A familiar picture of the criminal justice system portrays harried and cynical prosecutors presenting sheaves of cases to lazy, soft-headed judges. Everyone involved has virtually given up trying to contain the rising rate of crime. Defendants are shown either as unregenerate hoodlums who mock the feeble system or as bewildered innocents herded into pleading guilty through fear of vindictive officials and crammed and violent jails. Victims and witnesses are numbed by endless, unexplained delays until they give up coming to court. And all the time crime gets worse, perhaps picking up impetus from the system’s incapacity to deal swiftly and justly with criminals.

Two new books show that these popular judgments are often considerably at odds with the results of scholarly inquiry. Hans Zeisel is a University of Chicago social scientist who during the 1960s was coauthor of a celebrated study of the American jury and whose work is a model of the lucid presentation of large themes.1 His new book is based on a research project which he directed some years ago for the Vera Institute of Justice in New York. That study collected data on the processing of approximately two thousand felony cases in New York City, in order to discover how many such cases result in conviction and punishment and why many are dropped on the way. At first impression the figures presented in Zeisel’s graph on this page seem to ratify popular beliefs.



As the graph shows, out of every 100 felony arrests forty-four are dismissed at an early stage and only twenty-seven offenders receive any custodial sentence; twenty-two of them go to jail for less than a year. Surely this reveals a system on the verge of collapse. Not at all, Zeisel says.

He makes two important points. First, the statistics in New York City today are hardly any different from those for felony prosecutions in American cities forty and fifty years ago. Second, they are about the same as those in Germany and Austria, countries usually thought of as having very effective court systems. The huge loss and dilution of felony prosecutions seem to be characteristic of many jurisdictions in modern times. They have several causes, none of which necessarily reflects discredit on the system.

The most important explanation for the very large percentage of dismissals is the backing off of the complaining witness, who often has a close relationship with the person arrested—for example, a wife who has been injured by her husband. In other cases the evidence turns out to be very weak or the prosecutor, in studying the details of the incident, decides that there are good reasons not to proceed. Many cases involve first offenders, or alleged crimes of no great seriousness or with some mitigating circumstances, so that dismissal or conviction for a misdemeanor is justifiable. The relatively few prison sentences that are imposed seem for the most part to be rationally connected with the gravity of the offense and the record of the offender. American sentences are certainly not light when compared with European ones.

Measured in this way, then, the system is not demonstrably less efficient than it used to be and seems about as efficient as some in Europe. We have to understand, as Zeisel puts it, that “like the steam engine in the course of transforming heat into motion, civilized law enforcement must allow a large part of its energy to go to waste.” The conclusion he asks us to accept is that the criminal justice system can do very little about reducing crime. This can best be achieved through measures to prevent crime and to protect people from it. Good street lights will avert more robberies than tinkering with the criminal justice system; against burglaries we should invest in locks; against larcenies we should carefully identify the objects we own and automatically deposit social security and welfare checks in the banks; safer cars and better highways promise to prevent accidents and injuries much more than stricter laws on drunken driving.

But at the end Zeisel strikes a quite different note. To the cool realism of his advice on crime prevention he adds a plea to return to the concerns of an earlier time about the causes of crime and the rehabilitation of offenders. There is, he argues, an unwavering connection between American crime and the fact that the young male members of particular racial or ethnic groups are socially impoverished. These groups may change their identity with the generations; what remains constant is that the young men who commit crimes come from groups that are distinctly deprived. If we do not expend resources in ways that may divert these young people from crime then, Zeisel suggests, the crimes that they commit must be taken to be part of the price that we are in some sense willing to pay in order to avoid the costs of improving their situation.


The confidence and clarity of Zeisel’s approach have great appeal, but the temptation to embrace it unreservedly must be resisted. The figures may not be as conclusive as his treatment suggests. To show that the courts dismiss cases at the same rate today as they did forty years ago hardly tells us everything about whether the system might not function better. We need to know whether the felony arrests that ended in dismissals or minor dispositions forty years ago included the same proportion of serious offenses of violence as those dismissed or pleaded down today, and this Zeisel does not tell us. Equal performance with Germany and Austria also means very little. Comparisons with foreign systems are tricky matters, and one might point to the fact that (as best one can extract such a conclusion from the official statistics) the dismissal rate for serious crimes in England seems to be less than 10 percent. It is also the case, as Zeisel points out, that London, with about the same population as New York City, has one robbery for roughly every twenty-seven committed in New York. Does this mean that the London police and the English court system are “better” than those in New York? Hardly, without much more inquiry. But it would be rash to assume that some element of superior performance in the English system is not in any way relevant.

Zeisel’s rhetoric about the need for programs for deprived young people is silent about exactly what programs are likely to reduce crime. In fact, when more money was being spent on social programs during the 1960s, the crime rate was going up faster than it is now. The correlation of crime with poverty is full of complexity and we know very little about the kinds of social reforms that will stop people from committing crimes.

