justify police intervention. The important role of survey research is obvious here, since it can help to determine whether various possible examples of "disorder" inspire offended mental states like fear, anxiety, and disgust in many observers or only in a few. Wesley Skogan's work, which identified a rough consensus that certain kinds of behavior are offensive to a broad range of people, has laid an important foundation in this area. (96) It would be useful to follow it up with more detailed survey evidence about reactions to common concrete examples of disorder. Note that the crucial question for this research is not whether the residents of neighborhoods that are generally high in disorder generally have high levels of fear. (97) Instead it is whether specific behaviors, in and of themselves, provoke offended mental states like fear, anxiety, and disgust in a wide range of people or only in the skittish few. Neuropsychological research that investigates the physical bases of reactions like fear and disgust in the brain may offer useful contributions to this inquiry by presenting relatively standardized ways of identifying offended states. (98)

With this background, consider how the offense principle might help to analyze New Haven's approach to public urination. The lenient treatment of the man who relieves himself "modestly" in a side alley raises complications to which I will return. But the offense principle clearly justifies police intervention against his flagrant counterpart. Even if that man's behavior is not positively harmful (if we put aside public health concerns, for example), it almost certainly qualifies as the kind of offensive behavior that justifies criminal justice intervention; we hardly need to wait for survey results to guess whether this behavior is broadly offensive. Moreover, although the need to relieve oneself somewhere is great--the conduct "has value" to the person engaging in it--the man who does so flagrantly has many alternatives that would cause less offense. At the least, he could find a side alley and exercise some degree of modesty.

The man who relieves himself "modestly" in a side alley presents different issues that help to clarify an important constraint on order maintenance policing. Philosopher Jeremy Waldron has offered a useful framework for analyzing this example and related order maintenance tactics that impinge on the homeless. Waldron focuses on the implications of the obvious fact that the homeless are defined by their lack of a home--that there "is no place governed by a private property rule from which [they] may not at any time be excluded as a result of someone else's say-so." (99) This observation has an important implication for our understanding of freedom: when an action is prohibited on public property, it is for all practical purposes prohibited entirely for the homeless.
 For a person who has no home, and has no expectation of being
 allowed into a private office building or a restaurant, prohibitions
 on things like sleeping that apply particularly to public places
 pose a special problem. For although there is no general prohibition
 on acts of these types, still they are effectively ruled out
 altogether for anyone who is homeless and who has no shelter to go
 to. The prohibition is comprehensive in effect because of the
 cumulation, in the case of the homeless, of a number of different
 bans, differently imposed. The rules of property prohibit the
 homeless person from doing any of these acts in private, since there
 is no private place that he has a right to be. And the rules
 governing public places prohibit him from doing any of these acts in
 public, since that is how we have decided to regulate the use of
 public places.... [S]ince freedom to perform a concrete action
 requires freedom to perform it at some place, it follows that the
 homeless person does not have the freedom to perform them. If
 sleeping is prohibited in public places, then sleeping is
 comprehensively prohibited to the homeless. If urinating is
 prohibited in public places (and if there are no public lavatories)
 then the homeless are simply unfree to urinate. (100)

In short, regulations that prohibit acts like urination, sleeping, lovemaking, and cooking in public eliminate (for the homeless) the freedom to "tak[e] care of these elementary needs themselves, quietly, with dignity, as ordinary human beings." (101) If such regulations do not violate a formal right enumerated in the constitution, they violate something very similar, and something that is bound up with formal rights as a precondition to their exercise; indeed they restrict the ability of the homeless do things that are "basic to the sustenance of a decent or healthy life, in some cases basic to the sustenance of life itself." (102) Moreover, because the homeless lack access to private space of their own, they lack "alternative times and places" in which to carry out these essential activities.

This analysis is important for the issue at hand because it indicates that outright prohibitions on public behaviors of these kinds fail Feinberg's standards for the proper application of the offense principle in two ways. Such bans may indeed prohibit something that is "offensive," but they impinge on behavior that has very great value to the person engaged in it, and (because the homeless lack access to private space of their own) they leave no alternative times and places where it can be carried out.

