The behavioral economics of concurrent drug reinforcers: a review and reanalysis of drug self-administration research

Ratio size and cocaine concentration effects on oral cocaine-reinforced behavior

Monkeys were given a choice between cocaine solutions and water under concurrent fixed-ratio reinforcement schedules. The operant response was spout contact. Six rhesus monkeys served as subjects. The cocaine concentration was varied from 0.0125 to 0.8 mg/ml, and the fixed-ratio value was varied from 8 to 128. Cocaine maintained higher response rates than did water over a wide range of conditions. Response rate and number of cocaine deliveries per session were inverted U-shaped functions of concentration. These functions were shifted to the right as the fixed ratio was increased. The number of cocaine deliveries was more persistent as fixed-ratio value was increased when the unit dose was larger rather than smaller. Cocaine consumption was analyzed as a function of unit price (fixed-ratio value divided by cocaine concentration), and unit price accounted for between 77% and 92% of the variance in cocaine consumption for individual monkeys. The current data support the claim that a drug's reinforcing effects increase directly with dose and underscore the need to gather parametric data when examining the effects of experimental manipulations on a drug-reinforced baseline.

Further analysis of choice between cocaine and food using the unit price model of behavioral economics

Behavioral economics defines unit price (UP) as the ratio of the response requirement to magnitude of reinforcer. When applied to drug self-administration, the UP model defines UP as the ratio of the response requirement to the unit dose of drug. This model makes two predictions about drug self-administration: increasing UP decreases consumption and consumption at a given UP will be constant regardless of the response requirement and dose that make up the UP. In previous experiments conducted in rhesus monkeys allowed to choose between an i.v. injection of cocaine and food, the UP model has failed to adequately predict drug consumption in that consumption varied (increased with dose) at a given UP. However, previous experiments have allowed a fixed number of choice trials/day, thereby imposing a procedural ceiling on consumption that may have influenced conformity to the UP model. In the present experiment, the number of choice trials available was varied in such a way that constant drug consumption was possible over the range of UPs tested. The response requirement for cocaine was varied between 15 and 1200 lever presses/injection and the dose of cocaine was varied between 0.05 and 0.2 mg/kg/inj, yielding UPs from 300 to 5600 responses/mg/kg. The response requirement for food was always 30. As predicted by the UP model, cocaine consumption decreased as UP increased. Moreover, in contrast to previous experiments, consumption did not vary significantly across the response requirement/dose combinations that made up a UP. A detailed analysis suggested that a decrease in magnitude of the alternative reinforcer (one rather than three food pellets), rather than the increase in trials, was responsible for the improved conformity to the UP model in the present experiment relative to previous experiments. Taken together with previous experiments, the present experiment suggests that conformity to the UP model of drug consumption in a choice situation is dependent upon the magnitude of alternative reinforcers that are available. Consumption was best predicted by the UP model when the magnitude of the alternative reinforcer was small.

Food and amphetamine self-administration by baboons: effects of alternatives

The effects of the availability of an alternative reinforcer on responding maintained by food pellets or fluid solutions were examined in 6 adult male baboons (Papio cynocephalus anubis). During daily 23-hr experimental sessions, baboons had concurrent access to both food pellets and fluid, with responding maintained under fixed-ratio schedules of reinforcement that varied between the two commodities. The fixed-ratio requirement, or cost, for pellets was increased when (a) no fluid, (b) a dilute dextrose vehicle, (c) 0.002 mg/kg d-amphetamine, or (d) 0.004 mg/kg d-amphetamine was available. When given nonrestricted concurrent access to food pellets and amphetamine at minimal cost (FR 2), baboons self-administered sufficient amphetamine to decrease pellet intake. Increasing the response requirement for pellets decreased pellet intake at a similar rate regardless of the available fluid and increased fluid intake in a variable manner among baboons such that there were no statistically significant increases in fluid intake. In contrast, when access to pellets was restricted to 70% of maximal intake under nonrestricted conditions, increasing pellet cost decreased pellet intake and increased fluid intake more rapidly when the high amphetamine dose was available. Thus, amphetamine was more effective as an economic substitute for pellets when access to pellets was restricted. The response cost for vehicle and both amphetamine concentrations was increased when baboons had nonrestricted and restricted access to pellets. Increasing the response requirement for fluid delivery decreased intake of all three fluids similarly under both pellet-access conditions. The results indicate that substitution between commodities with minimal commonalities can be studied under controlled laboratory conditions and is dependent upon reinforcement schedule and commodity restrictions.

Modeling drug dependence behaviors for animal and human studies

Laboratory models are available to study drug reinforcement in animals and humans, but few are available to study other drug dependence phenomena, e.g., difficulty stopping or use despite harm. The present paper is a first attempt to illustrate the feasibility of developing such models for use in both nonhuman and human research and discusses their possible utility in research to understand and treat stimulant and other drug dependencies.

College on problems of drug dependence meeting, Puerto Rico (June 1996) marijuana use and dependence

Discoveries concerning an endogenous cannabinoid system and observations of dramatic increases in marijuana use among youth in the United States have fueled a recent increase in basic and clinical research to better understand and treat marijuana dependence. At the annual meeting of the College on Problems of Drug Dependence (Puerto Rico, 1996) a symposium 'Marijuana Use: Basic Mechanisms, Epidemiology, and Clinical Issues' reviewed a number of important areas of ongoing research that address marijuana dependence. Overviews and original research were presented regarding the development of dependence (preclinical and clinical research), motivational effects (laboratory models), the epidemiology of dependence and its development, clinical management of marijuana use among patients seeking treatment for other drugs of abuse, and treatment for adult marijuana dependence. This paper summarizes the symposium presentations and provides discussion of recent scientific developments concerning marijuana use and dependence.

Addiction as choice? Yes. As melioration? Maybe, maybe not

Melioration is not superior and may be inferior to other extensions of behavioral choice theory to addiction studies. Progress in the addiction field will be accelerated by the demise of the disease model rather than by attempting to resolve the contradictions it has created.

Maximization should sometimes lead to abstinence

Heyman's model, paradoxically, predicts that whereas a maximizing approach to drug choice will prevent escalation of drug use it will never yield complete abstinence. We suggest an alternative model that overcomes this difficulty by focusing on changes in drug tolerance. A small modification allows maximization to predict either abstinence or moderation (e.g., social drinking).

Which behavioral consequences matter? The importance of frame of reference in explaining addiction

The target article emphasizes the relationship between a matching law-based theory of addiction and the disease model of addiction. In contrast, this response emphasizes the relationship between the matching law theory and other behavioral approaches to addiction. The basic difference, I argue, is that the matching law specifies that choice is governed by local reinforcement rates. In contrast, economics says that overall reinforcement rate controls choice, and for other approaches there are other measures or no clear prediction at all. The response also differs from the target article in that there is more emphasis on the finding that stimulus conditions determine whether choice is under local or overall reinforcement rate control.

The contradiction unresolved

Extreme sensations – thirst, pain – can focus attention on local consequences at the expense of the overall, perhaps for good evolutionary reasons. Maybe the same phenomenon evolves from prolonged use of addictive substances. The matching law explains mistaken choice, not how a person who has confronted personal catastrophe manages to ignore it in making a locally induced choice.

Addiction is not as puzzling as it seems

Heyman's target article seeks to resolve the apparent paradox surrounding the issue of control in relation to addictive behaviour. The present commentary argues that addiction is in fact less paradoxical and more easily understandable than Heyman supposes. A fully satisfactory model of addiction, however, requires a more multifaceted approach than that provided by the type of behavioural choice theory favoured by Heyman.

The Janus faces of addiction

Heyman proposes that external stimuli can promote a switch from a local to a global frame of reference for evaluating the consequences of behavior and that such a change might be critical to breaking the grip of drag addiction. Could incentive stimuli promote a switch in the opposite direction and thus contribute to relapse in the recovered addict?

Relationship between melioration and the controlling variables

Knowledge of specific instances of melioration does not imply knowledge of the variables controlling behavior. Drug self-administration has been successfully analyzed in terms of standard variables such as the immediacy and magnitude of the drug reinforcer. An appeal to melioration may not be necessary.

The behavioral economics of addiction: A comprehensive alternative

Heyman's target article makes a strong case for a behavioral approach to addiction, yet some important assumptions require justification, and promising behavioral alternatives to the author's melioration approach should be considered. In particular, the behavioral economic approach to addiction appears well developed and comprehensive. How does the melioration approach complement or improve on a behavioral economic account?

The ameliorating addict: An illusion reviewed

It is difficult to accommodate compulsive drug use that destroys regulation by consequences in a unitary framework of melioration or maximization, especially as both are strategies of behavioral regulation. The “contradiction” of addictive behavior despite aversive consequences is an illusory issue because consequences that do not decrease addictive behavior cannot be considered punishing.

Addiction requires philosophical explanation, not mere redescription

Heyman's model explains the irrationality of addictive behavior, but it does not satisfactorily answer the question of whether this behavior is voluntary because it does not address the issue of the choice of preference functions. Furthermore, although Heyman disconfirms the disease model of addiction, this does not resolve the issue of whether addiction should be classified as a mental illness.

Matching and melioration as accounts of reinforcement and drug addiction

Heyman's view that addiction can be viewed as a natural outcome predictable by melioration and the matching law is provocative. Remaining to be explained more fully, however, are exactly how his view is an improvement on other reinforcement-based accounts. Included in these elaborations should be an account of how different “bookkeeping schemes” are developed and controlled and what new approaches to treatment and prevention of drug addiction are indicated.

How do people choose between local and global bookkeeping?

The matching law accounts for both addictive behavior and the usefulness of a person's evaluating choices in overall categories. To explain why overall bookkeeping, once learned, does not easily win out over local bookkeeping, another implication of matching is needed: intertemporal bargaining. The role of melioration, though probably important for new addiction is separate.

Addiction: Taking the brain seriously

Heyman's target article is an analytical tour de force, but it makes too hard a distinction between voluntary and driven behavior. It is more fruitful to think about brain and behavior as shifting, interacting “agents,” represented by multiple neural circuits. This has the virtue of better connecting behavioral analysis with wet neuroscience.

Heyman's steady-state theory of addiction

Heyman's target article contributes to our understanding of addictions by offering solutions to several paradoxes and by recognizing the stable nature of addictive behavior. Previous classical and operant conditioning models have emphasized molecular processes, such as acquisition and extinction, and have failed to address the aggregate effects of long-term exposure to the contingencies of drug and alcohol use.

Matching observation to addiction theory

Over the years, many theories have been proposed to account for the aberrant behavior of drug dependent individuals. Heyman posits that the existing theories of drug dependence are inadequate to explain the complex processes inherent in human drug-taking. He proposes that incongruous behaviors that comprise addiction, such as continued drug use in spite of adverse consequences, can be explained by application of the matching law approach. While the matching law theory of addiction explains certain aspects of human behavior, its application to the area of addiction must be subjected to experimental verification.

The biological, social and clinical bases of drug addiction: commentary and debate

This article summarizes the main discussions at a meeting on the biological, social and clinical bases of drug addiction focused on contemporary topics in drug dependence. Four main domains are surveyed, reflecting the structure of the meeting: psychological and pharmacological factors; neurobiological substrates; risk factors (including a consideration of vulnerability from an environmental and genetic perspective); and clinical treatment. Among the topics discussed were tolerance, sensitization, withdrawal, craving and relapse; mechanisms of reinforcing actions of drugs at the behavioural, cognitive and neural levels; the role of subjective factors in drug dependence; approaches to the behavioural and molecular genetics of drug dependence; the use of functional neuroimaging; pharmaceutical and psychosocial strategies for treatment; epidemiological and sociological aspects of drug dependence. The survey takes into account the considerable disagreements and controversies arising from the discussions, but also reaches a degree of consensus in certain areas.

Behavioral economics of concurrent ethanol-sucrose and sucrose reinforcement in the rat: effects of altering variable-ratio requirements

These experiments examined the own-price and cross-price elasticities of a drug (ethanol mixed with 10% sucrose) and a nondrug (10% sucrose) reinforcer. Rats were presented with ethanol-sucrose and sucrose, both available on concurrent independent variable-ratio (VR) 8 schedules of reinforcement. In Experiment 1, the variable ratio for the ethanol mix was systematically raised to 10, 12, 14, 16, 20, and 30, while the variable ratio for sucrose remained at 8. Five of the 6 rats increased ethanol-reinforced responding at some of the increments and defended baseline levels of ethanol intake. However, the rats eventually ceased ethanol-reinforced responding at the highest variable ratios. Sucrose-reinforced responding was not systematically affected by the changes in variable ratio for ethanol mix. In Experiment 2, the variable ratio for sucrose was systematically increased while the ethanol-sucrose response requirement remained constant. The rats decreased sucrose-reinforced responding and increased ethanol-sucrose-reinforced responding, resulting in a two- to 10-fold increase in ethanol intake. Experiment 3 examined the substitutability of qualitatively identical reinforcers: 10% sucrose versus 10% sucrose. Increases in variable-ratio requirements at the preferred lever resulted in a switch in lever preference. Experiment 4 examined whether 10% ethanol mix substituted for 5% ethanol mix, with increasing variable-ratio requirements of the 5% ethanol. All rats eventually responded predominantly for the 10% ethanol mix, but total amount of ethanol consumed per session did not systematically change. In Experiment 5, the variable-ratio requirements for both ethanol and sucrose were simultaneously raised to VR 120; 7 of 8 rats increased ethanol-reinforced responding while decreasing sucrose-reinforced responding. These data suggest that, within this ethanol-induction procedure and within certain parameters, demand for ethanol-sucrose was relatively inelastic, and sucrose consumption was independent of ethanol-sucrose consumption. Demand for sucrose, on the other hand, was relatively elastic, and ethanol-sucrose readily substituted for it. The results are discussed in terms of applying a behavioral economic approach to relationships between drug and nondrug reinforcers.