Using data collected in Anchorage as part of the Arrestee Drug Abuse Monitoring (ADAM) program, this article reports the preliminary findings of a developmental study of the connection between the use of illicit drugs and criminal offending. The analytic approach used for this study differs from much previous research on the drug use-crime connection, which has tended to take a rather static view and focus on whether or not an individual was "under the influence" of one or more drugs at the time the offense was committed. In contrast, the analysis presented here focuses attention on arrestees' drug use over time. The time period examined here is the 12-month period preceding arrest. Results show that for each of five illicit drugs (marijuana, crack cocaine, powder cocaine, methamphetamine, and heroin), there were dramatic increases in: (1) arrestees' drug use levels, and (2) the number of arrestees who intensified their drug use, approximately 90 to 120 days prior to arrest. These findings suggest that analysis of drug use trajectories preceding crime events may be important for furthering our understanding of the drug use-crime connection.
The article is separated into six sections. The first section provides a brief overview of other sources of data on the drug use of criminal offenders, as well as a context for the data presented here. Then, in section two, there is a description of the ADAM program and the data it produced. Following this, in section three, is a presentation of the demographic characteristics of the respondents included in the present study. Section four presents the data on the drug use behaviors of the sample. Section five presents information on criminal offending for the sample. Finally, section six presents a summary and discussion.
Drug Use-Crime Data Sources
National Survey on Drug Use and Health. The National Survey on Drug Use and Health (NSDUH) collects drug use information from a nationally representative sample of non-institutionalized persons age 12 and older. Individuals not included in the survey are the homeless, prisoners, active duty military personnel, those residing in nursing homes, and those hospitalized for mental illness. Data is collected via face-to-face interviews with sampled respondents. Respondents are asked if they have used drugs/alcohol in their lifetime, in the past year, and in the past month. To measure criminal involvement, the NSDUH asks respondents 18 and over if they were arrested and booked for various crimes in the past year.
Monitoring the Future. Monitoring the Future (MTF) is an ongoing survey of American secondary students, college students and young adults. Each year a random sample of approximately 50,000 eighth-, tenth- and twelfth-graders in both public and private schools are surveyed about their use of drugs and alcohol. Systematically excluded from this sample are children not enrolled in school. (Students and parents are advised in writing in advance of the survey, and there is the option to decline to participate.) Participants are asked about their drug/alcohol use over their lifetime, the past year, and the past month. Questionnaires are mailed to a subsample of participants for several years following the initial data collection. To measure criminal offending, respondents are asked about their own criminal/deviant behavior and their experiences as crime victims.
National Crime Victimization Survey. The National Crime Victimization Survey (NCVS) is an ongoing study of U.S. household residents age 12 and older. Like the NSDUH, the NCVS systematically excludes the homeless and institutionalized populations. The survey has been conducted every year since 1973 and was designed with four objectives in mind: (1) to collect detailed data on correlates and consequences of criminal victimizations; (2) to estimate the "dark figure of crime"-that is, the number and types of crimes not reported to police; (3) to develop uniform prevalence measures of specific crime types; and (4) to provide a way to compare rates of crime over time and geography. To measure drug/alcohol involvement in crime, the NCVS asks victims whether or not the person or persons who victimized them were under the influence of drugs/alcohol when the event occurred. Use of drugs/alcohol by victims is not measured.
Survey of Inmates in Local Jails. Produced approximately every five years, the Survey of Inmates in Local Jails (SILJ) collects drug and alcohol use information from a nationally representative sample of individuals held in jail prior to trial, those serving sentences in local jails, and individuals awaiting transfer to state prisons. Inmates are asked about their drug and alcohol use for three time periods: ever, the month preceding the offense for which they were currently jailed, and their use of drugs and/or alcohol at the time of the offense for which they were currently jailed. In addition, respondents provide their perceptions of the drug/alcohol intoxication of victims.
Survey of Inmates in State and Federal Correctional Facilities. The Survey of Inmates in State and Federal Correctional Facilities (SISFC) collects drug and alcohol use information from a nationally representative sample of sentenced inmates in state and federal prisons. Participants in this study are presented with the same set of questions presented to those who participate in the Survey of Inmates in Local Jails: had they ever used drugs/alcohol, had they used in the month preceding the offense for which they were currently jailed, and were they under the influence of drugs/alcohol at the time they committed the offense for which they were now serving time. In addition, respondents provided their perceptions of the drug/alcohol intoxication of victims.
The Association Between Drug Use and Criminal Offending
The correlation between drug/alcohol use and crime is well established and not in dispute among either criminal justice practitioners or researchers. Regardless of the data sources used, drug use is shown to be associated with criminal offending. Consider, for example:
â€¢ A recent analysis of NSDUH data published by the Substance Abuse and Mental Health Services Administration revealed that 60 percent of those arrested for a serious crime in the past year also used at least one illicit drug; the comparable figure for illicit drug use for those who were not arrested for a serious crime in the past year was 13 percent.
â€¢ Data from the SISFC show that 32 percent of state prisoners and 26 percent of federal prisoners committed their offense under the influence of drugs. Roughly half of each group (56% and 50% respectively) reported using drugs within a month of their arrest.
â€¢ According to SILJ data, 28 percent of convicted individuals incarcerated in local jails were under the influence of drugs at the time of their offense; 54 percent used drugs within a month of their arrest.
â€¢ NCVS data show that one in four victims of violent crime reported that the person who attacked them was under the influence of drugs or alcohol.
However, the fact that drug use and criminal offending are correlated does not necessarily mean that they are causally related. In fact, there is a consensus among researchers that there is no empirical support for a direct causal relationship between drug use and criminal behavior. The connection between drug use and crime is much more complex.
The challenge for researchers is to document this complexity and develop empirically verifiable explanations for the drug use-crime connection. Just how the information collected by the ADAM program can contribute to this effort is outlined below.
Funded by the National Institute of Justice, the Arrestee Drug Abuse Monitoring (ADAM) program was a national, multi-site drug monitoring research platform designed to collect detailed information on alcohol/ drug use behaviors among those booked into local jails and charged with violation of at least one local or state criminal statute. (Those who were charged exclusively for federal offenses, as well as those booked into a facility for non-criminal violations, were not eligible for inclusion in the study.) Researchers at each site entered local jails for 14 consecutive days four times a year to conduct face-to-face interviews. Anchorage was an ADAM site from 1999 to 2003.
ADAM collected six broad categories of information: (1) arrest and charge information, (2) demographics, (3) drug/alcohol use, (4) drug/alcohol dependence and abuse, (5) drug market information, and (6) urine sample analyses. Arrest and charge information was obtained from official booking documents. All demographic, drug use, and drug market information was self-reported by arrestees.
Much of ADAM's utility for the study of the drug use-crime connection is not due to its sample population, its collection of lifetime, past-year, and past-month drug use data, or even the information it collects about the extent to which offenders were under the influence of illicit drugs when they committed crimes. (The NSDUH, MTF, NCVS, SILJ, and SISFC studies include all or some of these information categories.) Rather, ADAM's unique contribution was the use of a calendaring methodology to capture detailed month-by-month information on the intensity of drug use by arrestees for the entire 12-month period preceding their arrest. Because of its use of drug use calendars, ADAM was unique in its ability to capture critical information on individual drug use patterns prior to the crime event that precipitated an arrest. This data collection strategy enables researchers to better contextualize crime incidents and to examine the drug use-crime connection developmentally, as a dynamic process.
From a methodological standpoint, the use of a calendar is beneficial because it improves the reliability and validity of retrospective data. Calendaring accomplishes this by helping respondents to accurately recall events through the use of visual cues and related life events, and to sequence events over time. Researchers have successfully gathered drug use data going back as far as five years using this technique.
The ADAM drug use calendar (see Figure 1) was administered to all arrestees who reported using illicit drugs within the past year. At the outset, respondents were asked to note significant life events in the calendar-for example, religious holidays, birthdays, anniversaries, and major happenings in their life that occurred in the past year. These significant life events were then used as reference points by respondents to aid with recall of their drug/alcohol use. Arrestees were first asked if they had used a particular drug in a particular month, then if they responded in the affirmative, they were asked to provide their best estimate of the frequency of their drug use (see Figure 1 for use level definitions). Interviewers then recorded an arrestee's use level for each month. Interviewers worked through the calendar from top to bottom, from left to right. Arrestees were asked about each drug singly and in sequence.
In addition to providing self-reported drug use, arrestees were also asked to provide a urine sample. (Across the four-year period 2000-2003, an average of 86.4 percent of arrestees who were asked for a urine sample provided one.) These urine samples were immediately sent to an independent laboratory for enzyme multiplied immunoassay testing (EMIT), a technique designed to detect the presence of drugs or drug metabolites in urine. Separate immunoassays were designed to detect one particular drug or class of drugs. Although the ADAM program tested for ten drugs in all, this article focuses on arrestees' use of four of the five drug categories specified by the National Institute on Drug Abuse for routine workplace drug testing, commonly referred to as the "NIDA-5" panel: marijuana, cocaine (both crack and powder), methamphetamine, and heroin (opiates). (The fifth NIDA-5 drug is phencyclidine (PCP).)
The sample of arrestees examined here is limited to adult males who reported using one or more of the following illicit drugs within a year of their arrest: marijuana, crack cocaine, powder cocaine, methamphetamine or heroin. Table 1 provides a demographic portrait of this sample.
Age. Respondents in the sample were typically young. The three most frequent ages reported by this group of arrestees were 19 years (5.4% of the entire sample), 20 years (6.1%), and 21 years (5.8%). The average age was 30.8 years. Because the study was limited to adults, the youngest members of the sample were 18 (4.6%); the oldest respondent was 59 years of age.
Race/Ethnicity. A preponderance of arrestees (47.4%) characterized themselves as White; nearly a third (30.8%) reported Alaska Native/American Indian ancestry; one in every seven described themselves as Black or African American. Slightly more than two percent reported Native Hawaiian/Other Pacific Islander heritage; less than two percent identified themselves as Asian; and slightly less than four percent indicated some other racial group membership. In excess of 90 percent of the sample was non-Hispanic. More than 95 percent claimed American citizenship.
Educational attainment. The data presented in Table 1 reveal significant educational deficits among male arrestees in Anchorage. Nearly one-fifth of the sample had not earned a high school diploma or equivalent degree. An additional 48 percent of the sample had earned a high school diploma or GED, but had gone no further in their formal educational careers. One in ten reported completing some form of vocational training. Nearly 20 percent of the sample reported having some college education or two-year associate degree. A scant three percent of the sample had attained a four-year college degree or higher.
Employment. Only half of the sample were active participants in the labor market. More than 40 percent of the sample were able to work, but were not employed on the day of arrest. Notably, one out of every five of those who were unemployed had dropped out of the workforce entirely, having quit looking for work. Among those able to work, but not working, were small numbers of homemakers (0.7%), full-time students (0.9%), and retirees (0.3%). Just over five percent of the sample was unable to work due to some form of disability.
Marital status. Nearly two-thirds of the sample (61.6%) reported their marital status as single, never married. Approximately 18 percent of the sample was married (including common law marriage) on the day of arrest; another 18 percent were separated or divorced. Just over 1 percent of the sample was widowed.
Overall, the demographic composition of the adult male arrestees differed greatly from that of the overall population of adult males residing in Anchorage. Minority group members were disproportionately represented in the adult male arrestee population. This pattern is particularly striking for Alaska Natives, who represent approximately 30 percent of the adult male arrestee population, but less than 6 percent of all adult males in the city. African Americans were also disproportionately represented in Anchorage's adult male arrestee population, as were those of Hispanic background.
Arrestees also demonstrated notable human and social capital deficits when compared to the general adult male population. Those booked into jail were twice as likely to not have earned at least a high school diploma or equivalent degree. They were roughly one-tenth as likely to have completed a four-year college degree or higher. Compared to the general adult male population, arrestees were much less likely to be active participants in the labor force. Fully half of the arrestee sample was not working on the day of their arrest. (Eight percent of the sample had dropped out of the workforce entirely, reporting their work status as "unemployed, not looking for work.") Finally, arrestees were much less likely to be married than adult males more generally.
Detailed information for the three most-serious charges leveled against each arrestee was obtained from booking records and charge documents. Table 2 presents descriptive statistics for the seven offense categories most often observed.
In general, drug-using male arrestees were booked into jail for non-violent misdemeanors and other low-level violations (see Table 2). More than half (54.7%) of all adult males booked into jail in Anchorage were charged with a single offense, just over a quarter (27.8%) were charged with two offenses, and 17.5 percent of were charged with three or more. Approximately a third of all arrestees were accused of felonies. Between a quarter and a third were booked into jail for one or more violent offenses, and roughly one-fifth were arrested for at least one property crime. One out of every six arrestees was jailed for violating the terms of their probation or parole. In excess of 15 percent of male arrestees were booked into jail for driving under the influence. A substantial number of arrestees were jailed because a bench warrant had been issued for their arrest (most frequently for failing to appear in court). Notably, only about five percent of this sample of drug users were arrested for committing a specific drug or alcohol offense.
Arrestee Drug Use
The results of the EMIT screening results for marijuana, cocaine, methamphetamine, and heroin are presented in Table 3. These data depict the prevalence of use for each of these drugs at the time of arrest (see Table 4 for EMIT detection periods for each drug).
Marijuana was the most common illicit drug used by Anchorage's male arrestee population, with more than half of all male arrestees testing positive in 2003. Approximately a fifth of men booked into the Anchorage jail tested positive for cocaine. Less than ten percent tested positive for heroin (or other opiates-see note at bottom of table), and roughly one percent tested positive for methamphetamine.
In addition to illustrated overall prevalence rates, the data presented in Table 3 also show a pattern of increased drug use among male arrestees for each of the four drugs examined. Between 2000 and 2003 the percentage of arrestees testing positive for marijuana increased 41.8 percent, the percentage testing positive for cocaine increased 21.1 percent, the percentage testing positive for heroin (or other opiates) increased 125 percent, and the percentage testing positive for methamphetamine increased by 400 percent (though, importantly, the base rate was very small).
Self-reported drug use
Table 5 presents arrestees' self-reported use for each of the five drugs of interest at four points in time. From left-to-right, the table presents the percentage of the sample that used at least one of the five drugs at least once in their lifetime, in the past year, in the past month, and in the past week. These data demonstrate two empirically distinct drug use patterns. The first is what might be termed the "hierarchy of preference." Marijuana is, by far, the most frequently used drug among adult male arrestees who use drugs. Among those who used drugs in the past year, more than 97 percent reported using marijuana at least once in their lifetime. Between half and two-thirds of drug users reported using powder cocaine at least once. Slightly fewer had used crack cocaine. Roughly a third of drug-using arrestees had used methamphetamine on at least one occasion; about one in every seven had experience with heroin.
Drug use over time
We come now to the examination of arrestees' drug use patterns for the year preceding their arrest. The data presented here come from the drug use calendars administered to all arrestees who reported using illicit drugs within 12 months of their arrest. For each month in the preceding year, arrestees were asked to provide their best estimate of their drug use level ("0"= No drug use, "1"= drug use on 1-7 days of the month, "2"= drug use on 8-12 days of the month, "3"= drug use on 13-30 days of the month). Figure 2 presents the average use level for each drug, for each month preceding arrest.
On average, the marijuana users demonstrated the highest average use levels, surpassing even heroin users (except for the month of arrest when the average use level for heroin was highest). Crack cocaine users recorded the third-highest average use level, although by the month of arrest they had closed the gap on marijuana and heroin users. Powder cocaine and methamphetamine users had the two lowest average use levels.
The data presented in Figure 2 also show that average use levels increased for all five drugs. Relative to the other illicit drugs examined, marijuana demonstrated the most stable use pattern. Between months 12 and 1 before arrest, there was a 21.8 percent increase in the average use level for marijuana-a substantial increase to be sure. However, the average use levels for powder cocaine, methamphetamine, crack cocaine, and heroin increased by 122 percent, 99 percent, 75 percent, and 55 percent respectively over the same time span.
But these increases in drug use levels did not occur gradually or in a linear fashion. Instead, they accelerated rapidly in a short period of time. After remaining stable for most of the year, arrestee drug use levels grew exponentially in a period of about 90 days prior to arrest. In fact, nearly all of the observed increases in drug use levels over the 12-month period were compressed into this small 90-day window of time. Notably, this pattern appeared for every drug, not just one or two of them.
Did this pattern apply to all (or nearly all) arrestees, or was the sharp increase due to only a select few? To address this question, calendar data were used to calculate the percentage of arrestees who experienced an increase in their drug use level from month to month. These data are presented in Figure 3. These data show that for each drug, there was a marked increase in the number of users whose use intensified in the year leading up to their arrest. For all drugs except powder cocaine the increase was particularly pronounced within 90 days of arrest.
Taken together, these data presented in Figure 2 and Figure 3 suggest that it was a limited group of arrestees who experienced dramatic increases in their drug use levels in the weeks immediately preceding arrest, and that it was this group that was largely responsible for pushing up the overall sample averages.
Summary and Discussion
Using data collected in Anchorage for the Arrestee Drug Abuse Monitoring program (ADAM), this article has presented preliminary findings of a developmental study of the drug use-crime connection. Results showed an explosive increase in arrestee drug use approximately 90 days prior to arrest. Importantly, this pattern of increased use was observed for all five illicit drugs examined, not just a few. Further analysis showed that much of the growth in arrestee drug use was likely due to an intensification of use by a relatively small cadre of individuals, rather than a uniform increase in drug use for all arrestees.
How do these findings contribute to our understanding about the relationship between drug use and crime?
Because this study is ongoing, it is premature to make any firm declarations about the significance of this research. With that caveat in mind, I would like to highlight some potential contributions. First, this research demonstrates that there is substantial variation in the frequency of drug use over time among illicit drug users. This straightforward (and seemingly self-evident) finding provides an important contrast to a great deal of existing research that, due to its emphasis on cross-sectional correlations between drug use and criminal offending and comparisons between "users" and "non-users," implies invariance in drug use over time. The data presented here suggest that perhaps rather than focusing simply on drug use per se, those who seek to understand the drug use-crime connection should expand their perspective to include the frequency of drug use as well. In addition, this study shows that for roughly one in five arrestees, increases in drug use levels were not gradual or linear; they were sudden, and the magnitudes of the increases in drug use levels were exceptionally large. Related to this are the results pointing to the 90-day window preceding arrest, when drug use levels began their steep incline. This may be a period of particular importance in the development of individuals' drug use trajectories. Finally, this research highlights the potential utility of using the calendar data collected by the ADAM program for advancing research on the drug use-crime connection.
For more information on ADAM, see the Annual Report 2000: Arrestee Drug Abuse Monitoring release in 2003 and available at http://www.ncjrs.gov/pdffiles1/nij/193013.pdf.
Brad A. Myrstol is an Assistant Professor with the Justice Center.