Age and Crime
AGE AND CRIME
The view that involvement in crime diminishes with age is one of the oldest and most widely accepted in criminology. Beginning with the pioneering research by Adolphe Quetelet in the early nineteenth century, criminological research consistently has confirmed that (the proportion of) the population involved in crime tends to peak in adolescence or early adulthood and then decline with age. This age-crime relationship is remarkably similar across historical periods, geographic locations, and crime types.
That the impact of age on criminal involvement is one of the strongest factors associated with crime has prompted the controversial claim that the age-crime relationship is universal and invariant (Hirschi and Gottfredson). However, considerable variation exists among offenses and across historical periods in specific features of the age-crime relationship (for example, peak age, median age, rate of decline from peak age). A claim of "invariance" in the age-crime relationship therefore overstates the case (Steffensmeier et al., 1989).
Age-crime patterns for the U.S.
The F.B.I.'s Uniform Crime Report (UCR) data, particularly the Crime Index (homicide, robbery, rape, aggravated assault, burglary, larceny-theft, auto theft) document the robustness of the age effect on crime and also reveal a long-term trend toward younger age-crime distributions in more modern times. Today, the peak age (the age group with the highest age-specific arrest rate) is younger than twenty-five for all crimes reported in the F.B.I.'s UCR program except gambling, and rates begin to decline in the teenage years for more than half of the UCR crimes. In fact, even the median age (50 percent of all arrests occurring among younger persons) is younger than thirty for most crimes. The National Crime Victimization Survey (NCVS), self-report studies of juvenile and adult criminality, and interview data from convicted felons corroborate the age-crime patterns found in the UCR data (Steffensmeier and Allan).
Explaining the youthful peak in offending. In a general sense, physical abilities, such as strength, speed, prowess, stamina, and aggression are useful for successful commission of many crimes, for protection, for enforcing contracts, and for recruiting and managing reliable associates (for a review, see Steffensmeier and Allan). Although some crimes are more physically demanding than others, persistent involvement in crime is likely to entail a lifestyle that is physically demanding and dangerous. Declining physical strength and energy may make crime too dangerous or unsuccessful, especially where there are younger or stronger criminal competitors who will not be intimidated, and this might help to explain the very low involvement in crime of small children and the elderly.
However, available evidence on biological aging reveals very little correspondence between physical aging and crime's decline in late adolescence. The research literature on biological aging (see especially Shock) suggests that peak functioning is typically reached between the ages of twenty-five and thirty for physical factors plausibly assumed to affect one's ability to commit crimes (strength, stamina, aerobic capacity, motor control, sensory perception, and speed of movement). Although decline sets in shortly after these peak years, it is very gradual until the early fifties, when the decline becomes more pronounced (Shock). Other commonly mentioned physical variables like testosterone levels peak in late adolescence but then remain at or near peak level until at least the mid-forties. In contrast, the age curves for crimes like robbery and burglary that presuppose the need for physical abilities peak in mid-adolescence and then decline very rapidly. In short, although biological and physiological factors may contribute toward an understanding of the rapid increase in delinquent behavior during adolescence, they cannot by themselves explain the abrupt decline in the age-crime curve following mid-to-late adolescence (which, in particular, is observed in contemporary, postindustrial nations).
A variety of social and cognitive factors can help explain the rapid rise in age-specific rates of offending around mid-adolescence. Teenagers generally lack strong bonds to conventional adult institutions, such as work and family (Warr). At the same time, teens are faced with strong potential rewards for offending: money, status, power, autonomy, identity claims, strong sensate experiences stemming from sex, natural adrenaline highs or highs from illegal substances, and respect from similar peers (Steffensmeier and Allan). Further, their dependent status as juveniles insulates teens from many of the social and legal costs of illegitimate activities, and their stage of cognitive development limits prudence concerning the consequences of their behavior. At the same time, they possess the physical prowess required to commit crimes. Finally, a certain amount of misbehavior is often seen as natural to youth and seen as simply a stage of growing up ( Jolin and Gibbons; Hagan et al.).
