An Investigation of Determinants of Recycling Consumer Behavior

ABSTRACT - Previous research on recycling consumer behavior (RCB) has been largely descriptive. The authors posit a conceptual model wherein RCB is determined by its importance as judged by benefits-to-costs deliberations. These, in turn, are hypothesized to be impacted by social influence, personal values, felt norms, and external cues. A preliminary empirical study reveals that RCB is affected by all these factors although in various ways. The authors note possible improvements in subsequent research and claim that their results suggest environmentally protective consumer behaviors define an arena of fruitful research.

Citation:

Anita L. Jackson, Janeen E. Olsen, Kent L. Granzin, and Alvin C. Burns (1993) ,"An Investigation of Determinants of Recycling Consumer Behavior", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 481-487.

Advances in Consumer Research Volume 20, 1993      Pages 481-487

AN INVESTIGATION OF DETERMINANTS OF RECYCLING CONSUMER BEHAVIOR

Anita L. Jackson, Louisiana State University

Janeen E. Olsen, Louisiana State University

Kent L. Granzin, University of Utah

Alvin C. Burns, Louisiana State University

ABSTRACT -

Previous research on recycling consumer behavior (RCB) has been largely descriptive. The authors posit a conceptual model wherein RCB is determined by its importance as judged by benefits-to-costs deliberations. These, in turn, are hypothesized to be impacted by social influence, personal values, felt norms, and external cues. A preliminary empirical study reveals that RCB is affected by all these factors although in various ways. The authors note possible improvements in subsequent research and claim that their results suggest environmentally protective consumer behaviors define an arena of fruitful research.

INTRODUCTION

Over the past several decades, consumers have become aware of the environmental problems facing the planet. This growing concern has led many people to alter their lifestyles in a manner they see as more ecologically friendly. Nevertheless, our environmental problems are far from being solved. In fact, experts warn that environmental conditions have become worse over the years. Solid waste has increased as potential landfills become scare. Toxic waste has led to soil and groundwater contamination (Rice 1988). Carbon emissions are being blamed for depletion of the ozone layer that protects the Earth (Williams 1989). It is clear that even with growing numbers of people supporting the environmental movement, still more consumers must be encouraged to consider the environmental consequences of their consumption decisions.

Recycling programs are a voluntary environmental protection activity in which consumers are often encouraged to participate. For many years recycling efforts were hampered due to lack of reverse distribution channels, markets for collected items and inconvenient methods of collection (Barnes 1982). Many of these difficulties have been addressed recently as recycling centers and collection programs have become more popular. More and more, consumers are finding recycling a feasible alternative for the disposition of containers, packaging, spent products, and trash in general.

As noted above and elaborated on below, while some descriptive research has been applied to Recycling Consumer Behavior (RCB), little has been done in the way of explanatory investigation. Given that: (1) recycling is recognized as a means of alleviating environmental damage, (2) institutional structures such as recycling centers exist for the disposition of unwanted glass, aluminum, paper and other recyclable materials, and (3) RCB is voluntary product/packaging disposition behavior, it seems worthwhile to attempt research to better understand its dynamics. Accordingly, this paper presents an initial conceptual model of possible influences and determinants of RCB. It reports an exploratory empirical test of this model, and its discusses implications of the findings.

LITERATURE REVIEW

Academic interest in environmental issues is not new. It can be traced to the time the environmental movement itself began to gain momentum (Kotler and Zaltman 1971; Kassarjian 1971; Anderson and Cunningham 1972; Fisk 1973; Kinnear and Taylor 1973). The broad range of environmental issues that has been investigated reflects the variety of environmental problems facing consumers, including the needs for pollution control, energy conservation, and recycling (Zikmund and Stanton, 1971; Barnes 1982).

Over the years, many of the empirical studies that have been conducted have followed the format of segmentation research. The primary objective of this approach has been to identify environmentally concerned individuals and/or participants of pro-environmental behaviors (Anderson and Cunningham 1972). In keeping with the segmentation orientation, studies have found demographic variables to be useful predictors of environmental concern. For example, with regard to recycling, higher education levels are often associated with recycling activities, and age has also been useful in identifying people who participate in recycling activities (Mohai and Twight 1987; Vining and Ebreo 1990). Income was also found to be a significant predictor for recycling (Jacobs, Bailey and Crews 1984; Vivivg and Ebreo 1990). However, while previous research has demonstrated that demographics can be useful variables for segmentation studies, the specific findings themselves may quickly become obsolete. Furthermore, recent studies suggest that demographics are only modest or ineffective predictors of environmental concern (Manzo and Weinstein 1987; Samdahl and Robertson 1989; Granzin and Olsen 1991). Also, environmental concern can spread to new segments of the population quite rapidly, effectively diluting demographic correlates.

Thus, one could argue that a fruitful strategy may be to investigate internal predispositions in consumers. Accordingly, depictions of environmentally concerned citizens have also relied on psychological and personality variables (Brooker 1976; De Young 1986; Balderjahn 1988) and personal values (Dunlap, Grieneeks and Rokeach 1983; Rankin 1983; Neuman 1986). As in the case of demographics, the studies investigating psychological and personality variables have emphasized how the environmentally concerned consumers differ from those that are not, and moderate success has been the result (Granzin and Olsen 1991).

In summary, prior research has experienced modest success using demographics and/or internal predispositions hypothesized to relate to an individual's desire to participate in recycling programs. The next logical step is to develop and test more comprehensive models that may better explain environmental protection behavior such as recycling.

CONCEPTUALIZATION OF DETERMINANTS OF RCB

To identify salient components of RCB, we searched for a similar consumer behavior phenomenon which has been addressed both conceptually and empirically. We concluded that RCB can be compared to preventive health care behavior in that all three of the conditions noted earlier are present. That is, preventive behavior is a means of avoiding harm which may occur in the future; there are endorsed regimens and institutionalized programs for prevention, and the behavior is voluntary. Thus, we scrutinized the preventive health care literature for a theoretical structure as a beginning point in our modeling of RCB. The Health Belief Model (HBM) has received much conceptual and empirical attention (see for example, Becker 1985; Becker, Drachman, and Kirscht 1974a, 1974b; Becker et al 1977). Consequently, we adapted its components in developing an initial conceptual model of RCB. The HBM posits that behavior results from a benefits-to-costs deliberation on the part of the individual. The likelihood of preventive health care behavior is a function of this benefits-costs comparison. If benefits outweigh costs, then health care actions are likely, while if costs (e.g. inconvenience, side effects, etc.) are perceived to exceed benefits, health care actions are unlikely. At the same time, the behavior is influenced by three other factors. One is the perceived threat of whatever disease or illness is being contemplated, while another is cues in the forms of impersonal (i.e. media) warnings, and the third is social pressures. High perceived susceptibility and severity combine to determine perceived threat, and high threat can directly impact the likelihood of taking recommended action. Cues and social interaction also serve to influence the amount of perceived threat. In our judgement, perceived severity, susceptibility and threat are not reasonable as constructs underlying RCB; however, the other factors in the HBM are relevant, namely cues and social pressures. Admittedly, our justification for the use of the HBM as a conceptual framework here is not as compelling as one might wish, but it is a cohesive model with a research tradition which might afford insights into RCB.

FIGURE 1

A PRIORI MODEL

The HBM has been compared to Fishbein's (1967) behavioral intentions model by Oliver and Berger (1979) as well as Rosenblatt, Cusson and McGown (1984). In both instances, the authors concluded that a combining of key constructs in the two models is advantageous; consequently, normative compliance (behavioral norm) was drawn from this model. Felt norms appear to be a logical construct for possible explanation of voluntary RCB. Since no formal rewards accompany RCB, nor does any penalty ensue for individuals who fail to practice RCB, felt norms may well be an important determinant. In addition, personal values were incorporated as they have been found to be associated with environmental protection actions on the parts of consumers (see Granzin and Olsen 1991).

The a priori model adopted for this study appears in Figure 1. As can be seen, it posits RCB as a function of a blend of decision-making, internal predispositions, and external influence factors. RCB is modeled as the direct result of the considered importance of RCB. We have opted to use importance rather than intention for two reasons. First, intention may be preempted by situational factors such as lack of knowledge, storage constraints, transportation difficulties or other obstacles. Also, Rosenblatt, Cusson and McGown (1984) found importance to be a construct through which other factors influenced donation and personal health care behavior.

Consequently, we have opted to rely on importance as an mediating factor translating conscious decision making into a precursor for action. If consumers do not view recycling as important to themselves or society as a whole, we cannot expect them to exhibit sustained RCB.

Importance, in turn, is largely affected by three separate influences. First, it is postulated to be the result of a conscious weighing of the benefits and costs associated with the practices. Benefits are for instance, protection of the environment, energy conservation or less litter, while costs include storage of recyclables, necessary washing and bundling, and hauling. In other words, RCB is assumed to depend on a deliberation process on the part of the individual consumer where these benefits and costs are subjectively compared.

Further, we postulate these deliberations to be influenced by both internal factors and external influences. With regard to internal factors, both values and norms are assumed to be reflected in the decision making process. If a consumer values conservation, beauty, and preservation of the environment, for example, we would anticipate these values to enter into the deliberation process. Similarly, if a consumer nurtures a belief that recycling is expected of him/her, we anticipate this felt norm will also influence the deliberation outcome. As can also be seen, one would expect a high degree of comparability between the values harbored by a consumer and his/her subscription to certain behavioral norms. Similarly, social pressures mold normative compliance and affect one's value system. External forces exist both in the form of social pressures such as the urging or prompting of peers, family or significant others as well as informational cues garnered from non-personal sources. In our view, cues to RCB are a likely external influence as recycling has received and continues to receive attention in the forms of newspaper articles, televisions specials, and news broadcasts.

TABLE 1

ESTIMATED PARAMETERS FOR THE MEASUREMENT MODEL

To summarize our thinking, we have cast a person's voluntary recycling behavior as largely determined by how important that person considers recycling to his and/or society's well being. This importance level is a result of a conscious comparison of the benefits and costs (both personal and societal) associated with recycling. The deliberation outcome is, in turn, a reflection of the person's values and internalized norms of appropriate behavior. At the same time, external parties are issuing information and persuasion to encourage recycling. These cues are primarily mass media-based, and they may impact the benefits-costs deliberation ("It is easy to recycle"), the attribution of importance ("Recycling saves trees and energy"), or recycling behavior itself ("Put newspapers in a paper grocery bag for easy storage"). Similarly, social pressures can encourage any of these three pivotal factors as well. Social approval of RCB (a benefit), realization of widespread recycling (importance), or neighborhood drives (participation), for instance, all have credence as possible determinants of RCB. Finally, social influences are assumed to have a role in creating and reinforcing behavioral norms and the individual's values with respect to RCB.

TABLE 2

TEST STATISTICS FOR THE MODELS TESTED

METHOD

We opted to test the model with an exploratory empirical study. A questionnaire was designed to measure the constructs represented in the model. An attempt was made to measure the several constructs identified in the model with multiple indicators. Because no known scales have been developed to measure these constructs as they pertain to RCB, it was necessary to generate a number of items for each and to scrutinize them for reliability. Cronbach's alpha was used, and items were deleted based on low item-to-total correlations. In all cases a 7-point agree-disagree scale was used as the response scale. The final operationalizations of the constructs were as follows with number of items and alpha indicated: RCB (3; .83), importance (3; .83), benefits-costs (6; .72), values (3; .70), norms (3; .61), social influence (3; .82), and cues (2; .65). See Table 1.

The data was gathered using a quota sampling of 348 adults from a large metropolitan area in the western United States. The sample was designed to represent the general population with respect to age, sex, and socioeconomic characteristics of the most recent census for the area. Respondents completed the self-administered questionnaire in the presence of interviewers knowledgeable about the purpose of the study. The interviewers served as motivators, helped interpret the instructions when necessary, and monitored for compliance with instructions.

FINDINGS

The objective of our study was to develop and test a causal model of RCB. Table 1 gives the standardized parameter estimates for the measurement model and Table 2 shows the goodness of fit indices for the refined model. The model was formulated as a structural equation model and estimated with Lisrel VII (Joreskog and Sorbom 1989). The final model shown in Figure 2 is a refinement of Figure 1 after nonsignificant paths were deleted in order to obtain a more parsimonious model.

Evaluation of the maximum likelihood solution includes a chi square goodness-of-fit test, two goodness-of-fit indexes, and the root mean squared residual. Overall fit statistics based upon analysis of the correlation matrix show that the trimmed model fits the data with a GFI of .882, an AGFI of .852 and a RMR of .060. The other fit indices are high. The size of the standardized parameters of the structural and the measurement models, along with large t values for these parameters indicate which of the proposed relationships are significant. Table 3 gives the total effects of each of the constructs on the endogenous constructs of the model. Total effects sizes suggest that the underlying model is sound.

The results of the measurement model show that the factor loadings are high and consistent in sign with the theoretical constructs of the structural model (see Table 1). The t values of the measurement model parameters ranged from a low of 6.275 to 17.168 indicating that they are highly significant. The total coefficient of determination for the x variables it is .987, and for the y variables, it is .989. In addition, Cronbach alphas indicate that the reliabilities of the measures of the theoretical constructs are acceptable for exploratory research.

The trimmed structural equation model is shown in Figure 2 along with parameter estimates. The model explains a considerable amount of the variance in the relationships and is parsimonious and theoretically reasonable. Similarly, the size of the parameter estimates and their t values indicate that the relationships are strong. However, a few of the original hypothesized relationships are not significant. After dropping these paths the model was run again. The Chi Square, GFI, AGFI, and RMR were not changed significantly by the exclusion of these relationships. We feel that the resulting model is more parsimonious and still theoretically reasonable given the circumstances and constraints of our study.

DISCUSSION

Figure 2 and Table 3 afford interesting verification and possible refinement of the original conceptual model of recycling consumer behavior. In particular, RCB is a function of a myriad of influences as was initially hypothesized. Our results reveal that benefits-costs deliberations and importance are key translation constructs for personal and social forces affecting RCB. Further, an external influence on recycling behavior is media exposure. One can claim that cues and social influence act as stimulators of recycling behavior, and the findings suggest they operate through different modes. Social influences work through the individual's value system, plus they have a bearing on deliberations of the pros and cons of recycling, and they sway the individual's assessment of the importance of recycling. Cues, on the other hand, appear to have direct and unmodified effects on RCB. Finally, as suspected initially, the individual's felt norms and values enter into the cognitive decision process associated with evaluating the importance of RCB. In sum, we consider these preliminary results as encouragement for future work utilizing these constructs modeled according to our logic and/or findings.

FIGURE 2

FINAL TRIMMED MODEL

TABLE 3

TOTAL EFFECTS OF LATENT CONSTRUCTS ON ENDOGENOUS CONSTRUCTS

There are obvious improvements which we can point out at this time. One which immediately comes to mind is the cues construct. The cues used in this study were assumed to be the every day reporting of environmental issues in the media. Their significance in the model indicates the potential for including explicit promotions of recycling. Our findings reveal that nonspecific cues have a direct effect on recycling behavior, but no indirect effects. However, this result may reflect the type of cues we measured in this study. Our cues measure is an indicant of general knowledge rather than a measure of promotions targeted to convince the public to participate in recycling. By the same token, other cues to action could be included in future studies; for example the appearance of a rubbish pile in a residential neighborhood (i.e., illegal dumping) could be a cue that prompts recycling concern or behavior. In short, information or persuasion that signals a need for recycling or a danger to the environment could be a cue to action, and the fact that our general measure of cues proved significant as a determinant implies success for more specific, personally-relevant operationalizations of cues.

Another improvement concerns social factors. Social influence is a pervasive influence according to our findings. Inspection of the measurement model reveals that friends exerted the most influence, followed by others and then by family. Obviously, social dynamics are important in promoting and encouraging recycling behavior. However, precisely how social influences are communicated to the individual is unknown at this time. Possible improvements in future research include addressing how, when, and the circumstances of various social forces as they are perceived by the individual with respect to adopting recycling behavior.

Next, the operation of norms on recycling behavior bears close examination. We noted that a distinguishing aspect of recycling is its voluntary nature. As such, felt norms are a logical theoretical determinant, and this premise was verified in our exploratory study. Nonetheless, the origins of these norms and the forces working to mold them remain unclear. Better measurement of either or both constructs may resolve this relationship in future research. Finally, the decision making process must be examined further. What exactly goes into the decision to recycle or not recycle? Benefits-to-costs deliberations include the underlying items that go into the evaluation of recycling, but hidden at this time is the way a person subjectively weighs advantages and disadvantages of the complete recycling activity. A reasonable course for future research is to apply qualitative techniques such as focus groups, protocol analysis, or open-ended questions to attempt to map the deliberation process and its dimensions. This investigation may well alert researchers to a host of cognitive process constructs which operate in this area just as it should uncover the mechanisms by which consumers judge their product/packaging disposition behavior options.

In conclusion, we certainly do not view our research as definitive. Instead, we consider it only a first step in developing an understanding of why consumers do or do not engage in recycling which, in itself, is but one facet of environmentally protective consumer behavior. Clearly, the potential for much fruitful research exists in this arena.

REFERENCES

Anderson, W. Thomas, Jr., and William H. Cunningham (1972), "The Socially Conscious Consumer," Journal of Marketing, 36 (July), 23-31.

Balderjahn, Ingo (1988), "Personality Variables and Environmental Attitudes as Predictors," Journal of Business Research, 17 (August), 52-

Barnes, James H., Jr. (1982), "Recycling: A Problem in Reverse Logistics," Journal of Macromarketing, 2 (Fall), 31-37.

Becker, Marshall H. (1985), "Patient Adherence to Prescribed Therapies," Medical Care, 23 (May), 539-555.

Becker, Marshall H., Robert H. Drachman, and John P. Kirscht (1974a), "A New Approach to Explaining Sick-role Behavior in Low-Income Populations," American Journal of Public Health, 64 (March), 205-216.

Becker, Marshall H., Robert H. Drachman, and John P. Kirscht (1974b), "A Field Experiment to Evaluate Various Outcomes of Continuity of Physician Care," American Journal of Public Health, 64 (November), 1062-1070.

Becker, Marshal H., Louis A Maiman, John P. Kirscht, Don P. Haefner and Robert H. Drachman, (1977), "The Health Belief Model and Prediction of Dietary Compliance: A Field Experiment, Journal of Health and Social Behavior, Vol. 18, December, 348-366.

Brooker, George (1976), "The Self-Actualizing Socially Conscious Consumer," Journal of Consumer Research, 3 (september) 107-112.

De Young, Raymond (1986), "Some Psychological Aspects of Recycling: The Structure of Conservation Satisfactions," Environment and Behavior, (July), 435-449.

Dunlap, Riley E., J. Keith Grieneeks and Milton Rokeach (1983), "Human Values and Pro-Environmental Behavior," Energy and Material Resources: Attitudes, Values, and Public Policy, ed. W.D. Coon, Boulder, CO: Westview Press, 145-168.

Fisk, George (1973), "Criteria For a Theory of Responsible Consumption," Journal of Marketing, 37 (April), 24-31.

Granzin, Kent L. and Janeen E. Olsen (1991), "Characterizing Participants in Activities Protecting the Environment," Journal of Public Policy & Marketing, Vol 10, No 2 (Fall), 1-27.

Joreskog, Karl G. and Dag Sorbom (1989), Lisrel 7 A Guide to the Program and Applications, SPSS Inc., Chicago, Ill.

Kassarjian, Harold H., (1971), Incorporating Ecology into Marketing Strategy: The Case of Air Pollution," Journal of Marketing, 35 (July), 61-65.

Kinnear, Thomas C. and James R. Taylor (1973), "The Effect of Ecological Concern on Brand Perceptions," Journal of Marketing Research, 10 (May), 191-197.

Kotler, Philip and Gerald Zaltman (1971), "Social Marketing: An Approach to Planned Social Change," Journal of Marketing, 35 (April), 3-12.

Manzo, Lynne C. and Neil D. Weinstein (1987), "Behavioral Commitment to Environmental Protection: A Study of Active and Nonactive Members of the Sierra Club," Environment and Behavior, 19 (November), 673-694.

Mohai, Paul and Ben W. Twight (1987), "Age and Environmentalism: An Elaboration of the Buttel Model Using National Survey Evidence," Social Science Quarterly, 68 (December), 798-815.

Neuman, Keith (1986), "Personal Values and Commitment to Energy Conservation," Environment and Behavior, 18 (January), 53-74.

Oliver, Richard L. and Philip K Berger (1979), "A Path Analysis of Preventive Health Care Decision Models," Journal of Consumer Research, Vol. 6, September, 113-122.

Rankin, William (1983), "The Influence of Human Values on Conservation and Energy Shortage Beliefs," in Energy and Material Resources: Attitudes, Values, and Public Policy, ed. W. David Conn, Boulder CO: Westview Press.

Rice, Faye (1990), "Where Will We Put All That Garbage?" Fortune, (April) 96-100.

Rosenblatt, Jerry A., Alain J. Cusson, and Lee McGown (1984), "A Model to Explain Charitable Donation-Health Care Behavior," Advances in Consumer Research, Volume XIII, Richard J. Lutz (ed.), Association for Consumer Research, 235-239.

Samdahl, Diane M., and Robert Robertson (1989), "Social Determinants of Environmental Concern: Specification and Test of the Model," Environment and Behavior, 21 (January), 57-81.

Vining, Joanne, and Angela Ebreo (1990), "What Makes a Recycler?: A Comparison of Recyclers and Nonrecyclers," Environment and Behavior, (January), 55-73.

Williams, Maurice (1989), "Sustainable Development: A SID Perspective," Development, 2 (No.3), 7-9.

Zikmund, William G. and William J. Stanton (1971), "Recycling Solid Wastes: A Channels-of-Distribution Problem," Journal of Marketing, 35 (July), 34-39.

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