U.S. Department of Justice Office of Justice Programs National Institute of Justice National Institute of Justice R e s e a r c h i n B r i e f June 2001 Issues and Findings The Rise of Marijuana as the Discussed in this Brief: Trends in marijuana use detected through Drug of Choice Among Youthful urinalysis among booked adult arrestees at 23 locations across the Nation served by the Arrestee Drug Adult Arrestees Abuse Monitoring (ADAM) program By Andrew Golub and Bruce D. Johnson from 1987 through 1999 as well as trends within the mainstream popu- lation based on self-reports of past- Various surveys have identified a rapid locations across the Nation served by month marijuana use recorded by increase in marijuana use during the the Arrestee Drug Abuse Monitoring the National Household Survey on 1990s, especially among youths. This (ADAM) program—formerly the Drug Drug Abuse (NHSDA) and Moni- raises a variety of questions about the Use Forecasting (DUF) program—from toring the Future (MTF) programs. future of the Nation’s drug problems. On 1987 through 1999. Key issues: An epidemiological one hand, the gateway theory posits that In addition, this report identifies na- perspective is taken to place in youthful use of alcohol and/or tobacco context the increased use of mari- and marijuana tends to precede use of tionwide drug use trends within the juana among arrestees and the other illicit drugs like crack and heroin mainstream population on the basis of general population. The course (see “The Gateway Theory”). The recent self-reports of past-month use, a measure of the recent marijuana upsurge increase in youthful marijuana use has roughly parallel to the length of time in is compared with that observed fueled speculation that a new epidemic which marijuana can be detected by uri- for previous crack and heroin of hard drug abuse may be imminent1 and nalysis. Those trends were derived from epidemics in which four phases that the burden of drug abuse will be dra- data collected by the National Household with distinct variations in preva- matically increasing in the near future.2 Survey on Drug Abuse (NHSDA) and lence and age of users occurred. Monitoring the Future (MTF) programs. The analysis also compares time On the other hand, the start of this new (See “The Study’s Data Sources.”) trends in marijuana use across epidemic coincides with the decline of age groups, populations, and the crack epidemic. This suggests that Overall, study findings suggest the geographic locations. youthful subcultures may have shifted following: from the destructive nature of crack Key findings: abuse to the use of less dangerous drugs. Recent increases in youthful marijua- Increases in marijuana use na use followed a natural pattern sim- during the study period were lim- The recent upsurge in marijuana use is re- ilar to previous drug epidemics. Use ited primarily to youths. Starting ferred to as the New Marijuana Epidemic of a particular drug sometimes fol- around 1991, most ADAM loca- to distinguish it from widespread use of lows a wave of popularity: starting from tions experienced a rapid increase marijuana prevailing in the 1960s and a lull, expanding rapidly, leveling to a 3 in recent use among youthful 1970s. This Research in Brief examines plateau, and subsequently fading away. adult arrestees (ages 18–20), trends in marijuana use detected through Prior research with ADAM/DUF data urinalysis to track the progress of the suggests that the popularity of heroin recent epidemic among arrestees at 23 injection (which mostly peaked in the continued… R e s e a r c h i n B r i e f
Issues and Findings 1960s and early 1970s) and crack The New Marijuana Epidemic had …continued (which mostly peaked in the late 1980s) a larger impact on youthful adult followed this pattern. The current arrestees than on youths in the gen- from an average low of 25 analysis suggests that the recent wave eral population. The epidemic in percent in 1991 to 57 percent in 1996, as detected by urinalysis. of marijuana use has followed a simi- youthful marijuana use recorded by The MTF and NHSDA surveys also lar pattern so far, although unlike the the MTF and NHSDA programs start- recorded rapid but more modest previous epidemics, the increases in ed 1 year later, increased more slowly, increases in youthful marijuana use use were primarily limited to youths. and was less prevalent at its peak than within the mainstream population the epidemic among youths who tend- Local differences are important. starting in 1992 (1 year later than ed to get in trouble with the law as There were exceptions at a few ADAM among ADAM arrestees). Around recorded by the ADAM program. 1996, the rates of marijuana use locations to every one of the major among arrestee and mainstream regularities in the New Marijuana A conceptual model of the Epidemic. Some locations did not populations reached a plateau. New Marijuana Epidemic observe an epidemic. At other loca- With exceptions at a few tions, the epidemic either was not Much research suggests that drug ADAM locations, the pattern of limited to youthful adult arrestees epidemics tend to follow a predictable growth in marijuana use among (ages 18–20), expanded more slowly, course. This analysis employs a conceptual youthful adult arrestees was simi- lar to that observed previously expanded for a longer period, or was model that distinguishes the characteris- for heroin and crack: lull, rapid less prevalent at its peak. tics of four phases: incubation, expansion, expansion, and plateau. Use of plateau, and decline. This model was ori- In the 1990s, marijuana replaced ginally developed to explain the course of both heroin and crack is now crack cocaine as the drug of choice in decline. the Crack Epidemic.4 It has since been among youthful adult arrestees. used to study the Heroin Injection Epide- Arrestees born since 1970 have been Marijuana appears to be the mic and has been adapted for the study drug of choice for arrestees born increasingly likely to be detected as of the recent increase in marijuana use.5 since 1970, who seem much less recent marijuana users. Unlike their This study found that the dynamics of re- likely to progress to crack or heroin predecessors, however, few of them had cent increases in marijuana use followed injection than their predecessors. progressed to crack or heroin by 1998. a pattern similar to that of the Crack and This provides some evidence to suggest Target audience: Local law Heroin Injection Epidemics, suggesting that viewing marijuana as a gateway enforcement and public health that all three epidemics were the result drug may be inappropriate for this new officials, drug-crime researchers, of a comparable diffusion phenomenon. administrators of juvenile jus- generation. Ethnographic evidence tice agencies and youth depart- from New York City suggests that use Theoretically, the passing of each phase ments, and local criminal of marijuana by youths may be associ- of the New Marijuana Epidemic should justice policymakers. ated with strong cultural and subcul- result in a distinguishable pattern for the tural norms that militate against use prevalence of marijuana use detected by of more dangerous drugs. the ADAM program, particularly among The New Marijuana Epidemic had youthful adult arrestees and, to a lesser plateaued by 1996 at most affected extent, among the overall population of locations and by 1999 at all affected adult arrestees (ages 18 and older).6 locations. From 1996 to 1999, most Incubation phase. Historical evidence ADAM locations as well as the MTF suggests that a drug epidemic typically and NHSDA surveys identified stable, grows out of a specific social context; high levels of recent marijuana use the Heroin Injection Epidemic grew out among youths. of the jazz era7 and the Crack Epidemic started among inner-city drug dealers.8
2 R e s e a r c h i n B r i e f
In both cases, there was an initial T-shirts. In New York City, they also to youths. In contrast, the Crack and incubation phase during which the preferred to smoke their marijuana in Heroin Injection Epidemics spread new drug-use practice was developed a blunt (an inexpensive cigar whose first among adults and only afterward and nurtured among a relatively small, contents are replaced with marijuana). to youths. This dynamic suggests that cohesive group of adult users. Mari- The extent to which the New Marijua- the ADAM program would be expected juana use has been widespread since na Epidemic outside of New York City to detect rapidly increasing marijua- the 1960s; however, the prevalence is associated with blunt smoking is not na use among youthful adult arrestees of its use had been declining since clear. Unfortunately, major national during an expansion phase. The rate 1979.9 During the incubation phase, the surveys, such as MTF, NHSDA, and of use among all adult arrestees would ADAM program would be expected to ADAM, do not distinguish among ways be expected to increase more slowly detect relatively low levels of marijua- of consuming marijuana. A number and for a longer period as members of na use by adult arrestees, including of focus groups across the Nation on the New Marijuana Generation aged those in the youthful category. cigar use11 and reports by leading drug and came to constitute a larger portion abuse experts12 provide limited (but of the ADAM sample and as the re- Ethnographic research in New York far from conclusive) support for the newed interest in marijuana diffused City suggests that the reemergence idea that blunt smoking may be a to older arrestees. of interest in marijuana use was pio- national phenomenon. neered as part of the youthful, inner- Plateau phase. Subsequent to its expan- city, predominately black hip-hop Expansion phase. Eventually, mari- sion, the New Marijuana Epidemic movement.10 These youths celebrated juana use spread rapidly as part of a could be expected to enter a plateau marijuana use in their music and on newly emerging subculture indigenous phase at each ADAM location. During
The Gateway Theory
uch research has identified that disproportionately generate youths M a. Kandel, Denise B., “Stages in Adolescent most American youths tend to progress who get in trouble with both drug Involvement in Drug Use,” Science 190(1975): through as many as four stages of sub- abuse and the law.c Moreover, the gate- 912–914. stance use: nonuse, alcohol/tobacco, way sequence may no longer charac- marijuana, and other drugs including terize the experiences of mainstream b. Ibid. For a review of replications, see Kandel, cocaine and heroin.a Individuals who do youths. Calculations based on National Denise B., ed., Stages and Pathways of not use substances associated with one Household Survey on Drug Abuse data Involvement in Drug Use: Examining the Gateway stage rarely use those associated with suggest that youths coming of age in Hypothesis, New York: Cambridge, forthcoming. later stages, but not all users at one the 1990s were much less likely to c. Golub, Andrew, and Bruce D. Johnson, b stage progress to the next. Because of progress from marijuana to cocaine “Substance Use Progression and Hard Drug their intermediary role, alcohol, tobacco, powder, crack, or heroin than were Abuse in Inner-City New York,” in Stages and d and marijuana have come to be regarded youths born previously. Pathways of Involvement in Drug Use: Examining as “gateway drugs.” Today, policies per- the Gateway Hypothesis, ed. Denise B. Kandel, These recent studies suggest that youth- taining to substance use prevention seek New York: Cambridge, forthcoming. to forestall or delay youthful use of gate- ful substance use progression reflects cul- d. Golub, Andrew, and Bruce D. Johnson, way drugs to reduce the likelihood of tural or subcultural norms among youths “Variation in Youthful Risk of Progression From subsequent abuse of drugs like heroin about which substances are acceptable Alcohol and Tobacco to Marijuana and Hard and crack. and that these norms vary over time and across locations. Thus, it seems essential Drugs Across Generations,” American Journal of In contrast, several analyses suggest to monitor not just which substances Public Health 91(2)(2001):225–232. the gateway sequence may not be as re- youths are using but what that substance levant to the inner-city populations that use represents to them.
