Bulletin on Narcotics

Volume LIII, Nos. 1 and 2, 2001

Dynamic drug policy: Understanding and controlling drug epidemics

<<   contents   >>

Key variables and data requirements in modelling drug systems

E. SINGLE
Professor of Public Health Sciences, University of Toronto,
and Senior Research Associate, Canadian Centre on Substance Abuse, Canada

Abstract
Introduction: characteristics of systems of drug information
The limits of drug information systems
Data systems depend on well-articulated goals and performance indicators
Greater attention is required on economic aspects of illicit drug abuse
A proposed dynamic model for drug systems
Dependent variables in the model
Proximate causes of drug use and drug-related harm
Underlying determinants: the role of drug policy
Summary and conclusions
References

ABSTRACT

The author of the present article explores factors that account for variation in drug use and drug-related harm over time. A model of a drug system is presented, consisting of dependent variables, proximate causes and underlying determinants. The dependent variables concern drug use (level and pattern), adverse health consequences and drug-related crime. The major proximate causes of drug use and drug-related harm include drug availability (price and physical availability), attitudes towards use (fear of legal sanctions, health beliefs regarding risks and cultural beliefs) and alternatives to the illicit market (home cultivation and legal intoxicants for users, alternative career and income prospects for sellers). The underlying determinants that influence those proximate causes include policy (drug laws, preventive education and risk management programming) and environmental factors (geographical isolation, climate and fauna, and threat of acquired immunodeficiency syndrome (AIDS) and other infectious diseases). Despite recent improvements in the measurement of problematic patterns of drug use, there is still a paucity of data on patterns of use and drug-related harm. The viability of drug policy is thus often measured in terms of changes in levels of drug use and/or changes in the number of persons detected and charged with drug crimes. Until valid and reliable data are available on its key variables, any model of drug systems will have very limited applicability.

Introduction: characteristics of systems of drug information

The author of the present paper explores aspects of drug systems that may account for a substantial portion of the variation in drug use and drug-related harm over time in a particular setting. The utility of any model depends on the availability of data on key indicators. Before focusing on the key variables that have an impact on drug use and drug-related problems, the discussion begins with three observations concerning the characteristics of drug information systems.

The limits of drug information systems

Firstly, there is a wide variation in drug problems and responses to those problems. Therefore, even though uniformity in drug information systems can and should be promoted, there probably can never be a perfect information system for all countries. The nature of drug problems differs in different settings [1]. In regions such as South America and South-East Asia where a large share of the world's cocaine and heroin are produced, the illicit drug trade has created substantial underground economies with consequent problems of law enforcement and economic control. In the consuming countries in Europe and North America, the problems tend to be the adverse health, social and economic consequences of the use of illicit drugs.

Furthermore, there is also a wide variation in national drug policies. Responses to illicit drug problems range from strict enforcement of punitive drug laws to benign neglect. In parts of Australia, Europe and North America harm reduction policies have been implemented to reduce the adverse consequences of illicit drug users who cannot be expected to cease their drug use at the present time [1]. Even within countries, there are often cycles of panic over emerging drug problems, followed by periods of indifference when other pressing issues push illicit drugs to a relatively low place on the national policy agenda [2].

Although international treaties provide a common framework for drug policy and a certain degree of uniformity in social responses to drug problems, there is inevitably considerable variation in the nature and magnitude of illicit drug problems as well as the social responses to those problems, both between countries and regions and even within countries over time.

Data systems depend on well-articulated goals and performance indicators

The second observation is that without a clearly articulated set of goals and performance indicators, drug information systems will inevitably fail to meet their objectives. This may seem obvious, but the fact remains that the goals and performance indicators for national drug strategies are rarely well articulated. Focus is often placed on reducing the prevalence of any illicit drug use rather than specific indicators of drug-related harm. Examples of reasonably well-articulated performance indicators for national drug strategies are relatively uncommon. For example, Australia's National Drug Strategy has been subject to evaluation [3] and its goals and performance indicators have been articulated in a National Drug Strategic Plan [4].

Canada's Drug Strategy currently suffers from a lack of dedicated funding, but its goals have been articulated as follows [5]:

(a) To reduce the demand for drugs;

(b) To reduce substance-related mortality and morbidity;

(c) To improve the effectiveness of and accessibility to information, education and other interventions, including treatment and rehabilitation;

(d) To restrict the supply of illicit drugs.

