Behavioural spillover in the environmental domain: An intervention study

Journal of Environmental Psychology 40 (2014) 381e390

Contents lists available at ScienceDirect

Journal of Environmental Psychology journal homepage: www.elsevier.com/locate/jep

Behavioural spillover in the environmental domain: An intervention study Pietro Lanzini a, John Thøgersen b, * a b


Ca
Foscari, Cannaregio 873, 30121 Venezia, Italy Department of Management, Universita Department of Business Administration, Aarhus University, School of Business and Social Sciences, Bartholins All e 10, Aarhus, Denmark

a r t i c l e i n f o

a b s t r a c t

Article history: Available online 7 October 2014

This study tests hypotheses about behavioural spillover in the environmental domain as well as the impacts of monetary inducements and verbal praise on behavioural spillover by means of a field experiment. A sample of 194 students from a large university in Denmark were randomly allocated to a control group or to one of two experimental conditions where they were encouraged to purchase “green” products by means of either financial compensation and incentives or verbal encouragement and praise. Participants answered a baseline survey containing questions concerning a wide range of environmentally relevant behaviours and after a six weeks intervention period where they were requested to keep track of their purchases by means of a shopping diary they answered a second survey with the same content as the first. This allowed us to analyse the change in self-reported pro-environmental behaviours over the six weeks, to identify instances of behavioural spillover from “green” purchase behaviour to other pro-environmental behaviours and to investigate if such spillover was affected by the nature of the intervention. The study revealed a positive spillover from “green” purchasing to other pro-environmental behaviours. However, the spillover mostly affects low-cost behaviours. Not unexpectedly, the monetary inducement had a stronger direct impact on “green” shopping than verbal encouragement and praise. However, contrary to popular beliefs, the spillover effects of a monetary inducement on other proenvironmental behaviours are at least as strong as that of verbal encouragement and praise. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Behavioural spillover “Green” shopping Economic incentives Field experiment

1. Introduction The possibility that one pro-environmental behaviour might lead to another has gained increased attention in recent years, both among scholars (e.g., Evans et al., 2013; Thøgersen & Crompton, 2009; Whitmarsh & O'Neill, 2010) and policy makers (Defra, 2008). “Behavioural spillover,” as this is sometimes termed (Thøgersen, 1999), implies that acting in a pro-environmental way changes (i.e., increases or decreases) a person's likelihood or extent of performing another/other pro-environmental behaviour(s). Indeed, there is mounting empirical evidence suggesting that proenvironmental behaviours tend to be correlated in practice (e.g., € Thøgersen & Olander, 2006; Whitmarsh & O'Neill, 2010). Further, the evidence suggests that this relationship is usually positive (see also Berger, 1997; De Young, 2000; Frey, 1993; Maiteny, 2002; Marian, Chrysochou, Krystallis, & Thøgersen, 2014; Scott, 1977). However, despite growing research on this phenomenon, still no

* Corresponding author. E-mail addresses: [email protected], [email protected] (J. Thøgersen). http://dx.doi.org/10.1016/j.jenvp.2014.09.006 0272-4944/© 2014 Elsevier Ltd. All rights reserved.

consensus has been reached on the nature of spillover, its drivers, possible contingencies or practical relevance (Thøgersen & Crompton, 2009). There is also no consensus about how long time it takes for eventual behavioural spillover to develop, although it seems obvious that behavioural patterns need time to be re-shaped, depending on the features of the specific activity (in terms of complexity, familiarity and salience for the individual, and so on). For example, a study of possible spillover effects of a carrier bag charge found no evidence of a behavioural effect, but a strengthening of participants' environmental identity was registered, which the authors speculated might lead to spillover to other proenvironmental behaviours in the longer run (Poortinga, Whitmarsh, & Suffolk, 2013). It is an important weakness of research on behavioural spillover that the available evidence is mostly correlational. Among the noteworthy exceptions is a recent study by Evans et al. (2013) finding that participants in two experiments needing to discard a sheet of paper were more likely to put it into the recycling bin when environmental goals had been primed by another proenvironmental behaviour (car-sharing). Else, most of the

