Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 20 October 2021

Crime and Public Policy

Abstract and Keywords

This introduction to the book starts by defining crime. It explains that cross-national agreement about what should and should not be considered as criminal is narrower than is sometimes recognized. Examples are given of various acts that are regarded as criminal in some countries but not others. Crime and its control raises many instrumental, expressive, emotional, and moral issues, the text states. The layout of the book is outlined and some noncontroversial conclusions are offered. Next this introductory text looks at broad public policy approaches for dealing with crime: criminal law enforcement, prevention, harm reduction, regulation, decriminalization, and nonintervention. Next rates and trends are detailed and the costs associated are analysed. Finally this introductory section of the book looks at choosing the right policies.

Keywords: crime, crime control, public policy approaches, criminal law enforcement, harm reduction, prevention, regulation, decriminalization, nonintervention, rates and trends

Crime, like death and taxes, will always be with us. Stupidity, cupidity psychopathy, greed, intoxication, strong emotion, and impulsiveness are part of the human condition, much as we might wish it were otherwise, and all societies have to deal with their effects.

Not all harmful behaviors are criminal. What is harmful and what is not varies with time and space. Some behaviors, such as killings, rapes, robberies, serious assaults, and taking other people's property, are criminal in almost all places. Some behaviors, such as witchcraft, heresy, adultery, homosexuality, and gambling, are not now crimes in most Western countries, but in earlier times they were. Other behaviors, such as child abuse, domestic violence, drug use, prostitution, and environmental despoliation, were until recently generally not considered the law's business, though now they are. And still other behaviors, including cybercrime, identity theft, human smuggling, hate crimes, and money laundering, became significant crimes only when they recently became technologically possible or were legally proscribed.

Cross-national agreement about what should and should not be considered criminal is narrower than is sometimes recognized. It is conventional to contrast developed Western countries with others, but that is an oversimplification. In many Muslim countries heresy, apostasy, homosexuality, adultery, and failure to observe important cultural conventions remain crimes. In Turkey acts or words showing disrespect for the nation can be criminally prosecuted.1 Conversely, some behaviors that are criminal in the United States, such as paying fees to middlemen as conditions of winning commercial or government contracts (U.S. law like that (p. 4) of most developed countries considers such payments bribes), are commonplace and sometimes legal in much of the developing world.

However, it is a mistake to assume that even Western countries' notions of what counts as crime are congruent. In much of continental Europe Holocaust denial is a crime, and people are occasionally prosecuted and imprisoned for it; the First Amendment of the U.S. Constitution forbids such laws. In England, Scotland, and some other European countries private possession of handguns and public possession of knives with blades longer than seven inches are crimes; some Americans believe that the Second Amendment to the U.S. Constitution makes criminalization of handgun possession unconstitutional. English law on “antisocial behaviour” allows criminalization of just about any behavior a plaintiff can convince a court is antisocial, and prison sentences up to five years can be imposed for violation of an order to desist. Extreme cases include forbidding an individual “to enter the City of Manchester,” forbidding a prostitute to possess condoms, and forbidding an autistic child to stare at people. The law allows for imposition of prison sentences for engaging in prostitution (as a form of antisocial behavior), even though English criminal law long ago decriminalized prostitution, and for begging (as antisocial behavior), even though English criminal law does not authorize prison sentences for begging (Burney 2005).

Prostitution has been legalized in a number of European countries, and drug use and minor trafficking have been decriminalized. Any adult tourist (or other adult) in the Netherlands can legally buy marijuana in coffee shops, and addicts can obtain maintenance doses of other substances from approved medical sources. Italy and Spain likewise tolerate levels and forms of drug distribution that U.S. policy makers deplore. Sweden, by contrast, is as intolerant of drug use as is the United States, but is much less punitive about it. In much of Europe unlawful entry into the country is not a criminal act, though in the United States it is.

Some of the differences described in the preceding few paragraphs are the products of fundamental national differences in judgments about the importance of particular cultural values. Émile Durkheim ([1893] 1933) long ago explained that criminal law is an outgrowth of a society's most fundamental social norms, what he called the conscience collective, and should embody, express, and reinforce them. Dutch tolerance of drugs and prostitution, Austrian intolerance of Holocaust denial, and some Muslim countries' unwillingness to tolerate nontraditional styles of dress are outgrowths of deeply bedded cultural norms, however much outsiders may disagree with them. As cultural norms change, which over time they do, laws change with them, but the transitional periods are typically rocky.

Other differences described in the preceding paragraphs, however, do not so much reflect cultural differences about the desirability of behavior as they do divergent opinions about the most effective ways to deal with it. Dutch drug policies, for example, are predicated on the ideas that less aggregate social harm will result from tolerant than from repressive drug policies, and that treating addicts as unfortunates needing state support will convert them in the eyes of young people (p. 5) from romantic outlaws into pathetic social failures, object lessons of the dismal effects of a hedonistic lifestyle (Leuw 1991; Buruma 2007).

Dutch tolerance of drug use and low-level trafficking is as much a public policy about crime as is American intolerance and prohibitionism. Which is better cannot be answered in any absolute way. Crime and its control raise a motley of instrumental, expressive, emotional, and moral issues about which reasonable people differ. A purely rational approach to drug policy in the United States based on cost-benefit analyses would involve decriminalization of use, regulation of distribution, greatly reduced use of criminal sanctions, and use of greatly reduced criminal sanctions. Such approaches have been beyond reach in recent decades because for some U.S. policy makers and some citizens, drug use is morally wrong and tough drug laws are an expression of that belief. Sometimes prohibition laws are as much defenses of the values of dominant social groups as they are efforts to avoid harms associated with drug or alcohol use.2 Whether drug policies are effective or just may be less important to their proponents than that such policies make an unambiguous moral statement about the wrongfulness of drug abuse and abusers.

