In this classic piece, Stanton and Richard DeGrandpre review human and animal research against the claim that cocaine is such a powerful reinforcer that it invariably causes the organism with unlimited access to self-administer the drug to the exclusion of all other activity and reward, often until death. In place of this model, Stanton and Rich apply behavioral economic research and models which show that animals balance the opportunities for available rewards, among which cocaine appears to be a strong but far from overwhelming or unique example. They contrast their view with that of Nobel prize-winning economist Gary Becker, who rather than suggesting an economic model of behavior instead imagines that drugs create a biologically compelling state that drives the addict's behavior.

Further Reading


Published in Addiction Research, 6:235-263, 1998.
© 1998 Overseas Publishers Association.
Reprinted with permission from Gordon and Breach.

Cocaine and the Concept of Addiction:
Environmental Factors in Drug Compulsions

Stanton Peele
Morristown, New Jersey

Richard J. DeGrandpre
Dept. of Psychology, Saint Michael's College, Colchester, Vermont




Introduction: A Brief History of the Concept of Addiction

Animal Research With Cocaine

Studies of Huan Cocaine Use

Addictiveness as a Pharmacological Property of a Drug

The (Behavioral) Economics of Drug Demand

Policy Implications of the Economic Understanding of Addiction





Addiction is an evocative psychological and medical term whose meaning has changed significantly over time. For most of this century it has been described in terms of an abstinence syndrome (dependence and withdrawal) and associated with heroin use. In the 1980s, however, cocaine replaced heroin as the prototypical drug of abuse. Cocaine had heretofore not been considered to produce "physical dependence." Nonetheless, for both cocaine and heroin, current models of addiction — models widely propagated by the media — reduce drug use patterns to the properties of drugs and biological characteristics of the user. In creating this model, scientific and clinical debates along with public debates rely on the supposedly typical, inevitably addicting results of repeated cocaine consumption. Yet naturalistic human drug use and drug taking by animals in the laboratory instead reinforce the picture that use of all drugs depends on the user's environment. Indeed, even the most severe examples of compulsive drug use can be reversed when key elements in the setting are modified. Such findings should by now play a fundamental role in both scientific and public conceptions of addiction, but they do not.

Introduction: A Brief History of the Concept of Addiction

From antiquity into the nineteenth and twentieth centuries, the term addiction meant abandonment to a bad habit so that habitues totally ignored other life considerations. Addiction was not specifically associated with narcotics or with drugs at all (Peele, 1985; Sonnedecker, 1958). Around the turn of the twentieth century, addiction was appropriated by medical authorities as a property of narcotics (Berridge and Edwards, 1987; Isbell, 1958). The behavioral and psychological markers of addiction were codified as pathologic withdrawal and craving in a deterministic model that replicated the alcoholism-as-disease notion of drug-induced loss of control (Levine, 1978; see Peele, 1990).

For most of this century, the idea that addiction is a physiological process set off by heroin consumption dominated popular and pharmacological thinking (Musto, 1987; Peele, 1985). However, even the earliest efforts at systematic research with addicts did not find that heroin use conformed to this simple, billiard-ball causation (Light & Torrance, 1929; see Peele, 1990). While the public, media, and medical authorities assumed that addiction was a well-defined physiological construct, pharmacologists were compelled instead to construct descriptions of drug use in behavioral, phenomenologic, and existential terms. Thus the World Health Organization (WHO) Expert Committee on Addiction-Producing Drugs separated addiction into "physical dependence" and "psychic dependence" in the 1960s, claiming that psychic dependence is "the most powerful of all factors involved in chronic intoxication with psychotropic drugs...even in the case of most intense craving and perpetuation of compulsive abuse" (Eddy, Halbach, Isbell, & Seevers, 1965, p. 723). In a primary pharmacology reference text, Jaffe (1980) defined addiction as "a behavioral pattern of drug use, characterized by overwhelming involvement with the use of a drug [compulsive use], the securing of its supply, and a high tendency to relapse after withdrawal" (p. 536).

By the mid-1980s, cocaine had supplanted heroin as the drug said to be most dangerous and quickly addictive (for historical overviews of cocaine's public image, see Harrison, 1994; Jones, 1992). Cocaine came to be seen as the major public health menace in this country (before AIDS), and the imagery previously associated with heroin was usurped by cocaine:

Cocaine-driven humans will relegate all other drives and pleasures to a minor role in their lives.... If we were to design deliberately a chemical that would lock people into perpetual usage, it would probably resemble ... cocaine.... (Cohen, 1984, pp. 151-153)

Cocaine (and later crack) addiction became the major focus of government funding for research and treatment on illicit drug use. Since cocaine had not been classified as a drug capable of producing physical dependence, the experiential effects that compel continued drug use once more rose to the fore of theorizing about addiction:

[That] cocaine produces no gross physiological withdrawal symptoms...demonstrate[s] that subjective experiences or symptoms other than physiological discomfort are crucial in addiction to cocaine and to other substances of abuse.... [I]nvestigators are now exploring how psychological symptoms in drug withdrawal, particularly unpleasant mood states and craving for drug euphoria, maintain chronic drug addiction. (Gawin, 1991, p. 1580)

But this recognition of the centrality of lived experience in cocaine addiction did not halt speculation that "cocaine causes a neurophysiological addiction" or slow the search for an understanding of addiction based exclusively on "unraveling the neurophysiological mysteries of human experiences of pleasure and pain" (Gawin, 1991, pp. 1580, 1585; see also Wise, 1988). This view recombines the psychic and physical dependence categories into a single biological construct.

After a period of convoluted reasoning about cocaine's addiction-like properties stretching over the decade from the mid-1980s to the mid-1990s (cf. Peele, 1985), the elaborate distinctions drawn over the previous quarter-century between physical dependence and addiction were lost. Today, the director of the National Institute of Mental Health (Hyman, 1996, p. 611) may simply identify cocaine (and amphetamines, which mimic the effects of cocaine) as addictive in the same sense and as a result of the same changes in "molecular mechanisms" following chronic drug ingestion as heroin: "Repeated doses of addictive drugs — opiates, cocaine, and amphetamine — cause drug dependence and, afterward, withdrawal."

Our modern pharmacological era strives to understand addiction as foremost a biologic response of the organism. Even when researchers and clinicians claim to recognize the multideterminacy of cocaine and other drug addiction (as Gawin above does), they treat the environmental and experiential components in the equation as unmeasurable and unscientific and strive to reduce addiction to a drug's chemical structure and pharmacological effects. This article argues instead that addiction cannot be defined strictly in terms of the addicted organism and a chemical substance, and that drug problems — in this case, compulsive drug use — can never be isolated from cultural and other contextual factors and from the situation of the actor (DeGrandpre & White, 1996; Sidman, 1956; Zinberg, 1984). We review animal laboratory and human epidemiologic studies to show that environmental factors ultimately determine drug use, including the very addictiveness of a drug's appeal and the urge to continue compulsively consuming the substance (Peele, 1985). Such alternative attachments and rewards are powerful enough — or can be made so — to overcome the allure of any pharmacological substance (Reinarman et al., 1994).

If addiction is not identified with any chemical or biological process, moreover, then it can occur with a wide range of involvements in addition to drug use, such as love and gambling (Peele & Brodsky, 1975). Each type of addictive involvement, moreover, does not require a separate theory of addiction. The addictive process with different activities and in different individuals shares a variety of common elements and influences, even though the exact pathways to addiction will most certainly vary from person to person, time to time, and place to place (see Bry, 1996).

Animal Research With Cocaine

Research on human cocaine use refutes the dominant image of the effects of cocaine and of the typical behavior of regular cocaine users. Even researchers aware of these results can shade or ignore them, however, by proposing that the underlying biological reaction to cocaine is uncontrollable escalation of use and effects. While — in this view — human users may escape the inevitable addictive consequences of regular use, captive animal cocaine users accurately reflect this addictive pattern uncomplicated by superfluous environmental factors. This view, too, rather than accurately reflecting the research, has been disproven by the evidence.