For those in late adolescence or early adulthood (roughly ages seventeen to twenty-two, the age group showing the sharpest decline in arrest rates for many crimes), important changes occur in at least six spheres of life (Steffensmeier and Allan):
- Greater access to legitimate sources of material goods and excitement: jobs, credit, alcohol, sex, and so on.
- Age-graded norms: externally, increased expectation of maturity and responsibility; internally, anticipation of assuming adult roles, coupled with reduced subjective acceptance of deviant roles and the threat they pose to entering adult status.
- Peer associations and lifestyle: reduced orientation to same-age-same-sex peers and increased orientation toward persons of the opposite sex or persons who are older or more mature.
- Increased legal and social costs for deviant behavior.
- Patterns of illegitimate opportunities: with the assumption of adult roles, opportunities increase for crimes (for example, gambling, fraud, and employee theft) that are less risky, more lucrative, or less likely to be reflected in official statistics.
- Cognitive and analytical skill development leading to a gradual decline in egocentrism, hedonism, and sense of invincibility; becoming more concerned for others, more accepting of social values, more comfortable in social relations, and more concerned with the meaning of life and their place in things; and seeing their casual delinquencies of youth as childish or foolish.
As young people move into adulthood or anticipate entering it, most find their bonds to conventional society strengthening, with expanded access to work or further education and increased interest in "settling down." Leaving high school, finding employment, going to college, enlisting in the military, and getting married all tend to increase informal social controls and integration into conventional society (Steffensmeier et al., 1989). In addition, early adulthood typically involves a change in peer associations and lifestyle routines that diminish the opportunities for committing these offenses (Warr). Last, at the same time when informal sanctions for law violations are increasing, potential legal sanctions increase substantially.
Variations in the age curve
Although crime tends to decline with age, substantial variation can be found in the parameters of the age-crime curve (such as peak age, median age, and rate of decline from peak age). "Flatter" age curves (i.e., those with an older peak age and/or a slower decline in offending rates among older age groups) are associated with several circumstances:
- Types of crime for which illegitimate opportunities increase rather than diminish with age.
- Population groups for whom illegitimate opportunities and integration into adult society do not markedly increase with age (i.e., during young adulthood).
- Cultures and historical periods in which youth have greater access to legitimate opportunities and integration into adult society.
Crime types. The offenses that show the youngest peaks and sharpest declines are crimes that fit the low-yield, criminal mischief, "hell-raising" category: vandalism, petty theft, robbery, arson, auto theft, burglary, and liquor law and drug violations. Personal crimes like aggravated assault and homicide tend to have somewhat "older" age distributions (median ages in the late twenties), as do some of the public order offenses, public drunkenness, driving under the influence, and certain of the property crimes that juveniles have less opportunity to commit, like embezzlement, fraud, and gambling (median ages in late twenties or thirties). However, even these older age-distributions (e.g., fraud) have shifted toward younger peak ages in recent years (see below).
Those offenses with flatter age curves are often those for which the structure of illegitimate opportunities increases rather than disappears with age. For example, some opportunities for fraud exist for young people (such as falsification of identification to purchase alcohol or gain entry to "adult" establishments), but since they are too young to obtain credit, they lack the opportunities for common frauds such as passing bad checks, defrauding an innkeeper, or credit card forgery. Similarly, young people have more opportunities for some kinds of violence (for example, street fights or gang violence) but less opportunity for other kinds of violence (for example, spousal violence).
Older people may also shift to less visible criminal roles such as bookie or fence. Or as a spin-off of legitimate roles, they may commit surreptitious crimes or crimes that, if discovered, are unlikely to be reported to the authorities, such as embezzlement, stock fraud, bribery, or price-fixing. Unfortunately, we know relatively little about the age distribution of persons who commit these and related lucrative crimes, but the fragmentary evidence that does exist suggests that they are likely to be middle age or older (Shapiro; Pennsylvania Crime Commission). Evidence also suggests that the age curves for lucrative crimes in the underworld like racketeering or loansharking not only peak much later but tend to decline more slowly with age (Steffensmeier and Allan; Steffensmeier).