3 R e s e a r c h i n B r i e f this period, youths coming of age and were experiencing their decline phases. a rapid decrease in marijuana use getting involved with illegal drugs These drugs have been much less pop- among youthful adult arrestees but would use the current drug of choice, ular among youths coming of age in a slower, more drawn-out decline marijuana. During this phase, the the 1990s than among their predeces- among all adult arrestees. ADAM program would be expected sors. However, heroin injection and to detect stable and high levels of crack smoking are still quite wide- Results marijuana use among youthful adult spread because many older users have If data had conformed to the con- arrestees and slowly increasing rates persisted in their habits. By analogy, ceptual model described above, the of use overall. when the New Marijuana Epidemic expansion phase of marijuana use enters a decline phase, the ADAM should be readily distinguished by Decline phase. In the 1990s, both the program would be expected to detect Heroin Injection and Crack Epidemics steady increases in each year (perhaps
The Study’s Data Sources T he authors collected data from recruit samples of juvenile arrestees. This citations increased substantially. Several three major programs: Arrestee Drug study examined trends at the 23 locations locations experienced increases in detect- Abuse Monitoring (ADAM), National using information obtained from more ed marijuana use in 1999 inconsistent Household Survey on Drug Abuse than 300,000 arrestees between 1987 with the trend in previous years, includ- (NHSDA), and Monitoring the Future and 1999. ADAM samples typically are ing, most notably, Atlanta, Birmingham, (MTF). The absolute magnitude in the not representative of the general popula- Houston, Miami, Phoenix, San Diego, prevalence of recent marijuana use was tion in communities where data collection and San Jose. Continuation of the upward expected to differ among the foregoing occurs. Given the drug-crime nexus, ADAM trend in Los Angeles and Portland also data sources because of differences in data provide excellent information about may have been due to the change in sample populations (across ADAM loca- drug use among many of the most seri- sampling strategy. tions and across NHSDA and MTF surveys), ous drug abusers at each location. This differences in survey procedures, and the information is of particular interest to crim- Throughout the life of the DUF and use of urine tests for ADAM in contrast inal justice and other agencies. Analyses ADAM programs, urine testing and many to self-reports for NHSDA and MTF. of ADAM data may be of even broader core questions have remained constant, interest to the extent that drug use among allowing for analysis of trends over time. Arrestee Drug Abuse Monitoring arrestees tends to parallel or perhaps even Urine test results provide particularly valid program. In 1987, the National Institute lead trends in the general population. indications of recent marijuana use. Mari- of Justice (NIJ) established the Drug Use juana metabolites tend to remain in the Forecasting (DUF) program to measure At inception, the DUF program sought body. Marijuana consumption can be trends in illicit drug use among booked to monitor substance use among serious detected by the EMIT (Enzyme Multiplied arrestees in most large cities (or coun- offenders: individuals charged with a Immunoassay Testing) urinalysis screen ties) with a total population of at least 1 felony offense were oversampled; indi- used by ADAM up to 7 days after last use million, as well as in many smaller cities viduals charged with a citation offense for infrequent users and 30 days or longer for geographical diversity. In 1997, DUF were excluded from DUF samples at most for chronic users. In contrast, the drug de- evolved into the ADAM program, which locations; and individuals charged with tection period for opiates (such as heroin) plans to expand to 75 locations over the drug offenses were not allowed to exceed and cocaine is only 2 to 3 days. In 1996, next few years.a The program collects 20 percent of the sample.b As part of the the cutoff level for determining recent urine samples (along with self-reported transition to ADAM, NIJ phased in sam- marijuana use was lowered from 100 information) from about 300 adult ar- pling strategies so that ADAM samples to 50 nanograms.c More than 34,000 restees each quarter at each location. would be representative of arrestees pass- samples from 1995 were tested at both Female arrestees are oversampled at ing through the central booking facility at cutoff levels. Overall, the prevalence of many locations and constitute about 30 each location. Starting in 1998, the preva- detected marijuana use increased 5 to 7 percent of the total. Some locations also lence of arrestees for drug offenses and percentage points when the lower cutoff
4 R e s e a r c h i n B r i e f as large as 5 percentage points or on the speed at which marijuana use use among high school seniors declined more per year), the plateau phase dis- spread within the target population from a peak of 37 percent in 1978 to a tinguished by high rates of use in each represented by each survey sample. low of 12 percent in 1992 (an average year (perhaps varying by no more than decline of 1.8 percentage points per 2 percentage points from year to year), The findings from the NHSDA and year). NHSDA recorded a remarkably and the decline phase distinguished MTF surveys in exhibit 1 fit this over- similar decline in use among youthful by steady decreases in youthful mari- all pattern quite well, although the household members (ages 18–20) for juana use comparable in size to the rate of increase during the apparent this period. Then from 1992 to 1996, increases observed during the expan- expansion phase was modest. Nation- the rate among high school seniors sion phase. The rate of increase and wide, overall marijuana use had steadi- steadily increased to 22 percent (an subsequent decrease would depend ly declined from 13 percent in 1979 to average increase of 2.5 percentage a low of 4 percent in 1992. Marijuana
level was used. The difference was most 12 to 17. The authors used sample Juvenile Arrestees, Washington, DC: U.S. pronounced among very young arrestees weights in all calculations to obtain Department of Justice, National Institute (under age 15) and older arrestees (over unbiased estimates. of Justice, 2000, NCJ 181426. age 30), two groups that tend to use c. National Institute of Justice, Drug Use Monitoring the Future program. Each marijuana less frequently. Forecasting: Annual Report on Adult and spring since 1975, the University of Michi- Juvenile Arrestees, 1996, Washington, DC: National Household Survey on Drug gan’s Institute for Social Research has con- U.S. Department of Justice, National Institute Abuse program. NHSDA was established ducted a survey to estimate the prevalence of Justice, 1997, NCJ 176800. in 1971 to measure the prevalence and of drug use among high school seniors correlates of illegal drug use and monitor in the United States and to monitor trends d. Substance Abuse and Mental Health trends over time in the noninstitutional- over time.f The survey tends to under- Services Administration, National Household ized population of the United States.d sample many of the most serious drug Survey on Drug Abuse Series: H-10. Summary The survey tends to undersample many users, who are disproportionately likely of Findings from the 1998 National Household of the most serious drug abusers, who to drop out of school or be absent on the Survey on Drug Abuse, Rockville, MD: Office are prone to incarceration, residence in day of the survey. Students at selected of Applied Studies, 1999, U.S. Department of other institutions, and unstable living schools complete confidential question- Health and Human Services Publication No. arrangements. Eligible participants are naires at their own pace during a normal (SMA) 99–3295. visited and interviewed in their homes. class period. Analyses presented in this e. Ibid. Through 1998, respondents were given report are based on more than 350,000 f. Johnston, Lloyd D., Patrick M. O’Malley, a separate sheet to record their confi- responses obtained in the 1976–97 period and Jerald G. Bachman, National Survey dential answers to questions about drug and contained in a published report for Results on Drug Use from the Monitoring the use to help ensure disclosure of sensitive 1998–99.g The MTF program employs a Future Study, 1975–1998, vol. 1, Bethesda, information. The survey was conducted complex sampling design. The authors used MD: National Institutes of Health, National in 1971 and 1972 and then every 2 or sample weights in all analyses to obtain Institute on Drug Abuse, 1999, NIH Publi- 3 years until 1990, when it became an unbiased estimates of substance use. cation No. 99–4660. annual survey. Analyses presented in this g. Johnston, Lloyd D., Patrick M. O’Malley, report are based on more than 200,000 a. National Institute of Justice, 1998 Annual and Jerald G. Bachman, Drug Trends in responses available in public use data Report on Drug Use Among Adult and 1999 Are Mixed, press release, Ann Arbor, files for surveys conducted in 1979, 1982, Juvenile Arrestees, Washington, DC: U.S. MI: University of Michigan News and Infor- 1985, 1988, and 1990–97 and in a pub- Department of Justice, National Institute e mation Services, 1999. Available online at lished report for 1998. NHSDA employs of Justice, 1999, NCJ 175656. a complex sampling design and oversam- www.monitoringthefuture.org/pressreleases/ b. National Institute of Justice, 1999 Annual ples Hispanics, blacks, and youths ages 99drugpr.html. Report on Drug Use Among Adult and
5 R e s e a r c h i n B r i e f points per year); the rate among youth- the other Southeast and West Coast arrestees conformed to the conceptual ful household members rose more ADAM locations. The similarity in model for the New Marijuana Epidemic modestly to 17 percent. Relatively findings across ADAM locations sug- and that year-to-year distortions of the stable rates were subsequently record- gests that the New Marijuana Epidemic overall pattern were quite modest. From ed through 1999 for high school sen- was national in scope. Based on this 1988 to 1990, detected marijuana use iors and through 1998 for household finding, an ADAM program average among all adult arrestees declined, on members, which suggests that the epi- was calculated to facilitate presenta- average, from 35 percent to 19 percent demic in the general population may tion of the general characteristics of and declined among youthful adult have reached a plateau around 1996. the phenomenon by simply averaging arrestees from 44 percent to 24 per- findings across locations. This program cent. Subsequently, the rate among The trend data on individual ADAM average does not necessarily represent youthful adult arrestees increased locations were often somewhat ambigu- the average across arrestees nation- steadily from 25 percent in 1991 to ous. The increases in youthful marijua- wide. Furthermore, it is not necessari- 57 percent in 1996 (an average of na use detected during the expansion ly a good idea to focus on this type of 6.4 percentage points per year), sug- phase were sometimes unsteady and an average when determining the rate gesting that the expansion phase, on the year-to-year variations during the of detected use of such other drugs average, occurred among arrestees plateau phase were sometimes greater as cocaine/crack, amphetamines, and from 1991 to 1996. than 5 percentage points. There are nu- heroin because prevalence rates vary merous possible reasons for year-to-year more widely across locations. Indeed, This doubling of marijuana use among variations, including changes in polic- even marijuana use was affected by youthful ADAM arrestees provides ing priorities and random chance (sam- important local differences, which some of the strongest evidence to sug- ples for individual ADAM locations are are depicted in exhibit 2. gest that a New Marijuana Epidemic much smaller than MTF and NHSDA has occurred, primarily among youths samples). Sometimes this variation con- Exhibit 3 shows that, on average, the and especially among youths who founded the study’s efforts to precisely variation in recent marijuana use tend to get in trouble with the law. pinpoint the timing of the phases of the detected among youthful ADAM This increase preceded the ADAM New Marijuana Epidemic. In response, small variations from one year to the Exhibit 1. Variation in past-month marijuana use within the U.S. general next were often disregarded by the population, NHSDA and MTF surveys* authors as potentially attributable to Percentage reporting past-month marijuana use the limited precision of the ADAM 50 estimates. Such ambiguous trends are clearly identified in this report and the basis for an interpretation is pro- 40 vided. The most credence was placed on strong trends consistently affecting 30 marijuana use across multiple years.
Exhibit 2 depicts the status in 1999 20 of the New Marijuana Epidemic at each ADAM location across the Nation. 10 By 1999, the marijuana epidemic among arrestees had clearly reached 0 the plateau phase at ADAM locations 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 Interview year in the Northeast, Midwest, and South- west. Miami and San Diego did not Household members age 12+ (NHSDA) Household members ages 18–20 (NHSDA) appear to have observed epidemics. High school seniors (MTF)
The epidemic had shown signs of pos- *NHSDA: National Household Survey on Drug Abuse; MTF: Monitoring the Future sibly having plateaued at almost all of
6 R e s e a r c h i n B r i e f
Exhibit 2. Status of the new marijuana epidemic among ADAM* arrestees, 1999
Portland
Detroit Manhattan Chicago Philadelphia Omaha Cleveland Denver Indianapolis San Jose Washington, D.C.
St. Louis Los Angeles
San Diego Phoenix Atlanta Birmingham Dallas
Houston San Antonio New Orleans Fort Lauderdale Miami
Expansion Plateau/expansion Plateau No epidemic *Arrestee Drug Abuse Monitoring
Exhibit 3. Variation in recent program’s 1996 change in the stan- stabilization in the late 1990s. These marijuana use detected among dard for determining marijuana use findings, along with the ethnographic U.S. arrestee population, ADAM* and, therefore, could not have been information cited previously, strongly program average caused by that methodological improve- suggest that a new nationwide epidem- ment. From 1996 to 1999, the rate of ic in marijuana use passed through its Percentage detected as marijuana users 100 use among youthful adult arrestees expansion phase by 1996 and was in 90 held steady at around 60 percent and its plateau phase through 1999. the rate among all adult arrestees held 80 constant at around 37 percent. The There were several important differ- 70 modest increase of use among youthful ences across surveys. The increase in 60 adult arrestees to 62 percent in 1999 marijuana use started among youthful 50 may have resulted from changes in the adult arrestees (ADAM) about 1 year 40 ADAM sampling procedure. Thus, the before it started within the general pop- 30 plateau phase among arrestees appears ulation (NHSDA and MTF). In addition, 20 to have set in by 1996 and lasted, on the peak rate of reported past-month 10 average, at least through 1999. use among high school seniors occur- 0 ring during the plateau phase (about 87 88 89 90 91 92 93 94 95 96 97 98 99 All three major national surveys 22 percent) was far below the previous Interview year (NHSDA, MTF, and ADAM) recorded peak (37 percent) recorded in the late Youthful adult arrestees (18–20) a similar overall pattern in youthful 1970s. It was also far below the peak All adult arrestees (18+) marijuana use: a decline in the 1980s rate of detected marijuana use among reaching a low in the early 1990s, fol- youthful adult arrestees in the same *Arrestee Drug Abuse Monitoring lowed by a rise in the mid-1990s and period (about 57 percent) as well as
7 R e s e a r c h i n B r i e f
their rate of reported past-month use start date of the increase in the rate Philadelphia (about 60 percent).13 This suggests of use is difficult because the upward Percentage detected as marijuana users that the New Marijuana Epidemic trend was quite slow at first and a 100 started among those individuals who 1-year dip in youthful marijuana use 90 tended to get in trouble with the law occurred in 1993. Subsequently, the 80 and spread more widely within this popularity of marijuana among youth- 70 group than among youths in the gen- ful arrestees increased to a peak of 61 60 eral population. Conceivably, the preva- percent in 1996. From 1996 to 1999, 50 lence of marijuana use in the general the rate of marijuana use among youth- population could undergo another ful arrestees held steady at about 60 40 expansion if use diffused to other youth- percent, with the overall rate holding 30 ful subpopulations. Further research is at about 30 percent. 20 clearly needed to identify which groups 10 of mainstream youth have been most Manhattan 0 affected so far. 87 88 89 90 91 92 93 94 95 96 97 98 99 Percentage detected as marijuana users Interview year 100 The following sections examine Youthful adult arrestees (18–20) 90 geographic variation among ADAM All adult arrestees (18+) 80 arrestees across the program’s locations 70 in the following regions: Northeast, 60 Washington, D.C.—plateau since Midwest, Southeast, Southwest, and 1996. In 1990, only 7 percent of 50 West Coast. From 1996 to 1999, the all Washington, D.C., adult arrestees 40 majority of ADAM locations detected were detected as recent marijuana rates of marijuana use among adult 30 users. The rate increased rapidly and youthful adult arrestees close to 20 among youthful arrestees and then the ADAM program average shown in 10 among older arrestees. By 1996, about exhibit 3. However, five Midwest loca- 0 60 percent of youthful arrestees and tions (Chicago, Detroit, Indianapolis, 87 88 89 90 91 92 93 94 95 96 97 98 99 Interview year 35 percent of all arrestees were Omaha, and St. Louis) had substantial- detected as recent marijuana users. ly higher rates, while five Southwest Youthful adult arrestees (18–20) All adult arrestees (18+) and West Coast locations (Houston, Washington, D.C.
Los Angeles, Phoenix, San Antonio, Percentage detected as marijuana users and San Jose) had substantially Philadelphia—plateau since 1995. 100 lower rates. Marijuana use among all adult arrestees 90 in Philadelphia dropped precipitously 80 The New Marijuana Epidemic from 30 percent in 1988 to 16 percent 70 in 1990. From 1990 to 1993, mari- among ADAM arrestees in 60 juana use among youthful arrestees the Northeast 50 expanded rapidly and the rate among Manhattan—plateau since 1996. all adult arrestees returned to its former 40 Marijuana use in Manhattan had level. In 1993, the rate among youthful 30 dropped from 27 percent overall (i.e., arrestees appeared to have entered a 20 among all adult arrestees) in 1987 plateau at about 52 percent, but it sub- 10 to 16 percent in 1991. From 1991 sequently inched up to 59 percent in 0 to 1993, the popularity of marijuana 87 88 89 90 91 92 93 94 95 96 97 98 99 1995. The rate among youthful arrestees Interview year started to rise among youthful adult remained around 60 percent from 1995 arrestees (hereinafter referred to as Youthful adult arrestees (18–20) through 1999, and the overall rate All adult arrestees (18+) “youthful arrestees”). Assessing the held steady around 35 percent.
8 R e s e a r c h i n B r i e f
These rates remained relatively stable youthful arrestees began a steady rise Detroit from 1996 through 1998. (This loca- from 14 percent in 1991 to 72 percent Percentage detected as marijuana users tion did not collect a full sample in by 1998. The overall rate reached just 100 1999.) below 40 percent in 1997, where it re- 90 mained through 1999. The rate of ma- 80 The New Marijuana Epidemic rijuana use detected among youthful 70 arrestees in 1999 dipped slightly, sug- among ADAM arrestees in 60 gesting that the epidemic in Cleveland the Midwest 50 had entered a plateau in 1998. Chicago—plateau since 1996. Mari- 40 juana use among all adult arrestees in Cleveland 30 20 Chicago dropped from 48 percent in Percentage detected as marijuana users 1988 to 23 percent in 1991. In 1993, 100 10 however, the overall rate bounced up 90 0 87 88 89 90 91 92 93 94 95 96 97 98 99 to about 40 percent, where it approxi- 80 Interview year mately remained through 1999. The 70 rate of recent marijuana use detected Youthful adult arrestees (18–20) 60 among youthful arrestees rose dramat- All adult arrestees (18+) 50 ically from 27 percent in 1992 to 75 40 percent in 1996, where it approxi- Indianapolis—plateau since 1996. 30 mately remained through 1999. Marijuana use among all adult 20 arrestees in Indianapolis dropped Chicago 10 steadily from 41 percent in 1988 to Percentage detected as marijuana users 0 87 88 89 90 91 92 93 94 95 96 97 98 99 23 percent in 1991. Subsequently, 100 Interview year the rate among youthful arrestees 90 increased steadily from a low of 27 Youthful adult arrestees (18–20) 80 All adult arrestees (18+) percent in 1991 to 70 percent in 1996 70 and remained around that level through 60 50 Detroit—plateau since 1995. Mari- Indianapolis 40 juana use among all adult arrestees Percentage detected as marijuana users 30 in Detroit dropped from 32 percent 100 in 1988 to 13 percent by 1990. Sub- 20 90 sequently, the rate among youthful 10 80 arrestees increased steadily from 25 0 70 87 88 89 90 91 92 93 94 95 96 97 98 99 percent in 1990 to 73 percent in 1995. 60 Interview year The rate of marijuana use detected 50 among youthful arrestees fluctuated in Youthful adult arrestees (18–20) 40 All adult arrestees (18+) a broad range from 62 percent to 75 percent from 1995 through 1999. The 30 rate of recent marijuana use detected 20 Cleveland—plateau since 1998. among all adult arrestees inched up- 10 Marijuana use among all adult ar- ward from 38 percent in 1995 to 46 0 87 88 89 90 91 92 93 94 95 96 97 98 99 restees in Cleveland dropped from percent in 1999. 26 percent in 1988 to 11 percent in Interview year 1991. Subsequently, the rate among Youthful adult arrestees (18–20) All adult arrestees (18+)
9 R e s e a r c h i n B r i e f
1999. Overall, the rate of recent mari- St. Louis—plateau since 1996. The rate among youthful arrestees, how- juana use ranged from 39 to 45 per- Marijuana use among youthful ar- ever, did increase the most, reaching cent from 1995 through 1999. restees in St. Louis rose steadily from 69 percent in 1996. From 1996 to a low of 15 percent in 1990 to 72 per- 1998, the rate among youthful arrestees Omaha—plateau since 1996. Mari- cent in 1996, where it approximately drifted slightly downward to 62 per- juana use among all adult arrestees remained through 1998. The rate of cent. The rate of use among all adult in Omaha dropped from 45 per- overall use increased from a low of 14 arrestees also decreased, from 33 per- cent in 1988 to 21 percent in 1990. percent in 1991 to a steady 45 percent cent in 1997 to 25 percent in 1998. Subsequently, that rate rose steadily by 1996, where it remained through Both rates bounced back to new peaks to 39 percent in 1992 and to 49 per- 1998. (This ADAM location did not in 1999, suggesting the New Marijuana cent by 1996. The rate among youth- collect a sample in 1999.) Epidemic in Atlanta could still have ful arrestees rose from 25 percent in been in its expansion phase. On the 1990 and then held steady around St. Louis other hand, the relatively steady rate 55 percent from 1993 through 1995. Percentage detected as marijuana users observed from 1996 to 1998 suggests In 1996, the rate of marijuana use 100 that the epidemic might have plateaued detected among youthful arrestees 90 by 1996 and that the 1999 jump was jumped to 71 percent, where it approx- 80 an anomalous fluctuation. imately remained through 1999. This 70 change was probably not attributable Atlanta 60 to a change in the ADAM cutoff stan- 50 Percentage detected as marijuana users dard for determining recent marijuana 100 40 use. (The prevalence of marijuana use 90 30 among Omaha’s youthful arrestees in 80 20 1995 increased only slightly from 70 53 percent under the previous 100 10 60 nanogram cutoff to 56 percent under 0 50 the new 50 nanogram standard.) 87 88 89 90 91 92 93 94 95 96 97 98 99 Interview year 40 Omaha Youthful adult arrestees (18–20) 30 All adult arrestees (18+) 20 Percentage detected as marijuana users 100 10 90 0 The New Marijuana Epidemic 87 88 89 90 91 92 93 94 95 96 97 98 99 80 among ADAM arrestees in Interview year 70 the Southeast Youthful adult arrestees (18–20) 60 All adult arrestees (18+) 50 Atlanta—plateau/possibly expansion. 40 In 1990, the prevalence of recent mari- Birmingham—plateau/possible 30 juana use detected among youthful (6 percent) and all adult (3 percent) expansion. Marijuana use among all 20 arrestees in Atlanta was the lowest of adult arrestees in Birmingham dropped 10 any ADAM location. The rate among precipitously from 33 percent in 1988 0 87 88 89 90 91 92 93 94 95 96 97 98 99 all adult arrestees increased to 33 per- to 12 percent by 1990. Subsequently, Interview year cent by 1996. The epidemic did not the rate among youthful arrestees increased dramatically from 15 per- Youthful adult arrestees (18–20) appear centered on youthful arrestees All adult arrestees (18+) only; rather, the rate of recent mari- cent in 1990 to 64 percent in 1996. juana use detected increased among The overall rate reached 40 percent all adult arrestees as early as 1991. in 1996. In 1998, the rate among youthful arrestees declined modestly
10 R e s e a r c h i n B r i e f
to 57 percent and then jumped to 69 Fort Lauderdale no sustained trend in marijuana use percent in 1999. This suggests that Percentage detected as marijuana users has occurred among arrestees. Miami the expansion phase may have contin- 100 experienced neither a sustained decline ued through 1999. On the other hand, 90 in marijuana use among arrestees nor the lack of any increase in the rate of 80 the epidemic-like growth in use among use from 1996 to 1998 suggests that youthful arrestees observed at other 70 the epidemic may have plateaued by ADAM locations. 60 1996 and that the 1999 jump was an 50 anomalous fluctuation. The New Marijuana Epidemic 40 among ADAM arrestees in 30 Birmingham the Southwest 20 Percentage detected as marijuana users 100 10 Dallas—plateau since 1996. Marijuana 90 0 use among all adult arrestees in Dallas 87 88 89 90 91 92 93 94 95 96 97 98 99 80 had dropped steadily from 32 percent Interview year in 1988 to 17 percent in 1991. The rate 70 Youthful adult arrestees (18–20) of detected marijuana use among youth- 60 All adult arrestees (18+) ful arrestees subsequently increased 50 from 22 percent in 1991 to 57 percent 40 Miami—no epidemic. From 1988 in 1996. The overall rate increased to 30 through 1999, marijuana use among 38 percent. Both rates remained stable 20 all adult arrestees in Miami fluctuat- from 1996 through 1999. 10 ed around 30 percent. The rate among 0 youthful arrestees fluctuated within Dallas 87 88 89 90 91 92 93 94 95 96 97 98 99 a wider range—between 31 and 66 Percentage detected as marijuana users Interview year percent. The dramatic 1-year jump 100 Youthful adult arrestees (18–20) in marijuana use among youthful 90 All adult arrestees (18+) arrestees, from 45 percent in 1998 to 80 66 percent in 1999, may have been 70 Fort Lauderdale—plateau/possible caused by changes to the ADAM sam- 60 pling procedures. The data suggest expansion. Marijuana use among all 50 adult arrestees in Fort Lauderdale 40 dropped from 42 percent in 1988 Miami 30 to 20 percent in 1990. The rate of de- Percentage detected as marijuana users 20 tected marijuana use among youthful 100 arrestees started a very slow but 90 10 80 0 steady increase from a low of 28 per- 87 88 89 90 91 92 93 94 95 96 97 98 99 cent in 1990 to 63 percent in 1998. 70 Interview year The overall rate increased even more 60 Youthful adult arrestees (18–20) slowly, from 20 percent in 1990 to 38 50 All adult arrestees (18+) percent in 1998. The modest dip in 40 the rate in 1999 suggests that the epi- 30 Denver—plateau since 1994. In demic might have reached a plateau 20 in 1998. On the other hand, the rela- Denver, the rate of detected marijuana 10 tively slow expansion and a history of use among youthful arrestees rose rap- 0 2 previous years in which the expan- 87 88 89 90 91 92 93 94 95 96 97 98 99 idly from 26 percent in 1991 to 60 sion appeared to have halted (1992–93 Interview year percent in 1994, dropped modestly to 54 percent in 1995, and inched up to and 1996–97) suggest that the expan- Youthful adult arrestees (18–20) sion may not have plateaued by 1999. All adult arrestees (18+) 62 percent by 1999. The overall rate
11 R e s e a r c h i n B r i e f
rose more slowly, from 23 percent in Houston Phoenix—plateau since 1998. The 1991 to 41 percent by 1999. Percentage detected as marijuana users rate of detected marijuana use among 100 all adult arrestees in Phoenix dropped Denver 90 steadily from 42 percent in 1987 to Percentage detected as marijuana users 80 19 percent in 1991. Subsequently, the 100 70 rate among youthful arrestees entered a slow but steady expansion, increasing 90 60 from 22 percent in 1991 to 40 percent 80 50 in 1995. At that time, the marijuana 70 40 60 epidemic appeared to have entered 30 a plateau. However, youthful mari- 50 20 juana use jumped to 54 percent in 40 10 1998, where it remained in 1999. This 30 0 increase suggests that the marijuana 20 87 88 89 90 91 92 93 94 95 96 97 98 99 Interview year epidemic may have diffused in the 10 1997–98 period to another portion of 0 Youthful adult arrestees (18–20) 87 88 89 90 91 92 93 94 95 96 97 98 99 All adult arrestees (18+) youths who tend to get arrested. This Interview year change could have also been caused by changes in police priorities or Youthful adult arrestees (18–20) New Orleans—plateau since 1995. All adult arrestees (18+) Marijuana use among all adult arrestees ADAM sampling procedures. in New Orleans dropped precipitously Phoenix Houston—plateau since 1995. The from 46 percent in 1987 to 14 percent by 1991. Marijuana use among youthful Percentage detected as marijuana users rate of detected marijuana use among 100 all adult arrestees in Houston dropped arrestees subsequently increased from 17 percent (1991) to 54 percent (1995) 90 precipitously from 43 percent in 1988 80 to 14 percent by 1991. The rate among and then fluctuated in the 50 percent 70 youthful arrestees bounced back from to 60 percent range. The overall rate 60 a low of 19 percent in 1992 to 43 per- of detected marijuana use inched up to cent in 1995. In 1996 and 1997, the 35 percent by 1999, still well below 50 rate among youthful arrestees dipped the rate observed in the late 1980s. 40 to about 31 percent and then returned 30 to 49 percent by 1999. This increase New Orleans 20 in marijuana use among youthful Percentage detected as marijuana users 10 arrestees—well above the previously 100 0 established plateau level in 1995— 90 87 88 89 90 91 92 93 94 95 96 97 98 99 Interview year may have been attributable to changes 80 in ADAM sampling procedures. By 70 Youthful adult arrestees (18–20) All adult arrestees (18+) 1999, the rate of detected marijuana 60 use overall had returned to 31 percent, 50 still far below the rate observed in the 40 San Antonio—plateau since 1996. late 1980s and below the ADAM pro- 30 Marijuana use among all adult ar- gram average. 20 restees in San Antonio decreased 10 from 34 percent in 1988 to 18 per- 0 cent by 1991. The rate among youth- 87 88 89 90 91 92 93 94 95 96 97 98 99 ful arrestees then slowly increased Interview year from 20 percent in 1991 to 45 per- Youthful adult arrestees (18–20) cent in 1996, where it remained All adult arrestees (18+) through 1999. Overall marijuana
12 R e s e a r c h i n B r i e f use had increased to 32 percent by was caused by changes in ADAM from 47 percent in 1988 to 25 percent 1996 and fluctuated around this rate sampling procedures. If this was the in 1992. The rate of detected marijua- through 1999. case, the marijuana epidemic among na use among youthful arrestees sub- youthful arrestees in Los Angeles may sequently expanded from 28 percent San Antonio have plateaued as early as 1996. The in 1992 to 57 percent in 1999. The Percentage detected as marijuana users overall rate of marijuana use inched overall rate increased only modestly 100 up from 16 percent in 1991 to a high to a peak of 33 percent in 1998. 90 of 30 percent in 1999. 80 San Diego—no epidemic. Mari- juana use among San Diego’s youthful 70 Los Angeles arrestees remained steady and rela- 60 Percentage detected as marijuana users tively high from 1987 through 1999, 50 100 90 ranging from 37 to 55 percent. Mari- 40 juana use among all adult arrestees 80 30 exhibited a modest drop from 44 per- 70 20 cent in 1988 to 29 percent in 1991. 60 10 The rate then fluctuated around 34 0 50 percent through 1999. The rate of 87 88 89 90 91 92 93 94 95 96 97 98 99 40 Interview year detected marijuana use among youth- 30 ful arrestees exhibited a modest 1-year Youthful adult arrestees (18–20) 20 All adult arrestees (18+) increase from 37 percent in 1991 to 47 10 percent in 1992. The rate among youth- 0 ful arrestees subsequently fluctuated 87 88 89 90 91 92 93 94 95 96 97 98 99 in the mid-40-percent range. The mod- The New Marijuana Epidemic Interview year est dip and recovery in youthful mari- among ADAM arrestees on Youthful adult arrestees (18–20) juana use from 1989 to 1992 seem All adult arrestees (18+) the West Coast much too small to constitute a new Los Angeles—plateau/possible drug epidemic, although their timing expansion. It is difficult to determine Portland (Oregon)—expansion is consistent with that of the New the timing of a New Marijuana Epi- 1992–99. Marijuana use among all Marijuana Epidemic at other ADAM demic in Los Angeles because the rate adult arrestees in Portland decreased locations. Another steady but short of increase in detected marijuana use among youthful arrestees was very slow Portland San Diego in the early 1990s and because it took Percentage detected as marijuana users Percentage detected as marijuana users a dip in 1994, which suggests the rate 100 100 had plateaued. However, the increase 90 90 in detected marijuana use among youth- 80 80 ful arrestees from 22 percent in 1991 70 70 to 49 percent in 1996 strongly sug- 60 60 gests that a marijuana epidemic took 50 50 place. In 1997, the rate among youth- 40 40 ful arrestees declined modestly to 46 30 30 percent and inched up to 54 percent 20 20 by 1999. This continued increase sug- 10 10 gests that the epidemic may not yet 0 0 have plateaued by 1999. However, it 87 88 89 90 91 92 93 94 95 96 97 98 99 87 88 89 90 91 92 93 94 95 96 97 98 99 is possible that the modest increase Interview year Interview year in youthful marijuana use from 49 Youthful adult arrestees (18–20) Youthful adult arrestees (18–20) percent (1998) to 54 percent (1999) All adult arrestees (18+) All adult arrestees (18+)
13 R e s e a r c h i n B r i e f increase in youthful marijuana use 1990s who tend to get in trouble with withdrawal, overreaction to minor occurred from 1997 to 1999, when the the law in the same way that crack frustration, and antisocial behavior. rate among youthful arrestees inched had been the drug of choice previous- They emphasize that the treatment is up from 44 to 55 percent. Again, the ly. Analyses with two additional data- not for marijuana use itself but for an short period and rather modest increase sets of general population samples underlying problem that has marijua- suggest that this change was not part (the NHSDA and MTF surveys) fur- na abuse as one of its symptoms. They of a longer, sustained epidemic. ther confirmed the existence and tim- also suggest that the health risks of ing of this New Marijuana Epidemic. marijuana use are much less profound San Jose—plateau since 1995. Continued monitoring of drug use than those of cocaine or heroin use. Overall, marijuana use among San among arrestees is essential to deter- Jose arrestees was relatively stable at mine how long prevailing conditions A standing argument for controlling about 24 percent from 1989 through will persist. Some of the key issues marijuana use, based on the gateway 1998. The rate among youthful arrestees include the following: theory, is that it can lead to use of more increased from 21 percent in 1992 to dangerous drugs. As determined in this 43 percent in 1995, where it roughly How long will marijuana remain study, however, the drug of choice for remained through 1998. The sharp the drug of choice among youths persons born in the 1970s and coming increase to 56 percent in 1999 may coming of age who tend to get in of age in the 1990s has been marijua- be an anomalous 1-year fluctuation. trouble with the law? na. These youths have been much less prone to progress to other drugs than Will marijuana-using members San Jose their predecessors. This suggests that of the New Marijuana Generation Percentage detected as marijuana users the gateway theory may be less rele- continue to avoid use of other 100 vant to their substance use experiences, illicit drugs? 90 which would be good news. It would 80 To what extent will marijuana-using also be good news if the marijuana use 70 members of the New Marijuana were associated with a rejection of 60 Generation desist from such use crack and heroin due to their poten- 15 50 as they grow older? tially devastating consequences. 40 There are numerous ways to attempt This rejection of other drugs may not 30 to control drug abuse, including pre- be as characteristic of the broader 20 vention, treatment, interdiction, and population. From 1992 to 1997, the 10 law enforcement. In response to recent proportion of high school seniors re- 0 trends, drug abuse control policies porting lifetime use of LSD increased 87 88 89 90 91 92 93 94 95 96 97 98 99 might logically shift much of their from 8.6 percent to 13.6 percent, its Interview year focus to marijuana. However, this is highest recorded level since the start Youthful adult arrestees (18–20) not as simple as just targeting mari- of the MTF program in 1975.16 Use All adult arrestees (18+) juana use and users instead of crack of hallucinogens in England and the or heroin users. For one, the nature of United States has been frequently as- marijuana abuse is quite different, as sociated with the rave or dance party Conclusion noted by Grinspoon and Bakalar, who scene, typically involving white youths This study identified that the increase report that proportionately fewer mari- from middle- and upper-class subur- in marijuana use among ADAM ar- juana smokers become dependent ban enclaves.17 However, that is a restees in the 1990s generally con- than users of alcohol, tobacco, heroin, different story about a different popu- 14 formed to the conceptual model or cocaine. They suggest that psycho- lation of youths. described earlier for the diffusion of therapy may be the most appropriate a drug epidemic. Marijuana appears treatment for a troubled youth who It would appear that more has changed to have become the drug of choice uses marijuana frequently, such as one than the prevailing drug of choice among youths coming of age in the who manifests alienation, emotional among arrestees. Ethnographic studies
14 R e s e a r c h i n B r i e f
Distinguishing Age, Period, and Cohort Effects E xhibit A shows the ADAM given year, known as a birth cohort. The lence of marijuana use typically increased program average of detected marijuana entries in each row reflect changes in in their late teens. For example, among use as a function of both an arrestee’s a birth cohort’s level of marijuana use the 1977 birth cohort, the rate of detect- birth year and the year of the ADAM as arrestees age, to the extent that the ed marijuana use grew from 32 percent interview. The exhibit distinguishes three ADAM program recruits from roughly the at age 16 (in 1993) to 55 percent at age types of social factors that can influence same population in a similar way each 18 (in 1995). To facilitate identification of an individual’s drug use: age, period, and year. Age effects are those behaviors that age effects, the exhibit highlights three cohort effects.* Each row traces the mari- develop as people grow older. Among diagonals associated with ages 30, 25, juana use history of persons born in a the youngest birth cohorts, the preva- and 18, respectively.
Exhibit A. Age-period cohort analysis of detected marijuana use among arrestees, ADAM program average
Percentage Detected as Marijuana Users Interview Birth Year 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Total 18+ Year 1901–39 8 9 9 5 5 8 11 8 10 10 9 9 8 1940–44 22 19 12 9 7 9 12 12 13 15 11 10 12 1945–49 22 22 15 10 9 12 15 15 13 14 16 14 14 1950–54 28 30 18 15 13 16 16 16 18 21 19 18 18 1955–59 33 34 21 16 15 19 21 18 21 22 21 21 21 1960 37 39 23 17 17 20 24 19 24 26 23 23 23 1961 38 37 27 20 18 21 26 20 25 28 27 26 25 1962 44 37 26 22 18 22 26 20 25 28 24 25 25 1963 43 41 28 20 20 24 24 24 27 33 25 25 27 1964 47 43 26 21 20 25 26 25 26 28 27 24 27 1965 43 46 28 22 19 26 25 27 29 32 28 29 28 1966 44 45 29 24 22 28 32 26 32 33 27 30 30 1967 39 42 30 24 22 27 29 28 32 32 31 32 30 1968 52 48 34 26 25 29 31 26 33 36 33 30 32 1969 37 44 32 24 27 31 33 30 34 38 35 34 33 1970 53 39 30 26 24 32 33 33 35 38 38 36 33 1971 19 32 31 24 26 35 35 35 40 40 38 35 34 1972 6 30 20 24 33 38 40 43 47 40 39 36 1973 8 22 24 35 38 41 46 44 46 38 39 1974 4 14 30 39 43 47 50 48 44 43 1975 14 29 41 47 50 53 52 50 49 1976 23 36 46 53 54 51 51 51 1977 32 47 55 59 54 53 55 1978 39 51 58 59 57 57 1979 51 62 59 62 53 1980 55 57 59 62 1981 59 56 60 1982 54 Total 18–20 44 44 31 24 25 33 40 45 51 57 56 58 42 Total 18+ 34 35 24 19 19 24 28 29 33 37 37 37 30 Each cell percentage represents an average across up to 23 ADAM locations and reflects data only from ADAM locations in which at least 25 responses were recorded. Highlighted in white, three diagonals indicate the arrestee’s age at the time of the interview: age 30 (top diagonal), 25 (middle diagonal), and 18 (bottom diagonal). For example, when interviewed in 1990, persons in the 1960 birth cohort were 30 years old.
continued on page 16
15 R e s e a r c h i n B r i e f in inner-city communities suggest that In this regard, the potential for integrat- in distressed inner-city households there has been a dramatic shift in the ing persons from distressed inner-city with a greater stake in society may subculture of drug use and that inter- communities into mainstream culture help create a more productive labor personal interactions have become seems more promising than in the force and ensure further declines in more congenial and less violent.18 In 1970s and 1980s. Perhaps this is the drug abuse and its attendant crimi- this way, drug-using members of the time to deemphasize “tough” drug nality. If inner-city youths born in New Marijuana Generation are damag- enforcement policies in favor of indi- the 1970s who get in trouble with the ing themselves less physically and rect drug abuse control through the law could be transformed into fully socially than the preceding genera- reduction of the economic, educa- employable workers, their marijuana tions of crack smokers and heroin tional, and social barriers faced by use might also decline as they assume injectors. They are also causing much many inner-city youths in establishing conventional adult roles, just as mari- less harm to the broader population. a healthy and productive mainstream juana use tends to recede among mem- lifestyle. Providing youths struggling bers of the general population.19
continued from page 15 Distinguishing Age, Period, and Cohort Effects
The use of marijuana was affected by birth cohorts experienced sharp short- which had established their peak level of historical occurrences, or period effects, term declines and more modest long-term marijuana use by age 18, arrestees in the as shown by decreased levels of use declines from 1988 to 1993, except for 1972–76 birth cohorts exhibited a rise in (1988–90) and increased levels of use the oldest arrestees. Arrestees born in marijuana use in their early twenties. It (1991–93) among persons from each the 1901–39 period had a relatively low would appear that more and more of birth cohort. Overall, as recorded in the level of marijuana use of 9 percent in them became involved with marijuana bottom row of the exhibit, marijuana 1988, which declined to 5 percent in during the expansion phase of the New use among all adult arrestees (age 18+) 1990 but returned to a slightly higher Marijuana Epidemic. declined from 35 percent in 1988 to 19 level of 11 percent in 1993. percent in 1990 and 1991. This broad By 1996, a solid plateau in marijuana decline could have been the result of Some historical events permanently use had been established. Approximately increased drug law enforcement; greater affect individuals at an impressionable 60 percent of arrestees who reached age involvement with other drugs, such as age. Many persons who came of age 18 from 1996 to 1998 (the 1978–80 crack; or decreased availability of mari- during the Heroin Injection and Crack birth cohorts) were detected as marijua- juana. The overall rate subsequently Epidemics persisted in their habits na users in 1996 and as they aged sub- returned to 28 percent by 1993. This throughout much of their lives. In this sequently. From 1996 to 1998, the rate 5-year dip in usage (1988–93) was manner, the use of each drug became of marijuana use within each birth reflected in each birth cohort’s marijua- associated with members of a particular cohort remained relatively constant or na use experiences. For example, the birth cohort, a cohort effect. Marijuana declined modestly. marijuana use among arrestees born in use may have a similar effect on this new 1960 declined from 39 percent (1988) generation of drug users. The 1972 birth * Discerning age, period, and cohort effects is to 17 percent (1990) and then returned cohort reached age 18 in 1990, right at complicated by the multicolinearity of these to 24 percent (1993), where it remained the lull in marijuana use among arrestees. parameters—specifically, age = (interview year) relatively steady through 1998. The Their rate of marijuana use dropped from – (birth year). Hence, it is not possible to naive- nature of this dip suggests that for the 30 percent at age 17 (in 1989) to 20 per- ly include all three factors as independent 1960 birth cohort, the period effect had cent at age 18 (1990) but continually variables in an algebraic equation such as an immediate impact of reducing marijua- increased to 47 percent by 1996. For this employed in regression analysis. However, all na use by 22 percentage points (from 39 birth cohort, the period of lower mari- three types of effects result in a distinctive to 17 percent) and a somewhat smaller juana use led them to postpone, but pattern of birth cohort participation over time long-term effect of reducing marijuana did not forestall, their involvement with that can be discerned in a two-way table such use by 15 percentage points. The older marijuana. Unlike previous birth cohorts, as the exhibit.
16 R e s e a r c h i n B r i e f
Three Generations of Drug Use Among Arrestees I n a previous study,a the authors appears to have never experienced a 1970. Arrestees born since 1970, but used ADAM data for Manhattan to serious epidemic of crack use. especially after 1974, were likely to be identify three generations of arrestees detected as recent marijuana users and with distinct drug use patterns: the Findings based on all available ADAM very unlikely to report lifetime heroin Heroin Injection Generation (born data from 1989 to 1998 in Washington, injection or crack use. 1945–54), Cocaine/Crack Generation D.C., were typical (see exhibit A). The prevalence of lifetime heroin injection (born 1955–69), and Marijuana/Blunts a. Golub, Andrew, and Bruce D. Johnson, peaked with persons born around 1950 Generation (born 1970 and later). These “Cohort Changes in Illegal Drug Use and exhibited a sustained decline starting findings show variation across birth Among Arrestees in Manhattan: From the somewhere around the 1954 birth cohort, cohorts in reported lifetime (ever) heroin Heroin Injection Generation to the Blunts b reaching near zero among birth cohorts injection, lifetime (ever) crack use, and Generation,” Substance Use and Misuse c of the 1970s. The prevalence of lifetime detected marijuana use. The authors 34(13)(1999):1733–1763. performed a comparable analysis for crack cocaine use peaked among per- b. Lifetime heroin injection was approximately each ADAM location and produced sons born around 1960 and started to determined from persons who reported both similar findings at each, with only two decline around the 1964 birth cohort. lifetime heroin use and lifetime injection of exceptions.d In Phoenix, it appears that, Recent marijuana use exhibited a dra- illicit drugs. This calculation was necessary among arrestees, the Crack Epidemic matic and continuous increase with suc- because the ADAM questionnaire did not ask had not ended by 1998. San Antonio cessive birth cohorts starting around
Exhibit A. Three generations* of drug users among Washington, D.C., arrestees in the ADAM program
Percentage using indicated drugs 100
90
80
70
60
50
40
30
20
10
0 <40 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 Birth year
Heroin injection (self-report of lifetime use) Crack (self-report of lifetime use) Marijuana (current use as detected by urinalysis)
*Heroin Injection Generation: 1945–54; Cocaine/Crack Generation: 1955–69; Marijuana/Blunts Generation: 1970+
continued on page 18
17 R e s e a r c h i n B r i e f
Notes 8. Johnson, Bruce D., Andrew Golub, and continued from page 17 Jeffrey Fagan, “Careers in Crack, Drug Use, Three Generations of Drug 1. Office of National Drug Control Policy, Drug Distribution and Nondrug Criminality,” The National Drug Control Strategy: 1997, Crime and Delinquency 41(3)(1995):275–295; Use Among Arrestees Washington, DC: Office of National Drug Hamid, Ansley, “The Developmental Cycle Control Policy, 1997:23, NCJ 163915. of a Drug Epidemic: The Cocaine Smoking Epidemic of 1981–1991,” Journal of explicitly about heroin injection in all 2. Gfroerer, Joseph C., and Joan F. Epstein, Psychoactive Drugs 24(1992):337–348. years of the survey. At some locations, “Marijuana Initiates and Their Impact on many individuals reported injection drug Future Drug Abuse Treatment Need,” Drug 9. Johnston, Lloyd D., Patrick M. O’Malley, use but not heroin use (they may have and Alcohol Dependence 54(3)(1999):229–237. and Jerald G. Bachman, National Survey been injecting cocaine or ampheta- Results on Drug Use from the Monitoring the 3. Research in a wide variety of fields has mines) and others reported heroin use Future Study, 1975–1998, vol. 1, Bethesda, documented that new innovations often spread but no injection drug use (many were MD: National Institutes of Health, National within a population following a pattern similar Institute on Drug Abuse, 1999, NIH Publi- presumably sniffers). to a disease epidemic (see Rogers, Everett M., cation No. 99–4660; Substance Abuse and c. Variations in substance use across Diffusion of Innovations, 4th ed., New York: Mental Health Services Administration, birth years can be caused by age, period, Free Press, 1995). The term “epidemic” is National Household Survey on Drug Abuse or cohort effects. The authors confirmed employed in this report as a synonym for “dif- Series: H-10. Summary of Findings from the fusion of innovation” and refers to the rapid that heroin injection, crack smoking, and 1998 National Household Survey on Drug and broad spreading of a practice (such as marijuana use among ADAM-Manhattan Abuse, Rockville, MD: Office of Applied smoking marijuana) within a population or sub- Studies, 1999, Department of Health and arrestees were the result of period effects population (such as among 16- to 25-year-olds). Human Services Publication No. (SMA) in three separate age-period-cohort analy- 99–3295. ses for detected use of opiates, cocaine, 4. Golub, Andrew, and Bruce D. Johnson, and marijuana (see Golub, Andrew, and “A Recent Decline in Cocaine Use Among 10. Furst, R. Terry, Bruce D. Johnson, Eloise Bruce D. Johnson, “Cohort Changes in Youthful Arrestees in Manhattan (1987–1993),” Dunlap, and Richard Curtis, “The Stigmatized American Journal of Public Health 84(8)(1994): Illegal Drug Use Among Arrestees in Image of the ‘Crack Head’: A Sociocultural 1250–1254; Golub, Andrew, and Bruce D. Manhattan: From the Heroin Injection Exploration of a Barrier to Cocaine Smoking Johnson, Crack’s Decline: Some Surprises Across Generation to the Blunts Generation,” Among a Cohort of Youth in New York City,” U.S. Cities, Research in Brief, Washington, Deviant Behavior 20(2)(1999):153–181; Golub, Substance Use and Misuse). Similar con- DC: U.S. Department of Justice, National Andrew, and Bruce D. Johnson, “Cohort firmation was performed for each of the Institute of Justice, July 1997, NCJ 165707. Changes in Illegal Drug Use Among Arrestees remaining ADAM sites for detected use in Manhattan: From the Heroin Injection 5. Golub, Andrew, and Bruce D. Johnson, of marijuana (see Golub, Andrew, and Generation to the Blunts Generation,” Substance “Cohort Changes in Illegal Drug Use Among Bruce D. Johnson, “Monitoring the Use and Misuse 34(13)(1999):1733–1763; Arrestees in Manhattan: From the Heroin Marijuana Upsurge With DUF/ADAM Sifaneck, Stephen J., and C. Small, “Blunts Injection Generation to the Blunts Generation,” and Forties: The Drugs of Choice for the New Arrestees,” final report submitted to Substance Use and Misuse 34(13)(1999):1733– Generation,” New York: National Development the U.S. Department of Justice, National 1763; Johnson, Bruce D., and Andrew Golub, and Research Institutes, Inc., 1997, working Institute of Justice, Washington, DC, “Generational Trends in Heroin Use and manuscript; Sifaneck, Stephen J., and 2000) and for detected use of cocaine Injection Among Arrestees in New York City,” Charles D. Kaplan, “New Rituals of Cannabis in One Hundred Years of Heroin, ed. David (see Golub, Andrew, and Bruce D. Preparation and Self-Regulation in Two Musto, Westport, CT: Greenwood, forthcoming. Johnson, Crack’s Decline: Some Surprises Cultural Settings and Their Implications for Across U.S. Cities, Research in Brief, 6. In a more detailed report, the authors pro- Secondary Prevention,” New York: National Washington, DC: U.S. Department of vide analyses of marijuana use trends over time Development and Research Institutes, Inc., Justice, National Institute of Justice, within five mutually exclusive age categories. 1996, working manuscript. July 1997, NCJ 165707). See Golub, Andrew, and Bruce D. Johnson, 11. Department of Health and Human “Monitoring the Marijuana Upsurge With d. Golub, Andrew, and Bruce D. Services, Youth Use of Cigars: Patterns of DUF/ADAM Arrestees,” final report submitted Johnson, “Monitoring the Marijuana Use and Perceptions of Risk, Washington, DC: to the U.S. Department of Justice, National Upsurge With DUF/ADAM Arrestees.” Department of Health and Human Services, Institute of Justice, Washington, DC, 2000. 1999, publication number OEI–06098–00030. 7. Johnson, Bruce D., and Andrew Golub, 12. Community Epidemiology Work Group, “Generational Trends in Heroin Use and Identifying and Monitoring Emerging Drug Use Injection Among Arrestees in New York City.” Problems: A Retrospective Analysis of Drug Abuse
18 R e s e a r c h i n B r i e f
Data/Information, Bethesda, MD: National 15. Furst, R. Terry, Bruce D. Johnson, Eloise Dana, Rise of Hallucinogen Use, Research in Institute on Drug Abuse, 1999. Available Dunlap, and Richard Curtis, “The Stigmatized Brief, Washington, DC: U.S. Department of online at www.nida.nih.gov/cweg/retro.html. Image of the ‘Crack Head’: A Sociocultural Justice, National Institute of Justice, 1997, Exploration of a Barrier to Cocaine Smoking NCJ 166607. 13. The rate of self-reported past-month use Among a Cohort of Youth in New York City”; among youthful adult arrestees was calculated Golub, Andrew, and Bruce D. Johnson, “Cohort 18. Johnson, Bruce D., Andrew Golub, and separately to support this comparison. Across Changes in Illegal Drug Use Among Arrestees Eloise Dunlap, “The Rise and Decline of Hard all sites and interview years, most youthful in Manhattan: From the Heroin Injection Drugs, Drug Markets, and Violence in New adult arrestees (80 percent) detected as recent Generation to the Blunts Generation.” York City,” in The Crime Drop in America, ed. marijuana users via urinalysis also reported Alfred Blumstein and Joel Wallman, New York: past-month use. The nondisclosers were more 16. Johnston, Lloyd D., Patrick M. O’Malley, Cambridge, 2000:164–206. than offset by individuals who tested negative and Jerald G. Bachman, National Survey for recent marijuana use but still reported use Results on Drug Use from the Monitoring 19. Bachman, Jerald G., Katherine N. in the past month. the Future Study, 1975–1998. Wadsworth, Patrick M. O’Malley, Lloyd D. Johnston, and John E. Schulenberg, Smoking, 14. Grinspoon, Lester, and James B. Bakalar, 17. Parker, Howard, Judith Aldridge, and Fiona Drinking, and Drug Use in Young Adulthood: “Marijuana,” in Substance Abuse: A Comprehen- Measham, Illegal Leisure: The Normalization of The Impacts of New Freedoms and New Respon- sive Textbook, ed. J.H. Lowinson, P. Ruiz, R.B. Adolescent Recreational Drug Use (Adolescence sibilities, Mahwah, NJ: Erlbaum, 1997. Millman, and J.G. Langrod, 3d ed., Baltimore: and Society), London: Routledge, 1998; Hunt, Williams and Wilkins, 1997:199–206.
Findings and conclusions of the research Andrew Golub and Bruce D. Johnson authors’ research and their prepa- reported here are those of the authors and do not necessarily reflect the official position or are, respectively, principal investi- ration of this Research in Brief policies of the U.S. Department of Justice or gator and director, Institutes for were supported by NIJ under grant the National Development and Research Special Populations Research, at number 99–IJ–CX0020, awarded to Institutes, Inc. the National Development and the National Development and Research Institutes, Inc. The Research Institutes, Inc. The National Institute of Justice is a component of the Office of Justice Programs, which also includes the Bureau of Justice Assistance, the Bureau of Justice This and other NIJ publications can be found at and downloaded from the NIJ Statistics, the Office of Juvenile Justice and Web site (http://www.ojp.usdoj.gov/nij). Delinquency Prevention, and the Office for Victims of Crime.
NCJ 187490
Quick Access to NIJ Publication News For news about NIJ’s most recent publications, including solicitations for grant applications, subscribe to JUSTINFO, the bimonthly newsletter sent to you via e-mail. Here’s how: