Drug Use and Violent Crime Among Adolescents, Essay Example

Introduction

The relationship between crime and the use of alcohol and other drugs has received a great deal of attention in previous research (Banay, 1942; O’Donnell, 1966; Harper, 1976; Shanok, Pincus, & Glaser, 1979; McCord, 1981; Gary, 1980; Simonds & Kashani, 1980; Dawkins & Dawkins, 1983; Clayton & Tuchfeld, 1982; Lewis, Cloninger, & Pais, 1983; Anglin & Speckart, 1988; Collins, 1988; Wieczorek, Welte, & Abel, 1990). Studies have generally revealed a positive association between criminal behavior and abuse of alcohol and other drugs. An important aspect of this association has been the degree to which violent crimes are linked to the abuse of alcohol and other drugs. Among delinquents, the familiar categorization of “person” offenses versus “property” offenses (O’Donnell, 1966) further implies that differences exist between two types of offenders in terms of the nature of alcohol and drug involvement. For example, it is commonly assumed that violent crimes (offenses against persons) are more likely to be committed by persons who abuse “hard” drugs (e.g., heroin and cocaine), while less serious crimes (offenses against property) are committed by users of alcohol and marijuana. However, the research literature reveals conflicting evidence. For example, Simonds and Kashani (1980) found that the number of drugs abused by an offender rather than the specific type of drug was most predictive of being a member of the “person” offending group. Clayton and Tuchfeld (1982) and Anglin and Speckart (1988) provided further evidence that heavy drug users are more likely to be involved in violent criminal activity. In contrast, substantial research suggests that alcohol abuse is the most important substance-related factor in homicides and other types of violent crime (Harper, 1976; Harper & Dawkins, 1977; Gary, 1980; Dawkins, 1980; Harper, 1980; Dawkins & Dawkins, 1983; Collins, 1988; Wieczorek, Welte, & Abel, 1990). For example, Wieczorek, Welte, and Abel (1990) found that homicide offenders tend to drink heavily prior to committing their crimes.

In an effort to resolve the inconsistent findings of previous research, Yu and Williford (1994) examined and found support for a model which specifies the relationships among criminality and the use of alcohol and other drugs in terms of a developmental sequence. According to Yu and Williford, who studied individuals in jails, drinking-driver programs, and alcoholism treatment centers, the early onset of legal drug use (e.g., alcohol and cigarettes) induces the onset of illicit drug use (e.g., marijuana and cocaine) and eventually leads to involvement in criminal activity, including offenses against persons and property. Further, the use of alcohol and cocaine are mainly responsible for the increase in violent criminal behavior (Yu & Williford, 1994). Despite the model’s utility, these authors suggested the need for future research to focus on the direct relationship between specific types of crime and specific types of drugs. In addition, previous studies have been restricted to adult samples and a range of offenses too limited to specify types of violent and nonviolent crimes (for exceptions see: Tinklenberg & Woodrow (1974) and Simonds & Kashani (1980).

The purpose of the present study was to address the shortcomings of earlier studies and extend the research by examining the relationship between specific categories of alcohol and drug use and specific types of criminal offenses committed by adolescents, with a focus on the distinction between violent and nonviolent offenses. The importance of alcohol and other drug use relative to conventional criminal history factors as correlates of crime among adolescent offenders was also assessed.

Method

Data were collected from adolescent offenders at a juvenile training school in a northeastern state. The school included 416 incarcerated juvenile delinquents, and is the only facility in the state for adjudicated juvenile offenders. Questionnaires were administered to 342 (82%) of the offenders. The nonparticipation rate of 18% was due to refusals (6%), transfers (3%), and such administrative matters as court appearances, leaves, and work releases (9%). The present analysis includes only the male offenders (N = 312).

The variables included in the analysis were operationally defined as follows: Type of criminal offense variables were based on responses to the question, “One year before coming to the training school, how many times had you done the following?” Respondents were then given a list of twenty-one delinquency-related offenses. For each offense, the possible response categories were 5 = “five or more,” 4 = “three or four times,” 3 = “twice,” 2 = “once,” and 1 = “never.” The specific types of offenses included theft, general trouble with police, serious fights, alcohol abuse, trespassing, gang fighting, serious injury to a victim, stealing automobiles, stealing auto parts, parental problems, weapons used in crime, extortion, vandalism, petty auto theft, physical attack of instructor/supervisor, running-away, fights with father and mother, and arson.