382

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

experimental evidence pertaining to behavioural spillover has been generated within the so-called “foot-in-the-door” paradigm (Freedman & Fraser, 1966). Freedman and Fraser (1966) conducted two experiments, asking individuals to carry out a small request first and a larger one later. The experiments differed regarding whether the same or different person(s) asked the two requests and whether the requests were similar (except for their size) or different. In both experiments, people who carried out a small request were more likely to accept carrying out a larger request later. This “foot-in-thedoor” strategy has been applied and tested in a wide range of contexts, often showing a positive effect (e.g., Burger, 1999; Cann, Sherman, & Elkes, 1975; Pliner, Hart, Kohl, & Saari, 1974; Snyder & Cunningham, 1975), also with regard to pro-environmental behaviour (e.g., Scott, 1977; Souchet & Girandola, 2013). It has also been suggested that negative spillover is possible, for example if people perform a low-cost pro-environmental behaviour to get an excuse for not performing other (and perhaps more €rfer, 1998). Broad goals costly) behaviours (Diekmann & Preisendo such as the protection of the environment can be pursued by means of a wide range of different behaviours, entailing different financial, physical and/or psychological costs (Kaiser & Wilson, 2004). Arguably, it is rational for an individual striving to attain a specific goal to try easy behaviours towards that goal first, before more costly and inconvenient ones (Susewind & Hoelzl, 2014). Thøgersen and Crompton (2009) suggested that when the rational inclination to first do simple and easy things for the environment is combined with self-serving bias (a pervasive phenomenon, cf. Pieters, Bijmolt, van Raaij, & de Kruijk, 1998), the likelihood of performing more difficult behaviours after easy ones might dwindle. Negative spillover might also occur because acting in a proenvironmental way makes people feel they have a “license” to some moral leeway afterwards (Mazar & Zhong, 2010; Monin & Miller, 2001). Mazar and Zhong (2010) found that participants in an experiment who had shopped in a virtual store with mostly “green” products were afterwards more likely to cheat and steal for private gain than individuals who had shopped in a store with mostly conventional products. They attribute this effect to “green” purchasing giving a “license” to behave in an immoral way afterwards. Gneezy, Imas, Brown, Nelson, and Norton (2012) added an important caveat to the Mazar and Zhong study, however, finding that “moral licensing” only occurs when the moral behaviour is costless. When acting in a moral way is costly, participants in their study were more likely than the control group to perform other moral behaviours afterwards: a positive spillover effect. Since pro-environmental behaviour usually is costly, the latter suggests that pro-environmental behaviour is unlikely to lead to moral licensing in practise. In sum, the number of experimental studies documenting the possibility of behavioural spillover is still small, and they are mostly carried out in the artificial setting of the laboratory. Hence, there is a need for more experimental research, especially outside the lab, documenting the practical relevance of spillover for proenvironmental behaviour. It is the objective of this article to contribute filling this gap.

2. Theoretical foundations Behavioural spillover has been studied as “response generalization” at least since the 1970s (Stokes & Baer, 1977), but these early studies were broader in scope and did not focus specifically on pro-environmental behaviours (e.g., Ludwig & Geller, 1991). Spillover between pro-environmental behaviours, as described above, is consistent with several theories in psychology, including various consistency theories (Thøgersen, 2004), learning theories (Bandura, 1986), and goal theory (Dhar & Simonson, 1999).