It is possible to read Waldron's article as a critique of regulations like order maintenance policing, and it may have been intended that way. (103) It is better read, however, as an account of the constraints within which humane and legitimate order maintenance policies must operate. There are many essential human activities that the homeless must perform somewhere, even if they are offensive to others, so outright prohibition of them is problematic. Instead of having "zero tolerance" for activities like public urination, public drinking, and public sleeping or cooking, police should regulate them so that the homeless may perform them somewhere, but in a discrete way that contributes as little as possible to a sense of neighborhood disorder. This conclusion indicates the need for a nuanced approach to order maintenance that eschews general prohibitions and asks officers to consider context closely. New Haven's policy guidelines, at least, are consistent with this ideal, for they ask police to regulate the manner of public urination and other instances of "disorder" without prohibiting them entirely. By attending to context to determine how they should handle particular cases (Is there a nearby public restroom that the man might have used? Has he made an effort to be discrete?), they remain within the boundaries of legitimate regulation that Waldron implies. Moreover, since different communities offer different opportunities to meet these basic needs, their scope for legitimate regulation of offensive behavior differs as well: A community that provides adequate public restrooms has more legitimate discretion to prohibit public urination than one that does not, and a community with adequate public shelters is more justified in regulating public sleeping than one that does not. Such differences offer a principled basis for decisions to adopt different standards of "order" from community to community.

As already noted, Kelling has elevated these ideas to general principles, writing that good policing should not aim to "eliminate" many disorderly activities but instead try to "determine the conditions under which and how such activities can take place" and that appropriate standards may vary from one community to another. (104) Taken seriously, this conception of order maintenance policing need not entail any illegitimate intrusions on the quasi-rights Waldron identifies, and it lies within the boundaries of Feinberg's guidelines for application of the offense principle.

Order maintenance policing finds its best justification and its best guiding principles in this understanding of liberal theory. We misunderstand liberalism if we believe it only concerns itself with behavior that is seriously harmful in isolation. A sophisticated version of liberal theory should extend its concern to some harmless offenses and to some accumulative harms. Offenses and accumulative harms are more complex targets for criminal justice concern, so it is not surprising that there has been a tendency in criminal justice and academic circles to neglect them. It is the merit of order maintenance policing to call police attention to this difficult boundary area of problematic behavior--to remind police of the role they should play in regulating unacceptable kinds of behavior that make unfair use of public spaces, unduly undermine neighborhood livability, and give unnecessary or inappropriate offense to a wide range of people. At the same time, government should be careful as it moves away from serious harms and towards mere offenses and less considerable harms as objects of regulation. The latter do demand attention, but in a careful and moderate way that incorporates the kinds of considerations that Feinberg identified. It is essential to balance the extent and avoidability of an offense or minor harm against the value the offender's behavior has to himself, and to consider the availability of alternative times and places where the offender might engage in the behavior. (105) It is the merit of the most sophisticated brands of order maintenance policing--those that pay close attention to the discretion that police must exercise when they maintain order, and that offer detailed guidelines that shape the way police should use that discretion--that they regulate disorder in a way that is sensitive to these nuances. It is the flaw of cruder versions, like "zero tolerance" policing, that they ignore these important subtleties.


Before concluding, I need to consider one objection to this argument. Even if public order is a desirable goal, it is reasonable to ask whether police are the right institution to enforce it. Indeed, some scholars have argued that the task of order maintenance should be assigned to community institutions and informal community groups rather than the police. (106) Though this proposal deserves attention, it confronts two initial difficulties. First, considerable evidence suggests that informal sanctions by community institutions work badly without formal sanctions to back them up. (107) Second, it is not clear whether informal community institutions acting alone can be held sufficiently accountable for the exercise of this delicate social control function. In this connection, it is surprising to find a Foucauldian such as Harcourt proposing that community institutions rather than police should bear primary responsibility for controlling disorder. (108) Foucault was notoriously reluctant to state and defend his own normative views, but his general discomfort with disciplinary power exercised through the carceral network (i.e., institutions of social control outside the formal justice system, such as schools, factories, and social work agencies) was obvious. (109) One of his major contributions was to call attention to the dangers inherent in the informal and distributed exercise of power, which (unlike centralized juridical power) risks flying under the radar screen of our awareness. "[T]he carceral pyramid gives to the power to inflict legal punishment a context in which it appears to be free of all excess and all violence ...," Foucault wrote; for that reason, the power it exercises is "hidden," and "[a]s a result, justice no longer takes public responsibility for the violence that is bound up with its practice." (110) If society delegates the order maintenance function to community institutions like "park-keepers," "station guards," and "church organizations" (111) without formal involvement and oversight by more visible and arguably more accountable institutions like the police, there is a danger that justice will no longer take responsibility for the power bound up with its practice.


