1 Journal ofenvimnmento1 Psychology (1995) ls, loe-121 CD 1996 Academic Press Limited WHO RECYCLES AND WHEN? A REVIEW OF PERSONAL AND SITUATIONAL FACTORS P. WESLEY SCHULTZ, STUART OSKAMP AND TINA MAINIERI The Claremont Graduate School, 123E Eighth Street, Claremont, CA , U.S.A. Abstract Despite the societal importance of reusing waste materials, few empirical studies have specifically examined recycling behaviors as differentiated from attitudes and intentions. This paper reviews the empirical studies of recycling, summarizes research findings, and identifies areas for future research. The effects on recycling behavior of both personal variables (personality, demographics, and attitudes of environmental concern) and manipulable situational variables are reviewed. Results indicate that high income is a good predictor of recycling, whereas gender and age are not. General environmental concern appears to be related to recycling only when recycling requires a high degree of effort. However, relevant specific attitudes have consistently been found to correlate with recycling behavior. The seven situational variables reviewed (prompts, public commitment, normative influence, goal setting, removing barriers, providing rewards, and feedback) all produce significant increases in recycling behavior. However, there are several major limitations to the research. Results are based largely on single-variable assessments of recycling, and fail to consider interactions with characteristics of the environment or the population involved. Introduction The need to recycle used materials has become a pressing issue over the last 30 years (Ladd, 1990). This increasing concern is clearly evidenced in the proliferation of federal, state, and local legislation directed toward the implementation.of recycling programs. In 1993 in the U.S.A., 41 of the 50 states had in place laws specifying a minimum level of refuse that must be recycled (Grogan, 1993). Despite the increasing concern regarding conservation of natural resources, scant psychological research has been conducted on recycling or conservation behaviors as differentiated from attitudes, intentions, and beliefs. Although reviews of varied pro-environmental behaviors have previously been published (Geller et al., 1982; Dwyer et al., 19931, the topic of recycling behaviors has received relatively little attention. Given the recent explosion of communitywide recycling programs in the U.S.A. and other countries, a review of the scientific research is needed. This paper reviews the empirical psychological research conducted on factors that influence recycling behavior. This literature review focuses on two general strategies of investigation: personal and situational. The personal strategy for studying recycling 105 behaviors has attempted to identify characteristics of an individual that are associated with recycling behavior. Such characteristics include attitudes about the environment, demographic variables, and personality constructs. The situational strategy attempts to identify manipulable aspects of a particular environment that facilitate recycling behaviors. Two general classes of variables have been studied: antecedents and consequences (Geller, 1989). The term recycling can be defined as the process through which materials previously used are collected, processed, remanufactured, and reused (Ruiz, 1993). This paper focuses exclusively on the collection process. For a discussion of processing and reusing aspects of recycling, see Cichonski and Hill (1990). During the past 10 years, the types of materials collected and recycled have increased greatly, as have the strategies used to collect recyclable material. Folz (1991) listed 13 materials that were being collected and recycled in municipal recycling programs as of Data obtained through reports from a national sample of 264 municipal programs indicated that the five most commonly collected materials were: newspaper (96%), glass (94%), ahuninium (88%), plastics no. 1 and no. 2 (67%), and cardboard/corrugated paper
2 106 P. W. Schultz et al. (61%). In addition to the growing number of materials being recycled, there have been changes in the typical types of recycling programs, from short-term campaigns and drop-off programs, to voluntary curbside collection, and later to community-wide recycling, with recycled materials either being separated by the householder or commingled (placed in a single container, rather than separated by types>. Despite procedural differences, most recycling programs have one thing in common-reliance on individual participation. In attempting to develop effective and sustainable ways to reduce the amount of trash being buried in landfills, scientists, policy-makers, and community leaders need to understand the factors that lead people to recycle. It is important to note that caution should be used in generalizing research results from studies conducted 10 or more years ago. With the growth of environmental concern has come an increased interest in recycling programs. Early recycling programs were predominantly voluntary and restricted to the collection of one or two specific materials. This is in sharp contrast to modern curbside recycling programs that collect multiple materials. Also, because most states are attempting to divert a certain percentage of recyclable material out of the waste stream, modern recycling programs place more stress on broad participation. In essence, the recycling situation and procedures today are very different from those of ten years ago. Despite these changes, early recycling research provides a useful assessment of the relative effectiveness of various recycling interventions and is included in this review. Who Recycles? Research taking the personal approach to the study of recycling behaviors has explored several different types of variables. These personal variables can be broken down into four basic classes: environmental attitudes, knowledge, demographic variables, and personality variables. Each of these classes is discussed in turn (see Table 1 for a simple overview of the 6ndingsl. Environmental attitudes Research into attitudinal predictors of recycling behavior has examined both general concern for the environment and specific concern regarding a particular issue. The hypothesis that people who are more concerned with general environmental issues are more likely to recycle is a special case of TABLE 1 Summary table ofpersonal correlates of recycling Personal variable Number Research findings of studies Global environmental 8 Positive when effort is attitudes required Specific environmental 4 Positive relationship beliefs Knowledge about 3 Positive relationship recycling Demographic variables Age 5 Unclear Gender 5 Women more likely to recycle than men Income 4 Positive relationship Education 6 Unclear Ethnicity 1 Unclear Personality variables Responsibility 2 Recyclers more responsible Rigidity 2 Unclear Locus of control 1 Recyclers more likely to have internal locus of control the issue of correspondence between attitudes and behavior (Ajzen & Fishbein, 1977; Rokeach, 1979). Although attitudes often are not strong indicators of behavior, they frequently have been found to be significant predictors (Wicker, 1969). Research findings regarding the relationship between attitudes and recycling behaviors have been generally consistent with attitude-behavior theories. The majority of reported studies, investigating the ability of general environmentzil concern to predict recycling behaviors, have found significant, though relatively small relationships. Schultz and Oskamp (1994) reviewed eight assessments of the relationship between environmental concern and recycling behaviors. Of the eight assessments, five reported a positive relationship, whereas three reported no significant relationship. Analysis of the research indicated that the time period in which the study was conducted seemed to affect the results. All the studies conducted prior to 1980 reported a positive relationship, whereas three of the four studies reported in the 1990s reported no relationship. One possible explanation of this difference might cite the fact that articles reporting no relationship are usually not published unless they contest a previous finding. Thus, early studies in any area tend to report a significant relationship, and later studies may contest it. However, a theoretically more intriguing hypothesis
3 Who Recycles and When? 107 is that the relationship between environmental concern and recycling has decreased over the years. Prior to 1980, recycling programs were just getting started. There generally were few if any incentives to participate, and the amount of effort required to participate was high. Schultz and Oskamp (1994) argued that under conditions in which recycling requires a high amount of effort, only people who were environmentally concerned would recycle. However, the proliferation of recycling programs brought incentives, both monetary and social, along with a reduction in the amount of effort required to recycle. Because more people are recycling today, and doing so for more reasons than just altruistic concern for the environment, the relationship between general environmental concern and recycling seems to have diminished or disappeared. Despite the decreasing trend in the relationship between general environmental concern and recycling, concerns for specific related issues remain significant predictors of recycling. Several recent studies reported significant relationships between recycling and specific beliefs. A study by Howenstine (1993) that closely examined specific beliefs about recycling was conducted with 574 Chicago households. Respondents were classified as recyclers (n = 227) or nonrecyclers (n = 347). A questionnaire asked recyclers which materials they recycled, and asked nonrecyclers to rate each of 12 reasons for not recycling. Factor analysis of the 12 reasons rated by the nonrecyclers revealed three factors that accounted for 54% of the variance: nuisance, location, and indifference. The first factor that emerged from Howenstine s analysis of reasons for not recycling, nuisance, included ideas that recycling does not pay, it is too much trouble, it is too messy, and it requires too much space. Concurring results have been obtained in two other investigations. Vining and Ebreo (1990) found that nonrecyclers reported more concern with financial incentives and rewards, and with the inconvenience of recycling than did recyclers. Gamba and Oskamp (1994) found that infrequent recyclers were more likely than frequent recyclers to endorse personal inconveniences like no space for my recycling bin, and it s hard to move the recycling bin to the curb. Infrequent recyclers were also more likely to identify financial incentives as a motivation for recycling, stating that they recycled to earn money, or to reduce garbage costs. The second factor that emerged from Howenstine s (1993) factor analysis, labeled location, included beliefs that the recycling center was too far away, that not enough trash was generated to make recycling worthwhile, and lack of knowledge about where to take materials. The third factor that emerged from Howenstine s analysis of reasons for not recycling was labeled indifirence. Items loading on this factor were never thought about it, and it makes no difference. Gamba and Oskamp (1994) reported similar findings: both self-reported and observed recycling were significantly positively correlated with the belief that recycling is effective at preserving the environment. In sum, research on the relationship of specific attitudes to recycling behaviors has usually found consistent and significant results. Items measuring aspects of nuisance, location, and indifference have all been found to be related to recycling. However, there are not established scales for measuring specific attitudes pertaining to recycling, so the empirical examinations, though reaching similar conclusions, have developed separate questions with little or no theoretical guidance. Knowledge Knowledge about the recycling program has been found to correlate with recycling. In general, the more information a person has about which materials are recyclable, or where recyclables are collected, the more likely that person is to recycle. Oskamp et al. (1991) suggested that recycling behaviors may be less related to knowledge about global environmental issues than to knowledge about the specifics of recycling. Three studies found knowledge to differentiate recyclers from nonrecyclers. Vining and Ebreo (1990) argued that the greatest difference between recyclers and nonrecyclers is their knowledge of collectable materials. In their study of 197 Illinois households, they found recyclers to have significantly more knowledge about recycling than non-recyclers. Gamba and Oskamp (1994) and De Young (1989) found similar results. Demographic variables Before turning to the relationship of recycling and demographic variables, a brief mention should be made of the relationship of demographics to general environmental concern. Past research findings have indicated that people with the highest level of environmental concern tend to be young (Butte1 & Flinn, 1976; Mohai & Twight, 19871, female (McStay & Dunlap, 1983; Stern et al., 19931, better educated (Arbuthnot, 19741, higher earners (Van Liere & Dunlap, 19801, urban dwellers (Butte1 &
4 108 P. W. Schultz et cd. Flinn, 1976; Van Liere & Dunlap, 19801, and ideologically liberal (Dunlap, 1975; Schultz & Stone, 1994) (for an earlier review, see Weigel, 1977). Although these variables frequently have been correlated with environmental concern, their relationship to recycling behavior has been less consistent. In studies on recycling behavior, the four most often reported demographic variables are age, gender, income, and education. Five recent studies reported findings on age and recycling. In a study of commingled curbside recycling. Gamba and Oskamp (1994) reported a small significant negative correlation of age to self-reported recycling. Oskamp et al. (1991) found no relationship between age and selfreported recycling among community residents in a voluntary curbside recycling program. Vining and Ebreo (1990) and Lansana (1992) both reported a positive relationship, indicating that older residents are more likely to recycle. In a national sample of community recycling programs. Folz and Hazlett (1991) found that, across communities where recycling was mandatory, the median age was significantly negatively related to recycling as measured by the rate of waste diversion, but the two variables were significantly positively related in communities having voluntary recycling programs. Overall, the results of these studies are ambiguous as to both the existence and direction of the relationship between age and recycling. Education has been investigated as a possible predictor of recycling behavior. Of the six studies that reported on the relationship between education and recycling, three found no relationship (Hopper & Nielson, 1991; Oskamp et al., 1991; Gamba & Oskamp, 19941, whereas the other three reported a positive relationship (Webster, 1975; Vining & Ebreo, 1990; Lansana, 1992). The disparate results may be due to the range of education levels included in the samples. All three studies that failed to find a relationship between education and recycling were based on fairly affluent samples, whereas the three studies that found positive relationships were based on samples with a wider range in education levels. Research findings regarding the relationship of gender to recycling are clear. Five studies that studied the relationship between gender and recycling were unanimous in 6nding no significant relationship (Webster, 1975; Vining & Ebreo, 1990; Hopper & Nielson, 1991; O& et al., 1991; Gamba & O&, 1994). Thus, men and women are equally likely to recycle. Because recycling is often a household behavior, the person doing the recycling on a given occasion may be replaced by a person of the opposite gender on other occasions. Unlike gender, income has consistently been found to correlate positively with recycling behavior. Jacobs et al. (19841, Vining and Ebreo (1990), Oskamp et al. (19911, and Gamba and Oskamp (1994) all reported a significant positive relationship. People who make more money are more likely to recycle than people who make less money. The relationship of ethnicity to recycling has not received much research attention. Howenstine (1993) studied household recycling behavior reported by 574 Chicago college students in a sample whose ethnic diversity appeared representative of the surrounding community. Results indicated that 28% of Asians, 28% of Blacks, 12% of Hispanics, and 51% of Whites claimed to recycle. However, this assessment did not consider possible third variables (e.g. parental education, income, or occupaton). Overall, research on demographic variables has found that higher income appears to predict recycling behavior whereas the person s gender does not. Research findings on education are less uniform but suggest the possible existence of a relationship between greater education and recycling. The findings for age are contradictory, and too few studies have examined ethnic differences to reach any conclusions. Personality variables Few studies have assessed the relationship between personality constructs and recycling behavior, though there has been a limited attempt to define a recycling personality. In an early study, Webster (1975) argued that recyclers can be characterized as socially conscious consumers who have a high level of social responsibility. That is, recyclers participate in recycling programs because they believe they have a duty to society, and because they feel they can make a difference. In a sample of 250 urban households, Webster (1975) found that recyclers scored higher than nonrecyclers on both a socially conscious consumerism scale and a measure of social responsibility. Further, recyclers were more tolerant (as measured by the California Personality Inventory) than nonrecyclers, suggesting that recyclers may be less behaviorally rigid than non- recyclers. The relationship between mental rigidity and recycling, though, has yet to be adequately explored. Simmons and Widmar (1990) also supported the finding that recyclers are charcterized by a feeling of responsibility. In questionnaire data from 500 households, recyclers were more likely to feel a sense of responsible action than nonrecyclers. The authors argued, however, that responsibility is not
5 Who Recycles and When? 109 enough-only when it is coupled with knowledge about recycling programs does social responsibility predict recycling behavior. The idea that recyclers believe they can make a difference gains support from the research on environmental concern, although the link between environmental concern and recycling is not strongly established. Several studies (Arbuthnot, 1974; Borden & Francis, 1978; Hines et al ) have found concern for the environment to be associated with an internal locus of control, which can be defined as a belief that a person can determine his/her own destiny (Lefcourt, 1982). Although the relationship between internal locus of control and recycling apparently has only been directly tested by one study (Arbuthnot, 1974), it appears that the relationship may be positive. From the above review of personal predictors of recycling, there emerges a tentative list of demographic and personality variables that may be associated with recycling. However, the percentage of variance in recycling behavior accounted for by individual variables is probably small. In an analysis of community characteristics that predict recycling, Folz and Hazlett (1991) concluded: recycling success, as measured by participation and diversion, is clearly not dependent upon city socioeconomic characteristics or other political features of the community. What explained large portions of the variance in recycling performance among cities with different programs were the specific recycling policies adopted... and other features related to the program s operation. (p. 531) Thus, although some personal variables may be related to recycling behavior, it is necessary to examine situational variables to account for a larger portion of variance in recycling. Situational Factors-Antecedents Over the last 20 years, researchers have attempted to apply behavior analysis methods to problems pertaining to the environment, and more specifically to recycling behavior. The studies cited below are briefly summarized using an organizational framework earlier presented by Geller et al. (1982). These authors framework can be applied to any behavioral intervention research area, including recycling, litter control, energy conservation, or transportation use (see also Geller et al. (1990). Briefly, the scheme classifies behavioral intervention strategies into two groups: antecedent and consequence interventions. For other possible organizational schemes, see Cook and Berrenberg (1981), Gray (1985), and De Young (1993). Any intervention designed to increase recycling behavior by altering a variable prior to performance of the behavior (e.g. collecting recyclables, delivering recyclables to a collection center) is classified as an antecedent strategy. Five types of antecedent variables have been studied: prompting, commitment, normative influence, goal-setting, and removal of barriers to recycling. In the following review, when the data permitted it, effect size estimates have been computed for each intervention. The effect size estimates are reported either as Cohen s d, or as d. Cohen s d is a standard metric intended for comparing results across studies (Cooper & Hedges, 1994), defined as the difference between sample means divided by the average standard deviation. The other metric reported in the table, d, is simply the difference between two proportions, namely the treatment and the control conditions, and is reported in cases where information on standard deviations was not available. Many of the studies reviewed failed to provide sufficient information to calculate an effect size estimate, and in these cases other indications of quantitative findings are presented (e.g. experimental group significantly greater than control group). Because recycling has not attracted an abundance of research, few studies have been reported on each intervention. For these reasons, the effect size estimates are provided only as descriptive statistics. This is not a meta-analysis, and comparison across studies based on the available information would be unwarranted. (See Table 2 for a schematic summary of behavioral intervention studies; studies are listed in chronological order within each topical section of the table). Prompting Prompting represents the simplest, least expensive, and least intrusive of all the antecedent intervention strategies. In a prompting intervention, information (e.g. about the relevance of recycling to alleviating solid wast problems, or about the community s recycling program) is presented to potential participants before the recycling program begins (or continuing during the program). This information can be factual, persuasive, or merely reminders, and it can be delivered in writing, over the telephone, or in person. Twelve studies have examined the effects of prompting on recycling behavior. Three studies showed that a single prompt alone increased recycling (Jacobs & Bailey, 1982; Oskamp, 1986; Burn, 1991). These three studies focused on
6 110 P. W. Schultz et az. TABLET Studies with behavioral interventions to facilitate recycling Type of intervention* Dependent variablet and participants Type of Program and duration ReSUltSS Prompting interventions Raid et al. (1976) 1. Personal prompt + proximity of Witmer & Geller (1976) Arbuthnot et al. ( ) Jacobs & Bailey (in Celler et al., 1982) Jacobs, Bailey, & craws (1984) BumBOslmmp (1986) recycling bins 1. Prompts (flyers) (See under Reward) 1. Survey 2. Written appeal 3. Mailed letter 4. Pair-wise combinations 5. Survey + appeal + letter 1. In-person-delivered brochure 2. Telephone cab 1. Newspaper ad 2. Newspaper ad + door-to-door brochures 1. Written prompt 2. Individual commitment 3. Prompt + commitment 4. Control Amount of newspaper (Apartment residents) (See under Rewatd) Self-reported recycling at the city s households) Participation in newspaper recycling Participation Participation Viking & Ebreo (1989) 1. Community-wide Self-reported recycling multi-media participation, community educational pmgram Spaccarelh et al. (198%1990) Goldenhar & Connell ( ) Burn (1991) Austin et al. (1993) 1. Written prompt (door-to-door flyers) 2. Written prompt + verbal appeal (See under Feedback) (See under Block Leader) 1. Written pmmpt (close) 2. Written pmmpt (far) deyounget al. (1993) 1. Prompt (enviroxlmental) 2. Prompt (economic) 3. Prompt (both) 4. Control recorda on volume of material city-wide Participation (See under Feedback) (See under Block Leader) Amount of white paper recycled (College students and statf~ Self-reported waste reduction 3-week treatment Drop-off l-2 months & 18 months post intervention measure l-week treatment 11-week treatment B-week treatment, drop-off two measures over 36-month treatment 39 week treatment (See under Feedback) (See under Block Leader) lo-week treatment 1 > baseline (z = 4.6) 1 N.S. 5>(2+3)>allother treatments 1>2 2>1 d = 0.28 d = 0.31 d = > baseline (volume) 1 N.S. (participation) 2>1 (See under Feedback) (See under Block Leader) d = 0.54 d = 0.17 d = 0.36 d = 0.50 d = 1.11 commitmt?nt intanarlti01 McCaul & Kopp (1982) rls 1. Set goal of 4 cans/day 2. Public comnli~ent Cnan; appear in Number of cans recycled (College students) a-week treatment 1 > baseline Pardini&Katsev ( ) Burn h O& (1986) Katzav & Pardini ( ) 3. Private commitment (sign a pledge) 1. Oral commitment 2. written commitment 3. Information only (See above) 1. Written commitment 2. Reward bmldons) 3. commi&mi + ly?wanls 4. Control Participation and amount of materials collected (See above) Weekly participation and smount of newspapers Curbaide 2Aweek treatment; aweek follow-up (See above) B-week treatment; 3-week follow-up 2>1,3 (treatment eficts) (See above) 1,2,3 > 4 (amount) d = 0.27 d = 0.18 d = 0.27 (participation) (treatment e&&s)
7 Who F&cycles and When? 111 TABLE 2-contd. Study Type of intervention* Dependent variablet and participants Type of Program and duration Ilm3lt.a~ Commitment interventions-contd. Wang & Katzev 1. Group commitment (1990): I Wang & Katzev (1990): II 1. Group commitment 2. Individual commitment 3. Reward (retail coupons) 4. Control Amount of paper (Retirement center residents) Amount of paper and participation (CoIIege students) 4-week treatment; 4-week follow-up dweek treatment; 3-week follow-up 1 > baseline (treatment and follow-up e&3%) 1>4 2>1,4 3>4 (amount of paper) d = 0.39 d = 0.58 a=045 (participation) de Young et al. (1995) 3. Written commitment (See m&r Feedback) (See unoh Feedback) (See under Feedback) l-week treatment Normative influence interventions NeiIsen L Ehington (1983) 1. Block leader 2. Control Participation 5-month treatment Hopper & NeiIsen 1. Existing block leader Participation (1991) 2. New block leader + monthly prompt 3. Monthly prompt 4. Prompts twice during study 5. Control 24month treatment Burn (1991) 1. Writen prompt 2. Block leader 3. Control Participation Cmonth treatment O& et al. (1994) 1. Block leader Participation 2. Control 1>2 d = 0.32 d = 0.25 d = o-19 d = 0.08 d = 0.47 d = 1.16 d = Goal-setting interventions Hamad et al. ( ) 1. Group feedback (poster) 2. specified goal 3. Self-recorded recycling + reward (See above) Amount of newspaper (Elementary school students) X-week treatment 2 > baseline McCauI & Kopp (1982) (See above) (See above) (See above) Removing barriers Reid et al. (1976) Humphrey et al. (1977) Luyben & Bailey (1979) (See above) 1. Separate personal baskets 2. Divided personal baskets 3. Central collection bin 1. Proximity (See above) &?t? above) (See above) Ratio of recycled material to 1>3 discarded material lo-week treatment 2>3 (Office workers) Amount of newspaper (Mobile home residents) 9.5-week treatment 1 > baseline Reward interventions Geller et al. (1975) 1. Prompts (posters) 2. Reward (individual rattle) 3. Reward (contes& $X-dorm treasuryi 1. Prompts (postersj Amount of white paper recycled (College students) g-week treatment Witmer & GeIIer (1976) Amount of white paper 2. Reward (individuai (College students) I-week treaben~ 2>3>1 raffle) 3-we& follow-up B>baseIine 3. Reward (contest I>bafdine $15-dorm e#&te) Hamad et al. (1977) 1. Verbal prompt Amount of newspaper 2>1 2. Reward (food couponi (Elementary school students) 9.5weak treatment 2>1 3>1 1 N.S.
8 112 P. W. Schultz et al. TABLE 2-contd. Type of intervention* Dependent variablet and participants Type of program and duration Resultss Reward intmventions-contd. Luyben & Cummin5 1. Prompts + proximity Number of cans ( ) 2. Prompts + proximity + (College students) 4-week treatment reward (lottery) Jacobs & Bailey ( ) 1. Reward (lot&y) 2. Reward (monev) households) 4-month treatment 3. Informational kyers 4. Weekly pickups 6. Control (See above) (See above) (See above) Wang t Katzev (1990): II Diamond & Loewy Needleman (1992) & Geller 1. Info + reward Participation in glass (individual) recycling program l-week treatment 2. Infor + reward (group) (College students) twice during 3. Infor = individual academic year reward (lottery) 4. Information only 1. Prompt 2. Prompt (expert) 3. Rejection-retreaV 4. Foot-in-the-door 5. Goals + feedback 6. Reward (&Be) Amount of home-generated recyclables delivered to the work site (Office workers) 25-week treatment Feedback interventions Hamad et al. (See above) (See above) (See above) ( ) Goldenhar & Connell 1. Education (posters) Self-reported newspaper ( ) 2. Feedback (posters) 3. Education + Feedback recycling (College students.) B-month treatment 4. Control Katzev & Mishima 1. Group feedback Amount of paper (1992) (poster on weekly lbs. (College students) a-week treatment; paper collected) l-week follow-up de Young et al. (1995) 1. Specify group feedback Amount and quality of Centralized bin 2. General group feedback newspapers recycled a-month treatment 3. Written commitment (Apartment complex 4. Control residents) 2 > baseline d = 0.11 d = 0.05 d = 0.06 d = 0.06 (See above) d = 0.11 d = 0.23 d = > other 5 conditions (See above) d = 0.11 d = 0.25 d = 0.25 d = 2.47 (treatment 1 N.S. 2 N.S. 3 N.S. effect) * If a study used a control group, that group is listed. t Observed behavior unless noted as self-reported behavior. $ Results am for comparisons of intervention conditions unless noted as baseline comparisons. curbside recycling, and the prompt was typically delivered in writing. A larger group of studies, however, showed enhanced effects of combining different communication approaches. Jacobs et al. (1984) found the addition of brochures increased curbside participation to a level two to four times that produced by newspaper ads alone. Spaccarelli et al. ( ) found that an oral plea along with a written prompt resulted in a 22.1% increase in curbside participation, compared to a 2.4% increase in participation among residents receiving only the written message. Arbuthnot et al. ( ) found that prompts delivered as part of the Toot-in-the-door influence technique increased selfreported recycling at community drop-off centers among new residents by 88%, compared to less than 10% for each of the prompting strategies alone. Other studies showed that combining prompts with proximity of the collection bin increased recycling. Austin et al. (1993) found that providing containers in convenient places and delivering written prompts encouraged more people to recycle than using prompts alone. Reid et al. ( ) likewise found prompting and proximity together increased recycling in apartment complexes by loo%, 60%, or 50% (depending on the complex). How does prompting compare with other intervention approaches? Few studies applying a prompting strategy included another strategy for comparison. These studies (with one exception) agreed that prompting alone is not as effective as other approaches in increasing recycling behavior. For example, feedback Goldenhar & Connell, 1992) or rewards (Witmer & Celler, 1976) increased college
9 Who Recycles and When? 113 students recycling in special dormitory recycling drives, whereas prompting had no significant effect. In contrast, both block leaders and prompting increased participation in curbside recycling in Burn s (1991) study, although the block leader group participated at a higher rate than the prompting-alone group (average of 28% vs 12%, respectively). Burn and Oskamp (1986) found that prompting, commitment, and the two strategies combined produced similar increases in curbside participation (39%, 42% and 42%, respectively). Determining whether prompting can bring about enduring behavior change is difficult, because most prompting studies used short-term time frames without follow-up measures. Five studies, however, measured recycling over a period of four months or longer, and two of these employed separate treatment and follow-up periods. In these studies, prompting encouraged sustained participation in curbside and drop-off community programs, but not in the one study conducted in college dorms (Witmer & Geller, 1976). Vining and Ebreo (19891 reported a continued increase in the volume of materials collected during a 3-year, multi-media community campaign, and Spaccarelli et al. ( ) showed a sustained increase in curbside recycling participation over 39 weeks. Burn (1991) found that the increase in curbside recycling still remained 12 weeks later, and Arbuthnot et al. ( ) noted that the increased number of new recyclers continued 18 months following the implementation of the footin-the-door technique. Prompting may work better for some types of recycling than others. Most studies measured household recycling in community curbside or drop-off programs, except for one study conducted in apartment complexes and three studies conducted with college students or staff. The prompting strategies appeared to increase participation in community drop-off and curbside programs. In the apartment complex study, the effects of prompts alone could not be disentangled from proximity effects (distance to the recycling bin area), and the effects were further confounded by increased bin capacity. Two of the three studies of special drives in college settings were the only studies not to show a positive effect of prompting on recycling. One major limitation of all these studies is the lack of consideration of the individual characteristics of recipients of the promptsin all studies reviewed here, results of the interventions were reported across all participants; however, it seems likely that people with more knowledge or environmental concern would be more strongly affected by prompts than would participants low on these variables. In general, prompting interventions may be more effective with people who already have a favorable attitude toward recycling. Commitment interventions Commitment interventions are based on the principle that people become resistant to pressures to change their actions after making a decision to behave in a certain way (Oskamp, 1991). Seven studies have investigated the effects of commitment on recycling behavior, including public versus private commitment (McCaul & Kopp, 19821, written versus oral commitment (Pardini & Katzev, , and group versus individual commitment (Wang & Katsev, 1990). In most studies, commitment was initiated by requesting the research participant to sign a pledge or a statement. The two exceptions were (1) a public commitment condition, in which particpants were told that their names would appear in the college newspaper (McCaul & Kopp, 19821, and (2) an individual commitment condition, in which participants were asked in person if they would participate in the recycling drive (Wang & Katzev, 1990). Treatment generally lasted 2 to 6 weeks, with four studies including follow-up measures lasting from 2 to 4 weeks. After the seven studies, commitment produced increases in both curbside and special drive participation-not only during treatment, but also during follow-up. These findings indicate a potential for long-term effects; however, the longest treatment period was 6 weeks and the longest follow-up period only 4 weeks. This short-term nature of all the commitment strategy investigations precludes the ability to make statements with confidence about enduring effects of commitment on recycling behavior. Is there a difference among the types of conunitment (e.g. written, oral, individual, group)? Two studies actually compared various commitment approaches with each other. In these, written commitment produced greater increases than oral commitment in curbside participation and amount of recyclables collected (Pardini & Katzev, , and individual commitment yielded more participation than group commitment in a special recycling drive on a college campus (Wang & Katzev, 1990). Five studies, on the other hand, incorporated other interventions for comparison. Generally, commitment tended to produce longer lasting effects (i.e. on follow-up measures) than prompting or
10 114 P. W. Schultz et al. rewards. Pardini and Katzev ( ) found oral and written commitment groups recycled more newspaper and had higher curbside participation rates than the information-only group during both the intervention and follow-up periods. In contrast, Burn and Oskamp (1986) found commitment increased recycling 42% over baseline, but there was no significant difference between commitment and prompting as noted in Pardini and Katzev s study. Pardini and Katzev ( ) suggested that commitment strategies may work because people who make such pledges move beyond the external justification for recycling (signing or stating a pledge) and 6nd their own additional reasons for recycling. A competing explanation for the effectiveness of commitment interventions is that the changes are due to social pressure. All of these studies reported the effectiveness of commitments in terms of participation in a community recycling program, or of the amount of material collected. Both of these variables are socially visible-putting the bin at the curb to be recycled may make both the participation and the amount of materials observable to other residents. An interesting question is whether this change in behavior is internalized. If the changes in behavior are due solely to the social pressure of being observed, then they have not been internalized. This may imply that the change in behavior will be short-lived, and that it will not generalize to other recycling settings that are less visible (e.g. work, school, or travel). Because the commitment studies lasted only between two and eight weeks, conclusions about long-term changes in behavior cannot be made from these studies. Normative influence The use of social norms to encourage recycling behavior is a relatively new approach. One social psychological strategy is to enlist community members to model recycling behavior and to persuade their nonrecycling neighbors to participate in the recycling program. As a naturally occuming example, Oskamp et al. (1991) reported that participation in a curb side program was higher for people whose friends and neighbors recycled. Four studies have experimentally examined the effects of social influence on recycling behavior. These studies (with one exception) indicated that using peer support to establish community recycling norms can increase and sustain recycling behavior. For example, Nielsen and Ellington (1983) found a 26-W weekly curbside participation rate over 5 months among blocks with an identifiable recycling leader, as compared to 11.5% participation in blocks without a designated leader. Even when socioeconomic status and stability of neighborhood were held constant, results indicated that participation rates in blocks with block leaders were consistently higher than in blocks without leaders. Burn (1991) and Hopper and Nielsen (1991) found similar positive results. However, Oskamp et al. (1994) found no significant difference in the amount of recycled material, frequency of participation, or degree of contamination of the material recycled when a previously established block leader area was compared with that in a similar socio-economic area that did not have block leaders. As suggested earlier, the block leader approach has two potential sources of influence: information and personal contact. Two studies examined the effects of personal contact over information alone. Burn ( 1991) observed that block leader neighborhoods participated significantly more in curbside recycling, (58% of households recycled at least once during post-treatment) than the group receiving information left at the door (38% recycled at least once). The effects of both interventions did not diminish over the 12-week post-treatment period. Hopper and Nielsen (1991) found similar results. In sum, block leaders may be effective because they serve as initiators of social norms within their neighborhoods. The desire for social recognition may motivate nonrecycling neighbors to begin recycling, and this behavior may in turn be reinforced through social approval. Personal contact by the block leader may also prompt public commitment, which in turn could initiate recycling behavior. Although the block leader approach has yet to be thoroughly assessed, it appears to have the potential to produce longterm changes in recycling behavior, although the findings of Oskamp et al. (1994) suggest that the strategy may not always work. Furthermore, the use of volunteer block leaders in neighborhoods presents a cost-effective intervention for communities. As with the previously discussed interventions, studies conducted on social norms failed to consider any characteristics of the community. For example, residents who perceived themselves as part of the community may be more affected by this intervention than residents who feel isolated or alienated. It seems likely that rural residents may be more affected by social norms than residents of an urban community, and likewise home owners may be more affected than renters and apartment dwellers. The successful use of social pressure to induce recycling may be largely contingent upon the extent to which residents see themselves as part of the community.
11 Who Recycles and When? 115 Goal-setting Coal-setting involves the specification of a set target of material to be reycled. In his correlational study of community recycling programs, Folz (1991) found that cities which established a goal to recycle a specific proportion of the waste stream reported significantly higher levels of citizen participation in municipal recycling programs than cities which did not establish a goal. Only two studies have experimentally assessed the effect of goal-setting on recycling. These studies both found significant effects in increasing the amount of materials collected in special recycling drives at an elementary school (Hamad et al ) and a college (McCaul & Kopp, 1982). However, the persistence of behavior change was not tested, and since both studies used special populations, questions remain regarding the generalizability of the results to community residents. Students who spend many hours of the week together may develop a sense of cohesiveness that, in turn, may motivate efforts toward common goals. Making goals salient and important to members of larger groups, such as communities, may be a more difficult task. Removing barriers to recycling All recycling programs involve effort on the part of the participant. One of the most direct, but often overlooked, ways to increase recycling behavior is the removal of barriers to recycling. Simply stated, this strategy attempts to reduce response costs by minimizing the amount of effort required to recycle. Three barriers to recycling have been studied: distance of the collection location from the participant, method of collection, and sorting of recyclable materials. Distance. Most older recycling programs involved depositing materials in a central location. From an administrative perspective, this reduces the cost of the program. From the participants perspective, however, the use of a central collection location adds personal costs of extra time and effort involved in the transportation of recyclabales to the collection center. Three studies have experimentally examined the effect of increased bin proximity on recycling participation (Reid et al., 1976; Humphrey et al., 1977; Luyben & Bailey, 1979). While these studies are few in number, they consistently indicated that the closer participants are to the collection center, the more likely they are to recycle. For example, Luyben and Bailey (1979) found a 47% average increase in drop-off recyling participation in a mobile home park following the placement of six additional bins throughout each park. These additional bins increased each resident s proximity to a bin and thus reduced the extra costs of transporting materials. One limitation of research on proximity is the short-term duration of studies. Recycling was measured over periods ranging from 3 weeks in Reid et al. (1976) to a high of 10 weeks in Humphrey et al. (1977). On the other hand, a strength of this research is that proximity was shown to work in a variety of situations-apartment complexes (Reid et al., 19761, mobile home parks (Luyben & Bailey, 1979), and offices (Humphrey et al., 1977). Several other studies reporting nonexperimental findings support the claim that proximity to a collection center positively affects recycling. Witmer and Geller (1976) reported that students whose dorm rooms were closest to the collection center showed the highest level of participation in a paper recycling program. Cummings (1977) found that among 432 New York City residents, proximity to a voluntary recycling drop-off center was signscantly positively related to participation in the program. Collection method. A second barrier to recycling among home owners is the collection method. Folz (1991) examined differences between communities with a curbside collection program and ones using a drop-off location. His analysis revealed a large significant difference in estimated participation rates. Communities with vuluntary curbside collection had an estimated 49% participation rate, compared to 25% for communities with drop-off collection. This finding suggests that removing the need to transport materials to a central location can increase participation rates. The schedule of collection may also affect recycling participation. recycling programs collect recyclable materials on a fnted schedule, and in some cases the collection day for recyclables and for other refuse does not coincide. It seems likely that participation in recycling programs would be higher if both recyclables and other refuse were collected on the same day, and at frequent intervals. The single reported study on this topic found that cities with same-day pick-up of recyclables and other refuse did not report higher participation rates in recycling programs than cities with different-day collection schedules (Folz, 1991). However, this finding combined reports on mandatory and voluntary programs and had other possible confounding factors. Sorting. A third barrier to recycling is the effort required to sort materials. Asking participants to
12 116 P. W. Schultz et al. sort recyclable8 into different bins is common in home recycling programs. As participants begin to recycle more types of materials, they may find themselves separating those materials into numerous bins. In some German cities, for example, apartment complexes have up to seven different bins in which to place different materials. An alternative to having participants sort their recyclable8 is to use commingled recycling, in which participants place all recyclables mixed together in a single collection bin. The materials are then collected and sorted at a materials recovery facility, using both mechanical and manual techniques. This recycling method requires less effort by the participant. Gamba and Oskamp (1994) examined household participation rates in a city-wide commingled curbside recycling program. They found that over 90% of the households participated in the program at least once in five consecutive occasion-an amazingly high figure-whereas earlier, in a voluntary separated recycling program, less than 40% of city residents were estimated to take part. In contrast, Folz s (1991) correlational study across 264 cities concluded that requiring separation does not significantly decrease participation in recycling programs. In his analysis of municipal recycling programs, both mandatory and voluntary, there was no difference in the estimated average participation rates for programs that required separation and those that did not. The many other uncontrolled differences among cities in his study, however, make generalization of its findings questionable. Clearly the topic of separation requires further research. Summary In sum, the research data regarding antecedent intervention techniques indicate that many types of interventions have been successful in increasing recycling behavior for the duration of the intervention. Commitment, norms, prompts, goals, and the removal of barriers all can produce significant increases in recycling behavior. Several clear limitations, however, exist in the literature. First, the persisting effects of these strategies remain largely untested. Written personal commitment apparently increases recycling for a longer period of time than do extrinsic rewards, but the length of time that commitment affects recycling is unclear, because six weeks is the longest follow-up period over which its effects have been demonstrated. As De Young (1993) emphasized, the durability of program effects is a crucial issue for research intended to be relevant to public policy issues. Second, the relative effectiveness of different antecedent interventions has yet to be assessed. It seems likely that these interventions are more effective with people who already have favorable attitudes toward recycling. Third, almost all reported studies have employed single measures of recycling. As was pointed out above, interventions may have differential effects on different recycling variables (e.g. amount, frequency of participation, and contamination). Situational Factors-Consequence Variables Any intervention that attempts to modify recycling behavior by presenting a consequence (i.e. feedback of information, a reward, or a punishment) contingent upon the behavior is classified as employing a consequence strategy. The majority of empirical studies in this recycling area, however, examined the effects of rewards. No reported study has assessed the effect of punishment on recycling (probably for ethical reasons). Rewards Eight studies directly tested the effect of rewards on recycling behavior (see Table 2). This strategy is based on learning theory, which suggests that external contingencies or rewards will make a behavior more appealing and induce behavior change (Geller, 1989). All eight studies found that offering rewards (e.g. money, coupons, or lottery tickets) significantly increased the amont of material people will recycle. Furthermore, chances to win lottery prizes generally had stronger effects than did small cash payments, and individual rewards typically produced larger increases in recycling behaviors than did group rewards. Comparisons of reward intervention with other interventions suggests that rewards can produce larger changes in behavior. For example, rewards have been found to produce larger changes than prompting (Geller et al., 1975; Needleman & Geller, 19921, information (Diamond & Loewy, 19911, goal setting combined with feedback (Needleman & Geller, 19911, and group commitment Wang & Katzev, 1990). However, despite the potential of reward interventions, there are several drawbacks. First, the change in behavior produced by reward programs was short-lived, for after termination of a reward program, recycling behavior typically returned to baseline levels (cf. Katzev & Johnson, 1987). The incentives used in these studies may not have facilitated long-term behavior change for
13 Who Recycles and When? 117 several reasons. First, the rewards may have lost some of their novelty as time passed, and participants may have found that other factors, such as time and effort, outweighed the attraction of the reward. Second, the rewards may not have been meaningful to all particpants or substantial enough to catch participants interest; this poses the problem of developing attractive incentives for diverse groups of people. Third, the imposition of external motivators may have masked or reduced internal benefits derived from recycling behavior, as in the social psychological research literature on overjustification effects (e.g. Lepper & Greene, 1975). Furthermore, the cost of supplying rewards and organizing their advertisement and distribution often outweighs the economic benefits of recycling. Another issue concerning reward interventions is the extent to which behavior change produced for rewards of one material (e.g. aluminum cans) will generalize to other materials (e.g. newspaper). Needleman and Geller (1992) examined this issue in their study of recycling at a worksite setting. Employees were rewarded with an entry into a drawing each time they returned aluminum cans for recycling. Results showed that there was a significant increase in the amount of aluminum cans recycled, but no increases for other materials (e.g. newspaper, glass). This finding suggests that reward interventions are only effective in increasing behavior related to the specific material targeted with the reward, and that the changes in recycling behavior do not generalize to other materials. Although rewards appear to provide powerful short-term changes in recycling, two unanswered questions are evident. First, whose behavior is changing? Studies reviewed above on specific attitudes found nonrecyclers to be more concerned with financial issues than recyclers. This finding leads to the hypothesis that offering rewards will be more effective with people who are not currently recycling. Secondly, as with all previously reviewed studies, only single assessments of recycling have been studied in this literature. Most of the eight studies examined the amount of material collected, which was the rewarded behavior. However, if the rewards had been offered for frequency of participation, or quality of recycled material, rather than amount, different results might have been obtained. Feedback Another important aspect of consequences is the effect of feedback strategies on recycling. Presenting people with feedback about their behavior has been successful in decreasing energy and water consumption, typically by amounts in the lo%-15% range (Se&man & Darley, 1977; Se&-man et al., 1981). However, despite the success of the feedback technique in other arenas, only four studies have directly assessed its effectiveness in increasing recycling (Hamad et al., ; Goldenhar & Connell, ; Katzev & Mishima, 1992; De Young et al., 1995). Seligman et al. (1981) suggested that in order for feedback to be successful, several criteria must be met. First, people must be able to identify a relationship between their behavior and the feedback. This requirement is met most effectively by providing immediate feedback. For example, when thermostats are combined with meters that indicate energy consumption rates, people can observe the effects of their behavior. Second, the individual must be interested in change; feedback is not effective in changing behavior if the person has no desire to change. The desire to change is a strong mediating variable in the effectiveness of feedback interventions. The studies reported by De Young et al. (1995) and Hamad et al. ( ) failed to find a significant change in behavior, whereas Katzev and Mishima (1992) and Goldenhar and Connell ( ) did find a signif&.& increase in recycling. The two studies that reported a significant effect were conducted in the 1990s with college students, who are in general more liberal, educated, and higher SES than average. The study by De Young et az. was conducted on a sample of apartment residents, and the study by Hamad et al. was conducted in the early 1980 on school children-both groups being qualitatively different from college students. It seems likely that the feedback interventions were effective because the college student participants in the study were interested in change. Clearly more research is needed on the effects of feedback on recycling. Discussion Over the past 20 years, social scientists have attempted to identify effective techniques to encourage recycling. As the solid waste crisis continues to escalate, city and county officials are experiencing greater pressure to find ways to sustain a high level of waste diversion. Policy-makers and community leaders are asking social scientists which intervention, or set of interventions, produce the best results. A definitive answer to this question is d3ecu.h given the current state of the literature. Although
14 118 P. W. Schultz et uz. an understanding of who recycles, and when, is beginning to emerge, several clear limitations need to be addressed. First, the answer concerning which intervention is the best depends largely on the desired outcome of the intervention. To date, nearly all empirical investigations of recycling interventions have measured a single dependent variable, usually either the percentage of participants in the recycling program (new or continuing) or the amount of material collected. A third potential variable is the quality of the collected material. As the percentage of nonrecyclable material collected in the recycling program (termed contumination) increases, the usability and value of the collected material decreases. Only one study has examined all three dependent variables, and it found different effects for each variable (Oskamp et al., 1994). Overall, different recycling interventions may affect different aspects of recycling. For example, some prompting interventions (i.e. informational ones) may decrease the amount of contamination, but be ineffective at increasing either the number of recyclers, or the amount of material collected. Second, future research should examine the extended effects of various intervention strategies. many studies ex amining situational variables have measured behavior changes against a baseline, followed by a second baseline period, and then by a second intervention. That is, more than one intervention is often applied to the same sample. Using this method precludes the collection of follow-up data. As recycling programs become more prominent, social scientists will be asked how to produce long-term participation in recycling programs (De Young, 1993). This question is best answered by examining the effects of two or more interventions over an extended period of time, on randomly assigned groups. Using separate groups for each intervention allows comparisons across conditions, and also allows for follow-up measures of each intervention. Third, empirical investigations of recycling interventions to date have explored either personal OF situational variables. Such studies attempt to identify main effects: e.g. what type of person recycles, or what conditions are associated with more recycling. The next step for recycling research is to examine the differential effects of intervention strategies on various types of people. Research should look for interactions between the type of recycling program and characteristics of the individual-that is, interactions between person and situational variables. Fourth, the effectiveness of different interventions may depend largely on characteristics of the community in which the program is instituted. For instance, providing people with rewards for recycling may be more effective in increasing recycling among people low in environmental concern than among those high in environmental concern. People who are concerned for the environment are motivated to recycle for internal reasons; recycling makes them feel they are helping to protect the environment (Simmons & Widmar, 1990). People low in environmental concern, on the other hand, do not have this internal motivation. External incentives to recycle might provide them with a motive and cause an increase in recycling. Many other potential interactions have been mentioned throughout this review. The idea that the type of intervention should be selected based on the desired outcome and the characteristics of the target population is an integral part of social marketing concepts (Bloom & Novelli, 1981; Geller, 1989). What we have advocated here as an interactional approach to research has also been described as market segmentation, i.e. partitioning a potential market for the product (the intervention) into homogeneous subgroups based on common characteristics. According to Geller (1989), this technique provides a basis for selecting target markets and developing optimal promotional programs for individual target segments (p. 28). A fifth limitation of the current research is the unknown extent to which recycling one material predicts recycling of another (i.e. response generalization). Investigations of recycling behavior to date have examined the recycling of only one material, or of several materials combined-but not differential rates of recycling for different materials. Recycling behavior is ordinarily measured as the amount of newspaper, white paper, glass, plastic, or metal cans returned for recycling. It is implicitly assumed that both personal and situational variables found to predict recycling of one material will generalize to the recycling of another material. That is, people who recycle white paper are generally assumed to be more likely to recycle aluminium cans. This assumption is, in part, a reflection of the more general belief held by many researchers about the relationship among different proenvironmental behaviors. Investigators of proenvironmental behaviors (e.g. recycling, litter reduction, water conservation, energy conservation, and purchasing environmentally safe products) have often assumed that, for the most part, people who show a propensity for performing one proenvironmental behavior
15 Who Recycles and When? 119 are likely to show a similar propensity for another. That is, someone who recycles should also conserve water, be an environmentally conscious shopper, etc. Further, these behaviors are often understood as the manifestation of an environmental ideology (Dunlap & Van Liere, 1978; Commoner, 1990). In contrast, several studies have found that various proenvironmental behaviors are not closely related (e.g. Tracy & Oskamp, ). For instance, Siegfried et al. (1982) used attitudinal and demographic variables to predict each of four proenvironmental behaviors (lowering thermostats, using less hot water, purchasing environmentally safe products, and avoiding the use of unnecessary lights). Their analysis failed to reveal a consistent pattern of predictors for the four behaviors. The authors concluded generalizations from one specific proenvironmental behavior to other forms of behavior may be inappropriate (p. 288). Similar conclusions were reached by Oskamp et al. (1991). In planning and developing interventions to improve recycling programs, the costs of each intervention play a large part in determining which type of intervention is selected, and how it is implemented. However, despite the practical importance of intervention costs, very few scientific studies report on costs. Expenses for interventions come from two source: materials used, and their distribution. Materials for interventions can range from simple flyers in the case of prompting, to cash lotteries for reward interventions. Once the materials have been prepared, the intervention must be delivered to potential recyclers. Interventions that require contact with each potential recycler (e.g. rewards, individual feedback, and commitment) are typically more expensive. In contrast, interventions that can be delivered to recyclers without individual contact (e.g. group feedback, prompting, removing barriers) are less expensive. However, costs for interventions vary from setting to setting. A useful line of research would be an investigation of the amount of behavioral change produced per dollar, where dollar amounts include the expenditures for materials as well as a standardized estimate of the number of person hours required to develop and distribute the intervention. Conclusion Disposal of solid waste has become a serious societal issue. Recycling represents one attempt to reduce the amount of trash being buried in landfills. As of 1993, there were over 4000 community-wide recycling programs in the U.S. (Van Voorst, 1993). However, resolving the solid waste crisis involves more than simply implementing community recycling programs. In order for recycling programs to be an effective way to reduce the amount of trash we generate, changes must be made in our daily behavioral patterns. In order for behavioral scientists to help community leaders shape public policy, it is necessary to treat recycling as a multifaceted pattern of behaviors. Recycling is not simply the act of placing a can or bottle in a recycling crate; it involves performing such actions on a consistent basis across situations. Behavior change interventions designed to increase the effectiveness of recycling programs need to move beyond single-variable assessments of recycling, to consider interactions with the environment in which the program is based, and the population with whom the intervention is conducted. References Ajzen, I. & Fishbein, M. (1977). Attitude-behavior relations: a theoretical analysis and review of empirical research. Psychological Bulletin, 84, Ajzen, I. L Fishbein, M. (1980). Understanding Attitudes and Predicting Social Behauior. Englewood Ciiffs., NJ: Prentice-Hall Arbuthnot, J. (1974). Environmental knowledge and recycling behavior as a function of attitudes and personality characteristics. Personality and Social Psychology Bulletin, 1, Arbuthnot, J., Tedeschi, R., Wayner, M., Turner, J., Kressel, S. & Rush, R. ( ). The induction of sustained recycling behavior through the foot-in-the-door technique. Journal of Environmental Systems, 6, Austin, J., Hatfield, D., GrindIe, A. & Bailey, J. (1993). Increasing recycling in office environments: the effects of specific informative cues. Journal of Applied Behavioral Analysis, 26, Bloom, P. N. & Novelli, W. D. (1981). Problems and challenges in social marketing. Journal of Marketing, 46, Borden, R. & Francis, J. (1978). Who cares about ecology? Personality and gender differences in environmental concern. Journal of Personality, 46, Burn, S. (1991). Social psychology and the stimulation of recycling behaviors: the block leader approach. Journal of Applied Social Psychology, 21, Burn, S. M. & Oskamp, S. (1986). Increasing community recychng with persuasive communication and public commitment. Journal ofapplied Social Psychology, 16, Butte& F. L Flinn, W. (1976). EnvironmentaI politics: The structuring of partisan and ideological cleavages in mass environmental concern. Sociological Quarterly, 17, Cichonski, R. L Hill, T. (1990). Recycling Sourcebook. Detroit: Gale Research Inc. Commoner, B. (1990). Making Peace with the Planet. New York: Pantheon. Cook, S. W. & Berrenberg, J. L. (1981). Approaches to encouraging conservation behavior: a review and conceptual framework. Journal of Social Issues, 37, Cooper, H. & Hedges, L. (Eds.) (1994). The Handbook ofresearch Synthesis, New York: Russel Sage.
16 120 P. W. Schultz et al. Cummings, L. (1977). Voluntary strategies in the environmental movement: recycling as coopt&ion. Journal of Voluntary Action Research, 63-L De Young, R. (1986). Some psychological aspects of recycling: the structure of conservation and satisfaction. Environment and Behavior, 18, De Young, R. (1989). Exploring the differences between recyclers and nonrecyclers: the role of information. Journal of Environmental Systems, 18, De Young, R. (1993). Changing behavior and making it stick: the conceptualization and management of conservation behavior. Environment and Behavior, 25, De Young, R., Boerschig, S., Camey, S., Dillenbeck, A., Elster, M., Horst, S., KIeiner, B. & Thomson, B. (1995). Recycling in multi-family dwellings: increasing participation and decreasing contamination. Population and Environment, 16, De Young, R., Duncan, A., Frank, J., Gill, N., Rothman, S., Shenot, J., Shotkin, A. & Zweizig, M. (1993). Promoting source reduction behavior: the rule of motivational information. Environment and Behavior, Diamond, W. & Loewy, B. (1991). Effects of probabilistic rewards on attitudes and behaviors. Journal ofapplied Social Psychology, 21, X DunIap, R. E. (1975). The impact of political orientation on environmental attitudes and actions. Environment and Behavior, 7, DunIap, R. E. & Van Liere, K. D. (1978). The new environmental paradigm. Journal of Environmental Education. 9, Dwyer, W. O., Lemming, F. C., Cobern, M. K., Porter, B. E. & Jackson, J. M. (1993). Critical review of behavioral interventions to preserve the environment: research since Environment and Behavior, 25, Folz, D. H. (1991). Recycling program design, management, and participation: a national survey of minicipal experience. Public Administmtion Review, S& Foil, D. H. & Hazlett, J. M. (1991). Public participation and recycling performance: explaining program success. Public Administration Review, 51, Gamba, R. & O&, S. (1994). factors influencing community residents participation in commingled curbside recycling programs. Environment and Behavior, 26, GeIIer, E. S. (1989). Applied behavior analysis and social marketing an integration to preseve the environment. Journal of Social Issues, 45, GeIIer, E. S., Berry, T. D., Ludwig, T. D., Evans, R. E., Gilmore, M. R. & Clark, S. W. (1990). A conceptual framework for developing and evaluating behavior change interventions for injury control. Health Education Research: Theory and Pmctice, Geller, E. S., Chaffe, J. L. & Ingram, R. (1975). Promoting paper recycling on a university campus. Journal of Environmental Systems, 6, GeIIer, E. S., Far&, J. C. & Post, D. S. (1973). Prompting consumer behaviors for pollution control. Journal of Applied Behavior Analysts, 6, GelIer, E. S., Win&t, R. & Everett, P. B. (1982). Preserving the Environment: New Stmtegies for Behavior Change. New York: Pergamon. Goldenhar, L. M. & ConneIl, C. M. ( ). Effects of education and feedback interventions on recycling knowledge, attitudes, beliefs, and behaviors. Journal of Environmental Systems, 21,32X333. Gray, D. B. (1986). Ecological Beliefs and Behaviors: Assessment and Change. Westport, CT: Greenwood. Grogan, P. (1993). Legislative evaluations. In H. Lund, Ed., The McGraw-Hill Recycling Handbook. New York: McGraw-Hill, pp Hamad, C. D., Bettinger, R., Cooper, D. & Semb, G. ( ). Using behavioral procedures to establish an elementary school paper recycling program. Journal of Environmental Systems, 10, Hines, J. M., Hungerford, H. R. & Tomera, A. N. (1986)-1987X Analysis and synthesis of research on responsible environmental behavior: a meta-analysis. Journal of Environmental Education, 18,1-8. Hopper, J. & Nielson, J. M. (1991). Recycling as ahruistic behavior: normative and behavioral strategies to expand participation in a community recycling program. Environment and Behavior, 23, Howenstine, E. (1993). Market segmentation for recycling. Environment and Behavior, Humphrey, C., Bord, R., Hammond, M. & Mann, S. (1977). Attitudes and conditions for cooperation in a paper recycling program. Environment and Behavior, 9, Jacobs, H. E. & Bailey, J. S. ( ). Evaluating participation in a residential recycling program. Journal of Environmental Systems, 12, Jacobs, H. E., Bailey, J. S. & Crews, J. (1984). Development and analysis of a community-based resource recovery program. Journal ofapplied BehaviorAnalysis, 17, Katzev, R. D. & Johnson, T. R. (1987). Promoting Energy Conservation: An Analysis of Behavioral Research. Boulder, CO: Westview. Katzev, R. D. & Mishima, H. (1992). The use of posted feedback to promote recycling. Psychological Reports, 71, Katzev, R. D. & Pardini, A. U. ( ). The comparative effectiveness of reward and commitment approaches in motivating community recycling. Journal of Environmental Systems, 17, Ladd, A. (1992). The solid waste crisis and support for recycling: a research note. Sociological Spectrum, 10, Lansana, F. (1992). Distinguishing potential recyclers from nonrecyclers: a basis for developing recycling strategies. Journal OfEnvironmental Education, 23,X-23. Lefcourt, H. M. (1982). Locus of Control: Current 7 rervd.s in Theory and Research (2nd Edn.) Hillsdale, NJ: Erlbaum. Lepper, M. R. & Greene, D. (1975). Turning play into work: effects of adult surveillance and extrinsic rewards on children s intrinsic motivation. Journal of Personality and Social Psychology, 31, Luyben, P. D. 8z Bailey, J. (1979). Newspaper recycling: the effects of rewards and proximity of containers. Environment and Behavior, l& Luyben, P. D. & Cummings, S. ( ). Motivating beverage container recycling on a college campus. Journal of Environmental Systems, l& McCauI, K D. & Kopp, J. T. (1982). Effects of goal setting and commitment on increasing metal recycling. Journal of Applied Psychology, 67, McStay, S. & Dunlap, R. E. (1983). Male-female differences in concern for environmental quality. International Journal of Women s Studies, 6, Mohai, P. & Twight, B. (1987). Age and environmentai concern: an elaboration of the Butte1 model using national survey evidence. Social Science Quarterly, 68, Needleman, L. D. L Geller, E. S. (1992). Comparing interventions to motivate work-site collection of home-generated recyclables. American Journal of Community Psychology, 20,
17 Who Recycles and When? 121 Nielson, J. M. & Ellington, B. (1986). Social processes and resource conservation: a case study in low technology recycling. In N. Feiner & E. Geller, Eds., Environmental Psychology: Directions and Perspectives. New York: Praeger. Oskamp, S., Harrington, M., Edwards, T., Sherwood, P. L., Okuda, S. M. & Swanson, D. L. (1991). Factors influencing household recycling behavior. Environment and Behavior, 23, Oskamp, S., Zelezny, L., Schultz, P. W., Hurin, S., Burkhardt, R. & O Neil, E. ( Commingled versus separated curbside recycling and long-term participation. Paper presented at the annual meeting of the American Psychological Association, 1994, Los Angeles, CA. Pardini, A. U. & Katzev, R. D. ( ). The effect of strength of commitment on newspaper recycling. Journal of Environmental Systems, 13, Reid, D., Luyben, P. D., Rawers, R. J. & Bailey, J. ( Newspaper recycling behavior: the effects of prompting and proximity of containers. Environment and Behavior, 8, Rokeach, M. (1979). Understanding Human Values: Individual and Societal. New York Free Press. Ruiz, J. A. (1993). Recycling overview and growth. In H. Lund, Ed., The McGraw-Hill Recycling Handbook. New York: McGraw- Hill, pp ~15. Schultz, P. W. & Stone, W. F. (1994). Authoritarianism and attitudes toward the environment: field and laboratory perspective. Environment and Behavior, 26, Schultz, P. W. & Oskamp, S. (1994). Environmental concern and proenvironmental behaviors: Do only the concerned recycle? Paper presented at the annual meeting of the Western Psychological Association, 1994, Kona, HI. Seligman, C., Becker, L. J. & Darley, J. M. (1981). Encouraging residential energy conservation through feedback. In A. Baum and J. E. Singer, Eds., Advances in Environmental Psychology, Vol. 3, Energy: Psychological Perspectives. Hillsdale, NJ: Erlbaum. Seligman, C. & Darley, J. M. (1977). Feedback as a means of decreasing residential energy consumption. Journal ofapplied Psychology, 62, Siegfried, W., Tedeschi, R. & Cann, A. (1982). The generalizability of attitudinal correlates of proenvironmental behavior. Journal of Social Psychology, 118, Simmons, D. & Widmar, R. ( Motivations and barriers to recycling: toward a strategy for public education. Journal of Environmental Education, 22, Spaccarelli, S., Zolik, E. & Jason, L. A. ( ). Effects of verbal prompting and block characteristics on participation in curbside newspaper recycling. Journal of Environmental Systems, 19,4Ei-67. Stem, P., Diets, T. & Kalof, L. (1993). Value orientations, gender, and environmental concern. Environment and Behavior, 25, Tracy, A. P. & O&, S. ( ). Relationships among ecologically responsible behaviors. Journal of Environmental Systems, 13, Van Liere, K. D. BE Dunlap, R. E. (1980). The social bases of environmental concern: a review of hypotheses, explanations and empirical evidence. Public Opinion Quarterly, 44, Van Voorst, B. (1993). Recycling: stalled at curbside. Time, 18 October, Vining, J. & Ebreo, A. (1989). An evaluation of the public response to a community recycling education program. Society and Natural Resources, 2, Vining, J. & Ebreo, A. (1990). What makea a recycler? A comparison of recyclers and nonrecyclers. Environment and Behavior, 22, Wang, T. H. t Katsev, R. D. (1990). Group commitment and resource conservation: two field experiments on promoting recycling. Journal ofapplied Social Psychology, 20, Webster, F. (1975). Det.e rmhing the characteristics of the socially conscious consumer. Journal of Consumer Research, 2, Weigel, R. H. (1977). Ideological and demographic correlates of proecology behavior. Journal of Social Psychology, 103, Wicker, A. (1969). Attitudes vs actions: the relationship of verbal and overt behavioral responses to attitude objects. Journal of Social Issues, 26, Witmer, J. & Geller, E. S. (1976). Facilitating paper recycling: effects of prompts, raf5es, and contests. Journal of Applied BehaviorAnalysis, 9,3X-322. Manuscript received 26 October 1994 Revised manuscript received 28April 1996