Cocaine's Addictiveness

In contrast to an experiential model of addiction which focuses on sentient human beings and their involvement with their surroundings, the prevailing, reductionistic view of addiction points to evidence that animals will self-administer drugs such as cocaine as proof that human psychology and environment are irrelevant to addiction. The study of animal drug self-administration was begun by behavioral pharmacologists who implanted intravenous catheters so that animals could directly express a desire for drugs in measurable self-administered drug doses (Schuster & Thompson, 1969; Yanagita, Deneau, & Seevers, 1965). Behavioral-pharmacologic research with both animals and humans in laboratory settings since the 1950s has focussed on pharmacological and environmental conditions which maintain or modify an organism's response to various drugs. This work indicates that drug properties alone offer only a partial explanation for animal drug-taking behavior, to wit:

A drug is not a reinforcer because it has a certain molecular structure capable of exciting specific receptors. The structures and events so described yield, at best, only a potential for reinforcing action that is realized if, and only if, a set of additional conditions are satisfied. Neither is reinforcement a sensory experience that can be experimenter imposed. It is not a thing at all; it is a relational construct. (Falk, 1994, p, 48)

This quote indicates that mainstream pharmacologic research with animals can be conceived in a way very similar to an experiential model. But alarm about spreading cocaine use in the late 1980s challenged this viewpoint, and several key studies revealing dramatic animal cocaine self-administration and toxicity have been selectively and repeatedly cited to prove that cocaine has special pharmacologic properties that inevitably lead animals and humans to addiction. The results of these cocaine toxicity studies were welcomed for their anti-drug implications. Their impact was multiplied by their frequent citation in the media, in drug education programs (including films of monkeys self-administering cocaine and undergoing convulsions), and in scientific arguments about the unique reinforcing properties of cocaine. For example:

Cocaine's power of reinforcement produces its most notorious effects: the desire to keep taking it as long as the drug is available. In one series of experiments .... scientists let caged monkeys self administer ... cocaine until they died... . The drug made them monomaniacal. (Rolling Stone, February 9, 1989, p. 72; cited in Morgan & Zimmer, in press)


In laboratory animals, the intravenous injection of cocaine serves to initiate and maintain specific behaviors required to obtain additional injections. Such repetitive behaviors are ... the equivalent of human cocaine-seeking and compulsive use patterns ... (Cohen, 1985, pp. 151-152)

Some behavioral pharmacologists likewise attribute special addictive reinforcement properties to cocaine:

Cocaine appears to be a most potent reinforcer, and the self-administering organism is resistant to any attempts to decrease drug-taking. Indeed, the drug is so reinforcing that the organism self-administering it becomes totally preoccupied with drug acquisition. (Fischman, 1988, p. 7)

Animal Studies of Cocaine's Toxicity

The four primary studies cited in support of these views were published between 1969 and 1985. The studies showed that rhesus monkeys and rats will prefer cocaine to food (Aigner & Balster, 1978), will more often die from cocaine than heroin use (Bozarth & Wise, 1985), and will self-administer cocaine until death or near-death (Aigner & Balster, 1978; Bozarth & Wise, 1985; Deneau, Yanagita, & Seevers, 1969; Johanson, Balster, & Bonese, 1976). The stark picture presented by these studies, however, contrasts sharply with most other laboratory animal research on cocaine. Such studies typically manipulate factors commonly shown to affect human drug taking in natural settings, including (1) drug dose, (2) access to the drug, (3) effort required to obtain the drug, and (4) the presence or absence of alternative nondrug reinforcers. The four toxicity studies are notable for eliminating most such environmental variation, thus establishing baseline conditions under which animals can be induced to self-administer fatal doses of cocaine. The conditions — making potent doses of the drug available in an unlimited and rapidly-delivered intravenous (IV) fashion via an implanted catheter, requiring only minimal responding to gain drug reinforcement, and placing the animal in a highly restricted and impoverished context — are designed to limit the animal's behavioral repertoire and to program cocaine self-administration. However, even a change in basic elements of the cocaine-toxicity procedures engenders a qualitatively different picture of cocaine self-administration.

Cocaine dosage and access

Having access to an unlimited, direct flow of high concentrations of cocaine at all times at little or no cost (effort) is an unusual situation. On the other hand, animals readily "manage" their cocaine use within daily routines when they are given only periodic opportunities to inject cocaine. For example, rhesus monkeys with regular, intermittent access to cocaine "regulated their drug intake to a remarkable degree" (Johanson & Fischman, 1989, p. 24; cf. Wilson, Hitomi, & Schuster, 1971). Such results suggest that applying even modest environmental constraints to drug use, as opposed to uninterrupted access, dramatically affects cocaine self-administration.

Two studies of cocaine use following the unlimited-access studies have compared unlimited with somewhat limited-access conditions. Dworkin, Goeders, Grabowski, and Smith (1986) compared a group of rats that had unlimited access to cocaine and a group that were switched from unlimited access every hour to every other hour. Note that this second group of experienced cocaine-using rats still had considerable access to cocaine, which was fully available on alternating hours throughout the entire day. The former group of rats died within 28 days of being given unlimited access, while none of the rats in the limited-access condition died by the time the study was terminated (90-120 days).

Fitch and Roberts (1993) reported similar results in a study that varied both dose and schedule of cocaine self-administration. When varying amount of access to cocaine, they found cyclic (as opposed to uncontrolled or lethal) cocaine self-administration in a condition that scheduled up to four self-administrations per hour. At this point of frequent — but not unlimited — access to cocaine, patterns of self-administration were intense, but animals did not develop "outward signs of ill health." Even in the unlimited access conditions of that study, only the highest of three concentrations of cocaine dose led to erratic and self-destructive drug-taking behavior in the rats. At the middle dose, animals self-administered cocaine reliably, but intermittently or cyclically. At the lowest dose, they even failed to reliably self-administer the drug. When dose and access are constrained, as they would be in any naturalistic context of drug use, animals show patterns of drug taking typical for those of other drugs.

Cocaine versus other stimulant drugs

Cocaine self-administration has frequently been compared with patterns of use of other stimulants in the same studies. One of the four toxicity studies also offered rhesus monkeys unlimited self-administration of d-amphetamine (Deneau et al., 1969) which resulted in similarly intensive drug use, but without convulsions or death. Another toxicity study (Johanson et al., 1976), while reporting that two monkeys died from cocaine self-administration, also reported that six monkeys given unlimited access to d-amphetamine and d-methamphetamine self-administered remarkably large amounts of the synthetic stimulants. All six of these animals died.

The setting of toxicity studies

Consideration of the conditions in which cocaine self administration takes place is also crucial to understanding cocaine's toxicity. The settings of the toxicity studies with cocaine and amphetamines were highly impoverished relative to the settings in which these animals would live naturally, or the settings in which humans might acquire drug compulsions (Aigner & Balster, 1978; Deneau et al., 1969; Johanson et al., 1976). Drug administration was in small cubicles (1.3-1.5 meters) where the animals were attached to harnesses which held in place an implanted IV catheter. These harnesses require that the animals' mobility be curtailed, thus restricting other activities. Such conditions are, of course, dictated by the research goal of allowing the animals to express urges for the drug via direct drug self-administration.

Alternative research methods for studying drug reinforcement could offer a tradeoff between options for self-administering drugs versus conducting ordinary species activities. Alexander and his colleagues (Alexander, Peele, Hadaway et al., 1985) addressed this issue with narcotics by measuring oral morphine self-administration (a less powerful means for inculcating dependence) under different housing conditions. (Note that rats in self-administration studies have somewhat greater mobility than monkeys through the use of an extended tube for housing the implanted catheter; see Bozarth & Wise, 1985.) Rats exposed to the option of a sweetened morphine solution drank one-eighth as much morphine in a large cage they shared with other rats as did rats housed in small, socially-isolating cages. Further experiments by this group indicated that both the added space and companionship were critical factors affecting drug use. In other words, fundamental species activity played an essential role in drug reinforcement — when rats were free to roam and have sex, the morphine solution actually seemed to be negatively reinforcing, presumably because it interfered with these preferred activities.