Because offenses and accumulative harms are ambiguous and uncertain phenomena, there is a tendency for scholars and practitioners to neglect them. Where the concept of serious harm appears fairly clear-cut, the concepts of accumulative harm and offense raise difficult normative questions about which behaviors they encompass, when the desire to regulate them should give way to other interests, and what kind of regulations are appropriate when any are. It is tempting to try to avoid these questions by turning to the apparent certainty of strong causal reasoning--to ask not whether and how disorder that is offensive or harmful in accumulation should be regulated in and of itself, but whether it causes the kinds of serious harm that everyone already agrees should be reduced. We should resist this temptation for the reasons given in the first section of this paper. Strong causal arguments rarely succeed to the degree needed, (112) and even if they did, it is not clear that they would justify restrictions on otherwise innocuous behavior because it is usually inappropriate to punish someone for unintentionally causing someone else to commit a serious crime voluntarily. (113)

When the promises of strong causal reasoning fade, it will be necessary to revisit questions about the intrinsic merits of order maintenance policing. This paper has proposed one way of doing so one that focuses less on testing the strong causal broken windows theory and more on ethnographic description of order maintenance practice, together with normative analysis of the political ideals it implicates. It is not hard to see how this approach might apply more broadly to other criminal justice topics. In this respect, the order maintenance case offers a concrete example of how scholarship might contribute to policy debate by analyzing the intrinsic merits of criminal justice policies rather than their indirect effects on crime. In Rein and Winship's language, (114) the order maintenance case suggests by example how scholarship can move beyond strong causal reasoning and thereby avoid the dangers associated with its use.

Social scientists may be uncomfortable with this argument, which may seem to discount their central tool of causal analysis. (115) This paper's argument does not, however, imply that social scientists and other scholars can make no contribution to the study of order maintenance policing and other criminal justice policy issues. Most simply, I have not argued that causal analysis is inappropriate; I have only argued that what Rein and Winship call "strong causal reasoning"--causal analysis that aims to identify large indirect effects produced through complex causal chains--is unlikely to offer clear-cut advice for policy. That conclusion leaves substantial room for social science analysis in policy research. In particular, there remains an important role for "weak" causal analyses that do not rely on long causal chains--notably those that investigate whether order maintenance policing has a direct effect on order. (116) Moreover, although strong causal arguments may have limited implications for policy, they still hold considerable interest for basic science. Complex, surprising theories like the broken windows theory are intellectually fascinating, and such theories will always play a role in social science. I have only argued that we should not expect they will support clear-cut policy arguments; to the extent that they contribute to policy at all, they are more likely to inform careful and modest cost-effectiveness analyses rather than open-and-shut cases in policy argumentation. In these respects, this paper's argument leaves considerable room for the use of causal analysis in the study of order maintenance policing, as well as other policy questions in criminal justice.

Nevertheless, my argument does give reasons to conclude (notwithstanding the contrary views of Max Weber) (117) that causal analysis does not offer the only way for social scientists and other scholars to contribute to knowledge about social problems like crime and disorder. A full understanding of the intrinsic value of order maintenance requires two kinds of contributions that scholars are particularly well-situated to make, but that do not primarily involve causal analysis. The first is ethnographic research, which contributes to the analysis of order maintenance policing by systematically describing the kinds of situations that police might potentially treat as disorderly and the kinds of actions they might take to regulate it. Such thick description is an essential first step for normative analysis for the reasons I have given. It extends our understanding of the legitimate scope for order maintenance policing by helping to develop a moral taxonomy of order and disorder.(118) The second kind of contribution is normative analysis, which can aid our intuition when we are unsure whether and how police should respond to hard cases that may or may not justify criminal justice intervention. By drawing on a sufficiently sophisticated understanding of liberal political theory--one that recognizes reasons to regulate harmless behavior that is broadly offensive, or minor harms that accumulate to create more significant harms (119)--scholars can help determine whether and how police are justified in taking action against disorder. This kind of ethnographic study and normative analysis are the most pressing areas for future research about order maintenance policing, and indeed about many other areas of public policy.