Alcohol and other drug use variables were based on responses to questions related to frequency of use. Respondents were asked, “One year before coming to the training school, how often had you consumed alcoholic beverages including liquor, beer or wine?” Possible responses were “6 – “nearly every day,” 5 = “once or twice a week,” 4 = “once or twice a month,” 3 = “three to ten times a year,” 2 = “once or twice a year,” and 1 = “never.” The same question was asked two additional times substituting “used marijuana” and “used heroin” for “consumed alcoholic beverages.” The same response categories were used.

Criminal history variables were the number of self-reported previous arrests, convictions, and sentences.

Race was coded as a set of two dummy variables where 1 = white and 0 = not white; and 1 = black and 0 = non-black. The Hispanic category of the original variable served as the suppressed category in the dummy coding scheme.

Results And Discussion

The analysis involved (1) assessing the extent to which alcohol/drug use is associated with various types of crimes, with particular focus on violence-related criminal activity, and (2) examining the influence of alcohol/drug use as a predictor of criminal behavior relative to other predictors including race and criminal history variables. In the first part of the analysis, zero-order correlational techniques were employed (Pearson product-moment correlation coefficients) to describe the associations between alcohol/drug use variables and the criminal offense variables. In the second part of the analysis, multiple regression was used to examine the impact of the predictor variables on the various types of criminal activities.

Alcohol/Drug Use as Correlates of Violent and Nonviolent Acts

Among the three types of substances (alcohol, marijuana, and heroin), the strongest associations and the most consistent patterns were between alcohol use and criminal offenses. (See Table 1.) Alcohol use is significantly related to twenty of the twenty-one offenses (the exception being theft under $50.00; r = .08). Further, among the strongest associations were several where violence was either directly involved (serious injury to victims (r = .32); physical attack on instructor (r = .22); fight with father (r = .35) or may be implied (serious fights (r = .28), and arson (r = .27). Marijuana use was significantly related to twelve of the twenty-one offenses and is particularly important as a correlate of general trouble with police (r = .29), gang fighting (r = .29) and use of a weapon in the crime (r = .24). Several nonviolent, minor offenses are also strongly associated with marijuana use including store theft (r = .20), theft under $50.00 (r = .26), and petty theft from autos (r = .23). While heroin use is significantly related to twelve of the twenty-one offenses, the magnitude of the correlation coefficients tended to be lower than those reported for the associations between alcohol/marijuana use and criminal offenses. It should be noted that overall the level of heroin use (Mean = 1.61, SD = 1.41) is much lower than the use of alcohol (Mean = 5.04, SD = 1.55) or marijuana (Mean = 5.09, SD = 1.63).

Based on the correlational analysis, the findings indicate that alcohol use is more important as a correlate of criminal offenses, including violence-related and serious property offenses than the other substance use variables (marijuana and heroin). While marijuana use is also an important correlate, it tended to be more strongly associated with gang-related violence and less serious property offenses. Although heroin use is a significant correlate of selected offenses, in relative terms it is less important than alcohol and marijuana use. These findings, therefore, reflect the generally widespread acceptance of alcohol use among adolescents and the potential for alcohol use to play an important role in delinquent activities of a violent and nonviolent nature. The findings also suggest that marijuana use may be a part of delinquent subcultures, especially in terms of its link to gang-related violence. While these findings are revealing, they are based on bivariate associations which do not take into account control variables which may explain away these relationships.

Relative Importance of Substance Use in Predicting Offenses

Multiple regression analysis was employed to examine two issues: (1) Does substance use predict crime over and above criminal records (i.e., previous arrests, previous convictions, and previous sentences), and racial identity? (2) Are specific substance use variables (i.e., alcohol, marijuana or heroin) better predictors for certain types of offenses? The results of the multiple regression analyses are presented in Table 2.