Still less is known of the age distribution of "respectable" or upperworld offenders who commit lucrative business crimes like fraud, price-fixing, or bribery, since such data are not plentiful. However, data from New York Times articles on profitable business crimes (those involving gains of $25,000 or more) during the 1987–1990 period reveals a preponderance of middle-aged or older offenders, with a modal age between forty and fifty (Steffensmeier and Allan).
Minority differences. For black inner-city youths, the problems of youth described above are compounded by persistent racial discrimination and blocked conventional opportunity (Wilson, 1987; 1996). As inner-city blacks move into young adulthood, they continue to experience limited access to high quality adult jobs and are more likely to associate primarily with same-sex peers. As UCR data show, adult offending levels among blacks continue at higher levels than among whites, and the proportion of total black crime that is committed by black adults is greater than the proportion of total white crime that is committed by white adults (Steffensmeier and Allan). Arrest statistics for homicide/robbery from California further document the flatter age-crime curves among blacks than whites.
Cross-cultural and historical differences. In small preindustrial societies, the passage to adult status is relatively simple and continuous. Formal "rites of passage" at relatively early ages avoid much of the status ambiguity and role conflict that torment modern adolescents in the developed world. Youths begin to assume responsible and economically productive roles well before they reach full physical maturity. It is not surprising, therefore, to find that such societies and time periods have significantly flatter and less skewed age-crime patterns (for a review, see Steffensmeier et al., 1989).
Much the same is true for earlier periods in the history of the United States and other industrial nations, when farm youth were crucial for harvesting crops and working-class children were expected to leave school at an early age and do their part in helping to support their families. By contrast, "The typical job a teenager can get today provides neither the self-pride of economic independence not the socializing benefits of working alongside adult mentors. . .. Work relations seem to have been critical experiences for the socialization of many young men in the past. Such jobs integrated youths into adult society . . . instead of segregating them in a separate peer culture" (Coontz, p. 29).
Although youth has always been seen as a turbulent time, social processes associated with the coming of industrialization and the postindustrial age have aggravated the stresses of adolescence, resulting in increased levels of juvenile criminality in recent decades. The structure of modern societies, therefore, encourages crime and delinquency among the young because these societies "lack institutional procedures for moving people smoothly from protected childhood to autonomous adulthood" (Nettler, p. 241).
Unfortunately, reliable age statistics on criminal involvement are not available over extended historical periods. Nonetheless, we can compare age-crime distributions over the past sixty years or so in the United States and also compare these to early nineteenth-century age-crime distributions reported in Quetelet's pioneering study (see also Monkkonen). The age-crime plots for homicide clearly document the trend toward younger age distributions and younger peak ages.
The shift toward a greater concentration of offending among the young may be due partly to change in law enforcement procedures and data collection. Nevertheless, the likelihood that real changes have in fact occurred is supported by the consistency of the changes from 1830 to 1940 to 1980, and continuing into the mid-1990s. Support for the conclusion that real change has taken place over the past century also is found in the age breakdown of U.S. prisoner statistics covering the years 1890 to 1980 (Steffensmeier et al., 1989). As with the UCR statistics, the prison statistics show that age curves are more peaked today than a century ago and that changes in the age-crime curve are gradual and can be detected only when a sufficiently large time frame is used. Moreover, research shows that more-recent birth cohorts of juveniles are more violent than ones in the past (Tracey et al., 1990; Shannon, 1988).