With so many different kinds of crime, so many ways of thinking about them, and so many ways to deal with them, policy making is contested and complicated. This short article provides a bit of background. Section I sets out a typology of policy approaches (criminal law enforcement, prevention, regulation, decriminalization, harm reduction, nonintervention) that operate sometimes as alternatives, sometimes as complements. Section II discusses data on crime rates, patterns, trends, and costs.

Systematic knowledge and analysis concerning most types of crime are remarkably primitive compared with such subjects as the economy, public health, and education (and dentistry, as James K. Stewart, onetime director of the U.S. National Institute of Justice, liked to point out in the 1980s, contrasting his agency's $13 million budget with the National Institute of Dentistry's $165 million; Blumstein and Petersilia 1995). The amount of money devoted to research and evaluation of crime control policy, the size of the specialist research communities, and the quality of government data systems are meager compared with those relating to other policy subjects.

Nonetheless, a few relatively noncontroversial conclusions can be offered about crime and public policy:

  • Property crime (theft, burglary, motor vehicle theft) and homicide rates have declined in the United States and in most other developed countries since the early to mid-1990s, but no one has convincingly explained why. It is difficult to conclude that recent tough police practices and sentencing laws explain the American decline; no other country adopted comparable policies, and simultaneous declines occurred in many of them. Less serious violent crimes have almost certainly also been declining, but this is less clear because reduced tolerance of violence means that relatively more (p. 6) crimes are reported to police and to survey researchers than in earlier times. Seemingly higher rates of violence in some countries in recent years probably reflect higher rates of reporting and recording, sometimes more than offsetting, and thereby camouflaging, absolute declines.

  • Remarkably little is known about crime rates and trends for many crimes, including cybercrime, identity theft, environmental crime, organizational crime, organized crime, prostitution, human smuggling, child abuse, tax evasion, illegal gambling, and drug trafficking. Anecdotal and journalistic reports suggest that most of these behaviors are becoming more common, but we have no valid and reliable sources of data about them, so it is hard to know what to believe.

  • Remarkably little is known about the costs of crime and the cost-effectiveness of alternative crime control approaches. Some American and British academics have attempted to calculate the costs of common law crimes, but their approaches are so ideological and so dependent on calculation of intangible “pain and suffering” costs as not to be believable. For other types of crime, no one has seriously attempted to estimate costs (e.g., human smuggling, domestic violence, child abuse, terrorism, prostitution) or the estimates have a back-of-an-envelope quality and vary by orders of magnitude (e.g., money laundering, drug abuse, organizational crime, environmental crime).

  • Countries vary enormously in how they use their criminal justice systems. Some, such as the United States, England, and New Zealand, rely on deterrence and incapacitation to attempt to control crime. Others, notably the Nordic countries, are skeptical about these approaches and rely on punishment's moral educative effects coupled with strong social welfare approaches. Some countries have high imprisonment rates and others have low rates. Some countries send comparatively few people to prison but for long periods; others send comparatively many but for short periods. Some favor short sentences; others oppose short sentences as destructive. Some have experienced five- and sevenfold increases in imprisonment rates in recent decades; some have experienced slight or no increases. Finland and Japan experienced sharp decreases from 1950 to 1990, though Japanese rates have increased since then.

  • Every country uses a wide range of criminal justice, regulatory, preventive, and harm-reduction approaches to dealing with crime, and within countries there are typically major differences between crimes. Crime is a legal concept; strictly speaking, a behavior becomes a crime only after a legislature has defined it to be one. In individual cases, acts become crimes only tentatively when a prosecution is initiated and only conclusively when a conviction is entered. In some countries, and for some kinds of crime, policy makers have decided that preventive, regulatory, and harm-reduction approaches are more effective than criminal law enforcement.

(p. 7) I. Aims, Methods, and Functions

Six broad public policy approaches are available for dealing with crime: criminal law enforcement, prevention, harm reduction, regulation, decriminalization, and nonintervention. Sometimes these are alternatives, sometimes complements.

A. Criminal Law Enforcement

Criminal law enforcement is what the criminal justice system does. Police investigate crimes and arrest people. Prosecutors decide whether or not to initiate a prosecution and if so, for what. They negotiate guilty pleas, try cases, and make recommendations to judges about sentences. Judges preside over pretrial proceedings and trials, review and accept (or, rarely, reject) guilty pleas, and impose sentences. Corrections authorities of different sorts run prisons, jails, and probation and parole systems. Governors and the U.S. president have authority to issue pardons and to commute sentences. Legislatures enact and revise laws defining crimes and governing penalties. Sometimes these laws set ranges within which judges may make decisions (e.g., from probation through a prison sentence of ten years); sometimes, as with mandatory minimums and three-strikes-and-youʼre-out laws, judges ostensibly have very little discretion about the sentences they impose.

In describing roles of officials and agencies, the preceding description gives a false impression of U.S. criminal law enforcement because it implies a mechanistic system. To the contrary, the system is highly discretionary. Police may decide to make an arrest, or not. Prosecutors may charge suspects with serious crimes, less serious crimes, or no crime at all. They may divert obviously guilty suspects into treatment programs. They exercise discretion to make and withdraw charges and to make plea bargains that, constitutionally, is almost unreviewable. Judges sometimes have broad discretion to individualize sentences, and when they legally lack such authority they can cooperate with plea-bargaining lawyers to avoid application of mandatory minimum or three-strikes sentences they consider too severe. Judges do not always exercise discretion to avoid imposition of unjustly severe sentences, however; sometimes they impose them. The system is so riddled with discretion that wide disparities in how cases are handled are inevitable. This can be portrayed as good (officials can take into account ethically meaningful differences between cases) or bad (officials are not good at individualizing and often treat comparable cases starkly differently).

The criminal justice system is premised on legal threats. Criminalization broadcasts the message that certain behavior is forbidden, and sentencing laws threaten what may happen if citizens ignore the criminal law's adjurations. Conventionally it is said that criminal sanctions do their work through deterrence, incapacitation, rehabilitation, and moral education.