Each of those goals is further classified into sub-objectives with associated performance indicators. For example, performance indicators for the goal of reducing demand for drugs include not only reductions in levels of drug use, but also increases in attitudes that inhibit drug use (e.g. realistic health concerns), reductions in patterns of illicit drug use that are particularly likely to result in harm (e.g. regular use, injection drug use, needle-sharing) and particular targets for reductions in those indicators by high-risk groups such as native peoples, street youth, prisoners and other socially disadvantaged groups [5]. Similar sub-objectives and performance indicators are articulated for the other goals of the national strategy. The specific goals and performance indicators will vary between countries, as they should, because, as noted above, the nature of problems caused by illicit drugs varies between countries.

Thus, the task for internationally comparable drug information systems is to find a common set of goals and performance indicators that national drug policies can build upon and expand within their own national context. Data sources should also be identified, in order to determine information gaps and set research agendas. The key point is that the articulation of goals and specific performance indicators is required to create the framework for drug information systems that ultimately will provide the data to test any dynamic model of drug use and drug problems. Where the goals and performance indicators of a national drug strategy are not clearly articulated, drug information systems will be based on the availability of data rather than data requirements. The development of a dynamic drug model based on poorly articulated data systems may be doomed to failure.

Greater attention is required on economic aspects of illicit drug abuse

The third observation concerning drug information systems is that there is a strong need for more rigorous and comprehensive economic data on substance abuse to promote evidence-based decision-making and a more consistent response to substance abuse. The economic ramifications of illicit drugs are not well understood, either in the producing countries or the consuming countries and regions.

Four key questions need to be addressed to help policy makers to make well-informed decisions on drug issues [6]:

What is the cost of drug abuse to society?

What portion of those costs are realistically avoidable?

What and where should policy makers invest to avoid those costs?

How well are such investments performing over time?

Researchers have given substantial attention to the costs of alcohol and tobacco use [7]. In contrast, relatively little attention has been given to the costs that society bears as a result of the abuse of illicit drugs. That deficiency can be attributed largely to data problems inherent in any attempt to quantify the social costs of illicit drugs. It is difficult to quantify the production, consumption, import, export or price of illicit drugs. In addition, although significant information is available on the causal links between drug abuse and health, the causal links in other areas, crime in particular, are extremely difficult to quantify. For these reasons little quantitative information exists on the social costs of illicit drug abuse.

A proposed dynamic model for drug systems

The first step in modelling the dynamic aspects of drug systems is to identify the key variables and data requirements in a model that could explain variations in drug use and drug-related problems in a particular setting over time. It is proposed that a model of a drug system consists of dependent variables, proximate causes and underlying determinants:

( a ) Dependent variables:

(i) Drug use variables:

a. Levels of drug use;

b. Patterns of use: heavy, dependent use;

c. Injection drug use;

d. Associated risk behaviours such as needle-sharing;

(ii) Drug-attributable crime:

a. Drug violations;

b. Property crime attributable to drug use;

c. Violent crime attributable to drug use;

(iii) Other adverse health and social consequences:

a. Overdose, suicide;

b. Other health effects;

c. Abuse, family discord;

d. Poor school or work productivity;

( b ) Proximate causes:

(i) Drug availability;

(ii) Attitudes towards illicit drug use;

(iii) Alternatives to using or dealing in illicit drugs;

( c ) Underlying determinants:

(i) Environmental factors;

(ii) Drug policy:

a. Interventions with users:

i. Risk reduction among continuing users;

ii. Treatment and rehabilitation;

iii. Social welfare policy bearing on users;

b. Demand reduction;

c. Supply reduction.

The major proximate causes of drug use and drug-related harm include, but are not restricted to drug availability (price and physical availability), attitudes towards use (fear of legal sanctions, health beliefs regarding risks and cultural beliefs) and alternatives to the illicit market (home cultivation and legal intoxicants for users, alternative career and income prospects for sellers). The under lying factors that may influence those proximate causes include, but are not restricted to, policy (drug laws, preventive education and harm reduction programming) and environmental factors (geographical isolation, climate and fauna, threat of acquired immunodeficiency syndrome (AIDS) and other infectious diseases).