Cocaine as a Universally Addicting Reinforcer

That animals often control their self-administration of cocaine and that cocaine is no more lethal than other commonly-abused stimulants make the grounds for cocaine's unique reinforcement potency — or any idea of a universally enslaving drug reinforcer — dubious. As shown in the toxicity study comparing cocaine and heroin self-administration in rats by Bozarth and Wise (1985), although cocaine was more often fatal, all animals reliably self-administered morphine, while only 83 percent self-administered cocaine. Moreover, the daily intake for the cocaine-using rats — unlike those using heroin — was highly erratic, varying from none to extremely high intake even after many days of unlimited access (see Figure 1). This pattern is typical in studies of unlimited access to cocaine and other psychomotor stimulants in laboratory animals (Johanson et al., 1976; Thompson & Pickens, 1970).


Erratic daily drug intake...
Figure 1. Erratic daily drug intake for typical subject self-administering cocaine hydrochloride (mg/kg per infusion). Data are expressed as mean number of infusions per hour of testing, and error bars represent SEM. Figure derived from from Bozarth and Wise (1985)


Cocaine creates cyclic patterns of intake marked by periods when the animal administers little or no drug even though cocaine is constantly available. This does not support the idea that cocaine is an especially reinforcing pharmacological compound that leads to compulsive drug use, but rather suggests that the erratic self-administration of cocaine is a direct behavioral effect of the drug. In other words, drug activation is channeled into furthered cocaine use when (a) cocaine is freely available and (b) this is the only significant response available to the animal. Canadian researchers Fitch and Roberts (1993) expressed a similar view:

[Our] data suggest that when drug access is discontinued after several hours, allowing drug levels to clear from the system, the disruptive influence of the drug can be minimized. Conversely, if extended drug access permits continued accumulation of drug levels, then the influence of the drug may be carried forward to sustain self-administration behavior and disrupt other functions. (p. 120)

The accepted, classical view is that animals self-administer the drug to maintain the highest possible levels of drug reinforcement. If compulsive usage patterns serve to achieve such ends, then animals which have experienced unlimited access should respond more frenetically as access becomes constrained. For example, in the study described above by Dworkin et al. (1986), when access was changed from every hour to every-other hour, responding at the same rate would yield about one-half the previous levels of cocaine exposure. If maintaining high drug levels is the motivation, then animals in this condition should respond at about double the rate (a response rate which they can easily manage). Instead, the animals which had cocaine available every other hour reduced their responding and cocaine intake fell, not to one-half, but to one-fourth the exposure animals received in the unlimited-access condition. This finding affirms the view that toxic levels of cocaine self-administration occur more as a function of the direct behavioral effects of the psychomotor stimulant (including motivational effects on animal hunger) than as a result of any unique reinforcing properties inherent to cocaine.

Studies of Human Cocaine Use

Addicts' descriptions of their harrowing drug habits are typical media fare and, along with repeated references to animal cocaine-toxicity studies, comprise the major source for claims of cocaine's addictiveness (see DeGrandpre, 1996). "Cocaine addicts tend to go on binges, and monkeys hooked up intravenously will inject themselves repeatedly, rejecting food, sex and sleep until they die" (New York Times, June 14, 1992, p. 7; cited in Morgan & Zimmer, in press). For most of the 1980s, "few data exist[ed] on the characteristics of cocaine users who are not seeking advice or treatment" (Johanson & Fischman, 1989, p. 9). By the late 1980s and continuing to the present, however, a sizable amount of epidemiologic and community-based (as opposed to clinically-based) data on cocaine use have been accumulated which do not indicate that cocaine — whether snorted, smoked, or injected — is especially or inevitably addictive for humans (Erickson, 1993; Erickson & Alexander, 1989; Harrison, 1994).

Epidemiologic, Community, and Longitudinal Data

In 1995, WHO in conjunction with the United Nations Interregional Crime and Justice Research Institute published results from the "largest global study on cocaine use ever undertaken," conducted from 1992 to 1994. Among other findings, summarized in a WHO press release (March 20, 1995), the study reported that there is "an enormous variety in the types of people who use cocaine, the amount of drug used, the frequency of use, the duration and intensity of use, the reasons for using cocaine and any associated problems that users experience." The results of this comprehensive survey simply lent weight to what had already been established about cocaine and other drug use. Such evidence on human reactions to cocaine includes the following sources:

Drug use surveys

After a decade when cocaine use was reported to be rampant and uncontrollable for a sizable group of Americans, the 1990 National Household Survey of Drug Abuse (NHSDA) found that 11.5 percent of Americans reported ever using cocaine, 3 percent used cocaine within the past year, and 0.9 percent used in the last month (NIDA, 1991; see Harrison, 1994). Of current users (those who have used the drug in the last year), a third used the drug 12 or more times a year, and 10 percent used cocaine once a week or more. These results replicate another, earlier study:

Cocaine use appears to be experimental in nature and to involve few experiences for a substantial portion of those who report any lifetime experience with the drug. One-half (53%) of the male users and two-thirds (67 %) of the female users have used cocaine less than 10 times in their lives; 34% and 28 %, respectively, have used 10 to 99 times, 9% and 3% have used 100 to 999 times, 3% and 2% have used 1,000 or more times. (Kandel, Murphy, & Karus, 1985)

A Canadian survey found 5 percent of current users used monthly or more often (Adlaf, Smart, & Canale, 1991). But monthly and weekly use are far from addiction, and only 10-25 percent of regular users resemble clinical addicts, or about 1-2 percent of all current users (Erickson & Alexander, 1989).

Natural history or longitudinal data

Indeed, only a small minority of long-term cocaine users actually progress to addiction (i.e., compulsive use that leads to disruption in other life areas, such as health and work). Of the 50 regular users Siegel (1984) tracked for over a decade, only five became compulsive users at any point. The failure of most users to progress to addiction occurred even though average level of use increased during the study, seemingly because subjects — who were college students when first identified — had more disposable income. Studies of ongoing cocaine users in Canada, Scotland, Australia, and Holland identify controlled use as the most common usage pattern (Cohen, 1989; Ditton, Farrow, Forsyth et al., 1991; Fagan & Chin, 1989; Mugford & Cohen, 1989; Murphy, Reinarman, & Waldorf, 1989; see Harrison, 1994).

At the same time, as found in WHO's global study of cocaine, the level of use and appearance of problems vary with setting and lifephase. Problems connected with drug use of a variety of types — sleeplessness, nasal irritation, financial and family problems, unintended heavy use — do appear with many users. However, what is most notable is that, in response to these problems, heavy users in these studies rarely seek treatment and typically quit or cut back on their own (Erickson et al., 1987; Waldorf et al., 1991). In Holland, of 64 users of cocaine for five or more years, only one (2 %) actually underwent any treatment for cocaine use (Cohen & Sas, 1994). In these studies, the most common pattern for users who at some point experience problems is decreased or intermittent use (see Cohen & Erickson, unpublished results on 100 intensive crack/cocaine users, cited in Erickson, 1993; Cohen & Sas, 1994; Erickson et al., 1987). Thus, even in periods of heavy, repeated, and problematic use, users do not abandon themselves to their drug habit in a way characterizable as "loss of control" and depicted as the typical result of prolonged cocaine use (Cohen & Sas, 1992, reviewed in Cohen & Sas, 1994; Siegel, 1984).

Predictive variables in cocaine abuse

The NIMH (National Institute of Mental Health) Epidemiological Catchment Area (ECA) Program identified 1.8 percent of those at risk for cocaine use in urban areas who did begin or progress in their cocaine use (Ritter & Anthony, 1991).1 The relative risk (RR) of initiation/progression was 32 times as great for those who recently used marijuana and other illicit drugs; other risk factors were persistent depression (12), recent use of marijuana alone (10), recently gaining a job after being unemployed (5), and higher income (with a 1.0 increase in RR with each increment of income level). Risk was lower for married subjects and diminished with age. Longitudinal studies which model the factors that cause progression to cocaine abuse among adolescents identify prior histories of use of other drugs (such as alcohol, marijuana, cigarettes, amphetamines), disturbed family and social environments, and clinical profiles marked by antisocial personality dispositions and other psychological problems, such as depression, prior to cocaine use (Newcomb & Bentler, 1986a, b).