The evidence about deterrence is uncertain and contested. The fairest generalizations are that certainty and celerity of punishment are more important than (p. 8) severity and that the jury is out on the question of “marginal deterrence,” whether incremental increases in punishment have deterrent effects (Tonry 2008).3 Some reviews of the evidence conclude that the marginal deterrence hypothesis cannot be confirmed (e.g., von Hirsch et al. 1999; Doob and Webster 2003). Others conclude that a deterrent effect can be confirmed—sometimes, for some crimes, and under some circumstances—but that existing knowledge is so fragmentary that it has no practical relevance to policy (Cook 1980; Nagin 1998a).

There is no doubt that some sentences incapacitate some offenders, but there is doubt about the incapacitative efficiency of imprisonment (e.g., Nagin 1998b). Some prisoners will not reoffend in any case. Some, such as drug traffickers and gang members, will quickly be replaced; the crime will go on, the criminal will not. Some will age out of crime; “age-crime curves” instruct that crime, especially violence, is a young man's game and that most people age out of criminality at a young age (Farrington 1986). Even most career criminals desist by their thirties. A classic analysis of incapacitation concluded that its policy relevance is questionable (J. Cohen 1983).

The evidence about rehabilitation has been changing rapidly. In the 1970s the conventional view was that few programs could be demonstrated effectively to reduce later offending (e.g., Martinson 1974). Since the early 1990s the consensus view has been that well-managed, well-targeted drug treatment programs can reduce both drug use and offending among drug-using offenders (e.g., Anglin and Hser 1990); that finding is the premise for the modern drug court movement. Similar optimism exists, supported to some extent by research findings, that a wide range of other kinds of focused (e.g., intrafamily sexual abuse) and broad (cognitive-behavioral) treatment programs can reduce later offending (e.g., Gaes et al. 1999).

There is little evidence concerning the criminal law's moral educative effects. The philosopher H. L. A. Hart (1968, p. 6) believed in it; the answer to why acts are criminalized is “to announce to society that these acts are not to be done.” The Norwegian criminal law scholar Johannes Andenaes (1974), following Durkheim, believed moral education is the criminal law's most important and powerful effect. Contemporary Scandinavian punishment laws are premised on the view that the criminal law must reinforce basic social norms and, conversely, must take pains not to contravene them (Lappi-Seppälä 2001). It makes apriori sense to me, but weʼve no empirical evidence that it is true.

B. Prevention

There are three main forms of prevention: situational, developmental, and community. There is no doubt that situational prevention methods—physical measures ranging from bullet-proof glass, CCTV, improved lighting, exact change policies on buses and trams, and myriad other initiatives that make crimes harder to commit—can reduce crime rates, especially for kinds of crime that are often committed impulsively, and without complete displacement to other places (Clarke 2005, 2008).

(p. 9) Similarly, the evidence is overwhelming that a wide range of developmental prevention programs can reduce at-risk children's later likelihood of offending. These programs identify risk and protective factors in children's lives and attempt to weaken the former and strengthen the latter. Cost-benefit analyses show that public investment in developmental prevention is much more cost-effective (three to four times more) than investment in prison sentences (Aos, Miller, and Drake 2006; Farrington and Welsh 2007). Unfortunately, the investment pays off 10 to 15 years later, which is not the reference period that most politicians have in mind when proposing crime control legislation and voting for appropriations to pay for it.

Community prevention takes two main forms. The first is creation of community self-help crime prevention organizations, usually operating under some permutation of the name “neighborhood watch.” There is no credible evidence that such organizations, usually organized in low-crime neighborhoods, have any significant effects on crime (Bennett, Holloway, and Farrington 2006). The second is effectively situational crime prevention at community and architectural levels. Community-level initiatives include improving street lighting, altering traffic-flow patterns, and closing off streets; gated communities are an extreme form. Architectural initiatives include building “defensible space” with clear sight lines and plenty of surveillance opportunities that allow residents and others to see what is going on nearby. The evidence on the effectiveness of community prevention approaches is mixed, though some programs appear to be successful (Welsh and Hoshi 2006).

All of these prevention approaches focus on prevention of rather than reaction to crime. Situational and community prevention are premised on making crimes harder to carry out and on the notions that much crime is impulsive and that not all impulses are displaced to other targets. Although the financial choices are less acute than with developmental prevention, the logic is that investment in prevention will pay off more than investments in prosecutions and imprisonment (Welsh and Farrington 2005).

C. Harm Reduction

Jeremy Bentham ([1789] 1948), the inventor of utilitarian analyses of public policy, believed that the greatest good of the greatest number is the best justification of state policies and actions. In carrying out a utilitarian calculation (a “felicific calculus” he called it because felicity, or happiness, is the measure), everyone's interests must be taken into account. If crime prevention might be made more effective by punishing an individual, it should be done only if the suffering (unhappiness) averted for prospective victims was greater than the suffering to be imposed on individual offenders. The state should be parsimonious in the imposition of suffering, he reasoned. The “parsimony principle,” the notion that no suffering should be imposed in the name of crime prevention in excess of the minimum necessary to achieve valid public ends, remains influential (Morris 1974).

(p. 10) Harm-reduction strategies are Benthamism revived. Usually associated in the United States with public health professions, harm-reduction strategies aim to achieve policy goals at the lowest aggregate social cost. Dutch drug-control policies are a classic example. Dutch policy makers believe that their approach, combining effective decriminalization of marijuana, medical support and social marginalization of addicts, and law enforcement targeted on large-scale traffickers, manufacturers, and importers, minimizes the social costs of drug use. Dutch and German approaches to prostitution are also based on the goal of harm reduction. Legalizing prostitution but requiring registration of prostitutes and regular medical examinations are seen as ways to protect prostitutes and their customers from health risks and criminal harms associated with unlawful prostitution (Buruma 2007).