Rehabilitating offenders is an even more daunting problem. The very concept of rehabilitation is vague, embracing everything from general education through specific job training to various kinds of therapy designed to purge the criminal impulse. Though the goal of rehabilitation has recently become unfashionable it is not clear that either educational or therapeutic programs in prisons have been much diminished. Studies tend to show that such programs cannot be demonstrated to have any measurable effect in stopping recidivism. Certainly this may be because the programs have not been of the right kind, but we cannot be confident that increased efforts would have any effect on the crime rate.

The case for expending resources on improving opportunities for poor children ought to turn on general moral principles of fairness in social arrangements, and not on hopes about the impact on the crime rate. The claim for equality is too important to rest on a promise that will not bear its weight. Similarly the case for making vigorous efforts to rehabilitate criminals derives from the moral duty to give them every chance to acquire skills and insights that may lead to different choices; the case does not depend on demonstrating that such programs will limit recidivism. Zeisel falls here into the error of attempting to sweeten policies that should be morally obligating by making unprovable claims that they will yield popular dividends.

But if liberal criminologists have to contend with disturbing facts, so do conservative ones. Since the early 1970s the number of people in prison in the United States has doubled but the crime rate has continued to rise for most of that time. A dreadful consequence that does not seem to have attracted much attention is that we shall soon be turning out into the community twice as many prison graduates each year as we were a decade ago. Getting tough doesn’t seem to have helped very much either.

Perhaps what we must do is to take very seriously Zeisel’s first principle and stop evaluating the system according to the volume of crime. After all, we might do better at crime control if we had no criminal justice system whatsoever. Dragnets, Charles Bronson-type death squads, a “war on crime” à outrance might possibly (though not surely) lower the crime rate appreciably. But such a course would destroy the kind of society we profess to want to live in. The very name “criminal justice system” expresses a basic commitment to doing justice and, complicated though that notion is, the system must be judged by its fidelity to and fulfillment of our conception of that ideal. Criminal justice institutions make important statements—to offenders, to victims, to all citizens, and to the world—about our capacity to be serious and efficient in the service of justice. The low popular opinion of criminal courts in American cities, perhaps especially in New York City, is therefore very disquieting. Either waste, inefficiency, and idleness are assumed to abound or the courts are seen as starved of sufficient funds.


Malcolm Feeley’s book is a study commissioned by the Twentieth Century Fund that investigates the possibility of effective reform in the criminal justice system. Feeley evaluates four particular attempts at planned change in recent years—pretrial scrutiny of defendants to encourage bail, pretrial diversion of defendants into job or training programs, plans for speedier trials, and changes in sentencing laws. He concludes that none of these reforms has had anything like the results envisaged; yet at the same time neither are the criminal courts as inefficient and chaotic as they are popularly thought to be.

For example, critics in the press seem to believe that accepting a great many guilty pleas in order to move business along is a new corrupting phenomenon responsible for a scandalous clemency toward criminals. But Feeley argues that the percentage of cases ending with guilty pleas has not changed much since the end of the nineteenth century. Plea bargaining has become more explicit and organized, but this may be because the defendant is now routinely represented by a lawyer, whereas in guilty pleas and even in many trials thirty years ago the defendant had no counsel. The greatly increased recognition of defendants’ rights in the last twenty-five years has done much to civilize our criminal courts, but it has also led both to conservative complaints and to a heightening of liberal awareness and expectations so that, in Feeley’s words, “as things get better they appear to get worse.”

The popular view of court systems as bureaucratic tortoises is also mistaken, in Feeley’s view. Following Max Weber, social scientists generally conceive of bureaucracy as characterized by rationality, impersonality, and a centralization of authority. Courts resemble this model only marginally. Some court officials may be part of a bureaucratic organization but the main participants in criminal cases come from separate and often clashing interest groups—police, prosecution attorneys, private defense attorneys, public defenders, and judges. Thrown together in the criminal court system they look very little like the unified organization that is the model of a bureaucracy. Each group has separate interests and somewhat different conceptions of the strongest relevant principles.

Two results flow from these characteristics of American criminal courts. One is that business would hardly move at all unless accommodations were made. For the most part this is accomplished by “work groups” (judge, prosecutor, and public defender) who collaborate to clear most of the cases at an early stage. This process is commonly thought of as plea bargaining but in fact it is so perfunctory that it might be better to reserve that term for the process of negotiation that goes on in a small number of very serious cases. By contrast, the “work group” technique is a quick collaborative classification of a case according to the “tariff” that prevails in that court in that season. The agreed tariff for a first offender’s theft may be a guilty plea to a misdemeanor, saving the state the expense of a trial for a felony. Such classification may take two minutes while true plea bargaining may consume weeks. But it does dispatch cases out of the system with great speed and in that sense is very efficient.

The second consequence of the non-bureaucratic quality of criminal courts is that change is particularly difficult to introduce. Bureaucracies are traditionally slow to accept change but they are at least cohesive. As a result, sufficient pressure from their upper and middle echelons may eventually move the whole organization. But criminal courts are no more than places where largely autonomous groups briefly converge, so that attempts to change the procedures will work well only with the active cooperation of a number of these groups.

The federal government’s attempts in recent years to secure speedy trials (a right guaranteed by the Sixth Amendment) is a good example. A federal statute and local schemes have prodded and nagged the courts but with small success. This is because strong elements in the court system are either indifferent or hostile to speedy trials. Judges may agree in principle that speed is a desirable aim but they lose enthusiasm when it disrupts their in many cases already heavy schedules and forces them to put pressure on sullen members of the bar. Prosecutors may like the idea but defense lawyers are usually very hostile since speedy trial rules lock them into fixed dates and make it impossible for them to take on other cases. Defendants who are out on bail always hope that if a trial is delayed long enough the prosecution might forget about them or the world might somehow change, while defendants in prison are usually thinking about a plea and not about a trial at all.

For Feeley the lesson to be drawn is that in so complicated a system change must be introduced delicately and evaluated very carefully. Attempts at reform are usually marred by grandiose thinking, insufficiently careful research and planning, reliance on observations made at an early stage, and failure to reassess the reforms years later. What is needed is close knowledge of the detailed workings of the courts and a readiness to rely on slow and incremental changes of small procedures rather than to conceive of change as sweeping and new.

While Feeley’s book contains much valuable debunking and a healthy set of warnings about inflated expectations, there are also dangers in his “small is beautiful” approach to criminal justice reform. The “voice of moderation” risks dropping to a soothing murmur. The message that some readers will extract from Feeley’s book is that things aren’t nearly as bad as we thought and it’s very hard to do anything about the system anyway, so we had better go very slowly and carefully. But one could take a different view that draws some support from parts of Feeley’s own analysis.

Even if Feeley persuades us that the criminal justice system is not on the verge of collapse, still it can be viewed as deeply flawed in many ways. It is enormously discouraging to those who might wish to contest their guilt. It relies on mass processing of cases with outcomes that often conceal the nature of the criminal conduct for which the defendant was arrested. It makes use of dirty, disorderly courtrooms where the proceedings are incomprehensible to the public and often to defendants. The pace of appeals is so paralyzed that they often take longer than the sentence imposed. If all this is so, then the resistance to change is all the more to be deplored, and if that resistance is rooted in deep features of the system, then we must not shrink from exploring radical solutions.

An intriguing hypothesis, supported by Feeley’s analysis, is that the extreme dominance of lawyers in the American system is a prime obstacle to change. The so-called adversarial process, by which the legal outcome depends largely on a contest between opposing advocates—with a passive judge acting as umpire—puts the lawyers in a position of control that they are extremely reluctant to give up. If an independent profession is by its nature resistant to change then we may need to break the grip of that profession on the system.

This could be extremely difficult in a society where lawyers not only preside over the legal system but are powerful throughout political life as well. To prepare for such a reform we need to understand how the American version of the English adversarial system grew to dimensions so enormous as to dwarf its progenitor. Recent scholarship is giving us a clearer picture of how lawyers started to take over the English criminal trial in the late eighteenth century and how the English model then began to diverge sharply from the continental European type of criminal process, which is controlled not by adversary lawyers but by officials who are charged with determining the facts and applying the law.2 We need comparable work to understand how American lawyers came to exercise a role so dominant that it nowadays surprises their relatively muted English counterparts. An understanding of this history might make reappraising the system easier, and might render proposals for different institutional arrangements more persuasive.

Today the English criminal justice system arouses little of the disquiet so familiar in American cities. This is too crude a test to certify it as a better system, but there are aspects of it that would repay detailed study. The English decentralization of criminal justice in the lower courts, with seventeen magistrates’ courts in Inner London alone, offers an interesting model of criminal justice as a cottage industry. Why does London have physically well-kept criminal courts while New York does not? This obviously has to do with large cultural elements, but we might discover that it also has something to do with court management systems and court procedures. Then the English mode of disposing of the great mass of fairly serious criminal cases in the lower courts through a combination of the defendant’s choice and the magistrate’s assent needs to be carefully compared with American plea-bargaining procedures. Under the English procedure the original charge is not reduced; the defendant is not obliged to plead guilty to secure speedy disposition of his case with an attractively low maximum sentence.

The criminal justice system is complex, but merely to point to its complexity and to emphasize the difficulty of change is too simple a response. Historical and comparative research that reveals the possibility of quite different approaches is as important as modest internal reforms. These two valuable books clear away a good deal of underbrush but in different ways they over-simplify the tasks that ought to lie ahead.

This Issue

November 10, 1983