Goal theory (Dhar & Simonson, 1999) assumes that individuals have a set of broad underlying goals (e.g. living a healthy life, preserving the environment, enjoying pleasant activities) whose achievement requires the allocation of resources, both financial and others (e.g. time). It specifies that a person can make subsequent choices, for instance among multiple courses within a single meal, such that they reinforce each other and maximize their short-term effect on the achievement of a particular goal. An environmentally relevant example of applying this principle is the campaign to promote meatless days, which has been “described as an attempt to create and to highlight commitment to a shared, higher order goal” €sler, & Aiking, 2014, p. 121). (de Boer, Scho As previously mentioned, Evans et al. (2013) found that participants in two laboratory experiments needing to discard a sheet of paper were more likely to recycle the paper when a prior task had made them think about a completely different pro-environmental behaviour (car-sharing). However, the increased inclination to recycle only happened when the task had primed environmental goals (i.e., not when it primed financial goals). This suggests that behavioural spillover can be the product of one's proenvironmental actions priming broader environmental goals that, once activated, guide other behaviours (see also Spence, Leygue, Bedwell, & O'Malley, 2014). Hence, behavioural spillover can be the product of environmental goals that are relevant for a broad set of behaviours being activated by performing a pro-environmental behaviour. An even more popular theory in current empirical research on behavioural spillover is self-perception theory (Bem, 1972). Selfperception theory assumes that people form attitudes by observing and inferring from their own past behaviour and the contexts in which the behaviour took place. For example, it has been suggested that the foot-in-the-door strategy is effective exactly because individuals use their behaviour as a cue to their own attitudinal dispositions (Scott, 1977). Self-perception effects have been observed in a wide range of domains, including proenvironmental behaviour (Chaiken & Baldwin, 1981). Further, there is evidence that behaving pro-environmentally in a given domain might not only change the attitude towards that behaviour (Holland, Verplanken, & Van Knippenberg, 2002), but might also activate a general disposition in the individual, which might influence future behaviour also in other domains (Cornelissen, Pandelaere, Warlop, & Dewitte, 2008). That past behaviour influences a person's pro-environmental self-identity which might next increase the likelihood of performing completely different pro-environmental behaviours is also supported by a panel study finding evidence of spillover between a fuel conserving driving style and (intention to reduce) meat consumption, fully mediated through environmental self-identity (Van der Werff, Steg, & Keizer, 2014). In sum, it seems that individuals use their own behaviour as a cue to their broader dispositions and therefore a specific behaviour can potentially affect broader behavioural patterns and spark a positive spillover across behavioural domains. It has also been suggested that acting consistently across proenvironmental behaviours, leading to positive spillover from one behaviour to another, may be motivated by the desire to avoid cognitive dissonance and the uneasiness it bears (Thøgersen, 2004). According to Aronson's self-consistency revision of cognitive dissonance theory, the most disturbing inconsistencies are those threatening important elements of one's self-concept as a competent, moral, reliable person (Aronson & Carlsmith, 1962; Dickerson, Thibodeau, Aronson, & Miller, 1992; Thibodeau & Aronson, 1992). Moral standards may therefore influence whether an inconsistency is experienced as disturbing by the individual (Thøgersen, 2004). Other contingencies include specific constraints, either personal or contextual, on a specific behaviour (e.g., Guagnano, Stern, & Dietz,

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

€ 1995; Olander & Thøgersen, 1995), partly because inconsistencies attributed to forces outside individual control do not lead to cognitive dissonance (Collins & Hoyt, 1972; Festinger & Carlsmith, 1959). If a change in behaviour requires a big effort, people are likely to choose instead other routes to decrease dissonance, or simply to live with their perceived behavioural inconsistencies (Abelson, 1968; Festinger, 1957). Hence, for behavioural spillover to occur, the performance of the derived pro-environmental behaviour(s) should be as easy as possible. Learning theories stress the relevance of skills, awareness, knowledge and self-efficacy that can be developed by a learningby-doing process (Bandura, 1986). Maiteny (2002) argued that direct experience is relevant in driving long-lasting and generalizing pro-environmental behavioural changes, whereas shifts driven by means of incentives or regulations are more superficial, and bound to fade and revert back to old habits. Thøgersen, Haugaard, and Olesen (2010) found that consumers adopt a new eco-label more quickly the more they already know and use other eco-labels. Hence, learning theory and research predicts that engaging in a pro-environmental behaviour may facilitate the uptake of other consistent behaviours due to increased capacities or self-confidence in the environmental field, as well as strengthened awareness and overall knowledge of the issues at stake. However, others warn against excessive hopes in the role of knowledge in shaping environmental behaviour (and even more, in spurring spillover between different behaviours) (Kollmuss & Agyeman, 2002). Indeed, the track record of campaigns to increase awareness of environmental problems regarding pro-environmental behaviour change is notoriously poor (Staats, Wit, & Midden, 1996; € Olander & Thøgersen, 2014). The literature also identifies a range of factors that might moderate the spillover process. Characteristics of both behaviours (how easy or difficult are they, how similar) and individuals (such as values and norms, preference for consistency) are among the likely candidates. Although the identification of moderators is outside the scope of this article, we mention some of the most important ones for the sake of completion. For example, Thøgersen (2004) found that spillover is more likely to occur between behaviours perceived as similar than between behaviours perceived € as dissimilar and Thøgersen and Olander (2003) found that the likelihood of spillover is higher if the value domain that Schwartz (1994) termed “universalism” is given high priority or if a person possesses strong personal norms for environmental-friendly behaviour. Cialdini, Trost, and Newsom (1995) found that a personality trait that they termed “preference for consistency” moderates consistency-based behaviour and Guadagno, Demaine, and Cialdini (2001) found that the foot-in-the-door effect depends on preference for consistency. In sum, spillover processes have been studied in the framework of a range of different psychological theories suggesting different mechanisms through which spillover between different proenvironmental behaviours might occur, including the activation of broad pro-environmental goals, a change in self-perception or identity produced by past behaviour, the quest for consistency and learning-by-doing. These theories and mechanisms all predict a positive spillover across pro-environmental behaviours, but the suggested “path” of spillover and the driving forces differ. Although the amount of experimental research is increasing, it is insufficient to determine which of these theories and mechanisms best account for pro-environmental spillover, or whether the different proposed mechanisms supplement one another in producing this phenomenon. In the empirical analysis presented in the following, we focus on the identification of pro-environmental spillover as an empirical phenomenon in an experimental field setting, rather than sorting out which specific mechanism(s) produce(s) the spillover.