American public health officials have long urged a harm-reduction approach to gun injuries and deaths. Current policies focus primarily on increased penalties for criminal use of guns and on preventing felons and mentally ill persons from acquiring guns. The evidence cited above on deterrent effects of incremental changes in penalties offers little reason to believe that mandatory minimum sentences for gun crimes, or mandatory add-ons for crimes in which guns are used, significantly reduce criminal gun use. English and Scottish criminalization of private possession of firearms in the aftermath of the Dunblane school massacre in Scotland in 1996 is based on a harm-reduction rationale. Research on the effects of mandatory arrest policies for misdemeanor domestic violence shows that arrests result in increased violence by minority and unemployed men; a harm-reduction approach would look for other preventive means (Sherman, Schmidt, and Rogan 1992).

The drugs, prostitution, guns, and domestic violence examples are likely to raise hackles in the United States (albeit more for emotional, ideological, and moral than for practical or evidence-based reasons). Many people find harm-reduction proposals much more congenial for business regulatory crimes (e.g., van de Bunt and Huisman 2007). They provide the primary rationale for business and environmental regulatory strategies that focus on negotiation, collaboration, and self-monitoring (e.g., Braithwaite 2002).

Finnish criminal justice policy making explicitly incorporates harm-reduction principles: “The aim of minimalization (not ‘elimination’) emphasizes the costs and the harmful effects of criminal behavior instead of simply minimizing the number of crimes. By stressing that not only the costs of criminality but also the costs and suffering caused by the control of crime must be taken into account, the formula draws attention to the material and immaterial losses that arise from operation of the system of sanctions” (Lappi-Seppälä 2007, p. 231).

D. Regulation

Regulatory approaches to crime prevention are most often discussed in relation to business crimes, though regulation is much more widely used than that. Approaches to dealing with drugs and prostitution in the Netherlands and some other European countries contain major regulatory elements.

(p. 11) So do American approaches to gun crime. Many states, counties, and cities have laws that limit who can buy guns and under what circumstances. Almost all jurisdictions require waiting periods, all require identity checks, and some require completion of gun safety courses. None of these is especially effective at controlling gun violence (Cook and Ludwig 2001).

Many countries, including the United States, have replaced laws criminalizing many forms of gambling with newer laws and administrative arrangements for regulating gambling and attempting to minimize social costs and harms associated with it. Fifty years ago many states criminalized most forms of gambling, and every city had some form of the “numbers racket.” No more. Legalized gambling in the United States began with on-track horse and dog racing. In the past 40 years, however, state lotteries have become nearly ubiquitous, off-track betting parlors have become common, and casinos have proliferated. Vigorous law enforcement coupled with intrusive regulatory regimes have driven organized crime from casino gambling.

Regulatory approaches are better known in connection with organizational crimes, sometimes referred to as white-collar crimes, especially relating to the environment. Administrative law specialists have argued for decades about the best ways to shape organizational compliance with regulatory regimes (e.g., Yeager 1993; Braithwaite 2002; Hawkins 2003). A law enforcement approach emphasizes monitoring of compliance with regulatory standards, backed up by use of civil penalties and criminal prosecutions to punish past and deter future noncompliance. A “responsive” regulatory approach reserves civil and criminal penalties as measures of last resort, preferring negotiated and cooperative approaches (Braithwaite 2002; van de Bunt and Huisman 2007). Proponents of responsive regulation argue that heavier-handed law enforcement approaches are slower and less effective, partly because they are confrontational and condemnatory, which puts people's backs up, and because they invite stonewalling and litigation, which delay identification and implementation of solutions.

Political ideology often predicts whether political leaders favor law enforcement or responsive regulatory approaches. Political conservatives want to avoid shackling business and discouraging entrepreneurship. They typically favor relatively light regulatory regimes and also favor responsive approaches to law-breaking. However, conservatives tend to oppose adoption of regulatory regimes for drugs and prostitution. Political liberals tend to favor tougher law enforcement approaches for dealing with business crime but are typically more receptive than conservatives to proposals for regulatory approaches for dealing with drugs and prostitution.

E. Decriminalization

Decriminalization eliminates crime by definition. Sometimes it happens de jure, as when American legislatures repealed laws criminalizing heresy, adultery, and prostitution, or when the U.S. Supreme Court declared laws making flag burning criminal unconstitutional (and as could happen to city ordinances restricting gun ownership now that the Court has decided that the Second Amendment created (p. 12) a constitutional right of private gun ownership). Sometimes it happens de facto, as when prosecutors and courts in many U.S. jurisdictions in the 1970s stopped enforcing laws against private possession, use, and sometimes sale of small amounts of marijuana. In many U.S. jurisdictions, adultery, fornication, and homosexuality crimes remained in the statute books long after they ceased being enforced.

If harm reduction in other than business regulatory contexts ever becomes culturally and politically imaginable in the United States, decriminalization coupled with regulation of a wide range of behaviors will become viable. These might include prostitution, drug possession and use, minor drug trafficking, and euthanasia. For the foreseeable future, American moralism and insistence on expressive legislation concerning what are considered immoral behaviors make much decriminalization unlikely.

F. Nonintervention

One policy option that is always available is to do nothing. Edwin Schur (1973) a third of a century ago argued on harm-reduction grounds that the best response to juvenile delinquency often is to do nothing at all. His proposal was based on two empirical realities. The first is that studies of self-reported delinquency show that majorities of male and female adolescents admit to committing criminal acts, but also show that most soon desist from criminality. Delinquency is a common characteristic of adolescence and is in effect an aspect of a developmental stage that is part of growing up. The second is that labeling theory argued and research showed that being processed through the criminal justice system is in various ways criminogenic (children become stigmatized, making future delinquency a self-fulfilling prophecy; first-timers are put into contact with more deviant peers and are socialized by peer influences into further delinquency).