Together, these findings are consistent with the view that contemporary teenagers in industrialized nations are subject to greater status anxiety than in previous periods of history and that the transition from adolescence to adulthood is more turbulent now than in the past (Greenberg, 1979, 1982; Glaser). In comparison to earlier eras, youths have had less access to responsible family roles, valued economic activity, and participation in community affairs (Greenberg 1982). This generational isolation has fostered adolescent subcultures oriented toward consumption and hedonistic pursuits (Hagen et al., 1998; Hagan). The weakened social bonds and reduced access to valued adult roles, along with accentuated subcultural influences, all combine to increase situationally induced pressures to obtain valued goods; display strength, daring, or loyalty to peers; or simply to "get kicks" (Hagan et al., 1998; Steffensmeier et al., 1989).
Sex differences in the age-crime relationship. Although age-crime parameters differ as described above, there appears to be considerable similarity in the age-crime relationship between males and females (Steffensmeier and Streifel). UCR arrest statistics from the 1930s to the 1990s show that the age curves of male and female offenders are very similar within any given period and across all offenses, with the exception of prostitution. To the extent that age differences between the sexes exist, the tendency is for somewhat lower peak ages of offending among females—apparently because of their earlier physical maturity and the likelihood that young adolescent females might date and associate with older delinquent male peers. But overall, although male levels of offending are always higher than female levels at every age and for virtually all offenses, the female-to-male ratio remains fairly constant across the life span (Steffensmeier and Streifel). Also, the trend toward younger and more peaked age-crime distributions holds for both sexes.
The single major difference in the age curves of males and females is for prostitution (and to some extent vagrancy, often a euphemism for prostitution in the case of female arrestees), with females having a much greater concentration of arrests among the young. Although this difference may be due in part to more stringent enforcement of prostitution statutes when young females are involved, the younger and more peaked female age curve is also a function of the extent to which opportunity structures for sexual misbehaviors differ between males and females. Clearly, sexual attractiveness and the marketability of sexual services are strongly linked to both age and gender: Older women become less able to market sexual services, whereas older men can continue to purchase sexual services from young females or from young males (Steffensmeier and Streifel).
Variations in criminal careers
The youthful peak and rapid drop-off in offending that constitutes the most common societal pattern for conventional crimes is actually but one of a number of patterns identified when criminal careers are tracked for individual offenders (Jolin and Gibbons).
"Aging out" of crime. Research suggests that exiting from a criminal career requires the acquisition of meaningful bonds to conventional adult individuals and institutions (Sampson and Laub; Shover; Steffensmeier and Allan; Warr). One important tie to the conventional order is a job that seems to have the potential for advancement and that is seen as meaningful and economically rewarding. A good job shifts a criminal's attention from the present to the future and provides a solid basis for the construction of a noncriminal identity. It also alters an individual's daily routine in ways that make crime seem less likely. Other bonds that may lead people away from crime include involvement in religion, sports, hobbies, or other activities.
The development of conventional social bonds may be coupled with burnout or a belated deterrent effect as offenders grow tired of the hassles of repeated involvement with the criminal justice system, and the hardships of a life of crime. They may also have experienced a long prison sentence that jolts them into quitting or that entails a loss of street contacts which makes the successful continuation of a criminal career difficult. Or offenders may develop a fear of dying alone in prison, especially since repeated convictions yield longer sentences. Still other offenders may quit or "slow down" as they find their abilities and efficiency declining with increasing age, loss of "nerve," or sustained narcotics or alcohol use (Adler and Adler; Shover; Steffensmeier).
Older criminals. Older offenders fall into two categories: (1) those whose first criminal involvement occurs relatively late in life (particularly in shoplifting, homicide, and alcohol-related offenses); and (2) those who started crime at an early age and continue their involvement into their forties and fifties and beyond. What evidence is available on first-time older offenders suggests that situational stress and lack of alternative opportunities play a primary role. The unanticipated loss of one's job or other disruptions of social ties can push some individuals into their first law violation at any age (Jolin and Gibbons).