Dependent variables in the model

As mentioned above, there are three main dependent variables in the model: levels and patterns of drug use; adverse health and social consequences; and drug-related crime. Information is needed not only on the proportion of the population using illicit drugs, but also which drugs, in what combinations, by what mode of administration, in which settings and whether the drugs are used in a manner that minimizes or maximizes the chances that serious health and social consequences will occur. The key point concerning drug use variables is that a robust model requires more than simply information on levels of drug use. Without detailed information on patterns of use, relatively little variance in drug-related problems will be explained by the model.

Illicit drug use is associated with a variety of problems, including crime, a family dysfunction, workplace problems and health disorders [8]. Illicit drugs are among the leading causes of preventable death and illness among young persons and the social and economic costs related to illicit drug use are considerable. The problems of illicit drug use negatively affect many communities, making neighbourhoods unsafe, diminishing property values and diverting limited police resources from other pressing needs. The major health problems associated with illicit drug abuse are suicide, drug overdose and communicable diseases such as AIDS and hepatitis C. Other adverse health and social consequences of drug use include hospitalization or other treatment for drug dependency, lower economic productivity, poor school performance, child and spousal abuse and family discord.

Drug use is related to crime but the degree to which the relationship is causal is unclear. Chronic or dependent use of the so-called “hard” drugs—heroin, cocaine or crack, speed, lysergic acid diethylamide (LSD) and other strong hallucinogens—is often implicated as a contributory cause of property crime, in particular burglary and theft. Assault, homicide and other crimes of violence have resulted from “turf wars” in the illicit drug market. Illicit drug users are disproportionately involved in incidents of spousal and child abuse. Even cannabis use has been implicated as a contributory cause of crime, namely, the crime of impaired driving [9]. Criminal offenders have disproportionately high rates of illicit drug use [10, 11]. Up to 80 per cent of Canadian criminal offenders reported using illicit drugs during their lifetime, 50-75 per cent showed traces of drugs in their urine at the time of arrest and close to 30 per cent were under the influence of drugs when they committed the crime for which they were accused [12]. Drug addicts admitted to treatment often have criminal records [13].

There is clearly a strong relationship between illicit drug use and crime, but the fact that a crime is committed by someone using illicit drugs does not necessarily mean that the drug use caused the crime to be committed. The pharmacological effects of drugs and the need for addicts to commit crime to support their drug habit are at best only partial explanations for the link between drug use and crime. The majority of illicit drug users are not dependent and most users, even dependent users, do not commit property crimes [14]. Most addicts who commit crimes began doing so prior to becoming drug-dependent [15]. Addicts who commit property crimes tend to use drugs at very high levels, they have few legitimate sources of income and in the majority of cases they were engaging in criminal behaviour prior to drug use [12]. Furthermore, many former addicts continue to commit property crimes after they no longer use drugs [16].

A more plausible explanation for the strong association between illicit drug use and crime is that addicts adopt a deviant way of life that accounts for both their drug use and their criminal behaviour. A number of longitudinal studies have shown that drug use and criminality are related to a similar set of socio-demographic and personality variables, for example, poverty, poor future career or income prospects and low investment in social values [17-19]. There are undoubtedly many commonalities in the aetiology of both criminality and illicit drug use. Drug use and crime may well be mutually reinforcing, but according to that viewpoint, the real cause of both drug use and criminal behaviour is a complex set of underlying personality and social determinants.

Thus, the attribution of crime to illicit drug use is fraught with methodological difficulties and there is a lack of research on the proportion of all crime that can be causally attributed to illicit drug use [12, 20]. While there is some doubt concerning the extent to which drug use leads to crime, there is little doubt that crime results from systemic violence inherent in the illicit drug trade. Many crimes result from “turf wars” between rival distributors as well as arguments and robberies involving buyers and sellers on the illicit market [21]. Systemic violence in the illicit drug market is most common in economically and socially disadvantaged areas that have traditionally high rates of violence. It should also be noted that not only are drug addicts more likely to commit crimes, but they are also more likely to be victims of violent crimes [15].