(1) James Q. Wilson & George L. Kelling, Broken Windows: The Police and Neighborhood Safety, ATLANTIC MONTHLY, Mar. 1982, at 29-38.


(3) Thus one recent study concludes that "researchers should further evaluate the relationship between crime and disorder and should examine the effects that the police can exert on crime by policing social and physical disorder[]" because "[b]y doing so, they can determine whether quality-of-life policing is good public policy." Charles M. Katz et al., An Assessment of the Impact of Quality-of-Life Policing on Crime and Disorder, 18 JUST. Q. 825, 861 (2001).

(4) Martin Rein & Christopher Winship, The Dangers of "Strong" Causal Reasoning in Social Policy, 36 SOC'Y 38 (1999).

(5) E.g., MICHAEL BANTON, THE POLICEMAN IN THE COMMUNITY (1964); JAMES Q. WILSON, VARIETIES OF POLICE BEHAVIOR: THE MANAGEMENT OF LAW AND ORDER IN EIGHT COMMUNITIES (1968); Egon Bittner, Police Discretion in Emergency Apprehension of Mentally Ill Persons, 14 SOC. PROBS. 278 (1967); Egon Bittner, The Police On Skid Row: A Study of Peace Keeping, 32 AM. SOC. REV. 699 (1967); Elaine Cumming et al., Policeman as Philosopher, Guide, and Friend, 12 SOC. PROBS. 276 (1965).

(6) E.g., George L. Kelling, Acquiring a Taste for Order: The Community and Police, 33 CRIME & DELINQ. 90 (1987) [hereinafter Kelling, Acquiring a Taste for Order]; George L. Kelling, Conclusions, in THE NEWARK FOOT PATROL EXPERIMENT (1981) [hereinafter Kelling, Conclusions]; Gary W. Sykes, Street Justice: A Moral Defense of Order Maintenance Policing, 3 JUST. Q. 497 (1986); Samuel Walker, "Broken Windows" and Fractured History." The Use and Misuse of History in Recent Police Patrol Analysis, 1 JUST. Q. 75 (1984).

(7) Rein & Winship, supra note 4; David Thacher, Policing is Not a Treatment: Alternatives to the Medical Model of Police Research, 38 J. RES. CRIME & DELINQ. 387 (2001); Mayer Zald, Sociology as a Discipline: Quasi-Science and Quasi-Humanities, 22 AM. SOCIOLOGIST 165 (1991).

(8) 374 U.S. 483 (1954).

(9) Rein & Winship, supra note 4.

(10) Id. at 43.

(11) Id. at 43-44.

(12) Wilson & Kelling, supra note 1.


(14) E.g., MICHAEL MAXFIELD, EXPLAINING FEAR OF CRIME: EVIDENCE FROM THE 1984 BRITISH CRIME SURVEY (1990); WESLEY SKOGAN, DISORDER AND DECLINE: CRIME AND THE SPIRAL OF DECAY IN AMERICAN NEIGHBORHOODS (1990); Robert Sampson & Jacqueline Cohen, Deterrent Effects of the Police on Crime: A Replication and Theoretical Extension, 22 LAW & SOC'Y REV. 163 (1988).

(15) E.g., James Q. Wilson & Barbara Boland, The Effect of the Police on Crime, 12 L. & SOC'Y REV. 367 (1978).

(16) E.g., WILLIAM BRATTON & PETER KNOBLER, TURNAROUND: HOW AMERICA'S TOP COP REVERSED THE CRIME EPIDEMIC (1998); George Kelling & William Sousa, Do Police Matter? An Analysis of the Impact of New York City's Police Reforms, 22 CIVIC REP. 1 (2001).


(18) E.g., RALPH B. TAYLOR, BREAKING AWAY FROM BROKEN WINDOWS: BALTIMORE NEIGHBORHOODS AND THE NATIONWIDE FIGHT AGAINST CRIME, GRIME, FEAR, AND DECLINE (2001); Katz et al., supra note 3; Kenneth J. Novak et al., The Effects of Aggressive Policing of Disorder on Serious Crime, 22 POLICING 171 (1999); Robert J. Sampson & Stephen W. Raudenbush, Systematic Social Observation of Public Spaces: A New Look at Disorder in Urban Neighborhoods, 105 AM. J. Soc. 603 (1999).