383

3. Hypotheses In recent years, environmental campaigners have started considering how to include behavioural spillover in their campaign strategy. For example, the UK government's Department for Environment, Food and Rural Affairs, Defra (2008, p. 22) has been contemplating how “to promote a range of behaviours as entry points in helping different groups to make their lifestyles more sustainable.” Obviously, the first prerequisite for a spillover strategy to work is that one employs an intervention that will effectively promote the behaviour(s) used as entry point(s). In our empirical study, we test two interventions that aim at making consumers pay more attention to, and buy, “green” (e.g., eco-labelled or organic food) products in a context, Denmark, where “green” products are easily accessible. We test the following hypotheses about the direct effect of our interventions: H1.1 In a context where “green” (e.g., eco-labelled or organic food) products are easily accessible, an intervention that makes consumers pay more attention to such products will lead to an increase in “green” consumer purchases. H1.2 An economic inducement that at least compensates for the premium price for “green” products is more effective at promoting “green” consumer purchases than verbal praise for the “green” behaviour. H1.3 The interventions will lead to bigger behavioural changes for people who bought relatively little “green” products at baseline, compared to those that already bought them a lot. H1.2 is based on the general finding that one of the major barriers for consumers' purchase of “green” products is the premium price charged for these products (Hughner, McDonagh, Prothero, Shultz, & Stanton, 2007; Stall-Meadows & Davey, 2013). The effects of an intervention (and also possible spillover effects) being bigger for those that have more potential for change at baseline makes intuitive sense. To some extent, this can be considered a special case of the generally observed regression towards the mean (Cohen, Cohen, West, & Aiken, 2003). Although there is still no general consensus about the likelihood of spillover between pro-environmental behaviours, with some research even suggesting that a negative spillover is possible (e.g., due to moral licensing), most of the evidence in the reviewed literature supports the idea that pro-environmental behaviour can positively spill over from one domain to another (although there might be contingencies). Hence, we hypothesize: H2.1 An intervention that succeeds in making consumers increase their “green” purchases will have positive “knock-on” or “spillover” effects on other “green” behaviours, thus leading to a more sustainable consumption pattern. H2.2 Spillover effects will be stronger for behaviours that are relatively easy to change and where the change implies no or low costs (e.g., everyday water or electricity saving) than for more costly behaviours (e.g., changing to non-car travel modes or volunteering for a “green” cause). H2.3 The spillover effect will be bigger for people who performed such behaviours relatively little at baseline, compared to those that already performed them a lot.

