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).

Subscriber: null; date: 13 August 2020

Crime and Criminals

Abstract and Keywords

This article deals with the dynamics of macro- and micro-level crime. It is followed by the description of crime trends and reviews research on the factors associated with its rise and decline. The discussion concerns six categories, namely, demography and economic conditions; policing and incarceration; drugs, guns, and gangs; community and environmental factors; lifestyle and culture; and crime reporting and recording. The goal of this review is to provide an insight into the literature on crime trends. This article also discusses what is known and what needs to be known about the reasons for crime going up and down, and highlights some of the issues faced by those who take up the challenge. It discusses the factors associated with trends in crime and the technical issues involved and addresses their impact, and cross-national significance.

Keywords: macro-level crime, micro-level crime, crime trends, technical issues, drug, economic conditions

This Chapter takes as its organizing theme what we know—and do not know—about macro-social trends in crime. Following a brief description of crime trends in comparative perspective, the Chapter reviews research on the factors associated with its rise and decline. The discussion is organized into six categories: demography and economic conditions; policing and incarceration; drugs, guns, and gangs; community and environmental factors; lifestyle and culture; and crime reporting and recording. The goal of this review is to provide an entry point into the literature on crime trends, point out what is known and what needs to be known about why crime goes up and down, and highlight some of the issues facing those who take up the challenge. Reflecting the English-language literature, much of the following (p. 38) discussion is North American in orientation. However, the factors associated with trends in crime and the technical issues involved in addressing their impact are all of broad, cross-national significance, and could easily find application elsewhere.

I. Crime Trends Illustrated

Figure 1 depicts national trends in homicide and robbery recorded by police agencies in Japan, France, the United States, and England and Wales. The data come from a variety of official and research reports. Depending on the availability of the data, the trend lines start between 1946 (Japan) and 1974 (France). In every nation robbery is much more frequent than homicide, so their rates are illustrated using separate scale axes. The data illustrate a variety of ways in which officially recorded violent crime rose and fell over this period. In Japan, homicide rose during the late 1940s and early 1950s. It then dropped by two-thirds, before leveling off in the 1990s. Robbery also dropped, but after 1997 the rate proceeded to double; theft was up during this period as well. However, despite the most recent increase, in 2006 robbery stood at only 25 of its post-war high. In the UK, both homicide and robbery peaked in 2003 after rising steadily for more than 40 years. By that year homicide had risen by 30 over its 1958 low, and recorded robbery was up by a factor of 55. In France, homicide peaked in 1984, then declined thereafter—with an uptick in the mid-1900s. By 2006 it was down by 40 from its peak. Like Japan, French homicide figures include incidents classified by police as attempted murder, but in recent decades actual killings (which are also depicted in Figure 1) have been separately reported in France, and their trend closely parallels the broader number. In the United States, crime has waxed and waned fairly dramatically since the late 1950s, with police recording three new highs for homicide and robbery. What is distinctive about the most recent national peak in American homicide is not its height—there had been similar spikes in 1974, 1980, and earlier in the 1920s—but the extent to which the crime rate subsequently declined during the 1990s and then remained low in the 2000s. The national homicide trend depicted in Figure 1 did not descend quite to the lows registered during the post-war 1950s, but in late 1950s the nation recorded the lowest homicide rate since 1910, and 2007 came close to that figure.

 Crime and Criminals

Figure 1: Trends in Homicide and Robbery in Four Nations

Not surprisingly, such fluctuations are of perennial interest. Newly released crime figures are headline numbers virtually everywhere, and politicians and criminal justice officials have to be prepared to offer some explanation for their trend. They are quickest to step forward when the numbers are down. Social critics (p. 39) (p. 40) interpret rising crime as a reflection of deeper social tendencies that they view with alarm. Researchers have offered a variety of explanations for crime trends; these range from “a” (alcohol use, which is down in the United States) to “z” (zero tolerance policing, which is up). Researchers often conflict in their methods and differ in their views, but they are united in their concern for what the numbers actually measure.

Among the countries depicted in Figure 1, Japanese figures are interpreted as reflecting a distinct national culture, an aging population, a long period of economic success, and significant police-fiddling with criminal justice data. Recorded crime increases after 1990 appear to be due in part to a tightening-up of police recording practices by the National Police Agency following revelations of incompetence and nonfeasance (Leonardsen, 2006). In France, there are actually three official sources of homicide data. Only police statistics are presented in Figure 1; the others come from prosecutors and medical examiners, and they differ by a wide margin. About 60 of police-recorded homicides are actually attempted murders, and French researchers have pointed to police racism in the widespread practice of charging ethnic minorities with homicide in these cases. Econometric analyses of twentieth century data for England and Wales have found the long upward trend in predatory crime illustrated in Figure 1 to be driven by increasing affluence. However, the most recent British robbery numbers are widely discounted as reflecting changes in the way crimes are recorded in England and Wales. These changes were imposed on the police to ensure that all public complaints are taken seriously and are recorded more consistently across police forces. In contrast, the American data are locally collected and have avoided large-scale improvement efforts. It has been shown that trends in the data are, with careful controls, similar to those recorded in other systems, including the National Crime Victimization Survey and cause-of-death statistics gathered by public health officials. Differences in definitions and episodic changes in how crimes are recorded make comparing crime trends across nations even more perilous than examining them over time within nations, but even the latter requires extensive local knowledge of the data systems involved.

This Chapter reviews many of the research-based claims about why crime rises and falls within nations. The data behind these claims are more adequate for some than for others. Sometimes the data are not very good, or of only recent vintage, while other claims are based on relevant research which is handicapped by the absence of necessary data over time. It is important to note that there is not always a correspondence between the quantity or quality of the data and the potential significance of the claims. Sometimes there is no data at all, but this is not evidence of the unimportance of a particular claim. Rather, it is a consequence of what research has been funded and published, and whether or not some agency happened to begin collecting relevant data decades ago. In this Chapter, I discuss what I judge to be the serious claims.

(p. 41) II. Demography and Economic Conditions

One of the most fundamental features of common crimes is that offending is disproportionally concentrated among the young. Youths are by far the most likely to be arrested and to admit high levels of involvement in crime when they are surveyed by researchers. In the UK, Field 1999 found a 1 change in theft and burglary for every 1 change in the number of males age 15–20. Roberts and LaFree 2004 report that the dramatic aging of the Japanese population, along with its economic strength in earlier decades, is among the most important predictors of declining violence. Because of the overall aging of the American population during the 1990s—a result of the graying of the large post-WWII baby-boom generation—many have speculated that the national post-1991 drop in crime was due to a decreasing proportion of youths in high-risk age categories. The two did drop in unison until the mid-1990s. However, toward the end of the twentieth century, the 15–24 age group began to grow again, as the grandchildren of the boomers (the so-called “second echo” of the original baby boom) aged into higher-risk categories. For example, the percentage of the U.S. population aged 18 (a prime age for homicide offending) hit bottom in 1995, and then rose again. The post-1995 growth of the 15–24 age cohort was touted by forecasters as putting new upward pressure on crime rates. But while the pressure was doubtless there, crime rates did not in fact rise as predicted.

It is more accurate to say that upward pressure on the crime rate has generally come from the number of young males in the population, for one of the most universal findings of social science is that they are by far more likely to get in trouble. Except for certain sex-related offenses, men are also more likely than women to be victims of violent crime. During the 1990s, victimization rates for men and women in the United States both dropped, with the gap between the sexes remaining virtually unchanged for many types of crime. However, the rate of female commission of violent crime rose a bit relative to that of males. The rate of male offending dropped more than the female rate during the 1990s, so the female percentage of violent crime rose (see Heimer and Lauritsen, 2008). By 2007, females accounted for 22 of arrests for violent crime in the US, and in that year they accounted for a majority of those arrested for simple theft.

