Drug Abuse

First published in Contemporary Drug Problems Summer 1989

The undeterred cocaine user: intention to quit and its relationship to perceived legal and health threats

BY PATRICIA G. ERICKSON AND GLENN F. MURRAY

Both authors are researchers in the Policy Studies Department of the Addiction Research Foundation (33 Russell St., Toronto M5S 2S1, Ontario, Canada), where Dr. Erickson is Head of Drug Policy Research. They collaborated on The Steel Drug: Cocaine in Perspective (1987), and Dr. Erickson is also author of Cannabis Criminals (1980). Her training is in social administration, while Murray's is in criminology.

AUTHOR'S NOTE: The views expressed in this paper are those of the authors and do not necetsarily reflect those of the Addictaon Research Foundation. A preliminary version of this paper was presented at the annual meeting of the American Society of Criminology, Montreal, Canada, November 11-14, 1987.

Although our society more often deals with health-threatening behaviors by some combination of regulatory law and persuasion, the control of Illicit-drug use resides primarily in the criminal law. The assumption of the policy of criminal prohibition is that there is no acceptable level of illicit-drug use, and all transgressors of the norm are at risk of legal punishment. Yet recent research has been consistent in pointing to the importance of "extralegal" factors, such as perceived risks to health and social disapproval, rather than legal threats, in restraining illicit-drug use (see Erickson, 1980; Bachman et al., 1988). While many earlier studies have focused on marijuana, the growing tide of concern over cocaine use (White, 1988) raises similar policy-relevant questions as to how harmful use of this drug might be prevented.

Much of the emphasis in the perceptual-deterrence literature of criminology is on whether, as a function of risk perception, people do or do not start to engage in illegal activities. Researchers in the health field, however, have stressed that continuation and cessation of drug use are important phenomena that have been much-neglected topics compared with initiation of drug use (Goodstadt et al., 1986; Yamaguchi and Kandel, 1985). Not only starting drug-taking, but also decisions to continue or to cease it, may be related to a variety of legal and extralegal factors.

The purpose of the present paper is to investigate the intention to continue or to quit using cocaine among those who have akeady tried the drug~the undeterred. The dependent variable is a projected measure of intention to use cocaine in the next year. The relationship of this decision to ongoing cannabis use is explored, as well as its relationship to a variety of other factors, including patterns of use, peer support, drug availability, perceptions of legal sanction threat and perceived health risks.

Method

Although the validity of a projected measure of behavior has been viewed in the past with skepticism (Minor and Harry, 1982; Paternoster et a'., 1982; Saltzman et al., 1982), recent work (Lundman, 1986) provided grounds for the renewed examination of cross-sectional methods in perceptual-deterrence research. Williams and Hawkins (1986:557) have also called for more empirical investigation of "the correlation between projections of involvement and self-reports of actual involvement. ~A strong correlation] would justify the use of projections as a 'proxy' for actual behavior in a cross-sectional design."

Murray and Erickson (1987) did this in a two-wave study. They compared two models of future criminal behavior-one in which the dependent variable was projected criminality and another in which the dependent variable was actual subsequent criminality. Murray and Erickson (1987) concluded that prospective measurement provides a good estimate of subsequent criminal behavior, at least for the illegal activity-marijuana use-that was being investigated. Since marijuana use is the most frequently studied behavior in deterrence research (Lundman, 1986), these results should encourage further testing of projected measures.

The present paper will utilize a projected measure of cocaine use. A study that related intention to use cocaine with actual subsequent use showed a remarkably high concurrence: over a five-year longitudinal study, 19% of subjects reported a positive intention to try cocaine at Year 1; by Year 5, 20% of subjects were using cocaine (Newcomb and Bentler, 1986). The actual correlation coefficient was not reported, however, and the group who did eventually use cocaine may have consisted of different individuals than the group who intended to try cocaine.