Proximate causes of drug use and drug-related harm

The major proximate causes of drug use and drug-related harm consist mainly of drug availability, attitudes towards use and alternatives to the illicit market. Drug availability could be measured by survey questions regarding the ease with which respondents may obtain various illicit drugs or by tracking the street price per unit of potency for illicit drugs. In fact, this is rarely done. In the absence of better information on drug availability, seizure data are often taken as a surrogate mea sure of availability. Indeed, large expenditure for drug trafficking enforcement is typically justified purely on the basis of the number of arrests and amount of drugs seized (sometimes accompanied by a large amount of cash as well as weapons). The ultimate impact on drug availability is rarely assessed. In what may be the only systematic study on the effects of drug trafficking enforcement, it was found that arresting drug traffickers tends to have local, short-term and negative impacts on drug problems [22]. Typically, major drug arrests were followed by a temporary increase in street prices for illicit drugs until new sources emerged and led to more property crime by dependent users [22]. Long-term impacts were negligible. While one study is hardly conclusive evidence, it would appear that law enforcement data on trafficking arrests and drug seizures are a poor measure of drug availability.

Attitudes towards drugs represent a key determinant of drug use. While attitudes towards heroin or cocaine use are rarely measured, it is reasonably well documented that changes in rates of cannabis use appear to be more strongly connected with changing perceptions of health risks rather than fear of legal sanctions or any changes in the legal status of the drug [23]. For example, there is a clear relationship in the United States National Household Survey on Drug Abuse between perceived health risk and cannabis use from 1985 to 1997 among persons aged 12-17. One cannot be certain of the causal relationship without longitudinal data. Nonetheless, figure I shows that as perceived health risk increased in the late 1980s, rates of cannabis use declined. As perceived health risk diminished in the mid-1990s, rates of cannabis use began to rise again. This suggests that attitudes towards drug use play a key role in understanding trends in drug use. Attitudes are thus an important leverage point for drug prevention.

Figure I. Trends in cannabis use and perceived risk of cannabis use among persons aged 12-17 in the United States of America, 1985-1997
Figure I

A further proximate cause of trends in drug use and its patterns is the extent to which there are alternatives to the illicit drug market. In countries where there are favourable cultural traditions and/or climatic conditions for the cultivation of illicit drug crops, changes in availability of drugs from the illicit drug market will have less impact on drug use as users can turn to home cultivation for their drug supplies. Alternative career and income prospects are a vital aspect in understanding why some persons become involved in illicit marketing. A major appeal of drug trafficking for some residents of urban ghettos in the United States is the fact that there are few opportunities in legitimate activities to achieve the level of income that dealing in illicit drugs can provide.

Underlying determinants: the role of drug policy

Environmental factors play a role in influencing the major proximate causes of drug use. Geographical isolation may inhibit or slow the spread of new drugs or new patterns of drug use. Isolation also simplifies the enforcement of laws prohibiting the import or export of illicit drugs. As noted earlier, climatic factors have an impact on the nature and extent of cultivation of illicit drug crops. Another critical factor concerning the environment is the extent to which AIDS and other infectious diseases represent serious public health issues. In those countries where a substantial portion of the population is infected with communicable diseases that can spread via unsafe methods of injection drug use, there is a pressing need to institute risk reduction measures among injection drug users and thus there tends to be less support for purely abstinence-based approaches.

Nonetheless, the most critical underlying factor influencing the proximate causes of drug use and drug-related problems is drug policy, if only because this is the key leverage point in the model. A model developed for drug policy in British Columbia, Canada, is presented in figure II [24]. The framework components consist of the primary goal of the policy, strategies to achieve that goal, agencies responsible for interventions, strategic planning to develop programme priorities and performance indicators, research underpinning planning and evaluation, and funding. There is also a feedback loop in which performance indicators are monitored and the results are fed back into strategic planning for the next phase of the drug strategy.

Figure II (click for a larger view)

The framework outlined in figure II applies to any drug policy, regardless of whether its main goal is to eliminate drug use or if its major objective is harm reduction. The strategies to achieve either goal fall into three general categories: demand reduction, supply interdiction and interventions directed at drug users. The three major strategies are not entirely mutually exclusive: improving treatment effectiveness reduces drug demand and reducing drug demand can affect aspects of drug supply. By the same token, impacts on supply can influence demand. For example, Caulkins and colleagues have demonstrated how reductions in supply can increase the cost-effectiveness of drug law enforcement [25] and Moore has noted that drug law enforcement can enhance the effectiveness of treatment and prevention [26]. Nonetheless, these three strategies represent a reasonable classification of the major ways in which the goal of drug policy can be achieved.

Interventions to achieve these objectives consist of prevention programming to reduce drug demand, the enforcement of drug laws and interventions directed at drug users. The first two sets of interventions—supply and demand reduction—correspond to primary prevention, while the third set—interventions directed at drug users—corresponds to secondary and tertiary prevention. Interventions aimed at drug users consist of risk reduction, treatment and rehabilitation of drug users, and social welfare policies that support treatment and rehabilitation. In a drug policy that focuses on reducing overall levels of drug use, little or no emphasis is placed on risk reduction measures such as syringe exchange and drug maintenance programmes. A drug policy based on harm reduction, on the other hand, places more emphasis on use-tolerant measures aimed at users who cannot be expected to cease their drug use at the present time.

Law enforcement is primarily responsible for supply interdiction and health agencies are generally responsible for demand reduction, while interventions aimed at drug users are the responsibility of both health and social welfare agencies. Again, there is necessarily some overlap with regard to responsible agencies. For example, treatment can reduce both illicit drug demand and/or supply, as when a treated user-seller stops using and dealing [26]. Similarly, law enforcement contributes to prevention programming by stigmatizing illicit drug use through school-based educational programmes and social welfare agencies contribute to the reduction of illicit drug demand.

Ideally, the three major types of interventions should be well planned and coordinated with one another. In practice, this is made difficult by the multiplicity of government and non-governmental organizations involved, for example, health officials from different levels of government, different police forces, hospitals, health-care workers, addiction treatment agencies and academic and non-academic researchers. To ensure effective strategic planning, goals should be agreed upon, as well as strategic objectives and guiding principles. Decisions must be made concerning programme priorities and funding, and performance indicators must be specified and monitored.

The basis of good strategic planning is research. In order to make sound decisions on programme priorities, it is necessary to have scientifically credible basic research on the basic biological mechanisms of dependence, the psycho-social risk factors and the interplay of individual characteristics, pharmacological properties of psychoactive substances and the environment in which consumption occurs. Applied research on the effectiveness of interventions is also vital. The requirements to evaluate a use-reduction drug policy are more limited than the requirements to evaluate a harm reduction policy—generally, trends in levels of drug use are all that is required to provide a broad evaluation of the former, while much more detailed information on specific drug-related harms is needed to assess a harm reduction approach. In either case, however, research plays a key role in the evaluation of performance indicators, which in turn provides information needed for strategic planning of interventions in the future.

Political commitment to the drug strategy is the final and perhaps the most essential component of the framework. The degree of political commitment determines funding levels, which dictates the limits of what can be accomplished. The framework is a dynamic model in that it includes a feedback loop whereby performance indicators are monitored and that information is used to adjust programming and strategically plan the next phase of the drug strategy. The results of evaluation also have an impact on the level of political commitment: a strategy that is achieving its goals is more likely to receive continued political support. A strategy that is unable to demonstrate its effectiveness is less likely to receive continued funding.

Summary and conclusions

The author of the present paper has presented a general model for understanding trends in drug use and drug-related problems over time in a particular setting. The major dependent variables of concern are rates of drug use, problematic patterns of drug use and indices of specific drug-related harm. The proximate causes of drug use and drug-related harm are drug availability, attitudes towards drugs and drug use, and the availability of alternatives to the use and marketing of illicit drugs. It is further posited that the major underlying factors that may influence the proximate causes include environmental factors that have an impact on drug crop cultivation and/or support for drug policies. Most importantly, drug policy is viewed as a major underlying determinant. Enforcement of drug laws, preventive education, treatment and other interventions with drug users represent the key leverage point by which trends in drug use and drug-related problems can be influenced by drug policy.

Unfortunately, the proposed model is subject to the criticism that it may be impractical and idealized, given the current state of knowledge. While development of a sound model for drug systems requires specification of the interrelationships of the major sets of variables in the model, an even more primary issue is the availability of data. There have been recent advances in improved measurement of problematic patterns of drug use (e.g. estimation of the number of injection drug users), drug-related harm (e.g. deaths and hospitalizations attributed to drugs, such as opioid dependence and overdose) and drug-related crime. Nonetheless, there is still a paucity of data on patterns of drug use or drug-related harm, so the viability of drug policy is generally measured in terms of changes in levels of drug use and/or changes in the number of persons detected and charged with drug crimes.

There is an oft-told anecdote about an inebriated man who searches for his dropped keys under a street lamp, not because the keys were lost there but because that is where the light is best. Like the drunk futilely looking for his keys under the street lamp, undue focus has been placed on levels of use and drug charges as a surrogate for drug-related harm. Until valid and reliable data are available on all of its key variables, any model of drug systems will have limited applicability.

References

  1. United Nations International Drug Control Programme, World Drug Report (London, Oxford University Press, 1997).
  2. See, for example, P. Giffin, S. Endicott and S. Lambert , Panic and Indifference: the Politics of Canada's Drug Laws (Ottawa, Canadian Centre on Substance Abuse, 1991).
  3. E. Single and T. Rohl, The National Drug Strategy: Mapping the Future (Canberra, Australian Government Publishing Service, 1997).
  4. Department of Health, Housing, Local Government and Community Services, National Drug Strategic Plan (Canberra, Commonwealth Department of Health, Housing, Local Government and Community Services, 1993).
  5. E. Single, Performance Indicators for Canada's Approach to Substance Abuse (Ottawa, Health Canada, 1997).
  6. D. Collins and others, “Improving economic data to inform decisions in drug abuse control”, Bulletin on Narcotics, forthcoming.
  7. L. Robson and E. Single, Literature Review of Studies on the Economic Costs of Substance Abuse (Ottawa, Canadian Centre on Substance Abuse, 1995).
  8. D. English and others, The Quantification of Drug Caused Morbidity and Mortality in Australia, 1992 (Canberra, Commonwealth Department of Human Services and Health, 1995).
  9. E. Single and others, “The economic costs of alcohol, tobacco and illicit drugs in Canada, 1992”, Addiction, vol. 93, 1998, pp. 983-998.
  10. See, for example, Correctional Services of Canada, Final Report of the Working Group on Reducing Addictions (Ottawa, Correctional Services of Canada, 1990).
  11. S. Brochu and L. Guyon, “The issue of addiction among a sample of incarcerated women”, 37th International Congress on Alcohol and Drug Dependence, 1995.
  12. S. Brochu, “Estimating the costs of drug-related crime”, Second International Symposium on Estimating the Social and Economic Costs of Substance Abuse, Montebello, Quebec, 1995.
  13. W. Hall, J. Bell and J. Carless, “Crime and drug use among applicants for methadone maintenance”, Drug and Alcohol Dependence, vol. 31, 1993, pp. 123-129.
  14. B. Benson and others, “Is property crime caused by drug use or by drug enforcement policy?”, Applied Economics, vol. 24, 1992, p. 690.
  15. D. Kreuzer, “Drugs and delinquency”, European Criminology, vols. 5 and 6, 1993.
  16. R. Hammersley and others, “The relationship between crime and opioid use”, British Journal of Addiction, vol. 84, 1989, pp. 1029-1043.
  17. D. McBride and C. McCoy, “Crime and drug using behavior”, Criminology, vol. 19, 1981, pp. 281-302.
  18. J. Fagan, J. Weis and Y. Cheng, “Delinquency and substance use among inner city students”, Journal of Drug Issues, vol. 20, 1990, pp. 351-402.
  19. N. Brunelle et S. Brochu, “La prédiction de la délinquance et de la toxicomanie: les risques, les facteurs de risque”, 63ième Congrès de l'ACFAS, 1995.
  20. K. Pernanen, “The social cost of alcohol-related crime: conceptual, theoretical and causal attributions”, Second International Symposium on Estimating the Social and Economic Costs of Substance Abuse, Montebello, Quebec, 1995.
  21. J. Roth, Psychoactive Substances and Violence (Rockville, National Institute of Justice, United States Department of Justice, 1994).
  22. A. Sutton and S. James, Evaluation of Australian Drug Anti-trafficking Law Enforcement (Canberra, National Policy Research Unit, 1996).
  23. See, for example, E. Single, P. Christie and R. Ali, “The impact of cannabis decriminalization in Australia and the United States”, Journal of Public Health Policy, vol. 21, 2000, pp. 157-186.
  24. E. Single, “A harm reduction framework for drug policy in British Columbia”, paper commissioned by the British Columbia Harm Reduction Working Group, Victoria, Canada, November 1999.
  25. J. Caulkins and others, Mandatory Minimum Drug Sentences: Throwing Away the Key or the Taxpayers' Money? (Santa Monica, CA, RAND Corporation, 1997).
  26. M. Moore, “Supply reduction and law enforcement”, Drugs and Crime, M. Tonry and J. Wilson, eds. (Chicago, University of Chicago Press, 1990), pp. 109-158.
<<   contents   >>