(19) Sampson & Raudenbush, supra note 18. Note, however, that Sampson and Raudenbush's study did not in any sense disprove the broken windows theory. Broken windows offers a theory of policing, claiming that if police intervene in minor disorder they will stave off serious crime down the road. By contrast, Sampson and Raudenbush (to simplify only a little) offer an important criminological study of the natural relationship between disorder and crime, concluding that the two phenomena do covary in time and space, but that common social factors drive that covariation in the Chicago neighborhoods they studied. This conclusion does not directly challenge broken windows because it omits policing practice as an explanatory variable. The policy question posed by broken windows is whether police intervention into the natural process that links crime, disorder, and other social factors can alter the level of crime that results from that process. Since police practice did not vary across Chicago neighborhoods (or at least Sampson and Raudenbush did not study that variation), their research, while important for other reasons, does not speak directly to the broken windows hypothesis.

(20) Benjamin Bowling, The Rise and Fall of New York Murder, Zero Tolerance or Crack's Decline?, 39 BRIT. J. CRIMINOLOGY 531 (1999); John Eck & Edward Maguire, Have Changes in Policing Reduced Violent Crime? An Assessment of the Evidence, in THE CRIME DROP IN AMERICA 207 (Alfred Blumstein & Joel Wallman eds., 2000); Jeffrey Fagan et al., Declining Homicide in New York City: A Tale of Two Trends, 88 J. CRIM. L. & CRIMINOLOGY 1277 (1998); Judith A. Greene, Zero Tolerance." A Case Study of Police Policies and Practices in New York City, 45 CRIME & DELINQ. 171 (1999). But see Kelling & Sousa, supra note 16.

(21) Sampson & Raudenbush, supra note 18, at 638.

(22) E.g., HARCOURT, supra note 17; Katz et al., supra note 3; Loic Wacquant, Scrutinizing the Street: Poverty, Morality, and the Pitfalls of Urban Ethnography, 107 AM. J. SOC. 1468 (2002).

(23) E.g., WILSON, supra note 5; Kelling, Conclusions, supra note 6.

(24) E.g., HERMAN GOLDSTEIN, POLICING A FREE SOCIETY (1977); CARL KLOCKARS, THINKING ABOUT POLICE: CONTEMPORARY READINGS (1983); Kelling, Acquiring a Taste for Order, supra note 6; Carl Klockars, Street Justice: Some Micro-Moral Reservations: Comment on Sykes, 3 JUST. Q. 513 (1986) [hereinafter Klockars, Street Justice]; Sykes, supra note 6; Walker, supra note 6.

(25) See sources in note 5. "Broken Windows" itself emerged out of this earlier literature. On Kelling's account, the article arose out of an earlier evaluation of foot patrol activities in Newark, which he had overseen for the Police Foundation and which focused partly on the order maintenance function. After reading the description of order maintenance activities in the evaluation, Wilson contacted Kelling to discuss the broken windows idea. GEORGE L. KELLING & CATHERINE M. COLES, FIXING BROKEN WINDOWS: RESTORING ORDER AND REDUCING CRIME IN OUR COMMUNITIES 319 (1996). (26) Sykes, supra note 6, at 501.

(27) Id. at 501-02.

(28) Rein & Winship, supra note 4, at 39.

(29) HARCOURT, supra note 17, at 207.

(30) FEINBERG, HARM TO OTHERS, supra note 2, at 232-43; H.L.A. HART & A. M. HONORE, CAUSATION IN THE LAW 129-51 (1959).

(31) FEINBERG, HARM TO OTHERS, supra note 2, at 239. One important difference between the case at hand and the moviemaking case is that the social value of even innocuous disorder is often much less significant than the social value of free artistic expression. But the two cases are formally similar because both involve "prima facie innocent conduct" (insofar as broken windows purports to justify the criminalization of disorder that would otherwise be permitted) that is "a 'but-for cause' of the commission of a seriously harmful crime by independent parties" (i.e., the crime would not have happened "but for" the innocent conduct). See id. at 232, 239-40.