It might also be considered whether cocaine, a drug with a higher dependence liability than cannabis (Brady and Lukas, 1984), is less likely to be subject to the volition of the user. If a drug truly fits the "exposure model" of addiction, which views the user as powerless to control intake (Alexander, 1984), then intentions might indeed be irrelevant to later behavior. However, a number of nonclinical studies of cocaine users have questioned the extreme dependence-producing power of cocaine (Cohen, 1987; Erickson et al., 1987; Murphy et al., 1989). It seems clear that as with most illicit drugs, far more people try cocaine, use it only a few times, or use it infrequently, than progress to regular, heavy use (Smart and Adlaf, 1987). The respondents in the present study, moreover, were not addicted to or heavy users of cocaine. Thus, there is no reason to assume that cocaine use is less suitable for deterrence research than are other forms of criminal activity.

Sample

A purposive sample of adult cocaine users was screened on the criteria of being 21 years old or older, having been employed at least six of the past 12 months, and of having used cocaine in the past three years. A combination of snowball and advertising techniques produced a sample of 111 persons. Subjects were recruited in southern Ontario in 1983-84. This community sample of cocaine users shared sirnilar age and sex characteristics with a representative sample of cocaine users obtained in an Ontario provincial survey (Smart and Adlaf, 1984). The objective was to locate a fairly "typical" nondeviant group of users rather than users already in treatment or in prison.

Most of the participants in this study were relatively infrequent cocaine users. In the past year, 9% reported use at least 100 times, 33% between 10 and 99 times, and 58% fewer than 10 times. Over half of the respondents (55%) reported no cocaine use at all in the past month. All had used cannabis and alcohol, and about one-third were daily users of these drugs. A majority of respondents had lifetime experience with LSD (86%), amphetamines (85%), tranquilizers (59%), and barbiturates (52%); in addition, 45% had ever used PCP and 29% had tried heroin. These respondents may be better described as multiple-drug users than solely as"cocaine users," an observation that also describes cocaine users in representative surveys (Clayton 1985)

Results

Factors affecting cocaine use

described as multiple-drug users than solely as "cocaine users," an observation that also describes cocaine users in representative surveys (Clayton, 1985).

Of the total sample of 111 people, 24 said they would be "very unlikely" or "unlikely" to use cocaine in the next year, 86 responded that they would be "likely" or "very likely" to continue use, and one person did not know. Thus, 22% of the sample were considered "quitters" and 78% were coded as "continuers."

Four groups of factors were identified that potentially could affect the intention to use cocaine in the future. These were the perceived legal threats, the perceived health risks, the informal controls operating in the social context of use, and availability. These are components of a deterrence model that encompasses both legal and extralegal factors (Paternoster et a]., 1985). First the bivariate relationships will be examined.

The perception of legal threats usually includes both perceived certainty and perceived severity (Lundman, 1986). However, perceived certainty of arrest was so low in this sample-only 2% acknowledged any likelihood of detection by police-that it could not have any utility as an explanatory variable and was therefore excluded. On the other hand, nearly one-fifth of the respondents thought a jail sentence likely if they were caught for cocaine possession. Those who intended to quit were twice as likely (20.8%) to expect jail as those who planned to continue using cocaine (11.6%). This relationship between perceived severity and criminal behavior, although weak, is in the expected direction according to deterrence hypotheses.

Table 1. Intention to use cocaine by various factors specific to cocaine

Cocaine-related factors affecting intentions to use cocaine	Intend to continue using cocaine (5) (N=86)	Intend to quit using cocaine (%) (N=24)
Legal
Perceived severity = jail	11.6%	20.8%
Health
Perceived harm from regular use = great(1)	36.0	62.5
Uncontrollable urge to use cocaine most times/always after using	16.3	29.2
Informal controls
Half/all friends use cocaine (1)	61.6	33.3
Avoid using cocaine with nonusers	29.4	37.5
Care if people know they use cocaine	72.1	58.3
Availability
Some difficulty getting cocaine	32.6	45.8
Never purchased cocaine	18.6	58.3