Many have attributed the most recent decline in crime to the vigor of the American economy, which grew steadily between 1993 and 2001. During that period unemployment fell nationwide by almost a third, and real incomes rose for many groups. However, overall unemployment trends cannot explain much of the variation in national crime rates. Levitt's (2004) review yields an “elasticity” of 1.0 for unemployment. This means that, for every 1 change in the unemployment rate, crime (p. 42) shifts by 1. However, national changes in unemployment are typically relatively small, just a few percentage points except in the hardest of times, and cannot account for much larger changes in crime rates. Rosenfeld and Fornango 2008 find that a national index of consumer confidence is the best representative of American economic conditions in models of robbery and property crime. There also may be an effect on crime of income inequality, but that moved in the opposite direction during the 1990s drop in American crime, toward an increase in wealth at the upper end of the economic spectrum.

However, criminologists know that it is more important to look at the economic condition of youths and poor people. Research indicates both wages and employment possibilities primarily have their effect on crime in these segments of the labor market. Most of the crimes considered here are committed by young men with little education, few skills, and a checkered job history. Grogger 1998 reports that a 10 increase in real wages in this group leads to a 10 decrease in economically motivated crime. One study concluded that changes in wages, an increase in the minimum wage, and improving employment prospects among young, low-income youths could explain almost one-third of the decline in crime in the U.S. during the 1990s (Gould et al., 2002). Staying in school is also important for this group. Compared to school leavers, students who remain in school get into much less trouble with the law, and they are much more likely to succeed in the job market as young adults.

Around the world, immigration was certainly a big story of the end of the twentieth century, and will be in the twenty-first century, and this too could influence crime. For example, in the early 2000s there were about 37 million foreign-born residents of the United States. About 11.5 million of them were citizens; another 40 were permanent resident aliens, officially admitted refugees, and students. It is estimated that in 2006, 12 million American residents were unauthorized migrants, or about one-third of the total foreign-born population. During the 2000s, the unauthorized population has been increasing by about 500,000 to 800,000 persons per year (Passel, 2006).

There are many reasons to fear that immigrants could contribute disproportionately to the crime rate. Much in their background suggests that immigrants should be very prone to getting into trouble. Compared to others, they typically are young, unmarried, and disproportionately male. Unauthorized immigrants, in particular, are likely to come with limited language abilities, few skills, and little in the way of formal education. They gravitate toward poor and disorganized immigrant neighborhoods where they can retain their anonymity. However, this crowds them into a context dominated by poverty and weak community ties, and it locks them out of the informal networks that frequently are required to get a good job. New immigrants may also not share the values of the dominant culture, which in any event is not much interested in theirs (Lee and Martinez, 2009). Based on all of this, presumptions about the criminality of immigrants (p. 43) have become entangled in larger national debates over immigration, both in the United States and elsewhere.

It is at least a theoretical surprise, therefore, that the American research literature overwhelmingly concludes that immigrants offend at lower rates than others, and their neighborhoods have unexpectedly low levels of crime. Further, their presence may help “inoculate” non-immigrants from crime. Evidence for those claims has been reported for immigrants of different national origins, and in multiple cities and neighborhoods. Sampson et al. (2005) set the pacifying effect of immigration at a 25 reduction when it comes to violent crime in Chicago. Detailed homicide studies indicate that immigrant immunity is strongest against acquaintance, domestic, and non-stranger murders. There is more uncertainty about the intergenerational transfer of this tendency toward immunity from crime. Most studies find that second and third generation immigrants are more involved in offending, but some report that the children of immigrants can show continued resistance to getting into trouble with the law.

This “immigrant paradox” extends beyond the domain of crime, into arenas such as health and educational attainment. It appears that many ethnic groups bring with them more “family values” and social structure than expected. By the end of the twentieth century there was much more back-and-forth communication and even travel to their countries of origin than for previous waves of migrants, reinforcing their traditions. Clustering together in their new communities may help preserve migrants' cultures and deter their children from easily adopting delinquent lifestyles. In many American cities immigrants are the only part of the population that is growing, and—also strongly attached to the world of work—they have brought modest economic revitalization in their wake. Ironically, it appears that the culprit behind increasing criminality in future generations is assimilation, that long-sought goal of melting-pot theorists. The un-assimilated appear to fare better than those who succumb to the dominant culture.

III. Prisons and Policing

One of the liveliest debates in the crime policy community is over the impact of mounting national incarceration rates. In June 2007, 2.3 million people were locked up in the United States, two thirds of them in prisons and the remainder in city and county jails. Analysts differ in their methodologies, but they attribute a significant portion of the decline in crime in the U.S. during the 1990s to prisons alone. Levitt's (2004) estimate is one third of the total; Spelman (2000) used different methods and concluded it was 27.

(p. 44) Prison influences crime rates through incapacitation, which is the reduction in crime that occurs because inmates cannot commit offenses while behind bars. The deterrent effect of fear of going to prison also keeps people from offending in the first place, or from reoffending in the future. However, separating out the impacts of incapacitation, general deterrence, and specific deterrence is a very difficult problem, and most research simply aggregates their total effect. Many studies report property crimes to be somewhat more responsive than violent crimes to changes in incarceration rates, and to shifts in levels of policing. In part this may be because rational calculations of the threat of punishment play a reduced role in offenses such as “crimes of passion,” “going postal” (mass killings of co-workers), and violence fueled by alcohol. The consensus is that about 20 of the yearly change in U.S. prison populations is translated to subsequent changes in crime (Levitt, 2004; Donohue and Siegelman, 1998).

Very important recent research has documented that there are probably diminishing deterrent effects of prison as states push up their rates of incarceration (Liedka et al., 2006). This is probably due to lower payoffs from digging deeper into the pool of less chronic offenders and individuals caught up in simple drug possession cases, which many American states have done in order to continue to push up politically popular incarceration rates even though there has been a significant drop in serious violent and property crime. Averaging the overall deterrent effect of prison over a time series, which is the way that most research has been laid out, could disguise larger effects occurring at the beginning of a run-up in incarceration, when higher-rate offenders were more frequently the ones being targeted.

Growing jail populations could also account for a decline in crime. Research on crime reduction has focused on the impact of prisons, which house inmates who have been sentenced to long terms of punishment. However, many of those arrested by the police serve a short sentence or are released after being confined only briefly. In the American system, jails are places of short-term confinement either while awaiting trial or serving short terms (less than 12 months) of incarceration, and inmates come and go rapidly. A study of the Cook County jail in Chicago found that in a year it admits nine times the number of people it holds at any one time. Some admittees go on to prison. However, most are either quickly released on bail, diverted to electronic monitoring or day reporting programs, returned to community programs, or are fairly promptly sentenced to the time they have already served. There has been no applicable research on the preventive effects of jail on crime rates, which—because of short stays—would probably be due more to deterrence than incapacitation.

There are incapacitation issues that need to be monitored carefully, for they threaten to sustain rather than decrease crime. One is the adverse impact of the removal of large numbers of men from already fragile communities. While their disappearance may lend some degree of safety to the community, it is not without cost. Their stigma gets transferred to family, friends, and to the community as a whole. Breadwinners are lost, and many families have to reorganize and build new (p. 45) care networks if they are to survive. Children go fatherless, spousal relationships become unglued, and everyone may suffer a loss of self-worth. The return of large numbers of ex-offenders in turn further undermines the community, increasing the concentration of poverty, alienation from legal institutions, cynicism regarding conventional authority, and the stigma associated with “reentry recycling” (Clear et al., 2001). The fate of those who are released from prison is also a critical issue. Generally the fate of recently released inmates is strongly tied to the well-being of the communities to which they return, but this has disproportionally been to disadvantaged neighborhoods (Kubrin and Stewart, 2006). Often there are only limited programs for recent returnees, yet they have to succeed in the job market fairly quickly if they are to avoid getting into trouble again. However, their employment prospects are poor, and many do not succeed.

