Recent research on the functions of imprisonment has begun to provide quantitative, empirical knowledge of its rehabilitative and deterrent effects. Much less is known, however, about the incapacitative effect of imprisonment. While it has long been understood that the physical segregation of prison inmates prevents them from engaging in some criminal activity (as well as much non-criminal activity) during the period of their confinement, quantitative estimates of the size of this effect have been lacking. Leaving aside all deterrent or rehabilitative and counter-rehabilitative effects, it is of some interest to know whether the incapacitative function of imprisonment is large or small.

Crime and its prevention has become a much debated issue. Claims and counter claims are made as to the real extent and the possibility of controlling crime. Law makers are considering legislating heavier sentences, which raises questions about the role of prison sentences in crime prevention. Most discussions of crime and its prevention lack a way of generating quantitative estimates of what alternative policies would achieve and what they would cost. A recent paper by one of the authors (Avi-Itzhak and Shinnar, 1973) attempts to develop a mathematical model simple enough to make approximate quantitative estimates based on available statistical information. In this paper we try to simplify the model further and to present it in a way that will make it more accessible to the policy maker and social scientist. We also try to apply it to conditions in New York.

For Émile Durkheim, crime and punishment were integral features of organized social life. He considered the study of crime and punishment essential to the sociological enterprise precisely because these “social facts” revealed the inner workings of society and the mechanism through which societies change. In The Rules of Sociological Method Durkheim argues (1938:70) that “crime is … necessary, it is bound up with the fundamental conditions of all social life,” and that “it is no longer possible … to dispute the fact that law and morality vary from one social type to the next, nor that they change within the same type if the conditions of life are modified.” Because he viewed law as a reflection of basic social arrangements, Durkheim grounded his theory of social change in an analysis of comparative legal types. He assumed (1933:68) that “since law reproduces the principal forms of social solidarity, we have only to classify the different types of law to find therefrom the different types of social solidarity which correspond to it.” In his central theoretical work, The Division of Labor in Society (1933), Durkheim linked the changing nature of legal controls (repressive to restitutive) to transformations in the nature of social solidarity (mechanical to organic).

An impression widely held has it that American criminal courts are caught in a crisis. The manifestations of this presumably critical condition are well-known: ever-expanding workloads; indigent defendants held for weeks, months, and sometimes years of pre-trial confinement; officials and attorneys preoccupied with the mechanics of “plea bargains” instead of the intricacies of trials; protracted delays in settling some cases; disposition of other cases by means of arbitrary and prejudicial techniques; pervasive inequalities in sentencing decisions, and so on.

During the last five years, these symptoms have received attention from a number of respected journalists (James, 1971; Downie, 1972; Jackson, 1974). Criminal courts and their ailments have come under extensive study by several recent government commissions (President's Commission, 1967; Skolnick, 1969; National Advisory Commission, 1973). Social scientists have also shown increasing interest in these problems (Skolnick, 1967; Blumberg, 1970; Mileski, 1971; Levin, 1972; Mather, 1974).

In a recent article in these pages Professor W.L.F. Felstiner (1974) made a substantial contribution to the literature on dispute processing by analyzing the role of avoidance as a means of ending disputes, and by correlating the ubiquity of this technique with differing forms of societal organization. In brief, his insight—richly developed in the article—is that the tendency to walk away from contentious situations is greater in “technologically complex, rich societies” than in “technologically simple, poor societies.” This is because the organization of richer societies minimizes the importance and the intertwining (multiplexity) of family and other group ties and maximizes social, occupational and geographic mobility. In contrast, citizens of poorer societies are bound closely to the individuals in their communities and thus have a greater need and capacity for resolving—rather than walking away from—disputes with those individuals.

I have suggested (1974) that in societies such as the United States, where non-governmental institutionalized mediation and adjudication of interpersonal disputes are infrequent, some of the slack may be absorbed by avoidance. I implied that avoidance would generally have lower costs in the United States than in societies where such mediation and adjudication are widely available. Danzig and Lowy (1975: 676-682) argue that I have underestimated the costs of avoidance in the United States. They may be right. In any event, avoidance can be analyzed more precisely than it was in my original paper (see 1974: 70, 76, 79-80, 83-84). This response will try to provide such an analysis, including a more complete breakdown of avoidance costs. It will suggest why I appraise avoidance costs differently than do Danzig and Lowy. It will also re-evaluate some of the variance in avoidance costs between different societies. Finally, it will discuss avoidance and Danzig and Lowy's proposed neighborhood mediation.