Older offenders who persist in crime are more likely to belong to the criminal under-world. These are individuals who are relatively successful in their criminal activities or who are extensively integrated into subcultural or family enterprises. They seem to receive relational and psychic rewards (e.g., pride in their expertise) as well as monetary rewards from lawbreaking and, as a result, see no need to withdraw from lawbreaking (Steffensmeier). Alternatively, such offenders may "shift and oscillate" back and forth between conventionality and lawbreaking, depending on shifting life circumstances and situational inducements to offend (Adler and Adler). These older offenders are also unlikely to see many meaningful opportunities for themselves in the conventional or law-abiding world. Consequently, "the straight life" may have little to offer successful criminals, who will be more likely to persist in their criminality for an extended period. But they, too, may slow down eventually as they grow tired of the cumulative aggravations and risks of criminal involvement, or as they encounter the diminishing capacities associated with the aging process.
Effects of age structure on crime rates
The dramatically higher age-specific offending rates for young people suggest that shifts in the age-composition of the population could produce sizable changes in societal crime rates. The so-called baby-boom generation born between the end of World War II and the early 1960s brought a large, steady increase in the proportion of the population aged twelve to twenty-five—the most crime-prone age group—during the 1960s and 1970s, a period when the nation's crime rate was also increasing steadily. Ferdinand found that about 50 percent of the increase in the index crime rate during the 1960s could be attributed to population shifts such as the baby-boom generation's movement into the crime-prone years. Similarly, Steffensmeier and Harer found that virtually all the reported decreases in the UCR and NCVS Index crime rates during the early 1980s could be attributed to the declining proportion of teenagers in the population—that is, to a "baby-bust" effect.
More recently, Steffensmeier and Harer reported that the large impact of age-composition on crime rates during the 1980s had diminished during the 1990s, and that the broad decline in both the UCR and NCVS crime rates since 1992 (the years of the Clinton presidency) cannot be solely attributed to changes in population age composition. One explanation of the recent downtrend has attributed the decline to dramatic increases in incarceration rates that presumably incapacitate or prevent crimes by locking up high-frequency offenders who commit a disproportionate amount of all crimes. However, the rise in incarceration rates extends backwards to at least the late 1970s, and therefore considerably predates the 1990s drop in crime. Therefore it appears unlikely that higher imprisonment rates explain much, if any, of the recent drop in crime—just as they do not account for the rise in crime in the late 1980s.
Alternative explanations for recent downward trends in crime rates include the strong economy and low unemployment of the 1990s, an abatement of the crack epidemic of the late 1980s, and the wide variety of community-level criminal justice initiatives undertaken in the 1990s, such as Operation Weed and Seed, Pulling America's Communities Together, SafeFutures, and community policing. Also, Steffensmeier and Harer speculate that offenders may be shifting from risky, low-return offenses like burglary (also robbery) to others that are more lucrative (drug dealing) or less risky (fraud).
Criminologists have long recognized that age is a very robust predictor of crime, both in the aggregate and for individuals. The most common finding across countries, groups, and historical periods shows that crime tends to be a young persons' activity. However, the age-crime relationship is not invariant, and in fact varies in its specific features according to crime types, the structural position of groups, and historical and cultural contexts. On the other hand, the age-crime relationship seems to be fairly similar for males and females. Finally, although they constitute a very small group, relatively little is known about older chronic offenders. Clearly, the structure, dynamics, and contexts of offending among older individuals is a rich topic for further research.
See also Criminal Careers; Excuse: Infancy; Family Relationships and Crime; Juvenile and Youth Gangs; Juvenile Violent Offenders; Police: Handling of Juveniles; Prevention: Juveniles as Potential Offenders.
Adler, Patricia, and Adler, Peter. "Shifts and Oscillations in Deviant Careers: The Case of Upper-Level Drug Dealers and Smugglers." Social Problems 31 (1983): 195–207.
Coontz, Stephanie. The Way We Really Are: Coming to Terms with America's Changing Families. New York: Basic Books, 1997.
Ferdinand, Theodore. "Demographic Shifts and Criminality: An Inquiry." British Journal of Criminology 10 (1970): 169–175.