The issue of whether substance use predicts criminal offenses over and above criminal record and racial identity can be assessed by examining incremental change in variance “explained” ([R.sup.2]) by the inclusion of three sets of predictors in the regression equation. The specific variables in each set include black and white dummy variables in the racial identity set; and alcohol, marijuana and heroin use in the substance use set. Each offense was regressed on these three sets of predictors with the predictor sets entered into the equation in a forward, stepwise fashion. The variance attributed to the first set of predictors entered into the equation is unique for that set. When the criminal history variables are entered on the second step, the incremental change in [R.sup.2] can be assessed and the proportion of variance in the specific offenses uniquely explained by the criminal history variables can be derived. This same procedure was used for the set of substance use variables entered in the third stage of the process to yield the unique variance accounted for by the substance use variables.

The results show twelve offenses where the set of criminal history variables are the most important predictors (i.e., theft from store, theft under $50.00, trespassing, serious injury to victim, grand theft auto, weapon used in crime, vandalism, hit instructor/supervisor, runaway, arson, and fight with mother); six offenses where the set of substance use variables are the most important predictors (i.e., general trouble with police, alcohol abuse, gang fights, theft of parts from auto, extortion, and petty theft from auto); and two offenses where the set of racial identity factors accounted for the greatest proportion of unique variance explained (i.e., problems with parents and fights with father). In terms of offenses of a direct violent nature, criminal history variables counted for 28% of the variance in the serious injury to victim variable, while substance use contributed 5%, and race added 2%. For fights with father, criminal history variables contributed 21%, substance use added 6%, and race 6%. However, for gang fights, substance use contributed 10%, with criminal history and race adding 1% each to the overall explained variance. Substance use was also a relatively stronger predictor of general trouble with police where violence may be indirectly involved, contributing 8% to the overall explained variance versus 4% and 3% for criminal history and race, respectively. For problems with parents, substance use was the least important predictor, contributing 6% to the explained variance, with race accounting for 18% and criminal history accounting for 15%.

These findings indicate that when other factors are considered, substance use is a relatively strong predictor for only five of the twenty-one offenses included in this analysis. Criminal history variables are the most powerful predictors of the majority of the offenses (13), while race is the most important predictor of three of the offenses. However, it should be noted that substance use added to the explanation of variance in some of the instances where it was not the most powerful predictor. This suggests that the role of substance use in both violent and nonviolent offenses should not be underestimated.

In addressing the second issue outlined for this part of the analysis (the relative importance of specific substances in predicting various offenses), standardized regression coefficients (betas) were compared for the specific predictor variables in order to assess their relative effect in accounting for each offense. The results show that for many of the offenses that are directly or indirectly related to violence, specific criminal history variables tend to be relatively more important than specific substance use variables. For example, the standardized regression coefficients show that relative to the seven other predictors, previous arrest has the strongest effect within each of five equations for predicting specific offenses (including theft from store, theft over $50.00, trespassing, and vandalism); number of convictions has the strongest effect relative to the other independent variables in predicting serious injury to victim, use of a weapon in crimes, and fights with mother; and number of sentences has the strongest relative effect on only one offense (theft of parts from auto), while alcohol use is the most significant predictor for three offenses (serious fights, alcohol abuse, and extortion) with only one of the offenses being violence-related (serious fights). Marijuana use has the strongest effect relative to the other predictors on general trouble with police and gang fights.

While fewer substance use variables exerted the strongest effect on specific offenses relative to criminal history variables, in several instances substance use continued to be a significant factor although race and criminal history were more powerful predictors. For example, while the number of arrests and convictions had the strongest relative influence on the offenses involving serious injury to victims (betas = .24 and .25, respectively), the effect of alcohol use (beta = .18) was also substantial (albeit lower than arrests and convictions). The effect of each of the three variables (arrests, convictions, and alcohol use) was statistically significant (p [less than] .05). This type of pattern where substance use variables continue to be significant although racial identity or specific criminal history variables produce an equally strong or stronger effect can be found for several other offenses, including theft from store, theft over $50.00, problems with parents, use of weapon in crime, vandalism, petty theft from autos, hit instructor/supervisor, fights with father, and arson.

These findings, therefore, indicate that criminal history variables are relatively more important than race and substance use in predicting a wide range of offenses. Nevertheless, substance use variables, especially alcohol and marijuana use, are important in terms of their effect on some offenses – effects which persist after criminal history and race are introduced into the analysis. However, no clear pattern emerges with respect to the relative importance of substance use, criminal history, and racial identity in terms of predicting involvement in offenses directly or indirectly related to violence. One exception may be the consistent importance of marijuana use in gang fights and general trouble with the police, which tends to support assumptions regarding illicit drug use as part of the activities of delinquent subcultures.

Conclusion

This study examined the link between criminal offenses and the use of alcohol and other drugs. Of particular interest was the extent to which alcohol and drug use played a role in violent types of offenses and the relative importance of these substances when criminal history and racial identity of offenders are taken into account. The findings revealed that substance use, particularly alcohol use, is associated with a variety of offenses including those that are violent and nonviolent. However, when more conventional predictors of crime are introduced into the analysis, the association between substance use and criminal offenses is less direct. Although the criminal history factor, especially the number of previous arrests, tends to be more substantial overall in the impact on a wide range of offenses, substance use, particularly alcohol use, plays an important role in predicting a number of offenses.

It is also notable that alcohol and marijuana use tended to differ in their relative importance in predicting specific offenses. While alcohol use was a significant predictor of serious fights, alcohol abuse, serious injury to victim, problems with parents, extortion, vandalism, hitting instructor/supervisor, and fights with father, marijuana use was a significant predictor of store theft, general trouble with police, theft over $50.00, gang fights, use of weapon in crimes, and petty theft from autos. Therefore, while the findings do not suggest that substance use is more likely to be associated with violent rather than nonviolent offenses, the evidence does reveal that different types of substances may be more strongly linked to certain types of offenses. The implications of these findings for addressing the problems of crime and substance abuse among adolescents are that prevention efforts should take into account not only previous criminal history but the likelihood for the abuse of certain types of substances to lead to involvement in specific types of criminal activity. Since the present study was based on data from an offender population, future research is needed to further address these issues among nonincarcerated populations.

References

Anglin, M., & Speckart, G. (1988). Narcotics use and crime: A multisample, multimethod analysis. Criminology, 26, 197-233.

Banay, R. (1942). Alcoholism and crimes. Quarterly Journal of Studies on Alcohol, 2, 686-716.

Clayton, R., & Tuchfeld, B. (1982). The drug-crime debate: Obstacles to understanding the relationship. Journal of Drug Issues, 12, 153-166.

Collins, J. (1988). Suggested explanatory framework to clarify the alcohol use/violence relationship. Contemporary Drug Problems, 15, 107-121.

Dawkins, M. (1980). Alcohol and the black community: Exploratory studies of selected issues. Saratoga, CA: R & E Associates.

Dawkins, R., & Dawkins, M. (1983). Alcohol use and delinquency among black, white and hispanic adolescent offenders. Adolescence, 18, 799-809.

Gary, L. (1980). Role of alcohol and drug abuse in homicide. Public Health Reports, 95, 553-554.

Harper, F. (1980). Research and treatment with black alcoholics. Alcohol Health and Research World, 4, 10-16.

Harper, F. (1976). Alcohol abuse and black America. Alexandria, VA: Douglass Publishers.

Harper, F., & Dawkins, M. (1977). Alcohol abuse and the black community. Black Scholar, 8, 23-31.

Lewis, C., Cloninger, C., & Pais, J. (1983). Alcoholism, antisocial personality and drug use in a criminal population. Alcohol and Alcoholism, 18, 53-60.

Lewis, D., Shanok, S., Pincus, J., & Glaser, G. (1979). Violent juvenile delinquents: Psychiatric, neurological, psychological and abuse factors. Journal of Child Psychiatry, 18, 307-319.

McCord, J. (1981). Alcoholism and criminality: Confounding and differentiating factors. Journal of Studies on Alcohol, 42, 739-748.

O’Donnell, J. (1966). Narcotic addiction and crime. Social Problems, 13, 374-385.

Simonds, J., & Kashani, J. (1980). Specific drug use and violence in delinquent boys. Mexican Journal of Drug and Alcohol Abuse, 7, 305-322.

Tinklenberg, J., & Woodrow, K. (1974). Drug use among youthful assaultive and sexual offenders. Proceedings of the Association of Research on Nervous and Mental Diseases, 52, 209-224.

Wieczorek, W., Welte, J., & Abel, E. (1990). Alcohol, drugs and murder: A study of convicted homicide offenders. Journal of Criminal Justice, 18, 217-227.

Yu, J., & Williford, W. (1994). Alcohol, other drugs, and criminality: A structural analysis. American Journal of Drug and Alcohol Abuse, 20, 373-393.