Environment-Impacting Consumer Behavior: an Operant Analysis

Gordon Robert Foxall, University of Birmingham
ABSTRACT - A model of purchase and consumption based on the principles of behavior analysis is applied to the understanding of social demarketing interventions aimed at environmental preservation. The explanatory framework of the Behavioral Perspective Model of purchase and consumption (BPM) is outlined. The findings of behavior analysts on resource conservation and pollution control are interpreted in terms of the classification of consumer behaviors and contingencies proposed by the model. These environmentally-impacting consumer behaviors are elucidated through discussion of both their consequences and the nature of intervention to ameliorate them. An appropriate social marketing mix for each class of consumer behavior is suggested.
[ to cite ]:
Gordon Robert Foxall (1995) ,"Environment-Impacting Consumer Behavior: an Operant Analysis", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 262-268.

Advances in Consumer Research Volume 22, 1995      Pages 262-268

ENVIRONMENT-IMPACTING CONSUMER BEHAVIOR: AN OPERANT ANALYSIS

Gordon Robert Foxall, University of Birmingham

ABSTRACT -

A model of purchase and consumption based on the principles of behavior analysis is applied to the understanding of social demarketing interventions aimed at environmental preservation. The explanatory framework of the Behavioral Perspective Model of purchase and consumption (BPM) is outlined. The findings of behavior analysts on resource conservation and pollution control are interpreted in terms of the classification of consumer behaviors and contingencies proposed by the model. These environmentally-impacting consumer behaviors are elucidated through discussion of both their consequences and the nature of intervention to ameliorate them. An appropriate social marketing mix for each class of consumer behavior is suggested.

Applied behavior analysis has contributed substantially to knowledge of economic consumption as it adversely affects the physical and social environment (Cone and Hayes 1980; Geller, Winett and Everett 1982). It rests critically and selectively on Skinner's (1953) behavior theory which stresses that the causes of behavior are found in its environmental consequences. The applied field experimentation that characterizes this approach has explored how antecedent and consequential stimuli influence such consumer behaviors as excessive use of private transportation, over-consumption of domestic energy, littering and waste generation, and consumption of scarce resources such as water. The ultimate purpose has been to ameliorate these ecologically deleterious outcomes through the modification of consumer behavior (Winkler and Winett 1982). Despite this theoretical underpinning and practical direction, the empirical findings produced by applied behavior analysts lack systematic organization and theory-based generalization. Despite comprehensive reviews and attempts to draw lessons from these findings (Cone and Hayes 1980; Geller 1989), behavior analysis lacks an integrative model of consumer behavior and the effects of intervention based on a critical evaluation of behavior theory.

Thus the field has tended to fall back on cognitive frameworks devised by consumer researchers (Schwartz 1991) or the basic social marketing models derived from rudimentary marketing (Geller 1989). Yet both of these approaches contain serious conceptual, methodological and practical difficulties. Social marketing programs have been criticised on the grounds that they consist largely of informational and exhortative campaigns which do not generally use the full integrated marketing mix but rely to a disproportionate extent on social advertising. Further, the effectiveness of campaigns intended to change behavior by modifying attitudes has been questioned. Nor are these conclusions the result of theoretical preference or speculation uninformed by empirical evidence. There is abundant support from the research of cognitive psychologists themselves that, to the extent that they rely heavily on the use of persuasive communications to change pre-behavioral attitudes and values, social information campaigns have had little impact on consumers' conservation behavior. Despite vast general public knowledge about the potentially catastrophic consequences of failing to conserve energy, researchers report an inability to identify the required relationship between attitudes toward energy use and conservation; even those maximally informed are no more likely to save energy; nor does specifically informing people about the personal costs of current energy use and the benefits of reducing consumption affect behavior (Costanzo, Archer, Aronson and Pettigrew 1986).