Differences in Human and Animal Drug Use

The idea that experienced cocaine-using humans will be as irresistibly drawn to self-administer cocaine as are animals in the laboratory (Cohen, 1985; Wise, 1988) is not supported by results of human research in naturalistic environments (see also Morgan & Zimmer, in press). Compared to the toxicity studies, human drug use in natural settings is less passive, constrained, and irreversible. As well as being more actor-determined, human drug use also takes place within a set of values and a cultural milieu for which animals have no equivalent. When people quit smoking or other drug addictions, they typically cite family, career, or existential motivations (Peele, 1987). One study of heavy cocaine users, including crack users and freebasers, described how controlled users and former addicts balance urges to use cocaine intensively against "meaningful roles which provide[d] a positive identity and a stake in conventional daily life" (Waldorf, Reinarman, & Murphy, 1991, p. 153).

Cocaine elevates the mood of most human subjects in a way indistinguishable from amphetamines (Fischman, Schuster, Resnekov et al., 1976) and is now labelled addictive like heroin (Hyman, 1996). In an experimental study of preferences for oral doses of d-amphetamine among human volunteers (Johanson & Uhlenhuth, 1981), subjects reported clear elevation of mood (friendliness, elation, arousal) and strongly preferred the drug over placebo in initial trials. This preference disappeared over several trials, however, despite the subjects' continued identification of positive mood changes from drug use. Over time, these subjects were less interested in savoring the mood enhancement of the drug than in conducting their ordinary lives. Viewing reinforcement as a drug property would predict that the subjects continued taking the drugs; their actual behavior shows that drug use was competing against other activities that maintained a higher priority for users (see also Lamb et al., 1991).

Addictiveness as a Pharmacological Property of a Drug

Drugs which animals can be made to self-administer compulsively also have strong abuse potential for humans (Johanson, 1984). Thus, the World Health Organization (WHO) uses animal drug self-administration data as a baseline for assessing a drug's abuse liability (WHO, 1981). If addictive liability is a property of drugs, then their addictive potential — or "reinforcer efficacy" — can be scaled as invariant drug traits. Yet, behavioral pharmacology has produced a body of results that indicate reinforcement potency cannot be reduced into purely pharmacological terms, since the potency of a drug's reinforcement capacity depends so greatly on the conditions under which the drug is used (Hughes, Higgins, & Bickel, 1988; Katz, 1990). Pharmacological properties may make certain drugs suitable for compulsive use by humans, but these properties alone are insufficient to predict or explain the variability inherent in human usage patterns.

In fact, drug self-administration by animals alone does not predict the drugs most commonly abused by humans (Hartnoll, 1990). In Table 1, the psychoactive drugs most widely used by humans — caffeine, alcohol, nicotine, marijuana, and benzodiazepines — are those that animals are least likely to self-administer in the laboratory. Nor do animals respond compulsively to the substances to which Americans or Britons are most commonly addicted: caffeine (Hughes et al., 1992), nicotine, alcohol, and benzodiazepines.2 By WHO animal-testing criteria, therefore, tobacco, alcohol, and caffeine would be rated as low or nondependence-producing drugs with little potential for human abuse.

Obviously, a range of legal, psychological, social, and economic factors influence patterns of use of drugs. But these factors counterbalance subjugation to drugs among humans for even the most dependence-producing drugs, such as nicotine:

Non-pharmacological factors such as availability, relative cost, social pressures, legal consequences of use, and marketing practices provided the best explanation for the greater number of deaths associated with nicotine and the greater incidence of progression to addictive levels of intake than occurs with other addicting drugs. (Henningfield, Cohen, & Slade, 1991, p. 568)

Table I. Reinforcement potency of psychoactive drugs judged by laboratory and human usage patterns. This ordering of drugs was presented by Hartnoll (1990) for the United Kingdom.
Animal Clinical Epidemiologic
Readiness of Animal Self-Administration Prevalence of Human Clinical Problems Prevalence of Human Consumption
Cocaine Nicotine (tobacco) Caffeine
Amphetamines Alcohol Alcohol
Opiates Benzodiazepines Nicotine (tobacco)
Barbiturates Barbiturates Cannabis
Alcohol Opiates Benzodiazepines
Benzodiazepines Amphetamines/cocaine Amphetamines
Nicotine Psychedelics Psychedelics
Caffeine Cocaine
Cannabis Opiates

The (Behavioral) Economics of Drug Demand

Drug use by humans in natural environments reflects inherent costs associated with drug use (especially illicit use) and the tradeoffs between drugs and other available activities. One framework used by behavioral pharmacologists to explore the balancing of costs and benefits between different reinforcing practices is called the "behavioral economic" approach (DeGrandpre & Bickel, 1996; Carroll, 1993; Hursh, 1993). This model presents drug use as a preferential choice as in models of consumer behavior (Vuchinich & Tucker, 1988). Along with drug dosage, this model also incorporates factors affecting demand, such as income, work (cost) to obtain the drug, and the availability and cost of alternative reinforcers. Accordingly, drugs are considered an economic commodity and the organism's drug-taking is defined in terms of demand and demand elasticity, the latter of which refers to changes in drug consumption based on both drug cost and the availability/cost of competing reinforcers.

This economic framework provides a method of quantifying the dynamic interaction between competing drug and nondrug activities, relationships that have been investigated in some detail since the time of earlier research examining cocaine's toxicity. To wit, how is drug consumption affected by cost? Is demand for cocaine inelastic, as proposed in some versions of the addiction model (cf. Fischman, 1989)? Does availability of important alternative rewards depreciate cocaine-taking responses in animals? How does the organism balance the cost/dose of cocaine with the cost/magnitude of competing reinforcers?

Demand Curves for Drug and Non-Drug Reinforcers

Figure 2 displays five demand curves for drug and non-drug reinforcers tested under similar experimental conditions. In all cases, increasing the unit price of the reinforcer (i.e., responses required per unit of reinforcer) produces a positively-accelerated, nonlinear demand function (data are on log axes where slope equals elasticity; DeGrandpre, Bickel, Hughes, Layng, & Badger, 1993). Cost clearly exerts strong control over cocaine consumption and creates similar demand curves for cocaine and other potent reinforcers (whether cocaine, d-amphetamine, food, heroin, or nicotine) in a variety of species (including primates, pigeons, and humans)(see also Bickel, DeGrandpre, Higgins, & Hughes, 1990).

Consumption of a self-administered drug...
Figure 2. Consumption of a self-administered drug (top four panels) and food (bottom panel) is plotted as a function of unit price (unit price = responses required/reinforcer magnitude) in each of the five graphs taken from four studies. Each data point represents an individual experimental session. A line of best fit is shown for each data set that was derived from a multiple-regression equation (see DeGrandpre, Bickel, Hughes, Layng, & Badger, 1993; data are from: W. K. Bickel & R. J. DeGrandpre, unpublished data; Goldberg, 1973; Harrigan & Downs, 1978; Peden & Timberlake, 1984).

Cocaine Demand and Response-Rate Curves

The evidence derived from the behavioral-economic analyses is inconsistent with the idea that "Cocaine ... [is such a] potent reinforcer, ... [that] the self-administering organism is resistant to any attempts to decrease drug-taking" (Fischman, 1988, p. 7). Consider Figure 3, which shows response-rate functions (i.e., operant responding leading to drug infusions) specifically for cocaine self-administration by monkeys (rhesus or squirrel) in five studies covering two decades (see DeGrandpre et al., 1993). In all five studies, as unit price increases, response rate first increases, reaches a maximum, and then decreases.