(32) HARCOURT, supra note 17, at 208 cites Young v. New York City Transit Authority, 903 F.2d 146 (2d Cir. 1990) as a case in which a court relied on "broken windows" to justify an order maintenance ordinance (in this case an anti-begging ordinance in the subway). But in its opinion, the court did not rely on arguments about the indirect effects of disorder; instead it focused on the kinds of immediate harms and offenses discussed below, writing,
 [Subway] passengers perceive begging and panhandling to be
 "intimidating," "threatening," and "harassing." The conduct often
 involves "unwanted touching [and] detaining" of passengers. The
 police have great difficulty distinguishing between "panhandling
 and extortion." Begging is "inherently aggressive" to the "captive"
 passengers in the close confines of the subway atmosphere. Based on
 these facts, it is fair to say that whether intended as so, or not,
 begging in the subway often amounts to nothing less than assault,
 creating in the passengers the apprehension of imminent danger.
 Additionally, begging in the subway raises legitimate concerns about
 public safety. The conduct "disrupts" and "startles" passengers,
 thus creating the potential for a serious accident in the
 fast-moving and crowded subway environment. In short, the ...
 judgment that begging is alarmingly harmful conduct that simply
 cannot be accommodated in the subway system is not unreasonable.

Young, 903 F.2d at 158. Broken Windows' indirect causation plays no role in this analysis.

(33) Rein & Winship, supra note 4, at 40-41.

(34) Id. at 45.

(35) E.g., HARCOURT, supra note 17, at 127-248.


(37) E.g., SISELA BOK, LYING (1978). For a similar approach in just war theory, see MICHAEL WALZER, JUST AND UNJUST WARS: A MORAL ARGUMENT WITH HISTORICAL ILLUSTRATIONS (1978) [hereinafter WALZER, JUST AND UNJUST WARS]; Michael Walzer, Worm War II: Why Was This War Different?, 1 PH1L. & PUB. AFF. 3 (1971).

(38) JONSEN & TOULMIN, supra note 36.



(41) FEINBERG, OFFENSE TO OTHERS, supra note 2, at 10.

(42) Kelling, Conclusions, supra note 6; cf Bittner, supra note 5; Wilson, supra note 5.

(43) Bernard Harcourt, Policing Disorder, 27 BOSTON REV. 2, 20 (2002); cf. HARCOURT, supra note 17, at 48-49, 149.

(44) E.g., Greene, supra note 20.


(46) KELLING & COLES, supra note 25, at 188-89.

(47) Cf. BOK, supra, note 37; JONSEN & TOULMIN, supra note 36; WALZER, JUST AND UNJUST WARS, supra note 37.


(49) HARCOURT, supra note 17, at 149.

(50) Bent Flyvbjerg, Aristotle, Foucault, and Progressive Phronesis, in APPLIED ETHICS: A READER 11-27 (Earl Winkler & Jerrold Coombs cds., 1993); cf FLYVBJERG, supra note 40; Michel Foucault, Nietzsche, Genealogy, History, in THE FOUCAULT READER 76-100 (Paul Rabinow ed., 1984).

(51) KELLING & COLES, supra note 25, at 123.

(52) KELLING, BROKEN WINDOWS, supra note 45, at 30.

(53) Id. at 51.

(54) Id. at 33-35.

(55) HARCOURT, supra note 17, at 132.


(57) HARCOURT, supra note 17, at 127-84.

(58) Sykes, supra note 6, at 497.

(59) Klockars, Street Justice, supra note 24, at 513.

(60) I describe this kind of analysis in more detail in David Thacher, The Casuistical Turn in Planning Ethics: Lessons from Law and Medicine, 23 J. PLAN. EDUC. & RES. 269 (2004).

(61) JONSEN & TOULMIN, supra note 36.

(62) In the previous paragraph I suggested a less principle-driven alternative for analyzing these hard cases: That the cases should be analyzed using the methods of analogical reasoning, which aim to determine whether a hard case resembles a clear case of disorder that warrants police attention or a clear case of innocuous behavior that does not. Analogical reasoning of this kind is an important tool in ethical analyses of hard cases. See JONSEN & TOULMIN, supra note 36. In the final analysis, however, it cannot avoid engagement with more general principles altogether. We conclude that two cases are different (or similar) for a reason, even if that reason is not theoretically ambitious or given in advance. See CASS SUNSTEIN, LEGAL REASONING AND POLITICAL CONFLICT 65 (1996). I have discussed this issue at greater length in Thacher, supra note 60.


(64) JOHN RAWLS, A THEORY OF JUSTICE 19-21, 48-51 (1971); cf. RICHARDSON, supra note 39; John Rawls, Outline of a Decision Procedure for Ethics, 60 PHIL. REV. 177 (1951); John Rawls, The Independence of Moral Theory, in COLLECTED PAPERS (Samuel Freeman ed., 1999); David Thacher, Value Rationality in Policy Analysis (2003) (unpublished manuscript, on file with author).

