Data collected in the Arrestee Drug Abuse Monitoring (ADAM) program on alcohol and other drug use among arrestees provide a valuable opportunity to examine the relationship between alcohol use and violence. The data are used to explore the combined use of alcohol and other drugs among offenders and the relationships between substance use and the offenders’ demographic characteristics and offenses. These findings are used to identify changes in the offenders’ alcohol and other drug use over time. KEY WORDS: AODR (alcohol or other drug [AOD] related) crime; arrest; offender; AODR violence; self-report; AOD abuse; demographic characteristics; gender differences; ethnic differences; age differences

Although and crime is alcohol is legally available to adults age 21 and older and is the most widely used drug in the United States, the relationship between alcohol use the focus of less criminal justice research and receives less public attention compared with the research and attention devoted to the links between illicit drug use and crime. This article examines data collected from arrestees concerning their alcohol and other drug use prior to their offenses and explores the relationships among alcohol use, illicit drug use, arrestee characteristics, and charges filed at arrest as well as the implications of these findings.

Since the mid-1980s, the National Institute of Justice (NIJ) has tracked drug use by the persons arrested (i.e., detained by police) and booked (i.e., taken to central police facility for a preliminary judicial hearing) in the United States through its Drug Use Forecasting (DUF) system. In 1995 the survey instrument was modified and expanded. In 1998 the system was renamed the Arrestee Drug Abuse Monitoring (ADAM) program and increased from 24 to 35 cities with improved sampling procedures. Although the ADAM program continues to focus primarily on illicit drug use--including types of drugs used by arrestees, self-reported dependence, and the relationship between drug use and certain types of offenses-- self-reported alcohol-use information is also collected. Such data allow researchers to examine the association between recent alcohol consumption and intoxication (both alone and in combination with other drugs) and arrests for various types of crimes. This article explores those associations.

The ADAM data provide a unique perspective on the relationships between alcohol and other drug use and crime. Most other studies of alcohol and crime rely on the victims’ reports of offenders’ drinking (e.g., the National Crime Victimization Survey [NCVS], conducted annually by the Bureau of Justice Statistics [BJS]); self-report surveys of correctional populations or periodic BJS censuses of Federal, State, and local correctional facilities; arrest data from the Federal Bureau of Investigation’s Uniform Crime Reporting program; archival data from arrest records in specific jurisdictions; or select data on specific crimes.

Studies based on victimization data may understate alcohol’s presence, particularly in cases where the offender was a stranger, because the victim is less likely to have observed the offender drinking or to have known his or her drinking status. Studies based on correctional populations’ self-reports tend to suffer from problems with retrospective recall that may either underestimate or overestimate alcohol’s presence¹ (¹ Although most studies find that underreporting increases when the length of the recall period is more than 1 year, a BJS survey of prisoners calculated their blood alcohol concentration (BAC) based on self-reported drinking at the time of the offense and estimated that men had an average BAC of 0.27 percent and women had an average BAC of 0.37 percent (Greenfeld 1998).) and, subsequently, form a sample skewed toward the most serious offenders (i.e., people who have been convicted of crimes for which they have been incarcerated). Studies that rely on archival arrest report data also may understate alcohol’s presence, because the studies’ accuracy depend on how accurately police monitor and report the information, a process which is not consistent. Conversely, ADAM data are gathered from a representative sample of people who have been arrested for a wide spectrum of offenses and the arrestees’ reports on their alcohol or other drug use shortly after the substance-use occurs. In addition, the arrestees’ urine is tested for the presence of illicit drugs.

Methods

Arrestees who participate in the ADAM program are surveyed concerning their substance use and, at the conclusion of the interview, are asked to provide a urine specimen, which is screened for the presence of 10 illicit drugs. In 1998 data were obtained from 20,715 adult male arrestees who were interviewed at 35 sites and 6,699 adult female arrestees who were interviewed at 32 sites. The data collection area for each site included the county within which the city was located.² (² Two exceptions exist: (1) Anchorage data represent only that city, and (2) because New York City includes five separate counties, data are drawn from each of those boroughs.)

For each quarter in 1998, 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 hours. On average, each site sought to obtain a sample of 225 males and 100 females per quarter. All arrestees booked at a facility within the previous 48 hours were eligible for interviews. However, because more male arrestees were available than could be interviewed, the sample of men was selected on the designated interview days using the following order of priority based on reason for arrest: (1) nondrug felony, (2) nondrug misdemeanor, (3) drug felony, and (4) drug misdemeanor. Sites were asked to survey no more than 20 percent of the males charged with drug offenses. All women arrestees were eligible for an interview regardless of their charges.

Approximately 80 percent of eligible arrestees agreed to be interviewed. Of those who agreed to the interview, about 80 percent also provided a urine specimen. Only persons who agreed to the interview and provided a specimen were included in the data (i.e., nearly two-thirds of eligible arrestees). (For further information about data collection procedures and the data set, see the U.S. Department of Justice 1999.)

During the ADAM interview, arrestees were asked about their use of both legal and illegal drugs, including alcohol. Alcohol-related questions addressed the interviewed arrestees’ lifetime, past year, and past month alcohol use as well as alcohol use within the previous 72 hours (termed “recent alcohol use” throughout this article). Interviewers also asked arrestees if they were “under the influence” of alcohol at the time of the offense for which they were arrested. Similar questions were asked about each of a number of illicit drugs.

We analyzed two alcohol-related ADAM items: (1) recent alcohol use and (2) being “under the influence of alcohol” at the time of the offense (also termed alcohol abuse or intoxication). In this article, we present findings for both measures in order to provide a more complete picture of the role of alcohol in crime.

The ADAM survey instrument does not include questions related to quantity of alcohol consumed. Because the arrestees who reported recent alcohol use may have consumed an unspecified amount of alcohol and done so at any point during the 72 hours before being arrested, one cannot infer from this question that alcohol use affected the arrestee’s behavior. Asking arrestees whether they were “under the influence” yields a subjective self-report measure that may be systematically biased. In contrast to the self-reported alcohol measures, the illicit drug data are based on positive urinalysis. However, neither urinalysis nor breath testing for alcohol use is practical as part of the ADAM data collection, because alcohol is metabolized so quickly.

To address threats to the validity of the data that may be associated with self-reporting alcohol use, we employed three strategies. First we examined self-reports of persons whose responses to drug-use questions were inconsistent with their urine specimens and compared them with consistent reporters to determine if their levels of alcohol use differed. Second, we compared ADAM self-reported alcohol use and abuse data by age group with similar data from a nationally representative survey of alcohol and other drug use. Third, we reported both alcohol use and abuse, based on the assumption that recent use (i.e., drinking any alcohol in the 72 hours before the arrest) overstated the possible effect of drinking on criminal behavior, whereas self-reported “being under the influence” at the time of the arrest probably understated alcohol’s effects.

We examined the ADAM data to uncover subjects with discrepancies between urine test findings and self-reported use of marijuana, crack, and cocaine and labeled them as “inconsistent.” We then compared consistent and inconsistent arrestees regarding several measures of drug use. Using two cocaine-related variables, respectively, we identified (1) persons who admitted having used crack or cocaine but denied using it in the past 30 days although their urine tested positive (5.8 percent of the total sample) and (2) persons who denied ever using crack or cocaine in the previous month but whose urine tested positive (an additional 8 percent of the sample).³ (³ Although cocaine can be detected in urine for approximately 36 hours, metabolites of cocaine can be detected for up to 2 or 3 weeks in heavy users. For that reason, as well as to be consistent in measures across drugs, we adopted a conservative consistency measure for recent cocaine use.)  Similarly, two marijuana consistency measures identified (1) arrestees who denied use in the past month, admitted to having tried marijuana in the past, but tested positive (4.5 percent of the total) and (2) arrestees who denied ever using marijuana but tested positive for the drug (an additional 1.2 percent of the total). Three additional variables were created to identify arrestees who admitted ever using any of the aforementioned drugs but denied recent use (9.9 percent); arrestees who denied ever using either marijuana, crack, or cocaine but tested positive (9.5 percent of the sample); and arrestees who gave an inconsistent report on any of the validity checks (18.7 percent of the sample).

Like Harrison (1995) who studied DUF arrestees, we found reasonable congruence between self-report and urinalysis among respondents and drug-use underreporting based on social desirability. Thus, marijuana use was more likely to be reported than cocaine use, and ever use was more likely to be reported than recent use.

To assess how much arrestees underreported their alcohol use and abuse, we compared the self-reports of alcohol use and abuse for the respondents who gave consistent self-reports and urine with the inconsistent reporters. In our ADAM sample, no statistical difference existed in the percentage of respondents who reported recent alcohol use among those who had inconsistent self-reports and urine tests regarding recent or lifetime use of marijuana, cocaine, or crack and respondents whose drug tests and self-reports were congruent. With respect to being under the influence at the time of the offense, no difference existed between respondents who gave self-reports about recent drug use inconsistent with their urine tests and those who did not. However, respondents who denied ever using marijuana, cocaine, or crack but who tested positive for one of those drugs were significantly less likely to report being under the influence of alcohol (16.6 percent) than were the consistent reporters (21.1 percent). Because the number of persons in this group was quite small, however, we did not adjust the data or delete this group from the data reported here.⁴ (⁴ We estimate that the maximum number of persons who may have been dissimulating about alcohol intoxication was 107 respondents out of more than 27,000.)  To address possible greater underreporting of alcohol consumption by underage offenders than those persons legally permitted to drink, we compared ADAM data on drinking during the previous 30 days, broken down by age group, with similar data from the 1998 National Household Survey of Drug Abuse (NHSDA) (SAMHSA 1999). Although the percentage of respondents who reported using alcohol in the past month was consistently higher among ADAM arrestees than the house-hold survey respondents for each age category, the age-related patterns of use for both samples were similar. For example, in the NHSDA for 1998, 32 percent of 16- and 17-year-olds, 54 percent of 18- to 20-year-olds, 65 percent of 21- to 25-year-olds, and 53 percent of people age 35 and older reported drinking in the past month. Among the ADAM arrestee sample, the comparable figures are 54 percent for 16- and 17-year-olds, 67 percent for 18- to 20-year-olds, 73 percent for 21- to 25-year-olds, and more than 70 percent for arrestees older than 25. Thus the lower rate of self-reported alcohol use among persons less than age 21 is consistent with national data and is not likely to be caused by reporting bias.

Findings

The ADAM data provide an opportunity to explore the combination of alcohol use, drug presence, and offense data among recent arrestees; to examine possible links between alcohol use and abuse--alone and in combination with other drugs--and specific types of criminal behavior; and to identify changes in these associations since 1989, when the only DUF report containing alcohol data was published (Visher 1990). Given the ample literature indicating an association between alcohol use and violence (Murdoch et al. 1990; Parker and Rehbun 1995; Pernanen 1991; Roizen 1993; Lipsey et al. 1997), we focused specifically on the presence and influence of alcohol in violent crimes compared with other offenses, particularly property crimes.

A View of the Arrestees

The sociodemographic characteristics of the 1998 ADAM arrestees are shown in table 1.

Table 1 Characteristics of 1998 ADAM* Program Arrestees

 *ADAM = Arrestee Drug Abuse Monitoring. The ADAM program is a system used by the National Institute of Justice to track drug use by a representative sample of people in the United States who have been arrested and charged for a wide variety of criminal offenses.

 NA = not applicable; GED = general equivalency diploma.

 The arrestees tended to be young (i.e., median age of 30), but ranged in age from 15 to 82. Most arrestees were male (75.6 percent), African-American (45.6 percent), had limited education (48 percent had not finished high school), and were employed either full time (42.5 percent) or part time (18 percent) during the month prior to their arrests.

The vast majority of both male and female 1998 arrestees reported that they tried alcohol at some time during their lives (94 and 89 percent, respectively), and most had also consumed alcohol during the past year (80 and 70 percent, respectively). As shown in table 2, more than one-half of all male arrestees (54 percent), but somewhat less than one-half of the female arrestees (43 percent), reported drinking in the 72 hours prior to their arrests, and 22 percent of the males and 16 percent of the females reported being under the influence of alcohol at the time of their offenses.

Table 2

Alcohol and Cocaine Use/Abuse Among Percentages of 1998 ADAM¹  (¹ADAM = Arrestee Drug Abuse Monitoring. The ADAM program is a system used by the National Institute of Justice to track drug use by a representative sample of people in the United States who have been arrested and charged for a wide variety of criminal offenses.) Arrestees by Gender and Charge