The top response-rate graph in Figure 3 corresponds to the top demand curve graph in Figure 2, indicating that the nonlinear demand function results from a bitonic response-rate function. While the animal may at first attempt (partially) to maintain consumption by increasing bar-pressing, it eventually decreases its responses and accepts drastically depleted cellular cocaine levels. Thus, even animals consuming high levels of cocaine under impoverished laboratory conditions are unwilling to do the additional work of increased bar-pressing (which the animals can easily manage) to maintain their drug intake. Or, if viewed from the other direction, these data show that the animals could have responded more at the lower unit prices to increase their drug consumption; i.e., the animals were actually regulating cocaine intake at low prices by not self-administering as rapidly as they showed they could at somewhat higher unit prices. From either perspective, these studies do not reveal uncontrolled cocaine use even with easy (albeit not unlimited) access and the absence of competing alternatives to drug taking. This finding disputes the classical addiction model of inelastic demand and a monomaniacal increase in drug-seeking behavior.


Cocaine-maintained responding by monkeys... Figure 3. Cocaine-maintained responding by monkeys is plotted as a function of unit price (unit price = responses required/ reinforcer magnitude) for five different studies (Downs & Woods, 1974; Goldberg, 1973; Goldberg & Kelleher, 1976; Meisch, George, & Lemaire, 1990; Spear, Muntaner, Goldberg, & Katz, 1991). These data represent an attempt to compare cocaine-maintained responding across all studies that employed similar procedures in rhesus or squirrel monkeys (see DeGrandpre, Bickel, Hughes, Layng, & Badger, 1993).

Scheduled Alternatives: Demand for Cocaine Versus Food

We have noted that in the toxicity studies, animals were provided no significant alternatives to cocaine use. One toxicity study found animals preferred cocaine to the total exclusion of food (Aigner & Balster, 1978). But this result does not necessarily occur when food and cocaine are scheduled as competing reinforcers. For example, Carroll, Lac, and Nygaard (1989) assessed the effects of access to a sweetened (oral glucose/sucrose; G + S) solution on IV cocaine self-administration. In this study, animals doubled their cocaine demand when a previously available G + S solution was removed, while presenting a G + S solution significantly reduced the number of cocaine infusions for another group of rats already self-administering cocaine. Finally, a third group of rats in this study did not initially acquire cocaine self-administration when G + S was available. When G + S was replaced with water, however, they acquired and increased their demand for cocaine to over 400 infusions per day. Carroll and Lac (1993) also found that access to G + S prevented half of animals from acquiring self-administration of IV cocaine and significantly delayed it for many others.

In economic terms, raising the magnitude or dose of one reinforcer or increasing the work to acquire it should affect preferences for alternative reinforcers that can function as economic substitutes (Bickel, DeGrandpre, & Higgins, 1993; Bickel, DeGrandpre, & Higgins, 1995). Figure 4 shows the results of two cocaine self-administration studies in rhesus monkeys. Nader and Woolverton (1991; 1992) examined (1) the effects of magnitude of an alternative food reinforcer (banana-flavored food pellets) on cocaine intake, and (2) the effects of work requirements on choice between food pellets and cocaine. In each case, the demand for IV cocaine increased as the dose of cocaine increased — when the magnitude of the alternative stimulus and the work requirement for both remained constant. However, the demand for IV cocaine decreased (1) as the magnitude of the food stimulus increased (left graph), and (2) when the work requirement for cocaine increased (right graph). In fact, in the high-food-magnitude condition, animals never chose more than 50 percent of the cocaine choices, even at the highest cocaine dose.


Effects of increasing the magnitude...
Figure. 4. Effects of increasing the magnitude of an alternative source of reinforcement on cocaine choices in a rhesus monkey (M-8624; left graph), and the effects of increasing the work requirement for cocaine on cocaine choices in a rhesus monkey (M-8704; right graph). Effects of both manipulations are shown across several doses of IV cocaine. Data are taken from Woolverton (1992).


These results are consistent with the conclusion that:

Punishment, the availability of alternative reinforcers and response cost all determine whether or not cocaine has a reinforcing effect. In a sense, these are not surprising results.... Although current rhetoric would often have one believe otherwise, these results emphasize that the self-administration of cocaine is governed by the same laws that govern behavior maintained by other positive reinforcers. (Woolverton, 1992, p. 158)

Community Reinforcement Therapy

The economic model of drug dependence suggests the value of therapies which enhance alternative non-drug activities for addicts. The Community Reinforcement Approach (CRA), developed for alcoholics, includes a broad-based behavioral package of stress management, family counseling, job preparation, social skills training, recreational planning, and a buddy system (Sisson & Azrin, 1989). Outcomes of several clinical trials produced unusually strong results for CRA compared with standard hospital alcoholism treatments.

In addition, because of the low intensity of therapist intervention, one study found CRA to be by far the most cost-effective alcoholism treatment (Finney & Monahan, 1996). Nonetheless, CRA is not utilized in standard treatments (Miller & Hester, 1986). A CRA/contingency-reinforcement therapeutic model has also been utilized for cocaine dependence: Spouses or "buddies" and counselors were taught to socially reinforce clients for cocaine-free urine tests, patients were trained to recognize antecedents and consequences of drug use, and job counseling and recreational planning were provided. Most important, cocaine abstinence was reinforced with vouchers redeemable for recreational and consumer items. Studies of this approach have found far greater treatment acceptability, treatment retention, and success in achieving and maintaining cocaine abstinence for CRA than for comparison groups treated with a group support program based on the 12-step model (Higgins, Budney, Bickel et al., 1993; Higgins, Delaney, Budney et al., 1991).

Contingency-reinforcement models have also been used successfully with nicotine (Hall, Tunstall, Ginsberg et al., 1987), opiate (Higgins, Stitzer, Bigelow, & Liebson, 1986), benzodiazepine (Stitzer, Bigelow, Liebson, & Hawthorne, 1982), and amphetamine (Boudin, 1972) dependence. Therapies aimed at highlighting alternative reinforcers and the addict's ability to acquire them show clear promise in the addiction field, where few treatments have been subjected to systematic evaluation and fewer still have had substantial success across more than one addiction.

Policy Implications of the Economic Understanding of Addiction

Instead of a hardwired biological mechanism which is either "on" or "off," the behavioral-economic model portrays drug use as "more" or "less," even at extreme ends of consumption labelled addictive (DeGrandpre & Bickel, 1996). The economic model is thus consistent with the harm reduction model, which focuses on the many addicts who will not immediately abstain, but who could nonetheless reduce usage and harms while improving their overall functioning (Peele, 1994; 1996).

Behavioral economics also delineates the role of environmental factors in reducing drug use, a role described in the case of crack by prominent clinicians and scientists in a 1989 newspaper article subtitled: "In shift, importance of users' environments is stressed over the drug's attributes" (Kolata, 1989). In this view, "If crack were a drug of the middle and upper classes, we would not be saying it is so impossible to treat" (F. Gawin in Kolata, 1989, p. 1). On the other hand, the NIMH ECA Program identified increased income and recently gaining a job as cocaine progression risks (Ritter & Anthony, 1991). The same study also found age and marriage negatively predicted cocaine abuse, and prior use of other drugs was by far the best predictor of cocaine abuse. This suggests that when drugs are available, and when money is not imbedded in a larger set of attachments to family, career, and community practices, the simplistic economic prediction that more money equals greater drug use will be borne out (cf. Bickel & DeGrandpre, 1995).

This interaction between economic and social forces has created our nation's current drug epidemic. Although cocaine became more available and cheaper throughout the 1980s, overall usage in North America declined. However, a counterpattern of intensive use, including use of crack cocaine and heroin, simultaneously affected the most economically impoverished groups in America:

The question we must be asking now is not why people take drugs, but why do people stop. In the inner city, the factors that counterbalance drug use — family, employment, status within the community — often are not there. It is harder for people with nothing to say no to drugs. (D. Musto quoted in Kerr, 1987, p. 1)

These forces are difficult ones to change. Still, powerful life transitions which can influence addiction — marriage, maturation, personal resolve, religious transformation, changes in social networks — are regular features of ordinary human existence. Over 90 percent of 40 million American ex-smokers have quit without treatment (Fiore, Novotny, Pierce, et al., 1990), even though cigarettes are harder to quit (Kozlowski, Wilkinson, Skinner, et al., 1989) and more addictive than cocaine ("9 out of 10 people who tried cigarettes in real life went on to become addicted; 1 in 6 who tried crack became addicted"; J. Henningfield in Kolata, 1989, p. B7; cf. Henningfield et al., 1991). Natural remission is a powerful force in amelioration of drug addiction (Biernacki, 1986; Waldorf et al., 1991).

Finally, the basic behavioral-economic model applied here should not be confused with the "rational" model of addiction proposed by Nobel laureate economist Gary Becker (Becker & Murphy, 1988). Becker's theory posits drugs as overweening reinforcers so appealing that the only way to reduce demand is through eliminating drug availability. His economic modeling is actually based on the pharmacologically-deterministic view that wholly ignores the abundant data on actual drug use, such as the large majority of formerly addicted Vietnam veterans who took narcotics in their more rewarding home environments but did not become readdicted (Robins, Helzer, Hesselbrock, & Wish, 1980). This result startled the Vietnam drug researchers themselves since "laboratory experiments have shown it [heroin] to be a highly addicting drug" (p. 216). Actually, despite claims to the contrary, there is no disagreement between animal models of drug taking and naturalistic drug use: in both spheres all drug use depends on individual history and prevailing environmental circumstances.


  1. The results of this study should not be overinterpreted: (a) only 78 individuals who initiated or progressed in their cocaine use were analyzed, (b) the initial sample represented high-risk individuals, (c) the period covered in the research (the early 1980s) was an overall growth period for cocaine use, and (d) the initiation/progression categories were extremely gross, and lumped together new initiators with those who progressed from one to five use occasions to six or more occasions during follow-up.
  2. Benzodiazepine and caffeine addictions are often faceless, since this drug use is so readily accepted in our society and all of us know many nonaddicted users. When personal accounts of severe and problematic addiction were reported for tranquilizers in several first-person best sellers such as Barbara Gordon's I'm Dancing as Fast as I Can and Betty Ford's The Times of My Life, industry spokespeople pointed out that these cases were quite unusual. Indeed they are, as are cases of cocaine addiction. Nonetheless, it strikes most people as ludicrous to speak of tranquilizer and cocaine addiction in the same breath. One person for whom this is not the case is New York television newscaster Jim Jensen, who reported readily giving up a cocaine habit in treatment but being unable to shake his valium dependence: "Valium withdrawal soon plunged him into a massive depression that left him unable to eat or sleep. It took two more months in two hospitals for him to regain his mental and physical health" (Jensen, 1989, p. 67).


Adlaf, E. M., Smart, R. G., & Canale, M. D. (1991). Drug use among Ontario adults 1977-1991. Toronto: Ontario Addiction Research Foundation.

Aigner, T. G., & Balster, R. L. (1978). Choice behavior in Rhesus monkeys: Cocaine versus food. Science, 201, 534-535.

Alexander, B. K., Peele, S., Hadaway, P., Morse, S. J., Brodsky, A., Beyerstein, B.L. (1985). In S. Peele, The meaning of addiction: Compulsive experience and its interpretation (pp. 73-96). Lexington, MA: Lexington.

Becker, G. S., & Murphy, K. M. (1988). A theory of rational addiction. Journal of Political Economy, 96, 675-700.

Berridge, V., & Edwards, G. (1987). Opium and the people: Opiate use in nineteenth-century England (2nd ed.). New Haven, CT: Yale University.

Bickel, W. K., & DeGrandpre, R. J. (1995). Price and alternatives: Suggestions for drug policy from psychology. International Journal of Drug Policy, 6, 93-105.

Bickel, W. K., DeGrandpre, R. J., & Higgins, S. T. (1993). Behavioral economics: A novel experimental approach to the study of drug dependence. Drug and Alcohol Dependence, 33, 173-192.

Bickel, W. K., DeGrandpre, R. J., & Higgins, S. T. (1995). The behavioral economics of concurrent drug reinforcers: A review and reanalysis of drug self-administration research. Psychopharmacology, 118, 250-259.

Bickel, W. K., DeGrandpre, R. J., Higgins, S. T., & Hughes, J. R. (1990). Behavioral economics and drug self-administration. I. Functional equivalence of response requirement and drug dose. Life Sciences, 47, 1501-1510.

Biernacki, P. (1986). Pathways from heroin addiction: Recovery without treatment. Philadelphia: Temple University.

Boudin, N. M. (1972). Contingency contracting as a therapeutic tool in the deceleration of amphetamine use. Behavior Therapy, 3, 604-608.

Bozarth, M. A., & Wise, R. A. (1985). Toxicity associated with long-term intravenous heroin and cocaine self-administration in the rat. JAMA, 254, 81-83.

Bry, B. H. (1996). Psychological approaches to prevention. In W. K. Bickel and R. J. DeGrandpre (Eds.), Drug policy and human nature: Psychological perspectives on the prevention, management, and treatment of illicit drug abuse (pp. 55-76). New York: Plenum.

Carroll, M. E. (1993). The economic context of drug and non-drug reinforcers affects acquisition and maintenance of drug-reinforced behavior and withdrawal effects. Drug and Alcohol Dependence, 33, 201-210.

Carroll, M. E., & Lac, S. T. (1993). Autoshaping I.V. cocaine self-administration in rats: Effects of nondrug alternative reinforcers on acquisition. Psychopharmacology, 110, 5-12.

Carroll, M. E., Lac, S. T., & Nygaard, S. L. (1989). A concurrently available nondrug reinforcer prevents the acquisition or decreases the maintenance of cocaine-reinforced behavior. Psychopharmacology, 97, 23-29.

Cohen, P. (1989). Cocaine use in Amsterdam in nondeviant subcultures. Amsterdam: Instituut voor Sociale Geografie, Universiteit van Amsterdam.

Cohen, P. D., & Sas, A. (1992). Loss of control over cocaine: Rule or exception? Paper presented at the American Society of Criminology, New Orleans, November 3-7, 1992.

Cohen, P. D., & Sas, A. (1994). Cocaine use in Amsterdam in nondeviant subcultures. Addiction Research, 2, 71-94.

Cohen, S. (1984). Reinforcement and rapid delivery systems: Understanding adverse consequence of cocaine. In N. J. Kozel & E. H. Adams (Eds.), Cocaine use in America: Epidemiologic clinical perspectives (DHHS Publication Number ADM 85-1414; pp. 151-157). Washington, DC: U.S. Government Printing Office.

DeGrandpre, R. J. (1996). The impact of socially constructed knowledge on drug policy. In W. K. Bickel and R. J. DeGrandpre (Eds.), Drug policy and human nature: Psychological perspectives on the prevention, management, and treatment of illicit drug abuse (pp. 301-322). New York: Plenum.

DeGrandpre, R. J., & Bickel, W. K. (1996). Drug dependence in consumer demand. In L. Green and J. Kagel (Eds.), Advances in Behavioral Economics (Vol. 3, pp. 1-35). New Jersey: Ablex.

DeGrandpre, R. J., Bickel, W. K., Hughes, J. R., Layng, M.P., & Badger, G. (1993). Unit price as a useful metric in analyzing effects of reinforcer magnitude. Journal of the Experimental Analysis of Behavior, 60, 641-666.

DeGrandpre, R. J., & White, E. (1996). Drugs: In the care of the self. Common Knowledge.

Deneau, G., Yanagita, T., & Seevers, M. H. (1969). Self-administration of psychoactive drugs by the monkey: A measure of psychological dependence. Psychopharmacologia, 16, 30-48.

Ditton, J., Farrow, K., Forsyth, A., Hammersley, R., Hunter, G., Lavelle, T., Mullen, K., et al. (1991). Scottish cocaine users: Wealthy snorters or delinquent smokers? Drug and Alcohol Dependence, 28, 269-276.

Downs, D. A., & Woods, J. H. (1974). Codeine and cocaine reinforced responding in Rhesus monkeys: Effects of dose on response rates under a fixed-ratio schedule. Journal of Pharmacology and Experimental Therapeutics, 191, 179-188.

Dworkin, S., Goeders, N. E., Grabowski, J., & Smith, J. E. (1986). The effects of 12-hour limited access to cocaine: Reduction in drug intake and mortality. In L. S. Harris (Ed.). Problems of drug dependence (DHHS Pub. No. ADM 85-1414; pp. 221-225). Washington, DC: U.S. Government Printing Office.

Eddy, N. B., Halbach, H., Isbell, H., & Seevers, M. H. (1965). Drug dependence: Its significance and characteristics. Bulletin of the World Health Organization, 32, 721-733.

Erickson, P. G. (1993). Prospects of harm reduction for psychostimulants. In N. Heather, A. Wodak, E. A. Nadelmann, & P. O'Hare (Eds.), Psychoactive drugs and harm reduction (pp. 184-210). London: Whurr.

Erickson, P. G., Adlaf, E. M., Murray, G. F., & Smart, R. G. (1987). The steel drug: Cocaine in perspective. Lexington, MA: Lexington.

Erickson, P. G., & Alexander, B. K. (1989). Cocaine and addictive liability. Social Pharmacology, 3, 249-270.

Fagin, J., & Chin, K. L. (1989). Initiation into crack and cocaine: A tale of two epidemics. Contemporary Drug Problems, 17, 579-616.

Falk, J. L. (1994). The discriminative stimulus and its repetition: Role in the instigation of drug abuse. Experimental and Clinical Psychopharmacology, 2, 43-52.

Falk, J. L., Dews, P. B., & Schuster, C. R. (1983). Environmental control of behavior. In P.K. Levison, D. R. Gerstein, & D. R. Maloff (Eds.), Commonalities in substance abuse and habitual behavior (pp. 47-110). Lexington, MA: Lexington.

Finney, J. W., & Monahan, S. C. (1996). The cost-effectiveness of treatment for alcoholism: A second approximation. Journal of Studies on Alcohol, 57, 229-243.

Fiore, M. C., Novotny, T. E., Pierce, J. P., Giovino, G. A., Hatziandreu, E. J., Newcomb, P. A., Surawicz, T. S., & Davis, R. M. (1990). Methods used to quit smoking in the United States. Journal of the American Medical Association, 263, 2760-2765.

Fischman, M. W. (1988). Behavioral pharmacology of cocaine. Journal of Clinical Psychiatry, 49, 7-10.

Fischman, M. W., Schuster, C. R., Resnekov, L., Schick, J. F. E., Krasnegor, N. A., Fennell, W., & Freedman, D. X. (1976). Cardiovascular and subjective effects of intravenous cocaine administration in humans. Archives of General Psychiatry, 33, 983-989.

Fitch, T. E., & Roberts, D. C. S. (1993). The effects of dose and access on the periodicity of cocaine self-administration in the rat. Drug and Alcohol Dependence, 33, 119-128.

Gawin, F. H. (1991). Cocaine addiction: Psychology and neurophysiology. Science, 251, 1580-1586.

Goldberg, S. R. (1973). Comparable behavior maintained under fixed-ratio and second-order schedules of food presentation, cocaine injection or d-amphetamine injection in the squirrel monkey. Journal of Pharmacology and Experimental Therapeutics, 186, 18-30.

Goldberg, S. R., & Kelleher, R. T. (1976). Behavior controlled by scheduled injections of cocaine in squirrel and rhesus monkeys. Journal of the Experimental Analysis of Behavior, 25, 93-104.

Hall, S. M., Tunstall, C., Ginsberg, D., Benowitz, N. L., & Jones, R. T. (1987). Nicotine gum and behavioral treatment. Journal of Consulting and Clinical Psychology, 55, 603-605.

Harrigan, S. E., & Downs, D. A. (1978). Self-administration of heroin, acetylnethadol, morphine, and methadone in rhesus monkeys. Life Sciences, 22, 619-624.

Harrison, L. D. (1994). Cocaine using careers in perspective. Addiction Research, 2, 1-20.

Hartnoll, R. (1990). Non-pharmacological factors in drug abuse. Behavioural Pharmacology, 1, 375-384.

Henningfield, J. E., Cohen, C., & Slade, J. D. (1991). Is nicotine more addictive than cocaine? British Journal of Addiction, 86, 565-570.

Higgins, S. T., Budney, A. J., Bickel, W. K., Hughes, J. R., Foerg, F., & Badger, G. (1993). Achieving cocaine abstinence with a behavioral approach. American Journal of Psychiatry. 150, 763-769.

Higgins, S. T., Delaney, D. D., Budney, A. J., Bickel, W. K., Hughes, J.R., Foerg, F., & Fenwick, J. W. (1991). A behavioral approach to achieving initial cocaine abstinence. American Journal of Psychiatry, 148, 1218-1224.

Higgins, S. T., Stitzer, M. L., Bigelow, G. E., & Liebson, I. A. (1986). Contingent methadone delivery: Effects on illicit-opiate use. Drug and Alcohol Dependence, 17, 311-322.

Hughes, J. R., Higgins, S. T., & Bickel, W. K. (1988). Behavioral "properties" of drugs. Psychopharmacology, 96, 557.

Hughes, J. R., Oliveto, A. H., Helzer, J. E., Higgins, S. T., & Bickel, W.K. (1992). Should caffeine abuse, dependence, or withdrawal be added to DSM-IV or ICD-10? American Journal of Psychiatry, 149, 33-40.

Hursch, S. R. (1993). Behavioral economics of drug self-administration: An introduction. Drug and Alcohol Dependence, 33, 165-172.

Hyman, S. (1996). Shaking out the cause of addiction. Science, 273, 611-612.

Isbell, H. (1958). Clinical research on addiction in the United States. In R. B. Livingston (Ed.). Narcotic drug addiction problems (pp. 114-130). Bethesda, MD: Public Health Service.

Jaffe, J. H. (1980). Drug addiction and drug abuse. In A. G. Gilman, L.S. Goodman, & B. A. Gilman (Eds.), Goodman and Gilman's the pharmacological basis of therapeutics (6th. ed.; pp. 535-584). New York: Macmillan.

Jensen, J. (1989, September 4). A veteran TV anchorman's toughest story was his own — he had to beat drugs and depression. People, pp. 67-73.

Johanson, C. E. (1984). Assessment of the dependence potential of cocaine in animals. In J. Grabowski (Ed.). Cocaine: Pharmacology, effects, and treatment of abuse (DHHS Publication Number ADM 84-1326; pp. 54-71). Rockville, MD: U.S. Government Printing Office.

Johanson, C. E., Balster, R. L., & Bonese, K. (1976). Self-administration of psychomotor stimulant drugs: The effects of unlimited access. Pharmacology Biochemistry & Behavior, 4, 45-51.

Johanson, C. E., & Fischman, M. W. (1989). The pharmacology of cocaine related to its abuse. Pharmacological Reviews, 41, 3-52.

Johanson, C. E., & Uhlenhuth, E. H. (1981). Drug preference and mood in humans: Repeated assessment of d-amphetamine. Pharmacology Biochemistry & Behavior, 14, 159-163.

Jones, R. T. (1992). Alternative strategies. In G. R. Rock & J. Whelan (Eds.), Cocaine: Scientific and social dimensions (pp. 224-241). West Sussex, UK: Wiley.

Kandel, D. B., Murphy, D., & Karus, D. (1985). Cocaine use in young adulthood: Patterns of use and psychosocial correlates. In N. J. Kozel and E. H. Adams (Eds.), Cocaine use in America: Epidemiologic and clinical perspectives (DHHS Publication No. ADM 85-1414, pp. 76-110). Washington, DC: U.S. Government Printing Office.

Katz, J. L. (1990). Models of relative reinforcing efficacy of drugs and their predictive utility. Behavioural Pharmacology, 1, 283-301.

Kerr, P. (1987, August 30). Rich vs. poor: Drug patterns are diverging. New York Times, pp. 1, 28.

Kolata, G. (1989, August 24). Experts finding new hope on treating cocaine addicts. New York Times, pp. 1, B7.

Kozlowski, L. T., Wilkinson, D. A., Skinner, W., Kent, C., Franklin, T., & Pope, M. (1989). Comparing tobacco cigarette dependence with other drug dependencies. Journal of the American Medical Association, 261, 898-901.

Lamb, R. J., Preston, K. L., Schindler, C. W., Meisch, R. A., Davis, F., Katz, J. L., Henningfield, J. E., & Goldberg, S. R. (1991). The reinforcing and subjective effects of morphine in post-addicts. A dose-response study. Journal of Pharmacology and Experimental Therapeutics, 259, 1165-1173.

Levine, H. G. (1978). The discovery of addiction: Changing conceptions of habitual drunkenness in America. Journal of Studies on Alcohol, 39, 143-174.

Light, A. B., & Torrance, E. G. (1929). Opiate addiction. VI: The effects of abrupt withdrawal followed by readministration of morphine in human addicts, with special reference to the composition of the blood, the circulation, and the metabolism. Archives of Internal Medicine, 44, 1-16.

Meisch, R. A., George, F. R., & Lemaire, G. A. (1990). Orally delivered cocaine as a reinforcer for rhesus monkeys. Pharmacology Biochemistry & Behavior, 35, 245-249.

Miller, W. R., & Hester, R. K. (1986). The effectiveness of alcoholism treatment: What research reveals. In W. R. Miller & N. Heather (Eds.), Treating addictive behaviors (pp. 121- 174). New York: Plenum.

Morgan, J. P., & Zimmer, L. (in press). Animal studies of self-administration of cocaine: Misinterpretation, misrepresentation and invalid extrapolation to human cocaine use. In P. Erickson et al. (Eds.), New public health policies and programs for the reduction of drug related harm. Toronto: University of Toronto Press.

Mugford, S., & Cohen, P. (1989). Drug use, social relations and commodity consumption: A study of recreational users in Sydney, Canberra and Melbourne. Canberra, Australia: Research into Drug Abuse Advisory Committee, National Campaign Against Drug Abusers.

Murphy, S., Reinarman, C. & Waldorf, D. (1989). An 11-year follow-up of a network of cocaine users. British Journal of Addiction, 84, 427-436.

Musto, D. F. (1987). The American disease: Origins of narcotic control (rev. ed.). New York: Oxford University.

Nader, M. A., & Woolverton, W. L. (1991). Effects of increasing the magnitude of an alternative reinforcer on drug choice in a discrete-trial choice procedure. Psychopharmacology, 105, 169-174.

Nader, M. A., & Woolverton, W. L. (1992). Effects of increasing response requirement on choice between cocaine and food in rhesus monkeys. Psychopharmacology, 108, 295-300.

National Institute on Drug Abuse (NIDA). (1991). National household survey on drug abuse: Population estimates 1990 (DHHS Publication Number ADM 91-1732). Washington, DC: U.S. Government Printing Office.

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

Newcomb, M. D., & Bentler, P. M. (1986b). Cocaine use among young adults. Advances in Alcohol and Substance Abuse, 6, 73-96.

Peden, B. F., & Timberlake, W. (1984). Effects of reward magnitude on key pecking and eating by pigeons in a closed economy. The Psychological Record, 34, 397-415.

Peele, S. (1985). The meaning of addiction: Compulsive experience and its interpretation. Lexington, MA: Lexington.

Peele, S. (1987). A moral vision of addiction: How people's values determine whether they become and remain addicts. Journal of Drug Issues, 17, 187-215.

Peele, S. (1990). Addiction as a cultural concept. Annals of the New York Academy of Sciences, 602, 205-220.

Peele, S. (1994). Harm reduction and alcoholism treatment: What happens to the majority of alcoholics who won't abstain? The Lindesmith Center, New York, NY.

Peele, S. (1996). Assumptions about drugs and the marketing of drug policies. In W. K. Bickel and R. J. DeGrandpre (Eds.), Drug policy and human nature: Psychological perspectives on the prevention, management, and treatment of illicit drug abuse (pp. 199-220). New York: Plenum.

Peele, S., & Brodsky, A. (1975). Love and addiction. New York: Taplinger, 1975.

Pickens, R. (1968). Self-administration of stimulants by rats. International Journal of the Addictions, 3, 215-221.

Reinarman, C., Murphy, S., & Waldorf, D. (1994). Pharmacology is not destiny: The contingent character of cocaine abuse and addiction. Addiction Research, 2, 21-36.

Ritter, C., & Anthony, J. C. (1991). Factors influencing initiation of cocaine use among adults. In S. Schober & C. Schade (Eds.), The epidemiology of cocaine use and abuse (NIDA Research Monograph 110; pp. 190-210). Rockville, MD: U.S. Government Printing Office.

Robins, L. N., Helzer, J. E., Hesselbrock, M., & Wish, E. (1980). Vietnam veterans three years after Vietnam: How our study changed our view of heroin. In L. Brill & C. Winick (Eds.), The Yearbook of Substance Use and Abuse (Vol. 2, pp. 213-230). New York: Human Sciences Press.

Schuster, C. R., & Thompson, T. I. (1969). Self-administration of and behavioral dependence on drugs. Annual Review of Pharmacology, 9, 483-502.

Sidman, M. (1956). Drug-behavior interactions. Annals of the New York Academy of Sciences, 65, 282-302.

Siegel, R. K. (1984). Changing patterns of cocaine use. In J. Grabowski (Ed.). Cocaine: Pharmacology, effects, and treatment of abuse (DHHS Publication Number ADM 84-1326; pp. 92-110). Rockville, MD: U.S. Government Printing Office.

Sisson, R. W., & Azrin, N. H. (1989). The community reinforcement approach. In R. K. Hester & W. R. Miller (Eds.), Handbook of alcoholism treatment approaches (pp. 242-253). New York: Pergamon.

Sonnedecker, G. (1958). Emergence and concept of the addiction problem. In R. B. Livingston (Ed.), Narcotic drug addiction problems (pp. 14-22). Bethesda, MD: Public Health Service.

Spear, D. J., Muntaner, C., Goldberg, S. R., & Katz, J. L. (1991). Methohexital and cocaine self-administration under fixed-ratio and second-order schedules. Pharmacology Biochemistry & Behavior, 38, 411-416.

Stitzer, M. L., Bigelow, G. E., Liebson, I. A., & Hawthorne, J. W. (1982). Contingent reinforcement for benzodiazepine-free urines. Journal of Applied Behavioral Analysis, 15, 493-503.

Thompson, T., & Pickens, R. (1970, February). Stimulant self-administration by animals: Some comparisons with opiate self-administration. Federation Proceedings, pp. 6-12.

Vuchinich, R.E., & Tucker, J.A. (1988). Contributions from behavioral theories of choice as a framework to an analysis of alcohol abuse. Journal of Abnormal Psychology, 92, 408-416.

Waldorf, D., Reinarman, C., & Murphy, S. (1991). Cocaine changes: The experience of using and quitting. Philadelphia: Temple University.

Wilson, M.C., Hitomi, M., & Schuster, C. R. (1971). Psychomotor stimulant self-administration as a function of dosage per injection in the rhesus monkey. Psychopharmacologia, 22, 271-281.

Wise, R.A. (1989). The neurobiology of craving: Implications for the understanding and treatment of addiction. Journal of Abnormal Psychology, 97, 118-132.

Woolverton, W. L. (1992). Determinants of cocaine self-administration by laboratory animals. In G. R. Bock and J. Whelan (Eds.), Cocaine: Scientific and social dimensions (pp. 149-164). West Sussex: Wiley.

World Health Organization (1995, March 14). Publication of the largest global study on cocaine use ever undertaken. Press release, Brussels, Belgium.

World Health Organization (WHO) Expert Committee on Addiction-Producing Drugs. (1981). WHO Bulletin, 59, 225.

Yanagita, T., Deneau, G.A., & Seevers, M.H. (1965). Evaluation of pharmacological agents in the monkey by long term intravenous self or programmed administration. Excerpta Medica International Congress Series, 87, 453-457.

Zinberg, N. (1984). Drug, set, and setting: The basis for controlled intoxicant use. New Haven: Yale University.