Increasing incarceration rates are not the only widely touted explanation for declining crime. Beginning in the 1990s, American police leaders joined politicians in stepping forward to claim a share of the credit. Many cities hired more police. This was facilitated by the Crime Act of 1994, which set aside federal funds to support police salaries. Police chiefs also claimed credit because, they argued, they were policing “smarter.” They were focusing their efforts on crime hot spots, using computer technology to identify crime trends, and adopting community policing programs that brought them more cooperation from the neighborhoods they served.

A claim which has been evaluated is that increasing the size of American police forces helped account for declining crime during the 1990s and 2000s. The total did rise. In 1991 there were about 355,000 full-time officers serving in municipal police departments. By 2000 (when the decline in crime stalled) the total was 426,000, a 20 increase. That figure continued to increase, and by 2007 stood at 447,000. Nationally, there was a push to increase the number of uniformed officers assigned to community policing through federal police-hiring grants to cities.

How much of a decrease in crime could we expect from this quite sizeable increase in the number of police officers in the United States? There is not a clear consensus on the point, but the most prominently recommended estimates are that about half of the percentage increase or decrease in the number of police is translated into shifts in crime. Levitt 2004 concluded that between 1991 and 2001 the increasing number of police reduced crime by 5–6 across the board. Presumably, much of the deterrent effect of policing is through the experience—or fear—of being caught and possibly spending some time behind bars. However, there are no consensus estimates of how arrests independently affect the crime rate. Discussions of Japan's traditionally low crime rate usually allude to the extraordinarily high solution rates claimed by Japanese police. Roberts and LaFree 2004 found that the percentage of crimes that police claimed to solve, which began to deteriorate in the late 1990s, was among the strongest predictors of Japanese crime trends at the national and sub-national levels. However, crime solution or arrest rates are infrequently used (p. 46) in U.S. research. They are not reported by the FBI on a city-by-city basis, in order to avoid embarrassing apparently less effective police departments, and research indicates that there are vast differences in how agencies define and count these performance measures.

Like research on the link between crime and prisons, estimates of the impact of policing on crime are confounded by the reciprocal impact of crime (which creates a demand for greater security) on the size of police forces. To untangle this causal knot, analysts need to identify factors that independently influence the number of police officers and are not influenced by crime. In his statistical models, Levitt 2004 used police hiring increases leading up to local elections as a factor that independently affected police strength. Klick and Tabarrok 2005 used an even more powerful and focused event, the unexpected redeployment of police officers around Washington, DC, due to terrorism alerts. Interestingly, while their data showed that upsurges in police strength reduced property crime, its effect on violent crime was apparently zero.

Did crime go down because police are policing more intelligently? The world of policing was in ferment throughout the 1990s, in both the United States and other nations. It was a time during which both new policing strategies and refinements of tried-and-true tactics promised to increase their effectiveness. One of these refinements is focused “hot spot policing.” Using computerized crime mapping and data mining techniques, police began pouring resources into the relatively small number of places that in any city account for a disproportionate number of emergency calls and recorded crimes. They became increasingly sophisticated about the timing and management of crackdowns on street drug markets. Focused patrols and traffic stops aimed specifically at reducing gun carrying became more prominent. Targeted “quality of life policing,” which calls for aggressive arrest policies to counter seemingly minor crimes such as public drinking, graffiti writing, and panhandling, was credited by some as a theory-driven approach to reducing more serious crime (Kelling and Coles, 1996). Identifying and taking action at locations that are the source of repeat calls for police assistance became routine. Research indicates that one factor that makes neighborhoods high crime places is that an unusually large number of individuals and households there are repeatedly victimized. Targeting aid to first-time victims has been shown to reduce the extent of follow-on victimization, and thus area crime levels. During the 1990s a relatively new strategy, problem solving policing, became more popular, especially projects linking police with health, safety, housing, and other service and regulatory agencies, for this provided them with new tools for addressing chronic concentrations of crime. In addition, neighborhood-oriented community policing strategies were adopted on a large scale. This initiative called for community engagement and organizational decentralization, in addition to adopting a broad problem solving orientation.

Finally, in countries around the world, policing strategies of all kinds are increasingly guided by sophisticated police information systems that help managers (p. 47) discover and respond more nimbly to opportunities to prevent crime. Information technology is widely employed in managerial accountability systems (such as New York City's famous “CompStat” process) that put new pressure on police leaders to perform effectively. Collectively, these strategic initiatives could be characterized as the “smarter policing” of the 1990s and 2000s. Many attribute the drop in crime to increasing police effectiveness during this era.

The problem for this discussion is that there have been few evaluations of how well smarter and community-oriented policing work as city-wide initiatives, or how they influence the overall crime rate. Most of the research has been on the impact of experimental police teams in a few selected neighborhoods. Hot-spot policing, for example, has been evaluated by comparing before-and-after levels of crime in neighborhoods where police cooperated by trying out the program, with trends in a set of matched comparison neighborhoods where they did nothing new. Many neighborhood-oriented interventions, such as opening a storefront office, distributing newsletters, cleaning up neighborhoods, or instituting high-visibility foot patrol, have likewise been evaluated by trying them out in special test areas. However, the impact of the ensemble of the projects that make up a full-blown city program have not often been assessed, nor has the ability of the police actually to implement them on a city-wide basis. Further, the community policing movement has many important goals in addition to crime reduction, including reestablishing the legitimacy of the police in poor and minority communities, incorporating new immigrant communities into the body politic, improving the quality and civility of service (broadly defined) rendered by officers, and fostering civic involvement by residents. This greatly increases the range of outcome measures against which it needs to be evaluated.

A difficulty in evaluating claims that the diffusion of smarter policing or community policing around the country helped account for declining crime is finding ways to measure the breadth and depth of these diverse new policing strategies over the pre- and post-1990 era. An evaluation of community policing in Chicago found evidence that the introduction of the program in test districts reduced crime and fear, increased confidence in the police, and mobilized large numbers of residents around crime prevention projects (Skogan and Hartnett, 1997). Broadly focused national or multi-city studies do not even attempt to do this because there are no convenient over-time data on the quality, as opposed to quantity, of policing. A recent review of research on policing finds evidence supporting the effectiveness of many of the focused policing efforts described above, but those conclusions are based on the findings of city-by-city neighborhood evaluations (Skogan and Frydl, 2004). National-level and multi-city studies of the drop in crime focuses on the number of police, but it is surely what police do, rather than how many are on the payroll, that has an impact on the street. While their numbers went up, there is no reason to think what police did remained static during the 1990s, and the effects of adopting new tactics and technologies may by confounded with that of their increasing numbers.

(p. 48) IV. Drugs, Guns, and Gangs

This section examines the impact of three highly interrelated features of crime in America: drugs, guns, and gangs. There is good reason to consider them jointly, for together they lie at the core of big-city crime problems. A common narrative explaining the decline in crime which began in the early 1990s hinges on the rise and subsequent waning of interest in crack cocaine in American cities. It is argued that a crack epidemic during the late 1980s fueled an expansion of street drug markets and a broadening of the recruitment base for dealers and their assistants. This in turn precipitated wars among gangs over control of these markets, with the subsequent widespread diffusion of guns for both offensive and defensive purposes. Young African-American men in big cities were particularly drawn into the drugs-gangs-guns nexus. Killings by them (and of them) accounted for much of the upsurge in city violence that characterized the late 1980s. Then, it is argued, the crack market changed. What supposedly happened remains vague: crack markets are variously described as “maturing,” “stabilizing,” “waning,” “ebbing,” “becoming less lucrative,” and facing “diminished demand.” 1 This market shift presumably accounted for the sharp drop in homicides which followed.

The guns-gangs-drugs narrative described above is one plausible view of what happened. However, national or even multi-city studies typically do not have much independent information on drugs at all. Observers' conclusions appear to draw on the drop in crime and the changing profile of homicide victims and offenders—which is what they are explaining—rather than on systematic information about real changes in drug markets. Changes in drug markets may also affect crime through their impact on individual consumers. Drug abuse certainly remains at high levels among those arrested for violent and property crimes. Research on temporary police crackdowns on drug markets indicates one of their effects is to increase the frequency of robberies and burglaries nearby, presumably because a sharply constricted supply leads to upward spikes in drug prices (Sherman, 1990). Long-term studies have also documented that expanding heroin use leads to higher robbery rates through its criminogenic effect on consumers.

A difficulty with assessing the waning crack market argument is that the character and extent of drug markets are difficult to measure. Many studies primarily work with data on drug arrests. This is an enforcement measure, reflecting where and how police conduct their investigations, but there is evidence that the geographical distribution of drug arrests broadly mirrors the distribution of drug markets. The major alternative to police arrest reports is data on emergency room treatments and drug-related deaths.

(p. 49) What is certain is that there has been a tremendous increase in the number of drug-related arrests over this period. In Chicago, between 1991 and 1998, the number of drug arrests rose from 21,450 to 58,500, a 173 increase. After stabilizing, the drug arrest total rose a bit more, exceeding 59,000 by 2004. This growing number of arrests was particularly startling in light of the declining number of apprehensions in many other categories, reflecting the drop in crime. In 1991, drugs accounted for 9 of all the non-traffic arrests made by Chicago police; by 2004 drug offenses accounted for 30 of non-traffic arrests in Chicago. An important reason Illinois' prison population remained stable in the face of steadily declining violent and property crime during the 1990s was this new source of inmates. Another feature of this shift in crime control was that virtually all of these new arrests in Chicago targeted African Americans, and by 2004 they made up 79 of drug arrestees. When multiplied by the racially disproportionate prosecution, sentencing, and incarceration that takes place further downstream in the criminal justice system, this shifting enforcement pattern helped make Illinois one of the most racially disproportionate prison systems in the United States.

Gun availability and use also play an independent role in increasing the severity—and perhaps the rate—of violent crime. Zimring and Hawkins 1999 noted that American levels of non-gun violence do not particularly stand out in international league tables. Rather, it is what they describe as “the thin layer of lethal violence” that differentiates the United States from many other nations. A number of nations rank above the United States in terms of homicide rates, but none are in the same economic or political category. Why American homicide rates are so high has been widely discussed, with explanations ranging from racial divisions to its “cowboy culture.” However, it is obvious that the widespread diffusion of firearms, coupled with a readiness by too many to actually use them, is a factor that differentiates the U.S. from otherwise comparable nations. Guns intensify the consequences of violent encounters because they increase the likelihood of death. They may also increase the overall frequency of crime somewhat, because some crimes—high-payoff commercial, cargo, and bank robberies come first to mind—simply are not practical to carry out without a gun. However, the statistical evidence is that it is the severity of injury and risk of death, not the general frequency of offenses, that is driven by the availability of firearms (Cook and Ludwig, 2000). Increasing use of semiautomatic handguns was the most important component driving overall homicide rates in the late 1980s and early 1990s, because of their lethality.

Local level data on the extent of gun availability is difficult to assemble. However, there are direct measures of the readiness of individuals to carry guns, such as the number of guns seized by police. Firearms are seized under a variety of circumstances that empower police to make lawful searches. These circumstances range from traffic and pedestrian stops to arrests for committing other crimes, so seizures reflect what police find in the segment of the population that (p. 50) comes under suspicion. A second measure of gun carrying is arrests for weapons violations when individuals are found in possession of a gun but other more significant criminal charges such as robbery or assault cannot be laid against them. As a practical matter, rates of gun carrying are certainly a more significant factor than rates of ownership, given that a majority of shootings now take place in public spaces and involve people who have brought a weapon to the scene. Police officials point to this fact to justify aggressive use of pedestrian stops and vehicle searches, and when they do not find many guns despite these efforts, they see it as evidence that potential shooters have been convinced to leave their hardware at home.

Gangs are certainly another important cause of crime in the United States. Americans used to call them “youth gangs,” but they changed in character. During the 1980s, massive adult joblessness among African Americans extended the age profile of active gang members, because burgeoning drug markets provided them with a new way to make a living. Since then, members returning from prison have begun to rely on their old gangs for employment in the trade (Coughlin and Venkatesh, 2003). One of the functions of gangs is to recruit new members and steer them into the organization's many criminal enterprises. Depending on the city, gangs are typically implicated in drugs, gambling, extortion, theft, gun smuggling, reselling valuable parts from stolen cars, financing gypsy cabs, and selling “protection” from rival gangs. Another role of gangs is to recruit replacements for members who have been arrested and incarcerated. This contributes to a sad spiral of increasing overall criminality as a consequence of law enforcement efforts to counter the drug trade. Third, they cause crime to the extent to which they foster inter-gang violence and occasionally intra-gang violence when they end up making war on fellow gang members (Decker et al., 1998).

Gang homicide is different from other kinds of killings. While the many social and economic factors discussed here may have played a role in bringing about steadily declining levels of non-gang crime, a significant fraction of crimes in the United States are rooted specifically in the organizational dynamics and business environment of violent street gangs. In this world, disputes over honor and status can easily have violent outcomes. One attack leads to another. Killings lead to retaliatory killings, and violence ripples through the community, ricocheting among organizations, and sometimes catching bystanders in the line of fire. Between gangs, violence is a tool for settling disputes over drug markets and control of other illicit enterprises. Within organizations, violence is exercised in order to impose discipline, collect street taxes, and maintain the standing of power-holders (as when former kingpins return from prison demanding their share). In Chicago, the decline in homicide during the 1990s was driven by non-gang killings. Gang killings rose and fell, following their different logic. But because non-gang homicide fell precipitously, by 2005, gang killings comprised more than double the proportion of homicides they had a decade and a half earlier.

(p. 51) V. Community and Environmental Factors

A significant fraction of all criminological research focuses on the role of community factors in controlling crime. These factors range from the strength of informal bonds of trust among neighbors to organized crime prevention efforts by community groups. Communities struggle to control crime on several different levels. At the most private level lies the strength of family values and the ability of parents to socialize and control their children. Neighborhoods vary in how well their families do this. At another level lies shared norms and trust among neighbors, and a willingness in the community to intervene when things go wrong. Together, widespread trust and willingness to intervene make up what is known as “collective efficacy,” a community factor strongly linked to levels of crime (Morenoff et al., 2001). High-crime communities suffer from disrupted networks of friendship, kinship, and acquaintanceship, and this limits their capacity for mutual informal coordination and cooperation. Informal “pro-social” interventions and effective sanctioning behaviors are much weaker in these areas. Structural disadvantages, such as concentrated poverty and residential turnover, hurt communities in particular at the private and informal levels of cooperation. At a more overtly political level lie the groups and organizations that constitute civil society. There is some evidence that a strong infrastructure of organizations can sustain a community's capacity for self-healing social action. Controlling for other things, where organizations are strong, community residents are seen as more likely to take action and intervene, and this can compensate somewhat for weak informal ties among neighbors (Sampson, 2004).

Many decades of research have demonstrated that community factors are powerful determinants of levels of crime. However, little is known about whether community factors wax or wane over time. There has also been little research that actually addresses whether changes in community factors—as opposed to comparisons among communities that vary in those factors—are linked to changes in levels of officially recorded crime. In the main, collective efficacy is strongest in stable, white home-owning neighborhoods, so it could depend in part on whether such areas are growing or declining in number in a particular city (Sampson et al., 1997). Research on community factors typically stresses their complex and multifaceted character, and much of this research relies on specially collected local data. None of the prominent studies of crime trends in the United States have taken any notice at all of community factors. This void is certainly due in large measure to the absence of any relevant national, over-time data. For example, there is vigorous debate just over whether organizational involvement by Americans has gone up or down (Putnam, 1993; Paxton, 1999). Similarly, discussions of crime in Japan routinely point to the (p. 52) strength of informal social control as one explanation for the extraordinarily low levels of crime there. However, none actually advanced any measures of variations in social control across Japan, nor pointed to systematic evidence on whether or not informal control weakened as recorded crime (except for homicide) skyrocketed post-1995. Community has not figured in studies of trends in crime because the large body of richly theoretical research on this topic does not yield one or two index numbers which can be included in statistical models testing their role relative to other potentially important determinants of crime trends.

Another neighborhood factor could be the aggregate impact of housing and economic revitalization programs underway in many cities. These can be measured by such indicators as the number of building permits issued for residential and business projects, and the number of conventional residential mortgages registered each year. Housing and commercial investments are doubly important because they are cumulative, adding up over time. Trends in these indicators highlight the fact that many cities are far from moribund, and that new investments are being made in their future. What is unclear is whether this evident vitality is a cause of declining crime, or if it is a consequence of the drop in crime. Perhaps one reason for this quickening economic activity during the 1990s and early 2000s was the new sense of safety and security evident in neighborhoods in most of America's larger cities. Some of the best research on the relationship between crime and real estate activity has been done in New York City. It suggests that the decline in crime in that city sparked a real estate boom during the 1990s, and that crime decline accounted for about one-third of New York City's rise in real estate prices after 1994 (Schwartz et al., 2003).

Efforts in the United States to restore or demolish high-rise family public housing also began in earnest, toward the end of the 1990s. The Federal HOPE VI program in particular promised to replace these units with a mixture of public, subsidized, and private-market low-rise housing. Of course, the residents of the original apartment blocks did not go away; demolition just redistributed them somewhat. The limited research on the topic suggests their lives have taken a turn for the better. Residents leaving family high-rise projects with housing vouchers, or even unassisted, have mainly succeeded in moving to safer and substantially better-off areas with better-performing schools, and in surveys they report dramatically lower levels of concern about neighborhood drug markets, shootings, and other criminal activity. Relocatees who managed to move into private-market housing did better than did those who moved but remained in public housing (Popkin and Cove, 2007). There is broad agreement that the concentration of poverty created by massive public housing developments had bad consequences which were independent of the characteristics of the families living there. Gangs, drugs, and guns were an everyday feature of the lives of many residents, and there were few places to hide (Popkin et al., 2000). Now many of those developments are gone. In Chicago, crime in and around the sites of former high-rise developments is in rapid retreat, and there has been significant new construction and appreciating house prices in the neighborhoods of (p. 53) all of the major projects that were demolished. To date, research has appropriately focused on the former residents and their fate in the housing and job markets. We await follow-up work on the impact of demolition and relocation programs on crime prevention and displacement.

Finally, there may be some scope for environmental factors in causing crime and its trend over time. Crime is certainly highly correlated with pollution, measured (for example) by city levels of particulate matter in the air. This has long been used as an illustration of spurious correlation, because high crime and polluted air are often found together in declining rust-belt cities where the many causes of crime reviewed here abound. However, there are specific environmental factors that are known to have causal links to biological functioning, and through that to crime.

In particular, concentrations of lead in the human body have harmful effects on children's impulsivity, judgment, cognitive function, learning, and disciplinary behavior. Research has shown that the more lead there is in children's blood at a young age, the more likely they are to be arrested as an adult. The key point here is that the American states began requiring dramatically lower levels of lead in gasoline in the 1970s. They did so at different times and required these reductions at different rates, enabling researchers to examine with some confidence the lagged relationship between changes in lead exposure levels and subsequent declines in crime rates. Reyes 2007 reports that the reduction in childhood lead exposure in the late 1970s and early 1980s due to this policy was responsible for significant declines in violent crime in the 1990s, and may cause further declines in the future.

VI. Lifestyle and Culture

Lifestyle theories of victimization take note of the fact that crime has relational and situational features. Rather than just focusing on the supply of offenders, in this view crime occurs when motivated offenders come into contact with suitable targets in the absence of “capable guardians.” Target suitability is flagged by affluence and other features that make for attractive opportunities for crime. The finding by Field 1999 and others that burglary and theft in England and Wales was driven upward over time by increasing consumer affluence during the 1950–1970 period is interpreted in this light—that a noticeably growing stock of consumer goods meant there was more around to steal. Guardianship may be exercised by bystanders, shopkeepers, or other persons whose presence deters on-the-spot offending, and by physical barriers such as locked doors or hardened automobile ignition systems. Research has established clear relationships between the extent of victimization and a broad (p. 54) range of indicators of target suitability and guardianship. Further, there is evidence of fairly dramatic changes in many of these indicators, shifts that portend changes in crime patterns.

Some lifestyle indicators can be found in how people live and how they organize their households. The industrial world has witnessed declining family size, fewer family households, more people living alone, higher rates of divorce, and a declining popularity of marriage: all factors that expose people to an increased risk of victimization. People are more at risk of personal crime when they are out, and of victimization of their homes when no one is at home. Pursuing more work and leisure away from home, and taking more vacations, has the same impact. A very significant fraction of the total victimization of adult women in the labor force takes place at work, or on the journey to and from work. On the other hand, declining rates of marriage, replaced by increases in cohabitation, seem to lie behind the precipitous decline in spousal and intimate partner violence registered in the U.S. during the 1990s. Domestic homicides in particular declined at a tremendous rate during that period. Rosenfeld (2000) documents that declining marriage rates explained 40–65 of declining spousal homicide rates among high-risk groups. In reference to a trend of “declining domesticity,” he speculates that the ease with which failing relationships can come apart predicts the likelihood of one of the parties ending up divorced rather than dead.

Another lifestyle indicator, and one offering an intriguing explanation for declining crime, is alcohol consumption. In the United States, drinking has been on a nationwide decline on a per capita basis at the same time that violent crime has subsided. Research on alcohol use and its effects is very strong. The pharmacology of the product is well understood by criminologists. Physical testing of both offenders and victims documents that alcohol is frequently associated with violence. There have been many experiments and quasi-experiments gauging the impact of changes in alcohol distribution, at levels of analysis ranging from individuals to entire nations. This research finds the availability and volume of alcohol consumed is related to levels and trends in crime at the block, neighborhood, city, county, state, and national level. When it becomes harder for young people to acquire alcohol, youth violence (as well as traffic fatalities) subsides. A key point is that overall alcohol consumption has also been declining. Nationally, Americans' per capita beer consumption peaked in about 1976, and has since stayed about level. Consumption of spirits has plummeted since 1968, while wine consumption per capita has been fairly level since the mid-1980s. Parker and Cartmill 1998 find a two-year time lag in the relationship between declining levels of alcohol consumption and declining homicide rates.

Cultural explanations are frequently advanced to explain the relatively (and extraordinarily) low rates of crime in Japan. Japan is described as a homogeneous, group-based, other-directed, hierarchical society with a high consensus around conservative values that are reinforced by shaming practices which exact a high, if informal, price for non-conformity (Leonardsen, 2006). However, these features of Japanese life have proven difficult to capture statistically. They have been measured (p. 55) variously by divorce rates, female labor force participation, population mobility, and labor disputes. These are obviously very indirect indicators of what are reputedly very powerful social factors, so indirect that Roberts and LaFree 2004 describe most of them as measures of social disorganization rather than culture. It is also not clear if culture is changing, or why any such changes might currently dampen homicide rates but not other, rapidly increasing levels of crime in Japan.

Elsewhere, national culture is rarely invoked as an explanation for trends in crime. In the 1970s and 1980s, cultural shifts were identified by conservative critics as a cause of spiraling crime in the United States. They linked crime to the hedonism, self-indulgence, heedless consumerism, reckless individualism, unbridled self-expression, and standardless relativism that they claimed increasingly characterized a disintegrating liberal culture. However, I have no reason to believe that any such trends reversed their course dramatically following 1990, driving crime in the opposite direction. In any event, the drop in crime in the United States during the 1990s took place over such a short period of time that it would seem difficult to tie it to some cultural shift, because important values are unlikely to change so rapidly. Experienced urban ethnographers have reported that “the younger generation” (a group that grew more expansive with each passing year that crime declined) looked at the violence of the late 1980s, then decided in large numbers to turn away from a “culture of death” (Johnson et al., 2000). However, cultural claims about trends in crime will be particularly difficult to assess systematically.

It has also been claimed that changes in policy beginning in the 1970s reduced the number of “unwanted” children in succeeding generations, a reduction that is measured by state-level abortion rates. Donohue and Levitt 2001 argued that the remainder would be less likely to offend due to their more considered upbringing. Because they were “wanted,” fewer in number than they otherwise would be, and less likely to be born to high-risk mothers, Donohue and Levitt thought that post-abortion-era children would receive more attentive care and control. Tracking arrests over time, they reported finding suddenly decreasing offending rates among youth cohorts of the appropriate age, using pooled state-level data. Further, this decline coincided to some extent with the crime decline of the 1990s. However, the implications of research on this topic grew murky with the discovery of serious computational errors in the study. Some statistical critics think the claim was wrong (Foote and Goetz, 2008), but others have gotten results approximating those of the original study. I never thought their argument squared with data on the generation entering the high-risk age category in the early 1990s, for it was in worse social and economic shape than previous cohorts and showed precious little evidence of the presumed benefits of having been “wanted.”

There is also a view afoot that trends in crime are driven by “tipping points,” and that they somehow naturally come back down when they hit some threshold. This is a “what goes up must come down” theory of crime that is no theory at all in the absence of explanatory variables that predict the turnarounds. Without them it is just (p. 56) a description of what happened. A related claim is that declines in crime from their highs represent “regression” (Fagan et al., 1988) or “mean reversion” (Harcourt and Ludwig, 2007). Declines (and, I presume, increases) apparently are to be expected when cities or police precincts have unusually high (or low) crime rates, as they revert to levels that Fagan et al. characterize as “historically normal.” This view is a misreading of the well-known “regression artifact” issue in measurement theory. Regression artifacts threaten when researchers deliberately select a subset of cases that stand out based on their extreme scores and those scores include a significant random error term. On an ensuing measurement, scores for those extreme cases should indeed revert toward the middle of the distribution from which they were selected, to the extent to which the measures were dominated by random error. Credible statistical studies of crime decline are not seriously threatened by any of those factors. Unemployment does not come down because it went up without any intervening processes, and neither should we expect crime to follow that pattern for some artifactual reason.

More promising “turning point” concepts are “diffusion” and “contagion.” I argued above that crime is contagious in the context of American gang violence. That is a world in which bands of young men are ensnared in networks of power, status, and business relationships which propel them into episodes of tit-for-tat violence of the worst kind. Arms races followed by wars that later subside due to the exhaustion of the combatants could look like tipping points, except they tip upward again when everyone recovers. A key point is that ups and downs in gang-related contagious violence have been largely unrelated to the general, unpunctuated decline in non-gang crime, which continued apace on a much more massive scale throughout the period considered here. Outside of the insular world of street gangs, the idea that marshaling the bravado to commit crimes can be affected by the perception that “everyone is doing it,” and that there may be some “safety in numbers” when a lot of people are doing it, points toward a “critical mass” theory of crime causation. There is some evidence that crime diffuses like fads and fashions, fanning outward from big cities to the hinterlands. Police and journalists frequently refer to “copycatting,” another process by which social contagion may increase the overall crime rate. But while all of this may be promising, there has been no clear explication of the factors that dampen group-think or discourage copycatting and bring offending down, but that was the predominant trend during the 1990s.

VII. Crime Reporting and Recording

Perhaps increases or decreases in officially recorded crime reflect shifts in the willingness of victims to call the police, or variations in enthusiasm on the part of police (p. 57) to make an official record of crime reports that come their way. This could produce apparent shifts in the crime rate, not real ones. For example, there is speculation that negligence complaints against the Japanese police beginning in the late 1990s led them to record more completely reports of crimes that the public brought to police stations (which is how crime is reported in Japan), and that this increased attentiveness, plus more aggressive media coverage of crime encouraged more victims to step forward (Leonardsen, 2006). Likewise, in 1992 and 2002, changes were made to British crime recording practices which were intended to increase the fidelity with which public complaints—especially concerning common assault—were registered by police. However, in both cases it is clear from Figure 1 that robbery was increasing in both nations in advance of these shifts in record keeping, so they could be only part of the overall picture.

At least for the United States, recent evidence about trends in victim reporting of crime is to the contrary. The best over-time data is from the U.S. National Crime Victimization Survey, which interviews thousands of individuals each month. In the survey, victims are asked if their experiences came to the attention of the police. The NCVS finds that reporting of crime to the police actually went up during the 1990s, not down, as recorded crime dropped. Reporting of violent crimes rose by 6 during the period (Hart and Rennison, 2003). This put upward pressure on the official crime rate, leading it to perhaps understate the extent of crime decline. Likewise, there is speculation that community policing increases victim reporting, as public confidence in the police grows. However, little is known about the impact of changes in policing styles on crime reporting. A little research in the United States and abroad does suggest that effective neighborhood-oriented policing may encourage somewhat more crime reporting.

It is certainly plausible that immigrant communities may be particular reservoirs of unreported crime, and immigrant numbers are growing. When they are victimized, it is widely assumed that immigrants are less likely than others to report their experiences to the police. Refugees from many countries leave in part because of endemic corruption, violence, abuse, and incompetence among the police, whom they do not view as their friend. They may fear deportation if they are undocumented, and they may fear exposing family members and other undocumented people in their social network, even if they themselves are not. The problem is that police need the trust and cooperation of the community in order to respond effectively to crime. They rely on the willingness of victims and bystanders to cooperate with their investigations. To gain this cooperation, police need to remain in close and trusted contact with residents, and in many cities community policing units have taken on this responsibility (Skogan, 2009).

But there are counter trends. One is the new demands that are being placed on the police to become involved in enforcing immigration laws in the same communities where they are struggling to build trust in the immigrant community. Before 2002, it was a long-standing policy of the U.S. Department of Justice that local police (p. 58) should not be involved in enforcing immigration laws. State and local police were not empowered to arrest and detain violators; this was seen as the special responsibility of the (then) Immigration and Naturalization Service. Then, the Bush administration issued a memorandum announcing that local law enforcement officials have “inherent” authority to make arrests for civil immigration violations. They adopted a series of strategies aimed at expanding the role of local police in immigration enforcement (Skogan, 2009). In a study of Chicago, I found a dramatic difference in the rate at which English- and Spanish-speaking Latinos initiated encounters with the police and reported problems and concerns to them. Language and immigration factors played a very large role in shaping the relationship between the city's newcomers and institutions of government (Skogan, 2006). This also implies that official rates of crime—which depend upon ordinary citizens contacting the police to make official reports—may not adequately reflect the problems facing those living in heavily immigrant areas.

Police under-recording of the incidents reported to them may artifactually influence the crime rate as well. This has often been observed in individual American cities. For example, during the mid-1980s, Chicago police were “killing crime” by failing to write up official reports of huge numbers of offenses. An apparent run-up in crime in Chicago at the end of the 1980s was certainly due in substantial part to improvements in recording wrought by internal crime recording audits, which were instituted after the media became aware that detectives were cheating on their statistics. The breadth of this scandal makes it impossible to do meaningful statistical analyses of crime trends in Chicago that include the 1980s (Skogan, 2006). However, in my experience, outbreaks of recording malpractice are localized and scattered, and have become less common in recent decades. It is more likely that computerization, crime mapping, data-driven resource allocation, and the imposition of CompStat-like management accountability systems in police departments around the United States and elsewhere have improved the quality of crime data instead.

VIII. Conclusions

This Chapter has presented an item-by-item list of factors linked to macro-social trends in crime. However, the research on which it is based is multivariate, controlling for multiple factors in order to isolate the effects of particular variables of interest, and the list of significant coefficients is typically quite long. In my judgment, it is likely that the roots of the American drop in crime similarly lie in a mix of the factors described here, not just one, and this is likely true for significant, long-term shifts in rates of crime in any nation. Some or all of these social and economic trends (p. 59) and policies may be working in concert to reduce crime, each contributing to the end result.

A few of the factors discussed here could have pushed U.S. crime rates down through the entire period. Immigration proceeded at a (modern) record pace, and we have seen that this has paradoxically positive consequences for crime rates. Alcohol consumption dropped steadily as well, and that has long been known to be a feature of violence. The American economy was bad at the beginning of the drop, but improved a great deal later in the 1990s, including at the bottom end of the labor market. On the other hand, women became more criminally active. The proportion of young people in the population began to increase late in the 1990s, and the echo boom foretold increasing numbers of young men in the U.S. population from 1997 until about 2012. Causes may also simply run out of steam and no longer contribute to crime decline—witness the diminished effectiveness of incarceration as it reaches high levels relative to the pool of serious offenders.

Different kinds of crime can be sensitive to different factors. Property crime appears to be somewhat more “rational” than violence, in that it responds more directly to shifts in policing and incarceration. Upward and downward spirals of gang violence seem to occur in response to quite different forces than other kinds of violent crime, reflecting more the logic of arms races than social and economic trends. Interestingly, as gang killings have become a large percentage of the total homicide count in many places in the U.S., increasingly it is gang dynamics that drive the total. This disconnects overall homicide rates from the forces that previously drove them and the statistical models that apply to other types of crime. Likewise, over time, non-gun assault in the United States has dropped relative to gun violence, as domestic violence, fights in bars, and other kinds of relational and recreational violence subsided much more rapidly. As a result, in cities, aggravated assault figures increasingly measure shots that missed and are really indicators of poor marksmanship. In my own work I find that combining homicides (mostly shots that found their target) and gun assaults (a result of bad shooting) produces a “shootings” measure that is more robust than either component taken alone.

In this multi-causal view there is also room for programs and policies, as well as for demographics and economics. There is strong evidence that focused strategies like hot spot policing are effective, and surveys document that increasing proportions of police departments employ them. However, over time and aggregated or across-city measures of these developments have yet to be incorporated into influential models of crime rate trends. To argue that “the drop in crime began before they were introduced” is to reject the potential significance of programmatic innovations on an overly mono-causal basis. The crime drop of the 1990s in the United States continued to hold steady in the 2000s, and over an 18-year span it seems likely that combinations and reinforcing mixes of changing factors must have been at work.

(p. 60) I think the categories of factors considered here could describe research on crime trends in many nations. However, the specific indicators of (for example) the strength of communities would certainly vary with context. Demographic factors associated with crime will vary. For instance, what groups are socially and economically marginalized, and the nature of their condition, varies considerably from nation to nation. Lifestyle explanations of victimization point to factors such as housing patterns, family structure, and the role of women in the economy that can vary widely even among developed nations, creating an opportunity for comparative research on crime trends. Criminals also vary in their “modus operandi”: gangs in the United States play an increasing role in determining the homicide trend, but Japan's gangs seem more focused on business affairs and play a much smaller role in large-scale trends in homicide. The importance of seemingly comparable causal factors could vary as well. For instance, it seems likely that crime will vary less with shifts in macroeconomic conditions in societies with strong solidarity norms and social welfare policies that provide a true “safety net.”

The discussion here did not, on the other hand, encompass the effects of ruptures in the social order, which are rare in industrial democracies. For example, it was feared that the joining of the two Germanys in 1990 might push the social order of the East to a precipice over which it could tumble, as existing economic, social, and lifestyle patterns were threatened with discontinuous change. It does not seem to have worked out that way. Germany's crime statistics have included the East since 1993, and since then there has been a very steady downturn in well-reported crimes such as auto theft, burglary, gun crimes, and homicide (Bundeskriminalamt, 2007). The physical destruction and administrative collapse of the city of New Orleans in the face of a 2005 hurricane presents another example of a rupture in the social order, the criminogenic effects of which have not yet been sorted out.

A limitation of this presentation is that it has neglected the contributions of survey studies of criminal victimization. All of the countries depicted in Figure 1, and a long list of others, have conducted victimization surveys, and they yield somewhat different pictures of crime. Survey data encourages victim-centric explanations of crime, including the impact of their lifestyles and routine activities, but add additional methodological problems that are not considered here at all. Except for the American crime survey, none have been frequent enough at this point to support the kind of time-series oriented, macro-social research reviewed here. This research model is fairly demanding in terms of its data requirements, requiring many decades of regularly spaced and uniformly gathered information.

Of the factors considered in this Chapter, data gathering is needed most in the community domain. Many decades of criminological research have established the importance of community factors in controlling crime. Strong families, deep (p. 61) informal bonds among neighbors, and a willingness of residents to intervene to maintain order are linked to lower levels of crime. Even in the face of concentrated poverty and residential turnover, there is some evidence that a strong infrastructure of organizations can help sustain a community's capacity for self-healing as well. But little is known about whether these factors strengthened or weakened during the 1990s and 2000s, nor how they might have been affected by community policing programs that emerged across the country. There is an absence of evidence concerning their role in explaining declining levels of crime, due in large measure to the absence of systematic, over-time data on community factors.


Blumstein, A. (2000). “Disaggregating the Violence Trends,” in A. Blumstein and J. Wallman (eds.), The Crime Drop in America, New York: Cambridge University Press.Find this resource:

    Bundeskriminalamt (2007). Police Crime Statistics 2006, Wiesbaden: Bundeskriminalamt.Find this resource:

      Clear, T., Rose, D., and Ryder, J. (2001). “Incarceration and the Community: The Problem of Removing and Returning Offenders,” Crime and Delinquency 47: 335–51.Find this resource:

        Cook, P. and Ludwig, J. (2000). Gun Violence: The Real Costs, New York: Oxford University Press.Find this resource:

          Coughlin, B. and Venkatesh, S. (2003). “The Urban Street Gang After 1970,” Annual Review of Sociology 29: 41–64.Find this resource:

            Decker, S.H., Bynum, T., and Weisel, D. (1998). “A Tale of Two Cities: Gangs as Organized Crime Groups,” Justice Quarterly 15: 395–425.Find this resource:

              Donohue, J.J. and Siegelman, P. (1998). “Allocating Resources Among Prisons and Social Programs in the Battle Against Crime,” Journal of Legal Studies 27: 1–43.Find this resource:

                Donohue, J.J. and Levitt, S. (2001). “The Impact of Legalized Abortion on Crime,” Quarterly Journal of Economics 66: 379–420.Find this resource:

                  Fagan, J., Zimring, F.E., and Kim, J. (1988). “Declining Homicide in New York City: A Tale of Two Trends,” Journal of Criminal Law and Criminology 88: 1277–1324.Find this resource:

                    Fagan, J. and Wilkinson, D. (1998). “Guns, Youth Violence, and Social Identity in Inner Cities,” in M. Tonry (ed.), Crime and Justice: A Review of Research, 24.Find this resource:

                      Field, S. (1999). Crime Trends Revisited, Home Office Research Study 195, London: Home Office.Find this resource:

                        Foote, C.L. and Goetz, C.F. (2008). “The Impact of Legalized Abortion on Crime: Comment,” Quarterly Journal of Economics 123: 407–23.Find this resource:

                          Gould, E., Weinberg, A., and Mustard, D. (2002). “Crime Rates and Local Labor Market Opportunities in the United States: 1979–1997,” Review of Economics and Statistics 84: 45–61.Find this resource:

                            Grogger, J. (1998). “Market Wages and Youth Crime,” Journal of Labor Economics 16: 756–91.Find this resource:

                              Grogger, J. and Willis, M. (2000). “The Emergence of Crack Cocaine and the Rise in Urban Crime Rates,” Review of Economics and Statistics 82: 519–29.Find this resource:

                                (p. 62) Harcourt, B.E. and Ludwig, J. (2007). “Reefer Madness: Broken Windows Policing and Misdemeanor Marijuana Arrests in New York City, 1989–2000,” Criminology and Public Policy 6: 165–82.Find this resource:

                                  Hart, T.C. and Rennison, C. (2003). Reporting Crime to the Police, 1992–2000, Washington, DC: Bureau of Justice Statistics.Find this resource:

                                    Heimer, K. and Lauritsen, J.L. (2008). “Gender and Violence in the United States: Trends in Offending and Victimization,” in National Research Council (ed.), Understanding Crime Trends: Workshop Report, Washington, DC: National Research Council.Find this resource:

                                      Johnson, B., Golub, A., and Dunlap, E. (2000). “The Rise and Decline of Hard Drugs, Drug Markets, and Violence in Inner-city New York,” in A. Blumstein and J. Wallman (eds.), The Crime Drop in America, New York: Cambridge University Press.Find this resource:

                                        Kelling, G.L. and Coles, C.M. (1996). Fixing Broken Windows, New York: Touchstone.Find this resource:

                                          Klick, J. and Tabarrok, A. (2005). “Using Terror Alerts to Estimate the Effect of Police on Crime,” Journal of Law and Economics 42: 93–122.Find this resource:

                                            Kubrin, C. and Stewart, E. (2006). “Predicting Who Re-offends: The Neglected Role of Neighborhood Context in Recidivism Studies,” Criminology 44: 165–97.Find this resource:

                                              Lee, M.T. and Martinez, R. (2009). “Immigration Reduces Crime: An Emerging Scholarly Consensus,” in W. McDonald (ed.), Sociology of Crime, Law and Deviance, vol.16, New York: Emerald Publishing Company.Find this resource:

                                                Leonardsen, D. (2006). “Crime in Japan: Paradise Lost?,” Journal of Scandinavian Studies in Criminology and Crime Prevention 7: 185–210.Find this resource:

                                                  Levitt, S.D. (1996). “The Effect of Prison Population Size on Crime Rates: Evidence from Prison Overcrowding Litigation,” Quarterly Journal of Economics 111: 319–51.Find this resource:

                                                    Levitt, S.D. (2002). “Using Electoral Cycles in Police Hiring to Estimate the Effects of Police on Crime: Reply,” American Economic Review 92: 1244–50.Find this resource:

                                                      Levitt, S.D. (2004). “Understanding Why Crime Fell in the 1990s: Four Factors that Explain the Decline and Six That Do Not,” Journal of Economic Perspectives 18: 163–90.Find this resource:

                                                        Liedka, R.V., Piehl, A.M., and Useem, B. (2006). “The Crime-Control Effects of Incarceration: Does Scale Matter?,” Criminology & Public Policy 5: 245–76.Find this resource:

                                                          Morenoff, J., Sampson, R.J. and Raudenbush, S. (2001). “Neighborhood Inequality, Collective Efficacy, and the Spatial dynamics of Urban Violence,” Criminology 40: 213–30.Find this resource:

                                                            Parker, R. and Cartmill, R. (1998). “Alcohol and Homicide in the United States 1934–1995—Or One Reason Why U.S. Rates of Violence May Be Going Down,” Journal of Criminal Law and Criminology 88: 1369–98.Find this resource:

                                                              Passel, J. (2006). The Size and Characteristics of the Unauthorized Migrant Population in the United States, Washington, DC: Pew Hispanic Center.Find this resource:

                                                                Paxton, P. (1999). “Is Social Capital Declining in the United States? A Multiple Indicator Assessment,” American Journal of Sociology 105: 88–127.Find this resource:

                                                                  Popkin, S.J. and Cove, E. (2007). Safety Is the Most Important Thing: How HOPE VI Helped Families, Washington, DC: Urban Institute.Find this resource:

                                                                    Popkin, S.J., Gwiasda, V.E., Buron, L., Olson, L.M., and Rosenbaum, D.P. (2000). The Hidden War: Crime and the Tragedy of Public Housing in Chicago, New Brunswick, NJ: Rutgers University Press.Find this resource:

                                                                      Putnam, R.D. (1993). “The Prosperous Community: Social Capital and Community Life,” The American Prospect 13: 35–42.Find this resource:

                                                                        (p. 63) Reyes, J.A.W. (2007). “Environmental Policy as Social Policy? The Impact of Childhood Lead Exposure on Crime,” Berkeley Electronic Journal of Economic Analysis & Policy 7: Article 51, available at 〈〉.Find this resource:

                                                                          Roberts, A. and LaFree, G. (2004). “Explaining Japan's Post-war Violent Crime Trends,” Criminology 42: 179–209.Find this resource:

                                                                            Rosenfeld, R. (2000). “Patterns in Adult Homicide: 1980–1995,” in A. Blumstein and J. Wallman (eds.), The Crime Drop in America, New York: Cambridge University Press.Find this resource:

                                                                              Rosenfeld, R. and Fornango, R. (2008). “The Impact of Economic Conditions on Robbery and Property Crime: The Role of Consumer Sentiment,” Criminology, 45: 735–69.Find this resource:

                                                                                Sampson, R.J. (2004). “Neighbourhood and Community,” New Economy 11: 106–13.Find this resource:

                                                                                  Sampson, R.J., Morenoff, J. D., and Raudenbush, S. (2005). “Social Anatomy of Racial and Ethnic Disparities in Violence,” American Journal of Public Health 95: 224–32.Find this resource:

                                                                                    Sampson, R.J., Raudenbush, S.W., and Earls, F. (1997). “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy,” Science 277: 918–24.Find this resource:

                                                                                      Schwartz, A., Susin, S., and Voicu, I. (2003). “Has Falling Crime Driven New York City's Real Estate Boom?,” Journal of Housing Research 14: 101–35.Find this resource:

                                                                                        Sherman, L. (1990). “Police Crackdowns: Initial and Residual Deterrence,” in M.H. Tonry and N. Morris (eds.), Crime and Justice: A Review of Research, 12, Chicago: University of Chicago Press.Find this resource:

                                                                                          Skogan, W.G. (2006). Police and Community in Chicago: A Tale of Three Cities, New York: Oxford University Press.Find this resource:

                                                                                            Skogan, W.G. (2009). “Policing Immigrant Communities,” in W.F. McDonald (ed.), Immigration, Crime and Justice, London: Emerald Publishing.Find this resource:

                                                                                              Skogan, W. G. and Frydl, K. (eds.) (2004). Fairness and Effectiveness in Policing, Washington, DC: National Academies Press.Find this resource:

                                                                                                Skogan, W.G. and Hartnett, S.M. (1997). Community Policing, Chicago Style, New York: Oxford University Press.Find this resource:

                                                                                                  Spelman, W. (2000). “The Limited Importance of Prison Expansion,” in A. Blumstein and J. Wallman (eds.), The Crime Drop in America, New York: Cambridge University Press.Find this resource:

                                                                                                    Zimring, F.E. and Hawkins, G. (1999). Crime Is Not the Problem: Lethal Violence in America, New York: Oxford University Press.Find this resource:


                                                                                                      (1) This story has been in many places; see Blumstein 2000, or Fagan and Wilkinson (1998). For a data-driven version of the story, see Grogger and Willis 2000.