Glaser, Daniel. Crime in Our Changing Society. New York: Holt, Rinehart, and Winston, 1978.
Greenberg, David. "Delinquency and the Age Structure of Society." Contemporary Crisis 1 (1979): 66–86.
——. "Age and Crime." In Encyclopedia of Crime and Justice, vol. 1. Edited by Sanford Kadish. New York: Macmillan, 1982. Pages 30–35.
Hagan, John. "Destiny and Drift: Subcultural Preferences, Status Attainments, and the Risks and Rewards of Youth." American Sociological Review 56 (1991): 567–581.
Hagan, John; Heffler, Gerd; Classen, Gabriele; Boehnke, Klaus; and Merkens, Hans. "Subterranean Sources of Subcultural Delinquency Beyond the American Dream." Criminology 36 (1998): 309–342.
Hirschi, Travis, and Gottfredson, Michael. "Age and the Explanation of Crime." American Journal of Sociology 89 (1983): 522–584.
Jolin, Annette, and Gibbons, Don. "Age Patterns in Criminal Involvement." International Journal of Offender Therapy and Comparative Criminology 31 (1987): 237–260.
Monkkonen, Eric. "New York City Offender Ages: How Variable Over Time." Homicide Studies 3 (1999): 256–270.
Nettler, Gwyn. Explaining Crime. New York: McGraw-Hill, 1978.
Pennsylvania Crime Commission. 1990 Report—Organized Crime in America: A Decade of Change. Commonwealth of Pennsylvania, Conshohoken, PA., 1991.
Sampson, Robert, and Laub, John. Crime in the Making: Pathways and Turning Points through Life. Cambridge, Mass.: Harvard University Press, 1993.
Shannon, Lyle. Criminal Career Continuity: Its Social Context. New York: Human Sciences Press, 1988.
Shock, Nathan. Normal Human Aging: The Baltimore Longitudinal Study of Aging. Washington, D.C.: U.S. Government Printing Office, 1984.
Shover, Neal. "The Later Stages of Ordinary Property Offender Careers." Social Problems 30 (1983): 208–218.
Steffensmeier, Darrell. The Fence: In the Shadow of Two Worlds. Totowa, N.J.: Rowman & Littlefield, 1986.
Steffensmeier, Darrell, and Allan, Emilie. "Criminal Behavior: Gender and Age." In Criminology: A Contemporary Handbook. Edited by Joseph Sheley. Belmont, Calif.: Wadsworth, 1995. Pages 83–114.
Steffensmeier, Darrell; Allan, Emilie; Harer, MILES; and Streifel, Cathy. "Age and the Distribution of Crime." American Journal of Sociology 94 (1989): 803–831.
Steffensmeier, Darrell, and Harer, Miles. "Did Crime Rise or Fall During the Reagan Presidency? The Effects of an 'Aging' U.S. Population on the Nation's Crime Rate." Journal of Research in Crime and Delinquency 28 (1991): 330–359.
——. "Making Sense of a Recent U.S. Crime Trends, 1980–98: Age Composition Effects and Other Explanations." Journal of Research in Crime & Delinquency 36 (1999): 235–274.
Steffensmeier, Darrell, and Streifel, Cathy. "Age, Gender, and Crime across Three Historical Periods: 1935, 1960, and 1985." Social Forces 69 (1991): 869–894.
Tracey, Paul; Wolfgang, Marvin; and Figlio, Robert. Delinquency Careers in Two Birth Cohorts. New York: Plenum, 1990.
Ulmer, Jeffery, and Spencer, J. William. "The Contributions of an Interactionist Approach to Research and Theory on Criminal Careers." Theoretical Criminology 3 (1999): 95–124.
Warr, Mark. "Life-Course Transitions and Desistance from Crime." Criminology 36 (1998): 183–216.
——. When Work Disappears. Cambridge, Mass.: Harvard University Press, 1996.