(65) HARCOURT, supra note 17, at 127-84.

(66) Ronald Dworkin, Objectivity and Truth: You'd Better Believe It, 25 PHIL. & PUB. AFF. 113 (1996).

(67) HARCOURT, supra note 17, at 134.

(68) Id. at 158.

(69) Ronald Dworkin, Reply, BROWN ELEC. ARTICLE REV. SERV. (Apr. 9, 1997), at

(70) Nancy Fraser, Foucault on Modern Power: Empirical Insights and Normative Confusions, 1 PRAXIS INT'L 272, 286 (1981); cf. JURGEN HABERMAS, THE PHILOSOPHICAL DISCOURSE ON MODERNITY (1990).

(71) Kelling, Acquiring a Taste for Order, supra note 6, at 96.

(72) Sykes, supra note 6, at 497.

(73) James Q. Wilson, Foreword to KELLING & COLES, supra note 25, at xiii.

(74) Sykes, supra note 6, at 505.

(75) JOHN STUART MILL, ON LIBERTY (1978); Klockars, Street Justice, supra note 24, at 514; David Thacher, Equity and Community Policing, 20 CRIM. JUST. ETHICS 3, 5 (2001).

(76) PETER BERGER & RICHARD NEUHAUS, TO EMPOWER PEOPLE 68 (1996); cf. Kelling, Acquiring a Taste for Order, supra note 6, at 101.

(77) See, e.g., Philip Selznick, Social Justice. A Communitarian Perspective, in THE ESSENTIAL COMMUNITARIAN READER (Amitai Etzioni ed., 1996).

(78) Wilson, supra note 73, at xv.

(79) FEINBERG, HARM TO OTHERS, supra note 2; MILL, supra note 75.

(80) FEINBERG, HARM TO OTHERS, supra note 2, at 188-90.

(81) Wilson, supra note 73, at xiv-xv.

(82) FEINBERG, HARM TO OTHERS, supra note 2, at 225-32.

(83) See FEINBERG, OFFENSE TO OTHERS, supra note 2.

(84) FEINBERG, HARM TO OTHERS, supra note 2, at 229. Feinberg analyzes accumulative harms in illuminating detail, and in the process he shows that not all behaviors that become harmful through accumulation may legitimately be prohibited. Briefly, "the conduct should be permitted"--despite the fact that harm would result if large numbers of people engaged in the activity--"if in fact the likely number of persons who will engage in it falls below the threshold of harm, unless all have an interest in engaging in the conduct, but sufficient numbers refrain out of moral scruple or civic spirit." Id. at 244. This exception does not apply to the cases I will be discussing.

(85) KELLING, BROKEN WINDOWS, supra note 45, at 36.

(86) FEINBERG, HARM TO OTHERS, supra note 2, at 227-32, 244.

(87) As my parenthetical qualification suggests, the distinction between offenses and minor harms is often hard to draw, and I will not distinguish carefully between them in the discussion that follows. Feinberg himself describes the distinction as "to some degree arbitrary." Id. at 46. This imprecision is not important because Feinberg uses almost exactly the same language to develop many of the "mediating maxims" that he would apply to minor harms as he uses to develop the maxims appropriate for mere "offenses." Compare id. at 50, with FEINBERG, OFFENSE TO OTHERS, supra note 2, at 34. On the precise relationship between minor harms and offenses, see FEINBERG, HARM TO OTHERS, supra note 2, at 45-51.

(88) FEINBERG, OFFENSE TO OTHERS, supra note 2, at 26.

(89) Id. at 1.

(90) Id. at 22.

(91) Id. at 25-26.

(92) Id. at 25-47. The list that follows is condensed from page twenty-six, and all the quotations appear on that page.

(93) Id. at 26.

(94) Id. at 34. There is one important exception: Those with abnormal susceptibility to an offense deserve protection against those cruel souls who would intentionally exploit it. Id. at 33.

(95) Id. at 27.

(96) SKOGAN, supra note 14.

(97) See TAYLOR, supra note 18; see also MAXFIELD, supra note 14. Such research may, of course, be very important for other reasons.