(1) p <.05

The perceptions of possible damage to health from drug use have been related to frequency of use of a particular substance in that heavier users are more likely to downplay risks (Erickson, 1989; Johnston et al., 1984). In the present study, cocaine users who perceived "great" harm from regular use of cocaine were significantly more likely to quit than to continue use: 62.5~ vs. ~ A second health variable might be called the "fear of addiction." Respondents were asked if they had ever experienced a craving or an uncontrollable urge to use cocaine. Almost twice as many quitters as continuers reported feeling this urge most times or always after using cocaine: 29.2% vs. 16.3%. It appears that concerns about health and, specifically, about addiction act as disincentives to continuing cocaine use.

In addition to deterrence, other, more psychologically based explanations may also be considered for the observed relationship between quitting drug use and higher perceived risk of legal or health consequences. For example, the cognitive dissonance aroused by engaging in dangerous activities may be resolved by deciding to stop the behavior and only then fully admitting the risks to oneself. Conversely, to continue the activities may enhance efforts to deny or at least to minimize the risks to which one is exposed. Simply because the risky behavior in question is illegal, deterrence explanations should not be the sole possible interpretation of this empirical relationship between cessation and perceived risk.

Informal controls are those constraints operating within the cocaine user's social network (Maloff et al., 1979). A consistent finding in numerous studies has been that one of the strongest correlates of illicit-drug use is peer drug use (Dembo et al., 1981; Kandel, 1984). This sample did not differ in that regard. A significant positive relationship was found between intention to continue and having half or more cocaine-using friends (61.6%, as opposed to 33.3% of quitters). Two other possible social influences-avoiding the use of cocaine when with nonusers, and caring if people knew they used cocaine-were not significantly related to intention to use cocaine. It is of interest that a higher proportion of users cared whether people knew they used cocaine than said they avoided use with nonusers. Respondents generally restricted cocaine use to a small circle of fellow users and realized that social disapproval would be directed at them from many other quarters (Erickson et al., 1987).

A final set of variables related to availability. Access to substances, whether legal or illegal, has long been associated with higher rates of use (Johnston et al., 1984). Ease of obtaining cocaine (i.e. buying, receiving as a gift, trading) was not significantly related to intention to use, but the relationship was in the expected direction. While 45.8%o of those who intended to quit had some difficulty getting cocaine, 32.6% of continuers said they had difficulty. A stronger relationship was displayed for purchase of cocaine. Those who intended to quit were three times more likely never to have purchased cocaine than those who intended to continue (58.3% vs. 18.6%).

Table 2. Intention to use cocaine by various factors specific to cannabis

Parallel cannabis factors affecting intentions to use cocaine	Intend to continue using cocaine (5) (N=86)	Intend to quit using cocaine (%) (N=24)
Legal
Perceived severity = jail	2.4%	4.2%
Health
Perceived harm from regular use = great(1)	7.0	25.0
Informal controls
Half/all friends use cannabis (1)	98.4.4	85.0
Avoid using cannabis with nonusers	52.4	50.0
Care if people know they use cannabis	40.0	45.0
Availability
Some difficulty getting cannabis	6.2	20.0
Never purchased cannabis	0.0	5.0

(1) p <.05

Comparison of cannabis factors with cocaine factors

The above analysis of factors potentially related to intention to use cocaine showed that nearly all operated in the expected direction, and three perceived harm, friends using cocaine, and purchase of cocaine) were significant. When the bivariate analysis was repeated (see Table 2) with the parallel cannabis factors in relation to intention to use cocaine, the patterns were identical (with one exception), although the proportions were not.