However, applied behavior analysis indicates that simply informing people of the consequences of their actions is unlikely to modify their behavior unless they have been systematically exposed to those consequences in the past. The causes of behavior cannot be found in variables inferred from the behavior itself such as attitudes and these inferences are not therefore legitimate targets for intervention. Social demarketing based on applied behavior analysis can be systematically related to marketing mix management by a model founded on the behavior theory that underpins that analysis, elucidating the nature of both environmentally-impacting consumer behavior and interventions designed to ameliorate its deleterious ecological effects. First, the model and applied behavior analysis are described. Second, general propositions which must be fulfilled if the model accurately synthesises the findings of the applied research of behavior analysts are derived. Third, the major target behaviors of this research - transportation, energy consumption, waste disposal, water use - are described in terms of the model; the model's capacity to encompass and explain the incentives and feedback that control each of these behaviors is a prime concern. Finally, the role of the marketing mix in influencing each class of environment-impacting consumer behavior is discussed.

THE BEHAVIORAL PERSPECTIVE MODEL

The Behavioral Perspective Model of purchase and consumption (BPM) provides an account of consumer choice based on the antecedent and consequent learning contingencies which have featured in applied behavior analytic studies: antecedent setting variables, and consequent reinforcement and punishment (Figure 1). The BPM also clarifies the role of marketing management in the control of consumer behavior. The provenance and structure of the model have been thoroughly described by Foxall (1993, 1994); the following account is a summary. At the theoretical level, the purpose of the model and its developmental research program is not to supplant these or other structural approaches but to further understanding of the role of contextual influences and consumers' learning histories on their current actions as buyers and users of economic resources. The central explanatory mechanism of the model is the synomorphic consumer situation, the meeting place of the consumer's history of reinforcement and punishment (representing the personal variables responsible for current behavior) and the setting in which purchase and consumption occur (representing the contextual influences on behavior). The former incorporates the consumer's prior experience of purchase and consumption and the effects of the consequences of these acts on the probability of their performing similar consumer behaviors in the present and future.

FIGURE 1

SUMMARY OF THE BEHAVIORAL PERSPECTIVE MODEL

The settings in which consumer behavior occurs, such as a store, library, museum or passport office, can be described as relatively closed or relatively open. In sum, closed settings are those in which marketers or other providers exercise dominant control over what is available to the consumer, and are thus in a position to influence greatly his or her behavior; open settings are those in which the consumer has numerous choices, maximal discretion over what he or she will do (see Foxall 1993, 1994). Hence, closed behavior settings are those in which the setting is manipulated largely by persons or organizations other than the consumer and which encourage conformity to the behavior program ongoing in the setting. The physical and social surroundings which the consumer enters are particularly amenable to the achievement of such control. Banks, for instance, are arranged to maximize the orderly queuing by consumers and to minimize extraneous behavior patterns not connected with the efficient execution of current transactions. But a degree of closure may also be achieved symbolically as when a consumer feels under (benign) pressure to purchase a birthday present to repay a close relative's generosity; the source of closure here is the social rules which prescribe moral or material rewards for such reciprocity and, possibly, punishments for ignoring others' generosity. An open behavior setting is one from which such physical, social and verbal pressures are largely absent or in which their influence is less obviously traced and the customer is relatively free to determine his or her own rules for choosing among the products and brands on offer. An example is pre-purchase behavior for luxurious and innovative products: while social and other contextual influences are present, the consumer has discretion over which stores to visit, which products to examine and, if a purchase occurs, which particular product versions and brands to select. Behavior settings also perform the function of informing consumers of the consequences of their current behaviors - the reinforcing and/or punishing outcomes of purchase (or nonpurchase) and consumption.

Reinforcement takes two forms. In addition to the hedonic reinforcement customarily assumed by behavior analysis, which features the emotional and pleasurable consequences of behavior, the model incorporates informational reinforcement which is feedback on the level of performance of the consumer. Most significantly, it indicates the level of achievement or social status conferred by purchasing and consuming, and especially being seen to consume certain luxury and innovative products and services. Hedonic and informational reinforcements usually occur together but are conceptualized as independent influences on consumer choice; each can be high or low relative to the other or to itself at other times. When the consequences of behavior are remote from its performance, too small, too delayed or too improbable to affect it immediately and directly, verbal rules may act as discriminative stimuli, outlining the likely outcomes of behaving or failing to behave in specified ways, and providing motivation to act appropriately (Malott 1989). Antecedent (setting) and consequential stimuli (reinforcement and punishment), the rules that link them, and the consumer who brings a unique personal learning history to the setting combine to form the consumer situation.

The BPM proposes that consumer behavior can be described as Accomplishment, Pleasure, Accumulation or Maintenance depending upon the pattern of its reinforcing consequences (Figure 2). Accomplishment is personal achievement that results in relatively high levels of both hedonic and informational reinforcement. In an open setting, it might take the form of conspicuous consumption; in a more closed setting, the completion of a commercial training course or of a personal development seminar such as est or Insight, or even of gambling in a casino. Pleasure is behavior usually reinforced by entertainment, which is maintained by a high level of hedonic reinforcement and level of informational reinforcement which, by comparison, is relatively low. In an open setting, it might be exemplified by watching popular TV programs; in a closed setting, watching an inescapable in-flight movie. Also in this category are behaviors reinforced negatively by the removal of an aversive stimulus, such as taking aspirin for a headache.

Accumulation is the collection of reinforcers which have some hedonic content but which are principally informational, where sustained collecting is itself further rewarded. In an open setting, collecting trading stamps, making the sequential purchases necessary to qualify for a special deal or final prize, and saving up to buy a major item all qualify; accumulation occurs also in closed settings such as token economies in therapeutic contexts. Finally, maintenance is routine behavior necessary to sustain one's physiological being (e.g. eating, sleeping) and to function as a member of a social group, to be a citizen in society (e.g. paying taxes for enforced consumption). Maintenance behaviors are controlled by levels of both hedonic and informational reinforcement which, though far from unimportant, are lower than those associated with the preceding classes of behavior; often they are controlled negatively by the removal of a threat. In an open setting, maintenance would include regular purchasing of food products; in a closed setting it is noticeable in the payment of taxes for the compulsory 'purchase' of streetlighting, community health schemes and the armed forces, and obtaining a passport in order to travel abroad. Each of these broad classes of behavior can be interpreted as being maintained on its own schedule of reinforcement: variable ratio in the case of accomplishment, variable interval for pleasure, fixed ratio for accumulation and fixed interval for maintenance. Figure 3 gives examples of consumer behaviors classified by operant class and closed-open behavior setting.

Each class of behavior is maintained by positive reinforcers, whether hedonic or informational, but each has a downside which often does not impinge directly upon the consumer. Social and public costs accrue to the community as a whole and not specifically to the individual consumer who is responsible for their being incurred. Consumer behaviors frequently take the form of behaviors that are damaging to the environment: e.g., seeking ever-greater accomplishments that provide hedonic rewards and show forth one's status may result in the consumption of scarce and irreplaceable resources. Beyond a point, pleasure-seeking may also have a deleterious effect on the environment through indulgent energy consumption. Accumulation brings with it the concomitant need to dispose of packaging and, in an affluent societies marked by ever-shortening product life-cycles, the products themselves. And, finally,the consumption of the basic commodities of life themselves, such as water, now threatens further consumption by depleting stocks.

FIGURE 2

OPERANT CLASSES OF CONSUMER BEHAVIOR

FIGURE 3

CONSUMER BEHAVIORS BELONGING TO CLOSED AND OPEN BEHAVIOR STTINGS

ENVIRONMENT-IMPACTING CONSUMER BEHAVIOR

Applied behavior analysis has been concerned for the most part to assess the extent to which contextual factors control the demand for products and services which have deleterious effects on the physical environment. Often, the unrestricted acquisition of short term reinforcements by a limited number of individuals leads to long term aversive consequences for all users. The hedonically reinforcing consequences of behavior are encountered quickly and directly after the action is performed, whereas the environmentally deleterious results of the behavior are encountered, if at all, indirectly and only after a period of time has elapsed. The immediate reinforcement of behavior with ultimately deleterious effects is so great, especially the hedonic, and the aversive outcomes so remote, that the longer-term consequences can sometimes only be reduced or prevented through active self-management. The relatively open settings in which these behaviors typically take place and their maintenance by strongly hedonic reinforcers mean that some closure of the setting has been advocated in lieu of self-management in order to compel a degree of prosocial behavior.

Applied behavior analytic programs incorporate a variety of behavior-related antecedent and consequent stimuli. Antecedent stimuli have consisted of prompts, i.e. warnings, reasoned argument and facts, threats, pleas, etc. relating to the deleterious effects of actions that exploit or pollute the environment. Two varieties of consequential stimuli have been employed: feedback, i.e. information on the actual effects of individuals' actions, and incentives, i.e. financial bonuses, praise and encouragement. Antecedent prompts are intended to act as discriminative stimuli, signalling the aversive consequences of specific behaviors that impact the environment for ill. Feedback is essentially informational reinforcement, indicating the level of performance achieved by the consumer in, say, reducing his or her electricity consumption or private car mileage. Incentives are essentially hedonic reinforcers, rewarding consumers with additional consumption goods or the capacity to acquire them for their prosocial behaviors.

General Propositions

If the BPM interpretation of consumer behavior is accurate, then it should be possible to present a plausible interpretation of the consumer behaviors that have been the concerns of applied behavior analysis in terms derived from the model. (i) it should be possible to identify the contextual factors that control them in terms of hedonic and informational reinforcement and the setting variables that signal their availability. (ii) if specific classes of consumer behavior are maintained by defined patterns of high versus low hedonic and informational reinforcement and on schedules that can be consistently inferred, we would expect intervention to succeed when it either maintains current levels of reinforcement or increases the level of one source of reinforcement without reducing that of the other. (iii) for consumer behavior, we should expect hedonic reinforcement to play a broader role than informational and can predict that it would prove more effective in changing consumer behavior than either prompting or informational reinforcement alone. (iv) successful intervention should also feature changes in the nature of the behavior setting, opening or closing it further in order to make prosocial behaviors more probable. (v) an integrated program of antecedent and consequential stimuli should work best when rules that link behavior and its consequences with a degree of specificity (rather than through vague prompting) are provided and supported. These general propositions can be tested by reference to the literature on environmental intervention based on applied behavior analysis. Full references to the following literature survey can be found in Cone and Hayes (1980) and Geller et al. (1982). Space precludes full referencing.

Private Transportation as 'Accomplishment'

Of the environmentally-impacting consumer behaviors with which applied behavior analysts have been concerned, the use of private automobiles, often carrying a single individual to or from work, falls into this category. Such behavior is apparently maintained by high levels of both hedonic reinforcement - the fun of driving, control of one's journey - and informational reinforcement - speed, low and flexible journey times. In addition to these immediate sources of reinforcement, personal driving is powerfully maintained by intermittent reinforcements apparently available on a variable ratio schedule: social approval, personal safety, simplification of journey planning routines, all of which are contingent on the performance of a number of responses that varies among situations.

If this classification is correct, the research propositions developed above would lead to the following expectations with respect to successful strategies of behavior change. (i) since the current behaviors are maintained on high hedonic, high informational reinforcements, demarketing should attempt to replace the current behavior with alternatives similarly maintained, though hedonic reinforcement is likely to be the more effective. (ii) the use of aversive stimuli, punishing motorists through taxes, tolls and other uses of what essentially amounts to the price element of the marketing mix is likely to be counter productive since the high levels of both hedonic and informational reinforcement available from driving will compensate for attempts at punishment. (iii) the already open setting should be opened even further by the provision of effective competition to private driving: e.g. making buses more popular, comfortable and socially acceptable. (iv) general prompts alone are unlikely to work but discriminative stimuli, effectively linked to specific behaviors and their outcomes may be effective: these should stress the rewards for bus ridership in terms of the personal gratification this provides rather than vague predictions of a remote better environment.

Evidence for the classification of private motoring as accomplishment and for the efficacy of the above strategy in social demarketing is available from the findings of attempts to modify consumers' private transportation behavior which has been intended to reduce fuel consumption, urban congestion, and pollution by discouraging unilateral use of private cars and promoting public transportation. The most successful interventions have offered hedonic reinforcement in the form of financial incentives: provision of small monetary rewards for riding the bus has, for example, increased the number of users of public travel services by 50-180% (Geller et al. 1982). The need for principally hedonic reinforcement, albeit coupled with informational reinforcement in the form of continuous and effective feedback, is indicated by the relatively unattractive pre-intervention pattern of consequences for bus ridership. Riding the bus and other strategies which avoid private transportation (such as walking, car pooling and cycling) are at best minimally reinforced hedonically by social contact and, eventually, feelings of fitness, and informationally by cost savings. But they are punished by aversive consequences: slowness, discomfort, danger, exposure, crowding, noise, inflexibility, unpredictability, and lack of control.

Discouragement of car travel has reduced mileage travelled by between 10 and 50% (Cone and Hayes 1980). The provision of informational reinforcement plays a strong role in reducing driving but only in combination with hedonic reinforcement; however, the two forms of reinforcement cannot in this case be effectively separated since each relies on the provision of the other. While feedback alone (on the number of miles travelled, operating costs, depreciation, social costs, etc.) had no effect on mileage travelled, performance feedback influences behavior by allowing the driver to monitor his or her behavior in order to achieve the incentives. Although hedonic reinforcement once again emerges as the most effective single means of modifying behavior, its use in tandem with informational reinforcement has a mutually strengthening effect and provides a cost effective form of intervention.

In the marketing of alternatives to private car use, notably transportation by bus which for many drivers is likely to prove highly disruptive of their journey routines, hedonic reinforcement has been used almost exclusively and has taken the form of cash payments for riding the bus and, more cost-effectively, of tokens redeemable at stores and for additional bus trips. The use in this context of a variable person schedule of reinforcement (VP) in which every nth passenger is rewarded rather than every passenger not only reduces the costs of the transit program but indicates the relevance of a kind of variable ratio schedule to the maintenance of behaviors in this category. The evidence is that this opening of the setting further by providing genuine competition to private motoring can be effective and that prompts alone are most ineffective.

Domestic Energy Consumption as 'Pleasure'

Among environmentally-impacting consumer behaviors, pleasure is exemplified by the over-consumption of domestic energy derived from fossil fuels, notably electricity for heating and lighting. The hedonic reinforcements are high and closely related temporally to the responses that produce them - convenience, comfort. While informational reinforcement is less obvious, social approval may follow generous use of these resources in the company of others (meanness will certainly lead to social disapproval and loss of status). Punishments are real and may be severe (e.g. having to pay one's electricity and gas bills) but are remote in time and place from the usage situation and may be mitigated by staged payments direct from a bank account. The long term consequences are remote: e.g. depletion of resources, social disapproval. Consumption behaviors are apparently controlled by a variable interval schedule: comfort and satisfaction depend upon employing the source of heat or light for a time that varies from occasion to occasion with the individual's task requirements and state variables (e.g. cold, hunger).

If this analysis and classification are correct, the following should be expected of a successful strategy of behavior change. (i) since the current behaviors are maintained by high hedonic and low informational reinforcement, any attempt to punish that behavior by introducing aversive stimuli (very high prices) or reducing hedonic consequences, without a corresponding increase in hedonic benefits, is likely to fail. (ii) social demarketing should concentrate on making the behavior (including avoidance) more involving, encouraging the avoidance of high bills and a feeling of self-gratification at saving energy and reducing pollution. (iii) if a sufficient level of hedonic benefit can be guaranteed, the setting could be closed by increasing the costs of energy. Support for the classification of domestic energy use as pleasure and for the efficacy of this strategy comes from the attempted modification of consumers' domestic energy consumption which has used antecedent prompting, feedback, and incentives, separately and in combination. Alone, information relating to the environmental effects of pollution caused by high consumption of electricity at peak periods had little if any effect on peak usage. Greater effect was achieved by consumer self-monitoring of current energy usage: peak consumption reduced by up to 30% of mean baseline levels. Overall energy usage (i.e. peak and non-peak consumption) has also proved sensitive to informational feedback, even at times of steep increases in the price of energy. Combined feedback and monetary incentives have reduced peaking by about 65% of baseline, confirming the efficacy of combined consequences.

In line with the basic principle of operant conditioning that reinforcement must immediately follow the performance of a response in order for learning to occur, it has been demonstrated that daily feedback on overall energy usage, especially when combined with group feedback and mild social commendation for 'prosocial' behavior can be effective. More practically, weekly or monthly feedback corresponding with normal billing periods is particularly efficacious. The combination of prompts and feedback with incentives (e.g. payments of up to $5 per week for reductions of gas/electricity consumption by 20% or more of baseline mean) is even more effective. Comparisons of the individual effects of the separate elements of persuasion (prompts, feedback and incentives) indicate, however, that only incentives have an appreciable effect on behavior.

Waste Disposal as a Problem of 'Accumulation'

Waste generation is a consequence of accumulation but it is actually a problem manifested in the opposite of accumulation: disposal. Indiscriminate waste disposal has relatively few hedonic benefits other than convenience but its informational outcomes are extensive if subtle: it confers status through the assumption that someone else will clear up, and it may also imply conspicuous consumption. Such behaviors are maintained seemingly on fixed ratio schedules. Their long term consequences are also remote: gradual spoliation of the physical environment, accruing social disapproval.

The general research propositions developed above indicate that if this analysis and classification are accurate, the following will apply. (i) given the assumed pattern of current behaviors being maintained by high informational and low hedonic reinforcement, behavior change is likely to be accomplished by increasing such hedonic consequences as aesthetic pleasure while not reducing informational feedback. (ii) the personal element in reinforce should be especially effective. (iii) the encouragement of prosocial behavior can be achieved by paying people to return/recycle waste as long as the punishing consequence of doing so are moderate. (iv) closing the behavior setting by providing bins should be effective. (v) promotional appeals based on prompting in a general way would be ineffective but, modelling, showing the prosocial consequences of conformity would also be effective.

The findings of applied behavior analysis in this area do indeed confirm the analysis and classification. Attempts at reducing littering have relied heavily on the use of prompts. The results have been generally disappointing unless the prompts were accompanied by positive reinforcement, usually hedonic. Exhortations, lectures, and relevant general education have proved largely ineffective in this sphere. Even the attempt to reduce littering among children in a theatre by manipulating the physical environment (providing bags for waste) had little effect. Combined with messages pointing out the disadvantages of litter, the provision of bags had a moderate effect. However, when a reward of one dime was given for each bag of rubbish, the decrease in littering was massive. Another form of hedonic reward in the form of a ticket for a movie had a similarly substantial negative effect on littering. Similar results have been found in experimental studies of the reduction of littering in streets, and around and within buildings. Success is also apparent in the closure of the behavior setting, e.g. providing more litter bins and devising trash cans that are fun to use, and by ensuring the initial cleanliness and attractiveness of the environment; all of these strategies have had some effect by bringing behavior under stimulus control, but only the presentation of positive reinforcers in the form of payments has any dramatic effect on behavior. Prompts, used alone, have little if any effect, perhaps because of their reliance on punishment for unapproved behavior: the individual who litters nevertheless and avoids immediate punishment is actually likely to be reinforced for his or her littering. The relative effectiveness of prompts and incentives indicated by litter studies has been confirmed by experiments aimed at increasing consumers' willingness to conserve irreplaceable materials through recycling. Attempts at increasing consumers' purchases of returnable bottles are a typical example. The use of prompts informing customers of the savings to which such behavior would lead and that they would be contributing to the fight against pollution have had mixed effects . Giving consumers small financial rewards for the reuse of such items as egg cartons, milk containers and grocery bags, accompanied by in-store prompts and a pleasant and enthusiastic reaction by salespersons, has led to increases in custom.

Some attempts at increasing consumers' recycling behavior have had significant punishing consequences. The Bottle Laws enacted first in Oregon and subsequently adopted by several other states impose considerable transaction, inventory and time costs on retailers who pass them on to their customers (Guerts 1986). Both are penalized for their participation in the waste reduction campaign and, even though distributors are legally bound to comply, their consumers are in general unlikely to incur the costs involved in prepaying deposits and returning glass bottles unless they are adequately compensated for the punishing consequences of these prosocial endeavours. Experimental attempts to encourage the recovery of waste materials such as paper which can be recycled also indicate that prompts have minimal effects on behavior while the provision of hedonic and, to a smaller extent, informational reinforcers has a substantial reinforcing effect. Hence students offered prizes in contests and raffles are more likely to reduce wastage than those who are only exposed to educative prompting. The provision of convenient containers for the collection of recyclable waste is also significantly more effective than prompting on the promotion of appropriate prosocial behaviors perhaps because it achieves a degree of closure of the setting and the combination of prompts and suitable receptacles for the collection of waste has produced a combined effect on behavior greater than that expected from their individual contributions.

Domestic Water Consumption as 'Maintenance'

Maintenance is exemplified as an environmentally-impacting consumer behavior by the domestic over consumption of water. Both hedonic and informational reinforcements are low compared with those that control other the other classes of consumer behavior, though neither is absent: the luxury and status of having water continuously available on tap are easily taken for granted but being able to drink, clean, bathe and water the garden are indicative of comparative wealth and power; they are hedonic and informational benefits directly related to the consumer's state of deprivation. The consumption behaviors in question are apparently maintained on fixed interval ratios, most of the uses of water taking place at some time or other on most days or most weeks.

If accurate, this analysis and classification would suggest the following. (i) punishment, especially involving price would be especially efficacious in reducing consumption. (ii) metering, to provide general association between behavior and its contingent consequences and to provide accurate and quick feedback on the outcomes of consumption would be especially effective. (iii) closing the setting by reducing the time and place during which water can be consumed would be effective. There is less experimental evidence for the behavioral economics of water consumption and conservation than for the other commodities and products considered but the limited evidence suggests that this analysis and classification are correct. A study of the conservation of metered water in Perth, Australia (Geller et al. 1982) indicates that water consumption decreased by over 30% in both an experimental group provided with daily feedback on water use and a rebate proportionate to demand reduction, and a control group provided only with feedback, though change in climatic conditions may also have affected the results. The low elasticity of demand for water makes financial rebates less appropriate than for other classes of consumer behavior.

SOCIAL DEMARKETING

This survey indicates that the general propositions suggested above are accurate. The BPM provides a coherent and plausible model of the role of antecedent and consequential stimuli in the shaping and maintenance of environmentally-impacting consumer behavior. Hedonic reinforcement is, as expected, the single most important influence on such behavior but the three basic components of the model, used in optimal combination that varies depending on the class of behavior in question, exerts the greatest control. The most effective general strategy for behavior change indeed appears to be the maintenance of current levels of hedonic and/or informational reinforcement plus the enhancement of relatively low levels of reinforcement, plus the manipulation of behavior settings to signal the consequences of modified consumption. Finally, different marketing mix strategies can be extracted from the results of the applied behavior analyses for the four classes of environmentally-impacting consumer behavior identified and described by the BPM. The fulfillment of the specific propositions put forward for each of these consumer behavior classes suggests the following generalizations, each well-supported by empirical evidence and capable of serving as hypotheses for refinement and further testing.

Accomplishment

Modification of accomplishment behaviors, exemplified here by private motoring, requires the development of a radically more attractive product with strongly reinforcing hedonic and informational attributes: this may even necessitate the creation of a different product. Price may be important too, but only when the new or thoroughly revamped product has been successfully launched and established: the price of the original might then be raised to punish its use. Until this point is reached, however, such a price rise would have little overall effect on demand for the original product given the abundant hedonic and informational reinforcers it provides. Indeed, to the extent that private transportation is a prestige good, maintained by informational reinforcement that derives from conspicuous consumption, an increase in the costs associated with it might be counter-productive, encouraging rather than discouraging consumption. During the introductory phases of the new product, its price might be subsidized to ensure that consumers switched to its use: whether the price reduction has to be maintained indefinitely depends upon the effectiveness of the primary hedonic and informational reinforcers provided by the novel product. As far as promotion is concerned, prompts are unlikely to have a strong effect on demand, though coupled with effective consequential stimuli they provide a necessary informative and persuasive role. However, advertisements containing modelling of the prosocial behaviors advocated would probably both increase awareness of the campaign and encourage imitative responses. Finally, as far as place is concerned, the behavior setting should be opened further by increasing competition and making the new product widely and flexibly available.

Pleasure

Consumer behavior modification in the case of pleasure is more subtly changed through the provision of increased, relatively rapid and regular information on consumption. This information can be seen as part of the product provided by the utilities companies. The maintenance of hedonic reinforcement is important and, since the overall goal of the campaign is a reduction in energy use, this must be accomplished by the encouragement of personal and domestic arrangements which promote thermal savings (e.g. better insulation, the wearing of more heat-efficient clothing and the elimination of useless energy consumption such as the illumination of unoccupied rooms). These factors, which might be considered part of the place element of the marketing mix since they determine the location of consumption, contribute to the closure of the behavior setting. Price might also be used to deter over use of resources but, given the highly hedonic consequences of energy consumption, it is unlikely to have a strong independent effect on usage.

Accumulation

The single most cost-effective means of reducing littering is probably the closure of the behavior setting. Since litter is itself a discriminative stimulus for further littering, the provision of bins, bags and other containers that encourage disposal is likely to have a cumulative effect on behavior. Prompting alone also has some effect on litter disposal if it is directly related to the means of acting prosocially, e.g. by pointing out what to put, where to put it and when. The behavior setting for recycling can be closed by the provision of containers for bottles, plastics, papers, and so on in convenient positions for consumers to use. Hedonic reinforcement remains a strong influence on behavior though it will often be an expensive alternative: competitions and variable person schedules appear to be the most effective means of changing behavior, especially if coupled with promotional campaigns emphasizing modelled prosocial behavior. The costs involved in some prosocial behaviors presently punish the consumer - e.g. in the case of returning bottles and other packaging; either these costs must be reduced through the collection of waste materials or the financial recompense for their return must be expanded until behavior is economically controlled.

Maintenance

Finally, in the case of maintenance, exemplified by water consumption and conservation, it is important to control the behavior setting by installing water-conserving methods (e.g. smaller cisterns), by encouraging the use of rainwater for garden watering, and the opportunity to use water less expensive than fully-purified drinking water for some domestic purposes such as flushing toilets. The alternative place strategies (rationing, standpipes, etc.) are politically unacceptable and usually unnecessary except during emergencies, though metering is probably an essential prerequisite of most systems of behavior modification based on consequential stimuli whether informational or hedonic. Price might be used to overcome overuse, though again this would be politically acceptable only within close bounds.

REFERENCES

Cone, J. D. and Hayes, S. C., (1980) Environmental Problems/Behavioral Solutions. Monterey, CA: Brooks/Cole.

Costanzo, M., Archer, D., Aronson, E. and Pettigrew, T. (1986) "Energy conservation behavior: the difficult path from information to action," American Psychologist 41: 521-528.

Foxall, G. R. (1993) "Situated Consumer Behavior," Research in Consumer Behavior, 6, 113-152.

Foxall, G. R. (1994) "Behavior Analysis and Consumer Psychology," Journal of Economic Psychology, 15, 1-87.

Geller, E. S. (1989) "Applied behavior analysis and social marketing: an integration for environmental preservation," Journal of Social Issues 45: 17-36.

Geller, E. S., R. A. Winett and P. B. Everett (1982) Preserving the Environment, Elmsford, N.Y: Pergamon.

Guerts, M. D. (1986) "The 'bottle bill' effect on grocery stores' costs," International Journal of Retailing, 1, 12-17.

Malott, R. W. (1989) "The achievement of evasive goals: control by rules describing contingencies that are not direct acting." In: Hayes, S. C. (ed.) 1989, Rule-governed Behavior, New York: Plenum.

Schwartz, I. S. (1991) "The study of consumer behavior and social validity: an essential partnership for applied behavior analysis," Journal of Applied Behavior Analysis 24: 241-4.

Skinner, B. F. (1953) Science and Human Behavior, NY: Macmillan.

Winkler, R. C. and Winett, R. A. (1982) "Behavioral interventions in resource conservation: a systems approach based on behavioral economics," American Psychologist 37: 421-435.

----------------------------------------