(98) E.g., Andrew Calder et al., Neuropsychology of Fear and Loathing, 2 NATURE REVIEWS NEUROSCIENCE 352 (2001).

(99) Jeremy Waldron, Homelessness and the Issue of Freedom, in LIBERAL RIGHTS 313 (1993).

(100) Id. at 328-29.

(101) Id. at 334.

(102) Id. at 333.

(103) Waldron mentions order maintenance measures adopted in the New York City subways as an example of his concerns. Id. at 327-28.

(104) KELLING, BROKEN WINDOWS, supra note 45, at 36.

(105) Harcourt proposes a different way to analyze order maintenance policing, one that essentially balances the harms of disorder against the harms caused by efforts to police it. HARCOURT, supra note 17, at 209-14. In particular, he suggests balancing the "economic and aesthetic harms" of disorder against the harms that order maintenance itself causes--including the human and financial costs of misdemeanor arrests, the stigmatization of blacks, and the possibility that order maintenance exacerbates police brutality. Id. at 211. Harcourt's attention to the harms caused by order maintenance policing is important, but it needs to be supplemented with something like the more complex balancing process Feinberg suggests, which does not merely weigh harm against harm but instead offers a complex framework for comparing the interests of those who are harmed or offended by conduct with the interests of those who engage in it.

(106) E.g., id. at 221-24; Roger Matthews, Replacing 'Broken Windows', in ISSUES IN REALIST CRIMINOLOGY 19 (Roger Matthews & Jock Young eds., 1992).

(107) ROBERT K. YIN et al., U.S. DEP'T OF JUSTICE, PATROLLING THE NEIGHBORHOOD BEAT (1976); Janet Foster, Informal Social Control and Community Crime Prevention, 35 BRIT. J. CRIMINOLOGY 563 (1995); Tim Hope, Community Crime Prevention, in 19 CRIME & JUST. 21 (Michael Tonry & Norval Morris eds., 1995).

(108) HARCOURT, supra note 17, at 221-24.

(109) For an explication of this aspect of Foucault's critique see ANDREW POLSKY, RISE OF THE THERAPEUTIC STATE (1991).

(110) FOUCAULT, supra note 48, at 302, 09.

(111) Matthews, supra note 106, at 39.

(112) See Rein & Winship, supra note 4.

(113) FEINBERG, HARM TO OTHERS, supra note 2, at 232-43; HART & HONOREE, supra note 30, at 129-51.

(114) See Rein & Winship, supra note 4.

(115) On the centrality of causal analysis in social science, see MAX WEBER, Science as Vocation, in FROM MAX WEBER 129, at 144-45, 151 (H.H. Gerth & C. Wright Mills eds., 1958) and MAX WEBER, The Meaning of "Ethical Neutrality" in Sociology and Economics, in MAX WEBER ON THE METHODOLOGY OF THE SOCIAL SCIENCES 1, at 26 (Edward A. Shils and Henry A. Finch, eds., 1949) [hereinafter WEBER, The Meaning of "Ethical Neutrality "].

(116) E.g., Katz et al., supra note 3.

(117) See WEBER, The Meaning of "Ethical Neutrality", supra note 114.

(118) Cf. JONSEN & TOULMIN, supra note 36; see Thacher, supra note 60.

(119) See FEINBERG, HARM TO OTHERS, supra note 2; see also FEINBERG, OFFENSE TO OTHERS, supra note 2.

DAVID THACHER, Assistant Professor of Public Policy and Urban Planning, University of Michigan. I would like to thank Ryan Allen, George Kelling, and Martin Rein for helpful comments on an earlier draft.
COPYRIGHT 2004 Northwestern University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Thacher, David
Publication:Journal of Criminal Law and Criminology
Date:Jan 1, 2004
Previous Article:Capital punishment, execution publicity and murder in Houston, Texas.
Next Article:U.S. domestic prosecution of the American international sex tourist: efforts to protect children from sexual exploitation.

Related Articles
Unions may sue over rail work; HUMANS HELPLESS IN THE PATH OF NATURE'S FURY.
Diabetes after kidney transplant predicts graft failure and mortality.
Maximizing renewals begins with a 'back to basics' focus.
Work zone: reducing mishaps by 50%.
Mother told to pay up or risk losing her home.
Solihull News: Chain reaction on loo closures; MELL SQUARE: Council has re-think after storm of protest.
Mental causation; a nonreductive approach.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters