U.S. flag

An official website of the United States government, Department of Justice.

The Long View of Crime

Information collected over time provides valuable insight into criminal behavior.
National Institute of Justice Journal
Date Published
April 14, 2010

Many findings of longitudinal studies run counter to long-held beliefs about adolescent offending.

For a new edited volume, criminologist Akiva Lieberman assembled an impressive group of scholars to present conclusions from more than 60 studies on crime and delinquency in adolescence. He collected their essays into a book titled The Long View of Crime: A Synthesis of Longitudinal Research. Many of the findings challenge the conventional wisdom and could have significant implications for future policy and practice.

Longitudinal studies, which follow people for extended periods, are valuable because they provide information about offending behavior over time. By contrast, cross-sectional studies provide information about one particular period. It is the difference between a panoramic view and a snapshot. The wide-angle lens of longitudinal research is a powerful tool for sorting out some of the chicken-and-egg, "which came first" issues at the heart of criminal research.

Street Gangs: Why Do Gang Members Commit Crime?

No one disputes that gang members commit more crime than nonmembers. Two schools of thought have emerged to explain why the crime rate is higher among gang members. One theory, the "selection model," suggests that adolescents who are already predisposed toward delinquency and violence are the ones most likely to join gangs. The opposing theory, the "facilitation model," assumes that gang members are no more disposed toward delinquency and violence than others are and would not contribute to higher crime rates if they did not join a gang. However, when they do join a gang, peer pressures promote their increased involvement in delinquency.

Some longitudinal studies have tested these two competing approaches, and the uniformity of results is impressive. The studies show no evidence to support a pure selection model. The weight of the evidence favors the facilitation model, suggesting that street gangs promote delinquent behavior. A corollary finding that further supports the facilitation model is that delinquency (except that related to drug sales) typically declines after the member leaves the gang.

Longitudinal studies have exposed some other possible misconceptions about gang-related behavior. Several factors thought to be associated with gang membership, such as family poverty, family structure, low self-esteem and neighborhood crime, were not supported by the empirical evidence. Studies show that gang membership in "emerging gang" cities is transitory, typically lasting no more than a year. Traditional gang cities such as LA and Chicago were not included in the set of longitudinal studies and their gangs may function differently, with reports of long-duration and multigenerational gang membership.

These findings could have important implications for designing gang prevention and intervention programs. Delinquent behavior may stem more from gang membership than from any delinquent leanings of the gang members themselves.

Arrest and Sanctions: Do They Deter Delinquency or Make It Worse?

Two conflicting views have emerged about the effect of arrest on delinquent individuals. One is that arrest should deter or even end offending behavior because it makes them understand that their behavior is socially disapproved of and that they could be arrested again if they do not reform. The opposite view is that arrest increases offending behavior because arrestees begin to view themselves as bad people, which leads them to continue committing crimes. A similar debate rages about the deterrent effect of sanctions that might be imposed after arrest, such as fines, community service, making restitution, attending treatment programs or imprisonment.

The data from longitudinal studies on this question are robust and consistent. More than a dozen studies found that people who have been arrested are at least as likely to be arrested in the future as those who have not. Thus, rather than being a deterrent, arrest resulted in similar or higher rates of later offending. Fourteen studies that examined the effect of sanctions uniformly found that sanctions either had no effect on or increased later offending. Interestingly, as the severity of sanctions increased, later offending was flat or increased. In addition, several studies have suggested that arrest and sanctions have a negative effect on later employment and increase juveniles' chances of becoming high school dropouts.

Several factors may help to explain why arrest and sanctions do not have the expected deterrent effect. On release, most individuals return to the same risky environment that influenced their delinquency. They already have a well-established history of offending before their first arrest, and some have psychological characteristics that decrease their susceptibility to influence from a prior arrest.

The findings from these studies challenge some deeply entrenched notions about the deterrent effect of arrest and sanctions on offending adolescents. More longitudinal studies that employ samples from general populations and examine different kinds of persons who commit offenses (e.g., classified by age, sex, social class or stage of delinquent career) are needed to inform policy discussions about the possible benefit of more lenient interventions.

Hard Work: How Does Adolescent Employment Affect Offending?

Employment has long been viewed as a solution to the problems of crime and delinquency. However, studies have shown that the relationship between work and crime is far more complex than originally thought. Longitudinal studies now show that employment effects are likely to depend on the age of the person, and the importance of work varies for different groups (e.g., at-risk adolescents as opposed to older people who at one point were convicted of a crime) at different life stages.

One firmly fixed finding over the years has been that intensive work by adolescents (i.e., 20 or more hours per week) increases delinquent behavior. Researchers believed that intensive work made youths less engaged in school, less supervised by their parents and more likely to meet delinquent peers. However, even this formerly secure belief has been challenged by recent longitudinal studies. They suggest that sample selection — that is, the characteristics of the adolescents studied — may have skewed the findings. Perhaps adolescents who work intensively were already poor students, unsupervised at home and prone to harmful behavior and were therefore likely to be delinquent even without work.

Longitudinal studies also show that employment quality may be more important for crime reduction than the simple presence or absence of a job. Those at high risk for crime have many opportunities to earn money illegally. Recent longitudinal studies on adolescent employment and delinquency suggest the old rules about how jobs affect delinquency may be too simplistic.

Growing Up: Does Moving to Adult Roles Affect Delinquent Behavior?

One observation that has stood the test of time is that the prevalence of criminal offenses rises during adolescence and decreases in the early twenties. This age-related crime curve has led researchers to examine the transition to adult roles, such as marriage, cohabitation and parenthood, as a potential explanation for giving up delinquent ways. Theories abound about why taking on adult roles might reduce offending. Some suggest the transitions themselves are the cause for the change in offending behavior because they reorganize adolescents' lives in ways that limit unstructured socializing (i.e., time hanging out with deviant friends). Others postulate that a "cognitive shift" that precedes the transition, rather than the transition itself, causes the change in offending. They theorize that adolescents first must come to view their deviant lifestyles as undesirable. Only then can they embrace role transitions that will create conventional lifestyles.

Of all the role transitions examined, marriage most effectively and consistently reduces deviance. What is not yet clear is why this is so. Citing one study that showed that marriages to criminal spouses might increase offending, some have questioned whether the strength of the marital bond is as important as the spouses' own characteristics and conventionality.

Because studies show that marriage reduces offending, one might expect to see a similar effect from stable romantic partnerships. However, studies have found exactly the opposite: Individuals commit more crimes while living with romantic partners. One study suggests that this might be because adults who were delinquent youths choose antisocial romantic partners, which contributes to continued offending in young adulthood.

Recent longitudinal studies have also examined the "parenthood effect." Many theorists have expected a negative effect from parenthood on crime, either because new parents become invested in their children or, as with the marriage effect, because the demands of parenthood reduce unstructured socializing. However, the few studies that have looked at this role transition have not found that having children reduces offending. Although people intentionally enter marriage, the same is not always true of parenthood.

The Long View Ahead

Many of the longitudinal studies described in The Long View of Crime shed new light on or even skewer time-honored criminological theories. These findings may provide an impetus for further analysis of existing data. They may also spark a new wave of longitudinal studies that incorporate both advances in statistical methods and innovative designs. Structuring longitudinal studies to advance knowledge about the causes of delinquency could lead to a clearer understanding of the explanation, prevention and treatment of offending and antisocial behavior and more targeted policies to address them.

For More Information

  • Liberman, A.M., The Long View of Crime: A Synthesis of Longitudinal Research, Washington, D.C.: Springer, 2008.

About This Article

This article appeared in NIJ Journal Issue 265, April 2010.

Date Published: April 14, 2010