Gender differences in the association of alcohol intoxication and illicit drug abuse among persons arrested for violent and property offenses

Gender differences in the association of alcohol intoxication and illicit drug abuse among persons arrested for violent and property offenses

Journal of Substance Abuse 13 (2001) 563 – 581 Gender differences in the association of alcohol intoxication and illicit drug abuse among persons arr...

151KB Sizes 0 Downloads 40 Views

Journal of Substance Abuse 13 (2001) 563 – 581

Gender differences in the association of alcohol intoxication and illicit drug abuse among persons arrested for violent and property offenses$ Susan E. Martin*, Kendall Bryant National Institute on Alcohol Abuse and Alcoholism, NIH, DHHS, 505 Willco, 6000 Executive Boulevard, Suite 505, MSC 7003, Rockville, MD 20892-7003, USA

Abstract Purpose: To explore the associations between violent and other crimes, and alcohol intoxication and recent use of cocaine, marijuana, and other drugs among men and women arrestees and examine gender differences in these relationships. Methods: We conducted a secondary analysis of 1998 using Arrestee Drug Abuse Monitoring (ADAM) system data using a sample of 9242 male and 2594 women arrested for violent and property offenses in 35 cities. Logistic regression was used to predict arrest for a violent offense (rather than a property crime) from drug- and alcoholrelated, and other variables. Results: Both gender and alcohol intoxication are significantly related to arrest for a violent offense. However, the intoxication effects (in the absence of cocaine) are more than three times as great for female (Exp(b) = 5.59) as male arrestees (Exp(b) = 1.74), while the combined effects of alcohol and cocaine predict a property offense for women but are insignificant for men. Implications: To achieve further reductions in violent crime, intervention strategies need to focus on reducing alcohol intoxication as well as illicit drug use. Research on the role of alcohol on women’s aggression and violence also is suggested. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Violent crime; Arrest; Alcohol intoxication; Illicit drug use; Gender

$ Revision of a poster presented at the annual meeting of the Research Society on Alcoholism, Denver, CO, June 2000. * Corresponding author. Tel.: +1-301-443-8767; fax: +1-301-443-8774. E-mail address: [email protected] (S.E. Martin).

0899-3289/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 8 9 9 - 3 2 8 9 ( 0 1 ) 0 0 1 0 0 - 6

564

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

1. Introduction It is generally accepted that the relationships between different types of drugs of abuse and specific types of crime are varied and complex. Nevertheless, there are relatively few studies that simultaneously explore the contributions of both alcohol and illicit drug abuse to violent and property crimes (for a review, see Parker & Auerhahn, 1998), and even fewer that also explore whether these relationships are affected by the gender of the offender (despite large gender differences in rates of both substance abuse and arrest). Similarly, the growing body of research regarding the substance abuse-related behavior of women criminal offenders focuses almost entirely on illicit drug use, leaving unclear the extent to which alcohol alone and in combination with other drugs may be associated with violence as well as other types of criminal behavior. Many studies document an association between alcohol use and violent behavior (these are reviewed by Collins, 1981; Lipsey, Wilson, Cohen, & Derzon, 1997; Murdoch, Pihl, & Ross, 1990; Parker & Rehbun, 1995; Pernanen, 1991; Roizen, 1993). For example, Roizen (1993, pp. 4–5) reports that in nearly 40 studies of violent offenders, and an equal number of studies focused on victims of violence, alcohol involvement was found in about half the events and people examined. Research focusing only on homicide also supports the notion that, overwhelmingly, alcohol is the substance most often associated with this form of violence (Abel, 1987; Spunt, Brownstein, Goldstein, Fendrich, & Liberty, 1995; Spunt, Goldstein, Brownstein, Fendrich, & Langley, 1994; Wieczorek, Welte, & Abel, 1990). Despite clear findings of an alcohol–violence association, there is disagreement over whether the relationship is a causal one. Findings from experimental research suggest that alcohol causes aggression (Bushman, 1997; Pihl, Peterson, & Lau, 1993; Taylor & Leonard, 1983). Room and Rossow (2000, p. 44) contend that ‘‘ample evidence now exists in the empirical literature regarding alcohol as a cause of violence,’’ based on studies at the aggregate or population level. However, the relationship usually is conditional. In contrast, Lipsey et al. (1997, p. 277), in a meta-analysis of studies of alcohol and violence conducted at the individual level, take a more cautious approach. They conclude that despite evidence ‘‘consistent with a causal interpretation,’’ the causal issue remains ‘‘cloudy and uncertain.’’ Similarly, others have argued that the relationship between alcohol and aggressive behavior is spurious (Collins, 1989), that both drinking and violence are caused by third factors such as personality (Jessor & Jessor, 1977), or that aggression and violence precede alcohol and drug abuse (White, Brick, & Hansell, 1993). In contrast to alcohol studies, the literature on illicit drugs and crime often finds strong associations between illicit drug use and both violent and property offending. Goldstein’s (1985) tripartite conceptual framework for the relationship of drugs and violence identified three mechanisms or factors through which they may be linked: (1) the drugs’ psychopharmacological effects; (2) the user’s economic needs; and (3) systemic violence associated with the distribution and control of illicit drugs. Many studies find that drugrelated violence is systemically related to the drug trade (Ball, 1991; Baumer, 1998;

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

565

Feucht & Kyle, 1996; Mieczekowski, 1994). Others (Inciardi, 1980; Johnson et al., 1985) have linked the need for cocaine and heroin to robbery (i.e., economic factor). These and other studies find only limited evidence of a psychopharmacological basis for an illicit drug–violence association, and that may include aggressive responses due to acute alcohol withdrawal (Goldstein, 1979; Mednick, Pollock, Volavka, & Gabrielli, 1982). Data from the nationally representative National Criminal Victimization Survey (NCVS) from 1993 through 1998, for example, indicate that while about one quarter of the victims of violent crime annually perceives their offenders in their victimization incidents to have been drinking, less than 10% stated that the assailant was under the influence of an illicit drug (Greenfeld, 1998). Many of the studies that seek to disentangle relationships between both alcohol and drugs and violence fail to distinguish among the many different drugs of abuse although different classes of substances have different effects (for reviews, see De La Rosa, Lambert, & Gropper 1990; Parker & Auerhahn, 1998; Reiss & Roth, 1993). They also overlook analyses of the combined effects of various drugs of abuse such as alcohol and cocaine. After reviewing the literature on the effects of heroin, amphetamines, cocaine, and phencyclidine on violence, Parker and Auerhahn (1998, p. 298) observe that, at best, the available results are ‘‘inconclusive,’’ and that even where the most extensive literature and strongest evidence find a possible link between cocaine use and violence, most researchers highlight a social rather than a pharmacological basis for this link. Another limitation of the studies of alcohol, drugs, and crime is their use of primarily or exclusively male samples. Men both drink more often and heavily (SAMHSA, 1999), and are arrested and incarcerated at rates far higher than women. However, in the past decade, a number of studies have focused on women offenders whose rates of arrest and incarceration have increased far more than the rates for men (Steffensmeier & Allan, 1996). In a review of studies of the links between substance abuse and violence focusing on women, Hien and Hien (1998) conclude that there are few gender differences in the prevalence of crack/ cocaine use among men and women of lower socioeconomic status (i.e., those most likely to be arrested). Other studies, focused on the growing number of women offenders who are addicted to crack, emphasize the role of violent victimization (particularly in childhood), and focus on the women’s involvement in prostitution and associated crimes as a ‘‘lifestyle,’’ without looking at the association of a particular crime and the use of specific substances (Mancuso & Miller, 2000). Others (Goldstein, Bellucci, Spunt, & Miller, 1991) observe that women who are heavy users of cocaine were more often victims of violence than occasional users, and that their violence tended to occur within a context of selfdefense in nondrug-related disputes with intimates. Both experimental and community-based survey studies suggest gender differences in aggression. However, until recently, there has been little specific information on gender differences in the role of alcohol in aggression in experimental research. The only laboratory study of gender differences in alcohol-related aggression that used both male and female subjects and targets found that alcohol increased both shock intensity and duration administered by male experimental subjects but only increased shock duration in female subjects (Giancola & Zeichner, 1995). This suggests that the men may respond

566

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

more aggressively than women when they have been drinking. A recent examination of gender differences in alcohol-induced aggression also found that women displayed considerable direct aggression when highly provoked. This occurred both when they were intoxicated and in the absence of alcohol, so that alcohol intoxication did not increase aggression as it did in male subjects (Hoaken & Pihl, 2000). A communitybased survey found that more than half of the violent incidents perpetrated by men involved alcohol compared to 27% of incidents with a woman assailant (Pernanen, 1991). Although this and other studies indicate that women are less frequently violent while intoxicated than men, since more than a quarter of incidents involved women assailants, their intoxicated violence response merits further examination. Moreover, none of the laboratory or community studies examined drug use alone or in combination with alcohol. Ethnographic research focusing on gender, drugs, and crime also suggests that there are gender differences in alcohol and other drug use. These differences may arise from differences in the reasons that men and women abuse alcohol and drugs but may also result in gender differences in the types of criminal behavior each commits. A study of women drug users in Miami notes that although men use drugs for ‘‘thrills or pleasure’’ or in response to peer pressure, women are more likely to drink or use illegal drugs for ‘‘self-medication’’ (Inciardi, Horowitz, & Pottieger, 1993, p. 25), since many women drug users, including those who enter the criminal justice system, have histories of extensive victimization (Fullilove, Lown, & Fullilove, 1992). Given the limited understanding of the contributions of alcohol and other drugs to violent and other types of criminal behavior, our study explores the associations of alcohol intoxication and drug abuse with violent and other offenses, as well as specific offenses within each crime type. It also examines gender differences in the presence and influence of alcohol–drug combinations on the types of offenses for which men and women are arrested. To address the issues of the associations among alcohol intoxication, illicit drug use, criminal offenses, and gender, we used the 1998 Arrestee Drug and Alcohol Monitoring (ADAM) data. The ADAM program was developed by the Department of Justice to track the associations of various drugs with different types of criminal offenses. The program involves a quarterly survey of substance use among arrestees that includes several questions about alcohol use, as well as collection of urine specimens, which are screened for the presence of 10 illicit drugs. The ADAM data set provides a unique perspective on the associations of alcohol and other drugs with various types of crimes. It permits us to explore alcohol, drug, and offense data in a large sample of both male and female recent arrestees who are being booked into a jail within 48 h of arrest. It includes self-reported alcohol use and findings from voluntary urine tests that detect drugs in the arrestees’ urine as well as the arrest offense and demographic data, allowing for analysis of the associations of diverse substances with various types of criminal behavior among a representative sample of urban arrestees and for a focus specifically on the presence and influence of alcohol and other drugs on violent crimes compared with property crimes.

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

567

2. Methods To predict the type of criminal charge from drug- and alcohol-related variables within an adult sample of arrestees and explore whether there were differences in the alcohol and drugrelated predictors of men’s and women’s offenses, a subset of the 1998 ADAM data was used in the analysis. Cases were drawn from the 20,715 adult male booked arrestees who were interviewed in 35 sites and from 6699 adult female booked arrestees at 32 sites. The data collection area for each site included the county within which the city is located with two exceptions. Anchorage data represent only that city, and New York City data come only from one borough, Manhattan. 2.1. ADAM data collection procedures Enrollment in the ADAM sample occurs during booking when arrestees are asked to participate in the study. The 1998 data were collected in central booking stations in the 35 cities for about 7 days each quarter. Trained local staff at each site obtained voluntary and anonymous interview data and urine specimens from detained arrestees who had been in the booking facility for not more than 48 h. In each site, a convenience sample of about 225 males and 100 females provided interviews and urine specimens. Although all arrestees booked into any facility within the previous 48 h were eligible to be interviewed, respondents usually were drawn from the largest booking facility in the jurisdiction. Since there were more male arrestees than could be interviewed, the sample of men was selected in the following order of priority on interview days: arrest for a nondrug felony, nondrug misdemeanor, drug felony, and drug misdemeanor. Sites were requested to survey no more than 20% of the males arrestees for drug offenses. All women arrestees were eligible for an interview regardless of charge. Data for women arrestees were not provided by three sites. Generally, detainees who were intoxicated, mentally disordered, or potentially violent were excluded. Response rates for the interviews were approximately 80% of eligible arrestees. Of those who agreed to the interview, about 80% also provided a urine specimen. Only persons who both agreed to the interview and also provided a specimen were included in the data set (i.e., nearly two thirds of eligible arrestees). Urine specimens are analyzed for 10 drugs: cocaine, opiates, marijuana, PCP, methadone, benzodiazepines (Valium), methaqualone, propoxyphene (Darvon), barbiturates, and amphetamines. (For further information about data collection procedures and the data set, see U.S. Department of Justice, National Institute of Justice, 1999.) No published studies have systematically compared the characteristics of arrestees who declined to participate with those who are included in the ADAM program. From the full data set, we included only African–American, Hispanic, and non-Hispanic white men and women arrested for violent or property offenses. We excluded the few Native American and Asian arrestees because their numbers were quite small and their patterns of alcohol and drug use were quite divergent. We also omitted those ADAM arrestees whose most serious arrest charge was for drug sale or possession, an alcohol-related offense including DWI and liquor law violations, and persons charged with ‘‘other’’ offenses

568

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

including flight, probation or parole violations, prostitution, and disorderly conduct. Persons arrested for alcohol- and drug-related charges were excluded from the analysis because of the implicit or explicit inclusion of use of the substance in the charge itself, which could confound the findings. Arrestees charged with miscellaneous other offenses were excluded for two reasons: first, the underlying charges were widely varied; second, many persons were arrested as fugitives from justice and for parole violations so that alcohol and drug use prior to the immediate arrest was not related to the underlying specific criminal offense. The ADAM data are not without shortcomings (see Baumer, 1998; General Accounting Office, 1993; White & Gorman, 2000). These include lack of standardization across sites in data collection procedures and criteria for case selection; varied geographic definitions among sites; the inability to determine the number of unique arrestees (rather than the number of arrests) occurring in a time period; and the absence of questions regarding quantity of alcohol consumed. Nevertheless, it also has several advantages. General population surveys are a very inefficient way to study the intersection of crime and illicit drug use since both are relatively rare events. Data based on a population of jail or prison inmates do not represent criminal offenders since it is biased toward the most serious offenders who have been convicted and incarcerated. Studies that rely on archival data, such as arrest reports, tend to understate alcohol’s presence since the data depend on how accurately police monitor and report alcohol’s presence, and such data are not reported consistently. Thus, ADAM permits a unique perspective for analyzing the associations of alcohol and other drugs with various types of crime. 2.2. Variables 2.2.1. Dependent variable: arrest for violent (rather than property) crime The dependent measure, the type of crime, was based on the most serious charge for which the arrestee was booked. The outcome could be either arrest for a violent offense (coded 1) or not (i.e., arrest for a property offense which was coded 0). Violent offenses include homicide, kidnapping, extortion/threat, sexual assault, robbery, assault, weapons, and family offenses; property crimes include burglary, forgery, fraud, larceny/theft, stolen property, and stolen vehicle. 2.2.2. Independent variables: alcohol- and drug-related variables The ADAM interview included questions about arrestees’ use of both illegal and legal drugs, including alcohol. Several alcohol-related items in the ADAM data that measure acute and chronic effects of drinking are included in the analysis. The arrestee’s self-report of being ‘‘under the influence of alcohol’’ at the time of the arrest offense (also termed alcohol intoxication) is the primary alcohol-related variable of interest. It was dummy coded 1 for those who reported being under the influence, 0 for those who were not. We also included a dummy coded variable measuring drinking within the past 72 h (alternatively termed ‘‘recent drinking’’ and coded 1 if yes, 0 if no drinking had occurred). Two items suggesting chronic alcohol use were used: frequent drinking in the previous 30 days, and needs alcohol treatment. ‘‘Frequent drinkers’’ included those arrestees who said they had consumed alcohol

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

569

in more than 10 in the past 30 days (coded 1 if yes, 0 if not). Persons were classified as in need of alcohol treatment (coded 1) if they stated that they needed to cut down on drinking, felt dependent on alcohol in the past year, or ‘‘could use alcohol treatment.’’ Examining both recent alcohol use and being under the influence provides a fuller picture of the likely acute effects of alcohol on the crime. With a 72-h window for recent use, one cannot infer that the arrestee’s behavior was affected by drinking. Reporting being ‘‘under the influence’’ is a subjective self-report measure that may be systematically inflated as a means of ‘‘deviance disavowal’’ (McCaghy, 1968) or denial of responsibility for one’s behavior while intoxicated, or depressed in response to perceived social desirability (Harrison, 1995). The three illicit drug use measures are based on positive urinalysis findings rather than self-reported use (which was substantially lower). The variables include positive urine for cocaine and/or crack, for marijuana, and for all other drugs (each dummy coded 1 for positive urine, 0 otherwise). To assess the possible intensified effects of combining alcohol and cocaine use, we created an alcohol by cocaine interaction variable coded 1 if urine was positive for cocaine and the subject also was under the influence of alcohol; 0 otherwise. 2.2.3. Other variables Demographic variables included in the analysis include gender, two race/ethnicity dummy coded variables (African–American and Hispanic each coded 1; non-Hispanic whites were the suppressed category), marital status (coded 1 if single and 0 for all other statuses), and age category (with two dummy variables coded 1, respectively, for those under 21 and those over 35; persons 21–35 were the suppressed category). There were two employment-related variables: employed full time in legal work in the past month (coded 1) or not (coded 0), and income derived primarily from illegal activities (coded 1) or not (coded 0). Offense seriousness was broken into felonies (coded 1) and misdemeanors (coded 0). Prior criminal history distinguished between those who had served time for a prior offense in the past year (coded 1) or not (coded 0). 2.3. Validity of alcohol self-reports To address threats to the validity of the findings arising from inflated or diminished selfreported alcohol use by all respondents, we compared arrestees’ responses on self-reported drug use questions with their urine specimens, differentiated persons who were consistent and inconsistent with respect to drug use data, then compared these groups with respect to their self-reported alcohol use. There were two measures of inconsistency between self-report and urinalysis data based on drug type (i.e., separate measures for discrepancies in the use of marijuana and cocaine, including crack). Each was examined for two time periods: ever used drug and use within the past 30 days (i.e., those who admitted ever using but denied recent use but tested positive). Next we compared the self-reported recent alcohol use and intoxication of three groups: those arrestees who reported no drug use and tested negative,

570

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

those who reported no use but tested positive on one of the drug use measures, and those who tested positive and admitted their drug use. Like Harrison (1995), we found reasonable levels of congruence between self-report and urinalysis, and greater drug underreporting closely related to social desirability. Marijuana use was more likely to be reported than cocaine use; ever having used a drug was more likely to be reported than recent use. Overall, 7.2% of the sample denied using marijuana but tested positive, including 4.5% who admitted having used it but denied past month use but tested positive; 13.8% denied using cocaine or crack, including 9.5% who denied using crack or cocaine in the previous 72 h but had urine that tested positive. Combining the marijuana and crack inconsistent responders, 9.8% gave inconsistent responses about recent use, 10.2% denied ever using either marijuana or crack/cocaine but tested positive, and overall 19.3% gave one or more inconsistent self-reports. We next compared self-reported alcohol use and intoxication across these groups. Overall, there was not a significant difference between the 19% who gave inconsistent answers and most respondents who were consistent regarding cocaine and marijuana use with respect to either recent drinking or being under the influence of alcohol. Additional analyses found that those who were inconsistent with respect to recent drug use were more likely to report being under the influence but the difference was not significant; those who were inconsistent regarding ever using either substance were significantly less likely to report being under the influence. Finally, we conducted separate analyses excluding the inconsistent responders from the models and found that this did not alter the findings. For that reason, we did not adjust the data or delete this group from the data reported here. 2.4. Analysis We began by conducting bivariate analyses of the associations of violent and property crimes by gender and other demographic variables; alcohol use and intoxication; and urinalysis data on marijuana, cocaine, and other drug positive by type of crime as well as by specific offense within crime type. Next, because the dependent variable is dichotomous, we conducted logistic regression analyses using the SPSS 10.0 for Windows statistical package to estimate the likelihood of being arrested for a violent rather than a property offense after introducing controls for the effects of other factors.

3. Results 3.1. Characterizing the arrestees The sociodemographic characteristics of the sample are shown in Table 1. As the table indicates, both the violent and property arrestee samples had similar characteristics. Both tended to be young (median age of 30), although they ranged from 15 to 82 years of age. Most (78%) were male, nearly half were African–American (48%), had limited education

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

571

Table 1 Characteristics of 1998 ADAM violent and property arrestees Violent Age Mean Median Gender Male Female Race/ethnicity Black (non-Hispanic) White (non-Hispanic) Hispanic Education < High school HS diploma/GED Some college Employment Work full time Other legal income Illegal income No income Marital status Single Other Recent alcohol use Under influence/alcohol Need alcohol treatment# Frequent drinkers Cocaine/crack-positive Marijuana-positive Other drug-positive Alcohol  cocaine-positive Served time past year Charged with felony

Property

Total

Number

Percent

Number

Percent

Number

Percent

6024

30.4 29

5777

30.4 29

11,801

30.4

4941 1103

81.8 18.2

4301 1491

74.3*** 25.7

9242 2594

78.1 21.9

2948 1728 1368

48.8 28.6 22.6

2792 1834 1166

48.2 31.7*** 20.1***

5740 3562 2534

48.5 30.1 21.4

2702 1818 1033

48.7 32.7 18.6

2548 1751 994

48.1 33.1 18.8

5250 3569 2027

48.4 32.9 18.7

2942 2448 264 364

48.9 40.7 4.4 6.0

2220 2517 606 418

38.5*** 43.7*** 10.5*** 7.3**

5162 3965 870 782

47.9 36.8 8.1 7.3

1768 4276 3222 1541 2508 1485 1600 2141 628 540 1210 3546

29.3 70.7 53.4 25.5 41.8 24.6 26.5 35.4 10.4 9.0 20.0 59.0

2148 3644 2754 908 2160 1491 2367 1898 507 470 1633 3518

37.1*** 62.9*** 47.6** 15.7*** 37.6*** 25.8 41.0*** 32.8** 08.8*** 8.2 28.3*** 61.1*

3916 8920 5976 2449 4668 2976 3967 4039 1135 1010 2843 7064

33.1 66.9 50.5 20.7 39.6 25.2 33.5 34.1 9.6 8.6 24.1 60.0

* P < .05. ** P < .01. *** P < .001. # Affirmative response to any one of the three statements: I could use alcohol treatment; I have been unsuccessful in cutting down or I feel dependent on alcohol.

(48% had not finished high school), were employed full (42.5%) or part time (18%) during the month prior to their arrest. Nevertheless, the table also shows that there are statistically significant differences between the two groups of offenders. Although the vast majority of persons arrested for both violent and property offenses reported that they have tried alcohol at some time in their lives (92% of each group), and most also had consumed alcohol in the past year (80% and

572

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

78%, respectively) (not shown), significantly more of the persons arrested for violent crimes (53%) than the persons arrested for property offenses (48%) reported recent drinking and being under the influence at the time of the offense (26% of the violent and 16% of the property offenders, respectively), as shown in Table 1. While arrestees in both groups selfreported the need for alcohol treatment, the proportion is slightly higher among violent (41.5%) than property offenders (37%), although there was no difference in the proportion that drinks frequently. Violent and property offenders also differed with respect to drug use. The property offenders were significantly more likely to test positive for cocaine/crack (41%) than the violent offenders (26.5%). Violent offenders were more likely than property offenders to be positive for marijuana and, conversely, less likely to be positive for other drugs, although these differences across offense type were much smaller. The proportion of arrestees who were both under the influence of alcohol and also tested positive for cocaine was small (9% of the sample), and did not significantly differ by offense type. A significantly larger proportion of violent than property offenders was employed full time, while a larger proportion of property offenders was supported primarily by illegal income or have no income. Conversely, significantly more property offenders (28%) than violent offenders (20%) had served time in the past year. 3.2. Charge at arrest, gender, and alcohol intoxication, and drug use There are important differences in arrestees’ recent alcohol and drug use and intoxication by gender and specific charge at arrest, as shown in Table 2. Focusing on gender, 54% of the male and 43% of the female arrestees reported drinking within 72 h of their arrest; 22% of the male but only 16% of the females reported being under the influence of alcohol at the time of their offense. These gender differences reflect societal consumption patterns whereby men drink greater quantities and more frequently than women (SAMHSA, 1999). With respect to type of arrest offense, both men and women are more likely to have been drinking prior to violent crime (57% and 44%, respectively) than a property offense (52% and 36%, respectively). They also are more likely to have been under the influence of alcohol at the time of a violent than a property offense for which they are arrested. Men are much more likely than women to report recent drinking (52% vs. 36%) and intoxication (18% vs. 9%) prior to an arrest for a property crime. In contrast, the gender gap is quite small in overall rates of intoxicated violence (26% and 23%, respectively), as well as for persons arrested for family violence (26% and 30%, respectively). The high rates of recent alcohol use by both property and violent offenders, particularly men, suggest that drinking may be part of the lifestyle of young, urban males; alternatively, it could be associated with a criminal lifestyle. However, the stronger association of alcohol intoxication and violent crimes by both men and women is similar to the findings of many other studies (Lipsey et al., 1997; Murdoch et al., 1990; Parker & Rehbun, 1995; Pernanen, 1991; Roizen, 1993; Wolfgang, 1958), and suggests that alcohol-related violence may be triggered by a pharmacological effect or permitted by alcohol expectancies.

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

573

In contrast to the patterns of alcohol use and intoxication among arrestees, as shown in Table 2, men and women arrestees are equally likely to test positive for cocaine whether they are arrested for a violent or a property offense (27% and 41%, respectively). One notable exception is that even more women (52%) than men (40%) arrested for robbery—an offense categorized as violent but usually economically motivated—were positive for cocaine. This association of cocaine and robbery is in marked contrast to the much lower rates of cocaine use for other violent offenses among both men and women. It also is notable that both men and women arrested for robbery were less likely to be under the influence of alcohol than persons arrested for assault and family violence. Thus, Visher’s (1990) observation that ‘‘cocaine use is more closely tied to income-generating crime than alcohol use’’ among male offenders continues to be the case and is true of women offenders as well. Also shown in Table 2, men were more likely to have urine positive for marijuana than women. But within gender, there were negligible differences in marijuana use by crime type. In contrast, gender differences in urine positive tests for various other drugs were small but both men and women arrested for property offenses were more likely to test positive for other drugs than those arrested for violent crimes. Particularly notable is the high rate of other drug use among women arrested for burglary. 3.3. Multivariate comparisons of arrestees’ alcohol intoxication and drug use Because many of the arrestees tested positive for 1 of 10 illegal substances and also reported being under the influence of alcohol at the time of their offense, we used multivariate logistic regression to explore the relationships of alcohol, cocaine, marijuana, and other drugs, alone and in combination, on the type of offense at arrest after controlling for a number of other variables. The result of this analysis is show in Table 3 in which we predicted the likelihood of an arrest for a violent (coded 1) rather than for a property offense (coded 0) from 17 independent variables. First, we entered 12 demographic and other control variables, then added a second block including 5 alcohol and drug use dummy variables: self-reported alcohol intoxication at the time of the crime; urine positive for cocaine; urine positive for marijuana; urine positive for another drug; and an alcohol–cocaine interaction term. In the first step, the strongest predictors of arrest for violent crime were being male, having a full-time job, and needing alcohol treatment; the strongest predictors of a property offense (log odds values less than 1) were being single, engaged in illegal work, and having served time in the previous year. Black and Hispanic arrestees were more likely than whites to be arrested for violent crimes as were persons between 21 and 35 years old. The addition of the alcohol and drug variables substantially improved the model. Being under the influence of alcohol (but not cocaine-positive) nearly doubled the likelihood of an arrest for violence, while being cocaine-positive (without alcohol intoxication) made it half as likely. Having marijuana-positive urine also increased the likelihood of arrest for a violent offense, while having urine positive for other drugs increased the likelihood of arrest for a property offense net of the demographic variables. The alcohol and cocaine interaction term was not a significant predictor of offense type. Tests for multicolinearity among the three alcohol-related variables indicated that their intercorrelation ranged from .28 to .35, well in the

574

Arrest charge a

Total cases (2594/9242) (%) Assault (592/1911) Homicide (19/173) Robbery (73/592) Sexual assault (1/203) Weapons (28/418) Family offenses (337/1345) Other violence (52/299) Total violence (1104/4941) Burglary (119/891) Theft (799/1814) Auto theft (89/523) Forgery/fraud (239/390) Other property (245/683) Total property (1419/4301)

Recent alcohol use

Under influence of alcohol

Cocaine-positive

Marijuana-positive

Other drug-positive

F

M

F

M

F

M

F

M

F

M

43 46 37 40 52 25 41 48 44 35 35 40 28 48 36

54 57 51 54 na 47 58 54 57 51 52 45 43 60 52

24 24 16 14 19 7 26 23 23 8 8 11 5 17 9

29 29 14 21 na 18 30 24 26 19 16 15 11 28 18

29 29 21 52 16 21 18 25 27 42 38 39 24 66 41

29 29 18 40 na 25 20 25 27 49 43 41 29 42 41

24 24 32 29 29 32 24 39 25 23 20 35 22 20 22

37 37 51 49 na 47 31 35 38 41 34 46 28 35 37

29 20 21 30 9 18 20 23 21 47 30 45 28 35 33

21 21 20 30 na 27 20 19 22 31 30 36 26 29 30

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

Table 2 Alcohol, marijuana, cocaine and other drug use/abuse by ADAM property and violent offenders by sex and most serious offense charge

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

575

Table 3 Logistic regression analyses predicting arrest for a violent (rather than property) crime for all ADAM arrestees: partial odds ratios and levels of significance Male Black Hispanic Under 21 Over 35 Single Served time past year Felony Full-time job Illegal work Frequent drinker Needs alcohol treatment Under alcohol influence Cocaine-positive urine Marijuana-positive Other drug-positive Alcohol  cocaine Constant

b

SE

0.457 0.161 0.191 0.163 0.162 0.327 0.394 0.096 0.198 0.717 0.104 0.21

0.047 0.045 0.054 0.053 0.044 0.042 0.046 0.039 0.041 0.081 0.047 0.041

1.580*** 1.175*** 1.211*** 0.849** 0.850*** 0.721*** 0.675*** 0.908* 1.218*** 0.488*** 0.898* 1.237***

0.185

0.063

0.835**

c2 = 511.313; df = 11 2log likelihood = 15,569.15 Nagelkerke R2=.057

Exp(b)

b

SE

0.420 0.293 0.247 0.291 0.072 0.324 0.363 0.082 0.151 0.505 0.231 0.108 0.642 0.615 0.237 0.237 0.140 0.130

0.049 0.048 0.056 0.055 0.046 0.043 0.047 0.040 0.042 0.083 0.050 0.043 0.054 0.047 0.046 0.052 0.100 0.068

Exp(b) 1.523*** 1.344*** 1.281*** 0.747*** 0.931 0.723*** 0.695*** 0.921* 1.162*** 0.603*** 0.793*** 1.115* 1.900*** 0.541*** 1.268*** 0.789*** 1.151 0.878

c2 = 951.154; df = 16 2log likelihood = 15,130.31 Nagelkerke R2=.105

* P < .05. ** P < .01. *** P < .001.

acceptable range. We retained them in the analysis to examine differences in the effects of acute and chronic drinking and note that frequent drinking became an even stronger predictor of a property offense with the addition of the intoxication and other drug variables to the model. Because the results were substantially different for men and women arrestees, and women made up a substantially smaller proportion of the arrestees in the data set, we ran separate regressions by gender, which are shown in Table 4. In the model with only men on the right, the significant predictors in the previous model remained significant: black and Hispanic men were significantly more likely than white men to be arrested for a violent crime as were men who were working full time. Predictors of a property offense (i.e., those that were significant and negative for a violent offense) included being under 21, unmarried, having served time in the past year, deriving most of one’s income from illegal work, and being a frequent drinker. Being under the influence of alcohol remains the strongest predictor of a violent offense; urine positive for both cocaine and other drugs was predictive of a property offense. The alcohol–cocaine interaction was not a significant predictor of crime type for men. In contrast, among women, blacks but not Hispanics, were more likely than whites to be arrested for a violent offense; neither age nor full-time work was related to offense type. Like the men, being unmarried and having an illegal source of income and criminal record

576

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

Table 4 Logistic regression analysis predicting arrest for a violent crime for male and female ADAM arrestees: partial odds ratios and levels of significance Women only Black Hispanic Under 21 Over 35 Single Felony Served time per year Full-time work Illegal work Frequent drinker Needs alcohol treatment Under alcohol Cocaine-positive Marijuana-positive Other drug-positive Cocaine  alcohol Constant

Men only

b

SE

0.431 0.207 0.101 0.008 0.382 0.099 0.324 0.020 1.108 0.290 0.026 1.717 0.483 0.371 0.497 0.548 0.320

0.104 0.130 0.126 0.101 0.096 0.086 0.122 0.096 0.209 0.134 0.103 0.201 0.115 0.109 0.113 0.257 0.123

c2 = 307.782; df = 15 2log likelihood = 3169.82 Nagelkerke R2=.153

Exp(b) 1.540*** 1.220 0.904 0.992 0.683*** 1.104 0.723* 0.980 0.330*** 0.748* 1.026 5.569*** 0.617*** 1.449** 0.608*** 0.578* 0.726*

b

SE

0.260 0.264 0.328 0.092 0.310 0.134 0.360 0.191 0.370 0.221 0.110 0.472 0.657 0.205 0.163 0.217 0.347

0.053 0.062 0.061 0.052 0.048 0.046 0.051 0.047 0.092 0.054 0.047 0.072 0.060 0.051 0.059 0.110 0.073

Exp(b) 1.296*** 1.302*** 0.720*** 0.912 0.733*** 0.875** 0.697*** 1.211*** 0.691*** 0.802*** 1.117* 1.603*** 0.518*** 1.227*** 0.849** 1.242 1.415***

c2 = 621.325; df = 15 2log likelihood = 11,883.075 Nagelkerke R2=.089

* P < .05. ** P < .01. *** P < .001.

predicted arrest for a property crime. Most notable among the women arrestees is the finding that being under the influence of alcohol (but not positive for cocaine) increased the likelihood of arrest for a violent crime more than fivefold (Exp(b) 5.597) while women with urine positive for cocaine and for other drugs were significantly more likely to be arrested for a property crime. Women who combined alcohol and cocaine were twice as likely as women who did not use either substance to have been arrested for a violent crime (Exp(1.717–0.483–0.548) = 1.96), but were more likely than women using alcohol alone to be arrested for a property offense.

4. Discussion Like other studies using the ADAM data to examine the associations of alcohol intoxication and illegal drug use and crime (Baumer, 1998; U.S. Department of Justice, National Institute of Justice, 1999; Valdez, Kaplan, Curtis, & Yin, 1995; Valdez, Yin, & Kaplan, 1997; Visher, 1990; White & Gorman, 2000), this study has found that both

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

577

alcohol and drug use are very widespread among arrested offenders regardless of offense type. No gender differences were found in the proportion of arrestees that tested positive for cocaine and other drugs, but as in the general population, among arrestees, women were less likely than men recently to have consumed alcohol or to be intoxicated at the time of the offense. Nevertheless, for both women and men, intoxication was significantly more likely to precede arrest for a violent crime whereas illegal drug use, particularly cocaine use, was significantly more likely to precede arrest for a property offense. When we reran the model using self-reported cocaine, marijuana, and other drug use, we found that the strength of the associations weakened slightly (and marijuana lost significance), but the model remained significant, indicating the robustness of our findings. Although these patterns of association between substance abuse and offense type hold for both men and women, the effect of alcohol intoxication as a predictor of a violent crime is more than three times as strong for the women as for the men (odds ratios of 5.59 vs. 1.61, respectively). This suggests that even among women offenders, many of whom are involved in illegal work and also use illicit drugs, when they are under the influence of alcohol, they are significantly more likely to be involved in assaultive behavior. Both psychopharmacological and sociocultural factors may help explain these findings. Goldstein (1985) argues that psychopharmacologically induced violence is most evident in alcohol intoxication as the alcohol reduces inhibitions and induces impulses allowing aggression. Parker and Rebhun (1995), in contrast, advance a selective disinhibition approach that seeks to explain why normatively prescribed behavior is disinhibited in relatively few instances. The occurrence of such violations of norms prohibiting violence depends on contextual factors in specific situations, the actors involved and their relationships, and the role of bystanders. Violations occur when people encounter conflicting sets of norms in a situation and they follow those allowing violence (e.g., self-defense). In addition, selective disinhibition may operate differently by gender with respect to the availability of conflicting norms or expectancies that ‘‘justify’’ use of violence (and intoxication). For men, drinking enhances masculinity, while defense of masculinity and power claims are used to justify violence (hence men’s involvement in intoxicated violence with nonintimate acquaintances). Aggressive men also may drink because they expect alcohol to disinhibit their behavior while intoxication also provides a justifiable ‘‘excuse’’ for alcohol-related violence. Women are less likely to become physically aggressive because it is socially viewed as unfeminine, is less likely to yield positive results, and may lead to injury. The one exception, based on experimental findings, is when women are provoked. Then their aggressive responses appear to occur regardless of the use of alcohol (Hoaken & Pihl, 2000). Such provocation seems most likely to arise in the context of a domestic dispute with a partner and result in violent defense of self and children. This supports Goldstein et al.’s (1991) observation that women drug abusers’ violence tended to arise from a nondrug-related dispute. In contrast, the psychopharmacological effect of drugs, such as cocaine, even with the addition of alcohol appear to work to suppress hostile feelings and aggressive behavior except during withdrawal so that most of the violence associated with drug use arises from economic need or ‘‘systemic’’ factors associated with the distribution and control of drugs (Goldstein, 1985).

578

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

The association of violent offenses with full-time employment may be related, in part, to the high proportion of such crimes that involve assault and domestic violence. These are expressive offenses that are not associated with a criminal lifestyle. In contrast, some of the arrestees who are not employed full time are motivated to commit property crimes out of economic need and are more likely to be committed to a criminal lifestyle. The high rates of alcohol intoxication and its strong association with violent offenses particularly among women offenders suggest avenues for both policy and research. First, to achieve further reductions in crime, particularly in violent crime, it is important to implement strategies that focus on reducing alcohol intoxication as well as illicit drug use. Interventions might include expanded programs to address the substance abuse treatment needs of those offenders under the control of the correctional system, as well as additional preventive intervention strategies that reduce crime by addressing environmental and situational factors that contribute to alcohol-related violence. For example, limiting alcohol availability by raising the minimum legal drinking age and increasing the excise taxes was found to have had the unanticipated salutary effect of reducing violent crime (Cook & Moore, 1993; Parker & Cartmill, 1998). Additional controls on alcohol availability might further reduce violent crime. Our findings regarding the association of alcohol intoxication and violence among women also suggest the need to shift or expand the focus of research on women and crime. While crime related to illicit drug use has been a major contributing factor to women’s greatly increased arrest and incarceration rates, it also is necessary to explore more fully the contribution of alcohol to women’s offending, with a particular focus on violence including that of drug-involved women. Isolating the effects of alcohol intoxication alone, in contrast to the combination of alcohol and cocaine, highlights the differences between men and women with respect to specific arrest offense. Only 8% of the women arrestees was under the influence of alcohol alone and another 8% was using alcohol and cocaine together. Those women combining cocaine and alcohol were more likely to be arrested for theft (6%) or another property crime (11%) than the women intoxicated by alcohol alone (2.5% and 6%, respectively), and less likely to be arrested for assault (11%) and family violence (8%) than the women using only alcohol (15% and 18%, respectively). Within the sample of women, more than 70% of arrests was for one of these four offenses. In contrast, men were arrested for a wider range and different mix of offenses. A higher proportion of men than women in the ADAM data set was arrested for robbery (6% and 3%, respectively), sexual assault (2% vs. none), as well as for burglary (10% vs. 5%, respectively); fewer were arrested for theft (31% compared with 10%). The men also were more likely to have been under the influence of alcohol alone (16%) than women (8%). Thus, what appears to be a gender effect also is related to offense mix and specific offense. Caution is necessary in interpreting our findings given the lack of information about the situational aspects of the arrest offense such as the victim–perpetrator relationship, and limitations of the ADAM program’s sampling procedures and its questionnaire’s alcoholrelated items. There is no information on the quantity of alcohol consumed either in the period immediately prior to the offense or the past month, or offenders’ typical pattern of

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

579

drinking or alcohol-related problems, while the measure of intoxication (i.e., being under the influence) is subjective and inconsistently reported. Despite these limitations, the ADAM data permit a useful comparison of the effects of various alcohol and drug abuse combinations in their relationships with violent and property crimes, and lend support to other studies that highlight the role of alcohol use and intoxication, particularly in violent crimes. Further analyses might explore the association of alcohol with specific types of violent offenses, probe trends in the association of alcohol and with these offenses, and consider and assess the effects of various approaches to further reducing the alcohol–violence nexus.

Acknowledgments We wish to thank Deborah Dawson of the Biometry Branch, NIAAA, for helpful comments on an earlier draft of this manuscript.

References Abel, E. L. (1987). Drugs and homicide in Erie County, New York. International Journal of the Addictions, 22 (2), 195 – 200. Ball, J. C. (1991). The similarity of crime rates among male heroin addicts in New York City, Philadelphia, and Baltimore. Journal of Drug Issues, 12, 225 – 239. Baumer, T. L. (1998). Assessing the need for treatment in the arrestee population. Journal of Crime and Justice, XXI, 173 – 190. Bushman, B. J. (1997). Effects of alcohol on human aggression: validity of proposed explanations. In: M. Galanter (Ed.), Recent developments in alcoholism, vol. 13 ( pp. 227 – 244). New York: Plenum. Collins, J. J., Jr. (1981). Alcohol use and criminal behavior: an empirical, theoretical, and methodological overview. In: J. J. Collins Jr. (Ed.), Drinking and crime: perspectives on the relationship between alcohol consumption and criminal behavior ( pp. 288 – 316). New York: Guilford Press. Collins, J. J., Jr. (1989). Alcohol and interpersonal violence: less than meets the eye. In: N. A. Weiner, & M. E. Wolfgang (Eds.), Pathways to criminal violence ( pp. 49 – 67). Newbury Park, CA: Sage. Cook, P. J., & Moore, M. J. (1993). Economic perspectives on reducing alcohol-related violence. In: S. E. Martin (Ed.), Alcohol and interpersonal violence; fostering multidisciplinary perspectives. NIAAA Research Monograph, vol. 24 ( pp. 193 – 212) (Publ. No. NIH 93-3496, Rockville, MD). De La Rosa, M., Lambert, E., & Gropper, B. (Eds.) (1990). Drugs and violence: causes, correlates, and consequences. Rockville, MD: National Institute on Drug Abuse. Feucht, T. E., & Kyle, G. M. (1996). Methamphetamine use among adult arrestees: findings from the Drug Use Forecasting (DUF) program. Washington, DC: National Institute of Justice. Fullilove, M. T., Lown, A., & Fullilove, R. E. (1992). Crack ’hos and skeezers: traumatic experiences of women crack users. Journal of Sex Research, 29, 275 – 287. General Accounting Office. (1993, June). Drug use measurement: strengths, limitations and recommendations for improvement. Washington, DC: Government Printing Office (PEMP-93-18). Giancola, P. R., & Zeichner, A. (1995). An investigation of gender differences in alcohol-related aggression. Journal of Studies on Alcohol, 56, 573 – 579. Goldstein, P. J. (1979). Prostitution and drugs. Lexington, MA: Lexington Books.

580

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

Goldstein, P. J. (1985). The drugs/violence nexus: a tripartite conceptual framework. Journal of Drug Issues, 15, 493 – 506. Goldstein, P. J., Bellucci, P. A., Spunt, B., & Miller, T. (1991). Volume of cocaine use and violence: a comparison between men and women. Journal of Drug Issues, 21, 345 – 367. Greenfeld, L. A. (1998). Alcohol and crime: an analysis of national data on the prevalence of alcohol involvement in crime. Washington, DC: U.S. Department of Justice (NCJ 168632). Harrison, L. D. (1995). The validity of self-reported data on drug use. Journal of Drug Issues, 25, 91 – 111. Hien, D., & Hien, N. (1998). Women, violence with intimates, and substance abuse: relevant theory, empirical findings and recommendations for future research. American Journal of Drug and Alcohol Abuse, 24 (3), 419 – 438. Hoaken, P. N., & Pihl, R. O. (2000). The effects of alcohol intoxication on aggressive responses in men and women. Alcohol and Alcoholism, 35 (5), 471 – 477. Inciardi, J. A. (1980). Youth, drugs and street crime. In: F. Scarpitti, & S. K. Datesman (Eds.), Drugs and the youth culture ( pp. 175 – 204). Beverly Hills, CA: Sage. Inciardi, J. A., Horowitz, R., & Pottieger, A. E. (1993). Street kids, street drugs, street crime: an examination of drug use and serious delinquency in Miami. Belmont, CA: Wadsworth. Jessor, R., & Jessor, S. L. (1977). Problem behavior and psychosocial development. A longitudinal study of youth. San Diego, CA: Academic Press. Johnson, B. D., Goldstein, P. J., Preble, E., Schmeidler, J., Lipton, D. S., Spunt, B., & Miller, T. (1985). Taking care of business: the economics of crime by heroin abusers. New York: Lexington Books. Lipsey, M. W., Wilson, D. B., Cohen, M. A., & Derzon, J. H. (1997). Is there a causal relationship between alcohol use and violence? A synthesis of the evidence. In: M. Galanter (Ed.), Recent developments in alcoholism, vol. 13 ( pp. 245 – 283). New York: Plenum. Mancuso, R. F., & Miller, B. A. (2000). Crime and punishment in the lives of women alcohol and other drug (AOD) users: exploring the gender, lifestyle and legal issues. In: C. M. Renzetti, & L. Goodstein (Eds.), Women, crime and criminal justice ( pp. 93 – 110). Los Angeles: Roxbury. McCaghy, C. H. (1968). Drinking and deviance disavowal: the case of child molesters. Social Forces, 16, 43 – 49. Mednick, S. A., Pollock, V., Volavka, J., & Gabrielli, W. F. (1982). Biology and violence. In: M. E. Wolfgang, & N. A. Weiner (Eds.), Criminal violence ( pp. 21 – 80). Beverly Hills, CA: Sage. Mieczekowski, T. (1994). The experiences of women who sell crack: some descriptive data from the Detroit crack ethnography project. Journal of Drug Issues, 24, 227 – 248. Murdoch, D. D., Pihl, R. O., & Ross, D. (1990). Alcohol and crimes of violence: present issues. International Journal of the Addictions, 25, 1065 – 1081. Parker, R. N., & Auerhahn, K. (1998). Alcohol drugs and violence. Annual Review of Sociology, 24, 291 – 311. Parker, R. N., & Cartmill, R. S. (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 – 1398. Parker, R. N., & Rehbun, L. A. (1995). Alcohol and homicide: a deadly combination of two American traditions. Albany: SUNY Press. Pernanen, K. (1991). Alcohol in human violence. New York: Guilford Press. Pihl, R. O., Peterson, J. B., & Lau, M. A. (1993, September). A biosocial model of the alcohol – aggression relationship. Journal of Studies on Alcohol (Suppl. 11), 128 – 139. Reiss, A., & Roth, J. (Eds.) (1993). Understanding and preventing violence. Washington, DC: National Academy Press. Roizen, J. (1993). Issues in the epidemiology of alcohol and violence. In: S. Martin (Ed.), Alcohol and interpersonal violence: fostering multidisciplinary perspectives. NIAAA Research Monograph, vol. 24 ( pp. 3 – 36) (NIH Publ. No. 93-3496; Rockville, MD). Room, R., & Rossow, I. (2000). The share of violence attributable to drinking: what do we know and what research is needed. Paper presented at Alcohol and Policy XII Meeting, Washington, DC, June 12, 2000. SAMHSA. (1999). Summary of findings from the 1998 National Household Survey on Drug Abuse. Rockville,

S.E. Martin, K. Bryant / Journal of Substance Abuse 13 (2001) 563–581

581

MD: Department of Health and Human Services (available on the web at http://www.samhsa.gov/OAS/ NHSDA/98SummHtml/NHSDA98Summ.htm). Spunt, B., Brownstein, H. H., Goldstein, P., Fendrich, M., & Liberty, H. J. (1995). Drug use by homicide offenders. Journal of Psychoactive Drugs, 27, 125 – 134. Spunt, B., Goldstein, P., Brownstein, H. H., Fendrich, M., & Langely, S. (1994). Alcohol and homicide: interviews with prison inmates. Journal of Drug Issues, 24, 143 – 163. Steffensmeier, D., & Allan, E. (1996). Gender and crime: toward a gendered theory of female offending. Annual Review of Sociology, 22, 459 – 487. Taylor, S. P., & Leonard, K. E. (1983). Alcohol and human physical aggression. In: R. G. Green, & E. I. Donnerstein (Eds.), Aggression: theoretical and empirical reviews, vol. 2. San Diego, CA: Academic Press. U.S. Department of Justice, National Institute of Justice. Arrestee Drug Abuse Monitoring (ADAM) program in the United States, 1998 (computer file). ICPSR version. Washington, DC: U.S. Department of Justice (producer) (1999). Ann Arbor, MI: Interuniversity Consortium for Political and Social Research (distributor) (1999). Valdez, A., Kaplan, C. D., Curtis, R. L. Jr., & Yin, Z. (1995). Illegal drug use, alcohol and aggressive crime among Mexican American and white male arrestees in San Antonio. Journal of Psychoactive Drugs, 27, 135 – 143. Valdez, A., Yin, Z., & Kaplan, C. D. (1997). A comparison of alcohol, drugs and aggressive crime among Mexican – American, black and white male arrestees in Texas. American Journal of Drug and Alcohol Abuse, 23 (2), 249 – 265. Visher, C. (1990). Reported use of alcohol by arrestees. Drugs and crime 1990: annual report of drug use forecasting. Washington, DC: National Institute of Justice. White, H. R., Brick, J., & Hansell, S. (1993). A longitudinal investigation of alcohol use and aggression in adolescence. Journal of Studies on Alcohol, Suppl. 11, 62 – 77. White, H. R., & Gorman, D. M. (2000). Dynamics of the drug – crime relationship. In: Criminal justice 2000, vol. VI ( pp. 151 – 218). Washington, DC: National Institute of Justice. Wieczorek, W., Welte, J., & Abel, E. (1990). Alcohol, drugs, murder: a study of convicted homicide offenders. Journal of Criminal Justice, 18, 217 – 227. Wolfgang, M. E. (1958). Patterns of criminal homicide. New York: Wiley.