Far fewer respondents perceived either the risk of jail (for cannabis possession) or the threat to health (for regular use of cannabis) to be as high as the comparable cocaine factors. The proportion of cannabis-using friends was greater than that of cocaine-using friends. Users were less likely to care if people knew they used cannabis than cocaine, but they were more likely to avoid use with nonusers of cannabis than they were with non users of cocaine-the only reversal in the data. This may be explained by these respondents' oft-voiced comment that with cocaine you are not "high," but actually are more in control. Perhaps the effects of cannabis are thought more difficult to conceal; also, unlike cocaine, cannabis is not odorless. Virtually all respondents had purchased cannabis and reported no difficulty in obtaining it.

It has been suggested that the cannabis-cocaine connection may become stronger (Murray, 1984). However, these two contrasting sets of data (Table 1 and Table 2) indicate that cannabis and cocaine cannot yet be equated in their legal, health, social and market profiles. These two drugs are viewed somewhat differently by the same group of users. Cocaine lacks the social integration, acceptability, and ease of access accorded to cannabis and is perceived to pose greater legal and health risks.

Multivariate analysis of future cocaine use

The eight cocaine-related factors that may have influenced the intention to use cocaine in the future were analyzed in a correlation matrix. Only seven of the 56 bivariate relationships reached 0.20: intention to use with perceived harm, peer use, and purchase; peer use with perceived severity; uncontrollable urge with purchase; and peer use with avail-ability and purchase. Compared with declared quitters, those who intend to continue cocaine use are less likely to perceive the harm of regular cocaine use as great, are more likely to have friends who use cocaine, and are more likely to have purchased cocaine.

In order to assess the relative and combined impact of the explanatory variables on intention to use cocaine, an ordinary regression analysis (OLS) was performed.2 Only two variables-ever purchasing cocaine and perceived harm of regular use-were found to be significant at the .01 level (see Table 3).' Experiencing a craving or uncontrollable urge, however, approached significance. In total, 25% of the variance in projected intention to use cocaine was explained in this analysis.

Table 3. Regression analysis (OLS) predicting intention to use cocaine

R² = .254
(1) See Table 1 for description of variables

Quitting cocaine and continuing cannabis

We asked respondents both about intention to use cocaine and about intention to use cannabis in the next year. Originally we envisioned a parallel analysis of intention to use either cocaine or cannabis in relation to the same factors as they related to each drug. However, we discovered that none of the respondents intended to stop using cannabis.

This was in contrast to cocaine, for which 22% of the sample expressed an intention to quit.

This pharmaco-specific example of cessation immediately placed us in a quandary. If the quitters of cocaine are not also going to quit cannabis, what is special about cocaine? While so much of the current drug literature has emphasized the progression from cannabis to cocaine (Clayton, 1985), we had apparently unearthed an example of "de-escalation."

The discovery that all persons m this sample intended to continue cannabis use thus merits further consideration. It seems pertinent to ask whether those who intend to quit cocaine but continue cannabis are from a particular part of the cannabis-use spectrum. If cocaine quitters were lighter cannabis users, this could signify a general turning away from, or cutting down on, illicit-drug use. In contrast, if cocaine quitters were heavier cannabis users, this could imply a potential preference; that is, cannabis is their drug of choice, rather than cocaine. However, neither of these interpretations is clearly favored by the data.

When cannabis use in the past year was assessed (Table 4), those who had used cannabis at least 100 times were slightly less likely to quit cocaine use than those who had used cannabis fewer than 100 times (20% and 23%, respectively). When the relationship of cannabis use in the past month to intention to use cocaine was considered, the opposite pattern was displayed (Table 4). Those in the more frequent cannabis use category of at least 10 times were slightly more likely to quit cocaine use than those who used cannabis less often (25~o and 19%, respectively). That the differences are so small, and in the opposite direction, for yearly and monthly cannabis use suggests that our understanding of quitting cocaine use has not been enhanced by examining cannabis

A possible interpretation of these findings is that once illicit drug use practices have been established, cannabis is unrelated to further cocaine use. Put another way, cannabis use may be important only in predicting initiation to cocaine use, not in predicting continuation or cessation. Other factors, such as those described earlier in Table 3, may be more important in the decision as to continued cocaine use. Whatever the precise details, these data indicate that "deescalation" of cocaine use may occur without a retreat from cannabis use, and that the intention of a user of both cocaine and marijuana to stop cocaine is not explained by his/her use of marijuana.

Conclusion

This study of perceptual deterrence has utilized a projected measure of a health-threatening crime-cocaine use. Our results suggest that purchasing cocaine is an indicator of commitment to continued cocaine use. Unlike those respondents who were content to share cocaine at parties, some of the users interviewed were prepared to make an effort to seek out cocaine so as to be more assured of a supply in their control. These results also suggest that concerns about harmful effects of cocaine and the possibility of addiction may act as barriers to continuing cocaine use. The apparent irrelevance of the formal controls implies that legal threats are perceived to be remote.

Policy makers who are concerned with prevention, then, might more realistically focus on the more salient health risks than rely on criminal prohibition. Moreover, the objective of prevention of harmful drug use need not focus only on initiation to drugs, but may also be informed by research that considers influences on quitting drug use at a later stage of the cycle of a drug-use career. Drug use may be sustained for long periods of time, and the impact of legal and health threats may vary, depending on the stage of the drug-using career and the particular drug.

The issue of the influence of informal social controls does, of course, extend much beyond the particular measures used in our study. Informal social controls are a complex and important determinant of substance use, but much remains to be learned about how they operate for a given drug in a specific context (Maloff et al., 1979). Many of the user's perceptions of health risks, risk of addiction and legal threats are socially transmitted and mediated through contacts with family and friends. Although our measures of informal control are limited, this area is obviously one that requires better formulation if noncoercive strategies for prevention are to be improved.

And finally, these data indicate that cannabis and cocaine were perceived quite differently by experienced users of both drugs. Compared with cannabis, cocaine use was thought to be more of a threat to health and more likely to result in a jail sentence, was shared by fewer friends, was more subject to social disapproval, and was more difficult to obtain. These findings help to explain why users are more prepared to give up cocaine than cannabis. Since the intention to quit cocaine was not related to current frequency of cannabis use, it is clear that an understanding of these divergent patterns requires a model that incorporates social, health-oriented and cultural factors, not one based on some notion of preordained pharmacological progression. This research serves as an important reminder that steppingstones can go down as well as up.

Notes

1. A similar correlation matrix was considered with cannabis replacing cocaine. The size and the direction of the correlation coefficients were very similar to the one with cocaine.

2. A log linear analysis was also conducted, since it is a more appropriate procedure for analyzing dichotomous categorical variables. Because the findings were the same as for the regression analysis, however, only the latter is presented here.

3. A fairly high correlation (.39) between purchase and peer use explains why peer use did not make a significant contribution in the regression analysis.

References

Alexander, B. K. 1984. When experimental psychology is not empirical enough: The case of the exposure orientation. Canadian Psychology 25:84-95.

Bachman, J. G., L. D. Johnston, P. M. O'Malley and R. H. Humphrey. 1988. Explaining the recent decline in marijuana use: Differentiating the effects of perceived risks, disapproval, and general lifestyle factors. Journal of Health and Social Behavior 29:92-112.

Brady, J. V. and S. E. Lukas (eds.). 1984. Testing Drugs for Physical Dependence Potential and Abuse Liability. NIDA Monograph No. 52. Rockville, MD: National Institute on Drug Abuse.

Clayton, R. R. 1985. Cocaine use in the United States: In a blizzard or just being snowed? In: Nicholas J. Kozel and Edgar N. Adams (eds.), Cocaine Use in America: Epidemiologic and Clinical Perspectives. NIDA Monograph No. 61. Rockville, MD: National Institute on Drug Abuse.

Cohen, P. D. A. 1987. Cocaine use in Amsterdam in nondeviant subcultures. Paper presented at the International Council on Alcohol and Addictions Congress, Lausanne, Switzerland, May 31-June 5.

Dembo, R., 1). V. Babst, W. Burgos and J. Schmeidler. 1981. Survival orientation and the drug use experiences of a sample of inner city junior high school youths. The Jnternational Journal of the Addictions 16:1031-1047.

Erickson, P.6.1980. Cannabis Criminal:: The Social Effects olPunishment on Drug Users. Toronto: ARF Books.

______( 1989. Living with prohibition: Regular cannabis users, legal sanctions and informal controls. The International Journal of the Addictions, 24:forthcoming.

Erickson, P.6., E. M. Adlaf, 6. F. Murray and R. 6. Smart. 1987. The Steel Drug: Cocaine in Perspective. Lexington, MA: D. C. Heath.

Goodstadt, M. S., 6. C. Chan, M. A. Sheppard and J. C. Cleve. 1986.

Factors associated with cannabis nonuse and cessation of use:

Between and within survey replications of findings. Addictive

Behaviors 11:275-286.

Johnston, L. D., P. M. O'Malley and J. 6. Bachman. 1984. Highlights from Drugs and American High School Students 197S- /983. Rockville, MD: National Institute on Drug Abuse.

Kandel, D. B. 1984. Marijuana users in young adulthood. Archives of General Psychiatry 41:200-209.

Lundman, R. J. 1986. One-wave perceptual deterrence research: Some grounds for the renewed examination of cross-sectional methods. Journal of Research in Crime and Delin quency 23:37~389.

Maloff, D., H. S. Becker, A. Fonaroff and 1. Rodin. 1979. informal social controls and their influence on substance use. Journal of Drug Irsues 9:161-184.

Minor, W. W. and J. Harry. 1982. Deterrence and experiential effects in perceptual deterrence research: A replication and extension. Journal of Research in Crime and Delinquency l9:l9~203.

Murphy, S. B., C. Reinarman and D. Waldorf. 1989. An 11-year follow-up of a network of cocaine users. British Jo urn al of Addiction 84:forthcoming.

Murray, 6. F. 1984. The cannabis-cocaine connection: A comparative study of use and users. Journal of Drug Issues 14:665-675.

Murray, 6. F. and P.6. Erickson. 1987. Cross-sectional versus longitudinal research: An empirical comparison of projected and subsequent criminality. Social Science Research 16:107-1 18.

Newcomb, M. D. and P. M. Bentler. 1986. Cocaine use among adolescents: Longitudinal associations with social context, psychopathology, and use of other substances. Addictive Behaviors 11:263-273.

Paternoster, R., L. E. Saltzman, T. Ci. Chiricos and Ci. P. Waldo. 1982. Perceived risk and deterrence: Methodological artifacts in perceptual deterrence research. Journal of Criminal Law and Criminology 73:1238-1258.

Paternoster, R., L. E. Saltzman, Ci. P. Waldo and T. Ci. Chiricos. 1985. Assessment of risk and behavioral experience: An exploratory study of change. Criminology 23:417-433.

Saltzman, L. E., R. Paternoster, Ci. P. Waldo and T. Ci. Chiricos. 1982. Deterrent and experiential effects: The problem of causal order in perceptual deterrence research. Journal of Research in Crime and Delinquency 19:172-189.

Smart, R. Ci. and E. M. Adlaf. 1984. Alcohol and Drug Use Among Ontario Adult: in 1984 and Changes Since 1982. Toronto: Addiction Research Foundation. .1987. Alcohol and Other Drug Use Among Ontario Adults 1977-1987. Toronto: Addiction Research Foundation.

White, H. R. 1988. Longitudinal patterns of cocaine use among adolescents. American Journal of Drug and Alcohol Abuse 14:1-15.

Williams, K. R. and R. Hawkins. 1986. Perceptual research on general deterrence: A critical review. Law and Society Review 20:545-572.

Yamaguchi, K. and D. B. Kandel. 1985. On the resolution of role incompatibility: A life event history analysis of family roles and marijuana use. American Journal of Sociology 90:1284-1325.