In many cases juvenile justice system intervention thus promised to make children's life chances worse and to increase the likelihood of future delinquency. Putting the two realities together, Schur argued, meant, perversely, that juvenile justice processes make future delinquents of kids who would in the ordinary course otherwise desist from delinquency. Were Schur's book written today, a third reality would be added: black kids are more often brought into contact with the juvenile justice system than are white kids, and at earlier ages (Feld 1999). The perverse effects Schur described disproportionately affect black children.

What I called decriminalization de facto is a form of radical nonintervention. Nonintervention may be the wisest public policy in connection with criminalized behaviors that raise some people's moral hackles: pornography, adultery, fornication, obscenity, homosexuality, drug use, gambling. In many cities in many times police have known where gambling and prostitution took place; for example, during Prohibition, speakeasies were difficult to hide. Those are examples of nonintervention policies. The great danger they pose is that they expose people to blackmail and create circumstances foreseeably conducive to police corruption. They also produce (p. 13) arbitrariness in law enforcement. Formal decriminalization is the safer policy option, but when that is not politically feasible the de facto form is better than none.

II. Rates, Trends, and Costs

Reasonably good evidence is available concerning rates and trends of serious violent and property crimes targeting individual victims. For a few other crimes some observations can be offered with some confidence on the basis of triangulation of longitudinal data from multiple sources. Both the National Family Violence Surveys (Straus and Gelles 1986) and the National Crime Victimization Survey (U.S. Department of Justice 2006), for example, show that domestic violence in which women are victims has been declining but that domestic violence involving male victims is not. For most of the crimes discussed in this volume, however, it is difficult to say much with confidence about rates or trends.

The authors in this volume discuss data sources and estimates of rates and trends for particular crimes. For the most part estimates are based on information that has come to the attention of the authorities. Seldom do we know how representative those cases are. Many health and welfare professionals who deal with children, for example, are obliged to report cases of suspected child abuse to state authorities, which in turn report them to federal authorities. Because the categories of professionals required to report have gradually been extended, and because professional reluctance to report may be declining, national data show steadily increasing numbers of abused children. It is not known whether more children are being abused or more incidents of child abuse are being reported.

Knowledge concerning the costs of crime is sparse and not very useful. There is a bit more credible data concerning the cost-effectiveness of alternative law enforcement, regulatory, and prevention strategies.

A. Rates and Trends

A good bit is known from official police crime data and victimization survey findings about rates and trends of the traditional common law offenses (e.g., Tonry and Farrington 2005). The pattern in the United States, which is paralleled by experience in many developed Western countries, is that crime rates began climbing some time in the 1960s, peaked in the early 1990s, and fell rapidly thereafter. The table “Rates per 100,000 Population, Police Data, Selected Crimes and Years, 1970–2006” shows aggregate rates for the seven traditional FBI index offenses (murder, rape, robbery, aggravated assault, burglary, theft, and motor vehicle theft) per 100,000 U.S. population from 1970 to 2006 and rates separately for murder, rape, robbery, burglary, and motor vehicle theft. (p. 14)

Rates per 100,000 population, police data, selected crimes and years, 1970–2006












All index


































































vehicle theft

Note: Three- and four-digit numbers rounded.

Source: Sourcebook of Criminal Justice Statistics Online, 31st ed., table 3.106.2006.

Crime rates in every category were much lower in 2006 than their peak levels in the preceding 36 years. U.S. crime rates began rising in the 1960s. They peaked in 1980–81, fell for five years through 1986, rose again for five years through 1990–91, and have fallen continuously since. By 2006 rates for all the crimes shown except rape were 45 to 50 percent below their peaks in 1980 and 1990. With the exception again of rape, rates for all individual offenses were below 1970 levels. Property crime declines have been most pronounced, and, because burglary and motor vehicle theft victims must generally file police reports to collect on property insurance, those rates are probably believable. Homicide rates, because they usually require bodies and because violent deaths can be validated from public health sources, are also generally viewed as reasonably reliable. Rates for robbery, rape, and aggravated assault have fallen least. There has, however, been a steady decline in public tolerance of violence, leading to higher rates of reporting incidents to the police. Apparent declines for these crimes are probably greater than appears because they are offset by higher likelihoods that victims reported incidents to the police.

The table “Household Victimization Rates, in Percentages, Selected Crimes and Years, 1994–2005” shows American household victimization data from 1994 to 2005 for all offenses and separately for rape, robbery, assault, burglary, and motor vehicle theft.4 Overall, the percentage of American households experiencing the crimes measured by the National Crime Victimization Survey declined from 25 percent in 1994 to 13.9 percent in 2005. Rates for individual offenses fell by half or more.

The pattern of declining crime rates in the United States characterizes most developed countries. The table “International Criminal Victimization Rates, ICVS Data, Various Years 1989–2004” shows results from the International Crime Victimization Survey (ICVS), which has been conducted five times in most developed Western countries, beginning in 1989. A standardized questionnaire is used in all participating countries. The table shows percentages of respondents reporting victimizations by particular crimes. Overall, and separately for auto theft, burglary, robbery, and sexual offenses against women (a broader category (p. 15) than “rape” alone), rates peaked in 1995 and have fallen continuously since. The authors of the most recent ICVS report observe, “According to ICVS data, the level of common crime in Europe reached a plateau around 1995 and has shown a steady decline over the past ten years. The level of victimization in Europe has now decreased to the level of 1990” (van Dijk, van Kesteren, and Smit 2007, p. 46).

Household victimization rates, in percentages, selected crimes and years, 1994–2005













Any crime








































































vehicle theft

Source: Sourcebook of Criminal Justice Statistics Online, 31st ed., table 3.27.2005.

International criminal victimization rates, ICVS data, various years 1989–2004

Crime category






Auto theft


















Sexual offenses against







Overall rate, nine






crimes measured all years

Source: van Dijk, van Kesteren, and Smit 2007, table A1.

Thus it appears, strongly and credibly, that crime rates for the most common violent and property crimes affecting individuals have been falling for a considerable time in most developed countries. We do not know why.

B. Costs and Cost-Effectiveness

There is little reliable evidence on the costs of crime. Systematic efforts to calculate the costs of common crimes are plagued by errors and conceptual and ideological blind spots. Efforts to calculate the costs of other crimes are plagued by the unavailability of reliable data and a need to make heroic assumptions.

(p. 16) Evidence on cost-effectiveness of alternative policy choices is in better shape, especially in relation to choices between investment in law enforcement and in developmental prevention or rehabilitative programs. That many developmental and treatment programs appear to be sounder public investments than law enforcement approaches is especially striking because the standard ways of making such calculations use data on the cost of crime that greatly exaggerate costs, thereby making benefits harder to cost-justify.

Studies of the costs of crime exaggerate costs of violent crimes enormously. That is because the most commonly used approach, based on work by the economist Mark Cohen (1988; Miller, Cohen, and Wiersema 1996; M. Cohen and Miller 2003), includes victims' pain and suffering as a cost of crime. Cohen developed his estimates from court files on contested cases in which crime victims filed suit for damages and won. This is inappropriate for two reasons.

First, the estimates are exaggerated. They are applied to all crimes falling within generic categories, such as murders, rapes, robberies, and assaults, even though contested civil damages cases are far from being representative of these categories. They are extreme cases, what statisticians call the right-hand tail of the distribution. They are the comparatively rare cases in which the defendant was solvent or insured, in which the prospect of winning was sufficiently high to attract plaintiffs' lawyers willing to work on a contingent fee basis, and in which the crime was so awful or the defendant so unattractive that juries were prepared to award large damages. Pain and suffering costs make up by far the largest share of costs in Cohen's calculations, especially for violent and sexual crimes. In one of Cohen's most widely cited reports, for example, pain and suffering were estimated to constitute $81,400 of the $87,000 total costs of a rape or sexual assault, and $1,910,000 of the $1,940,000 cost of a homicide (net of lost earnings; Miller, Cohen, and Wiersema 1996).

Cohen and others (e.g., Cook and Ludwig 2001) have used additional methods to estimate the costs of crimes. One, taken from environmental economics, is to ask people how much more they would be willing to spend on annual insurance premiums to reduce particular risks by 10 percent. Using the finding that people would be willing to pay an extra $100 to $150 per year to reduce risks of crime, M. Cohen and colleagues (2004) estimate the cost of a rape at $237,000 and the cost of a murder at $9,700,000.

The second problem is that Cohen's estimates are often used by him and others in cost-benefit studies of criminal justice system policies. The difficulty is that cost-benefit analyses take heavy account of victims' “intangible costs” and “lost earnings” but no account of offenders' intangible costs (and little account of their tangible costs such as lost earnings and reduced postprison earnings prospects, and no account of costs incurred by offenders' dependents). In explaining his felicific calculus Bentham made it clear that everyone's happiness and unhappiness should count in deciding whether a practice or policy is justifiable.

(p. 17) I have many times asked economist friends, including Mark Cohen, why offenders' costs are not taken into account, but I have never received a clear answer. Sometimes someone mumbles that offenders' interests are irrelevant because, after all, they are offenders. Omission of the costs of punishment to offenders and their families, including “intangible costs,” from cost-benefit calculations of the effectiveness of criminal justice policies, while counting such costs attributed to victims, is the equivalent of putting not a hand but a foot on one side of the scale, and foreordains the apparent cost-effectiveness of many punitive policies.

Cost-effectiveness studies do considerably better. Some of them take offenders' costs into account to some degree. Credible analyses for nearly twenty years have repeatedly shown that many developmental prevention methods and some treatment programs are better public investments than increased use of imprisonment (Welsh, Farrington, and Sherman 2001; Aos, Miller, and Drake 2006).

Many of the contributors to this book discuss cost-of-crime, cost-effectiveness, and cost-benefit analyses and generally suggest that they should be regarded with skepticism. Reuter (1984) long ago showed the tendency of efforts to estimate the costs of crime (in his case, drug abuse) to produce preposterously high numbers, as his title illustrates: “The (Continuing) Vitality of Mythical Numbers.”

III. Choosing Policies

Evidenced-based policy for many subjects is in vogue. It implies a rationalistic approach to policy making that takes account of evidence concerning the effectiveness, including cost-effectiveness, of alternative policy options, and takes account of intended and unintended foreseeable effects of policy choices. Efforts to prevent and control crime and minimize its harms and costs are, alas, not fully rational. They implicate economic, political, social, ideological, and moral issues that shape the limits of politically possible policies.

“Wars” on drugs and crime, for example, reflect moral entrepreneurship and political calculation at least as much as dispassionate assessment of how to prevent unwanted behavior and minimize related social costs. Recent policy attention to domestic violence has been motivated both to prevent and respond to harmful behavior and to elevate its visibility in the public mind and in political debate, and through doing so to strengthen the social, economic, and political positions of women.

People concerned single-mindedly about environmental damage or financial chicanery are deeply frustrated when they realize that others are equally or more concerned about the economic and social effects (for business) of vigorous law enforcement. Economic conservatives often see a bit of illegality and corruption (p. 18) as an inevitable and even necessary, if regrettable, effect of unleashing a capitalist economy. Tighter regulation of organizations might yield less unwanted behavior, but it might also yield lower rates of economic growth. Social liberals often see street crime and drug abuse as inevitable but tragic effects of child poverty, disadvantaged lives, and a weak U.S. welfare state, and prefer to try to prevent crime by improving criminogenic social conditions rather than primarily through deterrence, incapacitation, and rehabilitation.

People are not angels, and this side of the Pearly Gates they are never likely to be. People make mistakes, do stupid and cruel things, and commit sins (if you find that concept uncongenial, substitute: behave immorally or in ways that violate fundamental social norms of right conduct). Everyone's happiness and unhappi-ness should count. Laws and other public policies should be based on best-faith efforts to achieve the greatest public benefit at the least human cost. Finnish policy makers have a distinctive way to say this: crime policy should take all the costs of crime into account, including the costs of law enforcement (by which they mean not only the economic costs of crime and the criminal justice system, but also the economic and intangible costs offenders suffer because of what is done to them; Lappi-Seppälä 2007, p. 231). This is basically a restatement of Bentham's parsimony, the view that no suffering should be imposed on offenders that is not justified by greater suffering from which others are saved. The implication is that policy makers should try as best they can to restrain the influences of emotion and expressive considerations on public policy about crime. Evidence-based policies for dealing with crime, coupled with a sense of human frailty and the transience of life, may be a romantic aspiration, but it is the right aspiration.


Andenaes, Johannes. 1974. Punishment and Deterrence. Ann Arbor: University of Michigan Press.Find this resource:

Anglin, M. Douglas, and Yih-Ing Hser. 1990. “Treatment of Drug Abuse.” In Drugs and Crime, edited by Michael Tonry and James Q. Wilson. Vol. 13 of Crime and Justice: A Review of Research, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Aos, Steve, Marna Miller, and Elizabeth Drake. 2006. Evidence-Based Public Policy Options to Reduce Future Prison Construction, Criminal Justice Costs, and Crime Rates. Olympia: Washington State Institute on Public Policy.Find this resource:

Bennett, Trevor, Katy Holloway, and David Farrington. 2006. “Does Neighborhood Watch Reduce Crime? A Systematic Review and Meta-Analysis.” Journal of Experimental Criminology 2(4): 437–58.Find this resource:

Bentham, Jeremy. 1948. Introduction to the Principles of Morals and Legislation. Edited by Wilfred Harrison. Oxford: Oxford University Press. (Originally published 1789.)Find this resource:

Blumstein, Alfred, and Joan Petersilia. 1995. “Investing in Criminal Justice Research.” In Crime, edited by James Q. Wilson and Joan Petersilia. San Francisco: ICS.Find this resource:

Braithwaite, John. 2002. Restorative Justice and Responsive Regulation. New York: Oxford University Press.Find this resource:

Burney, Elizabeth. 2005. Making People Behave—Antisocial Behaviour, Politics, and Policy. Cullompton, UK: Willan.Find this resource:

Buruma, Ybo. 2007. “Dutch Tolerance: On Drugs, Prostitution, and Euthanasia.” In Crime and Justice in the Netherlands, edited by Michael Tonry and Catrien Bijleveld. Vol. 35 of Crime and Justice: A Review of Research, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Clarke, R. V. 2005. “Seven Misconceptions of Situational Crime Prevention.” In Handbook of Crime Prevention and Community Safety, edited by Nick Tilley. Cullompton, UK: Willan.Find this resource:

———. 2008. “Situational Crime Prevention.” In Environmental Criminology and Crime Analysis, edited by R. Wortley and L. Mazerolle. Cullompton, UK: Willan.Find this resource:

Cohen, Jacqueline. 1983. “Incapacitation as a Strategy for Crime Control: Possibilities and Pitfalls.” In Crime and Justice: A Review of Research, vol. 5, edited by Michael Tonry and Norval Morris. Chicago: University of Chicago Press.Find this resource:

Cohen, Mark. 1988. “Pain, Suffering, and Jury Awards: A Study of the Cost of Crime to Victims.” Law and Society Review 22: 537–55.Find this resource:

Cohen, Mark, and Ted R. Miller. 2003. “‘Willingness to Award’ Nonmonetary Damages and the Implied Value of Life from Jury Awards.” International Review of Law and Economics 23: 165–81.Find this resource:

Cohen, Mark, Roland T. Rust, Sara Steen, and Simon T. Tidd. 2004. “Willingness-to-Pay for Crime Control Programs.” Criminology 42: 89–109.Find this resource:

(p. 20) Cook, Philip J. 1980. “Research in Criminal Deterrence: Laying the Groundwork for the Second Decade.” In Crime and Justice: A Review of Research, vol. 2, edited by Norval Morris and Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Cook, Philip J., and Jens Ludwig. 2001. Gun Violence: The Real Costs. New York: Oxford University Press.Find this resource:

Donohue, John J., and Justin Wolfers. 2005. “Uses and Abuses of Empirical Evidence in the Death Penalty Debate.” Stanford Law Review 58: 791–846.Find this resource:

Doob, Anthony N., and Cheryl Marie Webster. 2003. “Sentence Severity and Crime: Accepting the Null Hypothesis.” In Crime and Justice: A Review of Research, vol. 30, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Durkheim, Émile. 1933. The Division of Labour in Society. Translated by George Simpson. New York: Free Press. (Originally published 1893.)Find this resource:

Farrington, David P. 1986. “Age and Crime.” In Crime and Justice: A Review of Research, vol. 7, edited by Michael Tonry and Norval Morris. Chicago: University of Chicago Press.Find this resource:

Farrington, David P., and Brandon Welsh. 2007. Saving Children from a Life of Crime. New York: Oxford University Press.Find this resource:

Feld, Barry. 1999. Bad Kids. New York: Oxford University Press.Find this resource:

Gaes, Gerald G., Timothy J. Flanagan, Lawrence L. Motiuk, and Lynn Stewart. 1999. “Adult Correctional Treatment.” In Prisons, edited by Michael Tonry and Joan Petersilia. Vol. 26 of Crime and Justice: A Review of Research, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Gusfield, Joseph R. 1963. Symbolic Crusade: Status Politics and the American Temperance Movement. Urbana: University of Illinois Press.Find this resource:

Hart, H. L. A. 1968. Punishment and Responsibility. Oxford: Oxford University Press.Find this resource:

Hawkins, Keith. 2003. Law as Last Resort: Prosecution Decision-Making in a Regulating Agency. Oxford: Oxford University Press.Find this resource:

Lappi-Seppälä, Tapio. 2001. “Sentencing and Punishment in Finland: The Decline of the Repressive Ideal.” In Sentencing and Sanctions in Western Countries, edited by Michael Tonry and Richard S. Frase. Oxford: Oxford University Press.Find this resource:

———. 2007. “Penal Policy in Scandinavia.” In Crime, Punishment, and Politics in Comparative Perspective, edited by Michael Tonry. Vol. 36 of Crime and Justice: A Review of Research, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Leuw, Ed. 1991. “Drugs and Drug Policy in the Netherlands.” In Crime and Justice: A Review of Research, vol. 14, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Martinson, Robert. 1974. “What Works? Questions and Answers about Prison Reform.” Public Interest 35(2): 22–54.Find this resource:

Miller, Ted R., Mark K. Cohen, and Brian Wiersema. 1996. Victim Costs and Consequences: A New Look. Washington, DC: U.S. Department of Justice, National Institute of Justice.Find this resource:

Morris, Norval. 1974. The Future of Imprisonment. Chicago: University of Chicago Press.Find this resource:

Musto, David. 1987. The American Disease: Origins of Narcotic Control. Expanded ed. New York: Oxford University Press. (Originally published 1973.)Find this resource:

Nagin, Daniel S. 1998a. “Criminal Deterrence Research at the Outset of the Twenty-First Century.” In Crime and Justice: A Review of Research, vol. 23, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

———. 1998b. “Deterrence and Incapacitation.” In The Handbook of Crime and Punishment, edited by Michael Tonry. New York: Oxford University Press.Find this resource:

(p. 21) Reuter, Peter. 1984. “The (Continuing) Vitality of Mythical Numbers.” Public Interest 78 (Spring): 135–47.Find this resource:

Schur, Edwin. 1973. Radical Nonintervention—Rethinking the Delinquency Problem. Englewood Cliffs, NJ: Prentice-Hall.Find this resource:

Sherman, Lawrence W., Janelle D. Schmidt, and Dennis P. Rogan. 1992. Policing Domestic Violence: Experiments and Dilemmas. New York: Free Press.Find this resource:

Sourcebook of Criminal Justice Statistics Online. 2007. 31st ed.

Straus, Murray A., and Richard J. Gelles. 1986. “Societal Change and Change in Family Violence from 1975 to 1985 as Revealed by Two National Surveys.” Journal of Marriage and the Family 48: 465–79.Find this resource:

Tonry, Michael. 2008. “Learning from the Limits of Deterrence Research.” In Crime and Justice: A Review of Research, vol. 37, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

Tonry, Michael, and David P. Farrington, eds. 2005. Crime and Punishment in Western Countries, 1980–1999. Vol. 33 of Crime and Justice: A Review of Research, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

U.S. Department of Justice, Bureau of Justice Statistics. 2006. National Crime Victimization Survey: Criminal Victimization in the United States—2005. Washington, DC: U.S. Department of Justice.Find this resource:

van de Bunt, Henk, and Wim Huisman. 2007. “Organizational Crime in the Netherlands.” In Crime and Justice in the Netherlands, edited by Michael Tonry and Catrien Bijleveld. Vol. 35 of Crime and Justice: A Review of Research, edited by Michael Tonry. Chicago: University of Chicago Press.Find this resource:

van Dijk, Jan, John van Kesteren, and Paul Smit. 2007. Criminal Victimisation in International Perspective: Key Findings from the 2004–2005 ICVS and EU ICS. The Hague: Boom Juridische uitgevers and Research and Documentation Center, Ministry of Justice, the Netherlands.Find this resource:

von Hirsch, Andrew, Anthony E. Bottoms, Elizabeth Burney, and Per-Olof H. Wikström. 1999. Criminal Deterrence and Sentence Severity: An Analysis of Recent Research. Oxford: Hart.Find this resource:

Welsh, Brandon C., and David P. Farrington, eds. 2005. Preventing Crime: What Works for Children, Offenders, Victims, and Places. New York: Springer.Find this resource:

Welsh, Brandon C., David P. Farrington, and Lawrence Sherman. 2001. Costs and Benefits of Preventing Crime. Boulder, CO: Westview.Find this resource:

Welsh, Brandon C., and Akemi Hoshi. 2006. “Communities and Crime Prevention.” In Evidence-Based Crime Prevention, rev. ed., edited by Lawrence W. Sherman, David P. Farrington, Brandon C. Welsh, and Doris Layton MacKenzie. New York: Routledge.Find this resource:

Yeager, Peter Cleary. 1993. “Industrial Water Pollution.” In Beyond the Law: Crime in Complex Organizations, edited by Michael Tonry and Albert J. Reiss Jr. Vol. 18 of Crime and Justice: A Review of Research, edited by Michael Tonry. Chicago: University of Chicago Press. (p. 22) Find this resource:


(1.) Americans should not feel especially virtuous and tolerant by comparison. Not long ago laws were passed making it a crime to burn the U.S. flag. A Supreme Court decision, Texas v. Johnson, 491 U.S. 397 (1989), declared such laws unconstitutional as infringements of the First Amendment's right of free speech.

(2.) For example, Gusfield's (1963) classic account of prohibition in the nineteenth century (established teetotaling Protestants from England anxious about incoming alcohol-drinking Catholic Irish and German settlers). Musto ([1973] 1987) offers parallel accounts of early twentieth-century criminalization of opiates (associated with Chinese immigrants), cocaine (associated with blacks), and marijuana (associated with Mexicans). This point is often made by contrasting policies concerning alcohol and nicotine, the recreational substances that generate the greatest social costs (in ill health, lost lives, monetary costs) but that are the recreational drugs of the majority population, with policies concerning marijuana and heroin.

(3.) There is, however, no credible evidence that capital punishment, which can be imposed in the United States only for murder, deters murders better than sentences that can otherwise be imposed (Donohue and Wolfers 2005; Tonry 2008).

(4.) Victimization rates have generally been declining for a much longer period, beginning not later than 1973, the National Crime Victimization Survey's first year. Major changes in the survey's design were effected in 1993, making the data series discontinuous.