4. Method For this research, we designed a panel study consisting of two waves of online surveys, administered before and after an intervention promoting “green” purchase behaviour. Participants were

384

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

research on pro-environmental behaviour and spillover (Thøgersen € & Olander, 2003). We also included questions about volunteering

Table 1 Sample profile. Total (N ¼ 194) Age

Mean 23.8 (18e44) Member of env. association? (%) Yes 5.7 No, never 91.2 Not anymore 3.1 Living with: (%) Alone 26.3 Partner 33.5 Parents 1.0 Other students 31.4 Other 7.7 Distance home-university: (%) <2 km 31.4 2 kme5km 45.4 5 kme15 km 17.6 >15 km 5.7

Males (NM ¼ 71)

Females (NF ¼ 123)

Mean 24.1 (19e36)

Mean 23.6 (18e44)

2.8 94.4 2.8

7.3 89.4 3.3

29.6 29.6 1.4 28.2 11.3

24.4 35.8 0.8 33.3 5.7

42.3 43.7 12.7 1.4

25.2 46.3 20.3 8.1

recruited via email in June and October 2012 among students at Aarhus University, School of Business and Social Sciences (Denmark). The recruiting message asked students about their willingness to participate in a study on consumer behaviour, which might imply the purchase of specific products for a limited period of time. 206 participants filled out some or all of the baseline survey, but 12 of these had to be dropped from the analysis because they did not complete the second survey or completed it insufficiently.1 The final sample hence consists of 194 participants with a mean age of 23.8. The gender composition is 123 female and 71 male. Further descriptive statistics on the sample are presented in Table 1. Participants were randomly assigned to experimental (105) or control (89) conditions. In the experimental condition, participants were randomly assigned to receive one of two interventions: either a monetary or a non-monetary inducement encouraging them to buy “green” products (a pro-environmental behaviour). Participants in the monetary condition received a sum compensating them for the premium price they had to pay when buying “green” products. The compensation was calculated based on average premium prices in Danish supermarkets for the product categories covered by the study. In addition, individuals in the monetary condition participated in a lottery draw for two attractive prizes: a cash sum of 2000 DKK and an iPad. The expected value of this incentive, given the number of participants in the draw, is around 70 DKK (close to 10V). Participants in the non-monetary condition received messages praising their effort for participating in the study and informational feedback about the benefits of buying “green” products. The questionnaire was developed in English and then translated to Danish. Both versions were pre-tested on a small sample of students to check the clarity of the questions and avoid misunderstandings and ambiguities. Besides introductory background questions, the questionnaire contained questions regarding the person's performance of 23 proenvironmental behaviours, in addition to questions not pertinent to the present purpose. The included behaviours cover the domains of travel-mode choice, recycling, energy/water conservation and green purchasing, which are the main macro-categories in most

1 Nine of these 12 were from the control group, but the difference in drop-out between groups is not statistically significant (Chi-square ¼ 5.177, 2 df., p ¼ .075). The dropouts did not differ significantly from the rest at the Bonferroni-corrected level (p < .002) on any at the behaviour items measured at baseline.

for a “green” cause (see Appendix 1 for a complete list of the behaviour items included in the study). For the battery of behaviours, participants were asked: “How often do you X [each behaviour]”, using a 7-point scale ranging from 1 (never) to 7 (always/ every time) and the midpoint being 4 (half the time), except for volunteering where a 5-point scale was used (see Appendix 1). We used exploratory factor analysis and item analysis to identify clusters of behaviour items that could be merged without loss of substantial information (see Appendix 1). For example, the five included types of “green” shopping are highly correlated and seem to represent a behaviour category in these consumers' minds (Cronbach's alpha time 1 ¼ .84, time 2 ¼ .82). The same holds for the recycling items (Cronbach's alpha time 1 ¼ .67, time 2 ¼ .68), the public transport items (Cronbach's alpha time 1 ¼ .73, time 2 ¼ .66), the biking items (Cronbach's alpha time 1 ¼ .84, time 2 ¼ .76), and the volunteering items (Cronbach's alpha time 1 ¼ .70, time 2 ¼ .69). Participants received the link to the online survey (using Qualtrics) by email and they could choose between a Danish and an English version. After filling out the first survey, establishing a baseline, experimental interventions were deployed. Participants receiving an experimental intervention (but not the control group) got an email with specific instructions for the continuation of the study where they were encouraged to buy organic food and other eco-labelled products for a period of six weeks. It was specified that environment-friendly products were to be considered those with an organic food label (in the case of food products) or products with eco-labels like the EU-flower or the Nordic Swan (for non-food products). We chose this as the source behaviour because it is a common and commonly promoted everyday behaviour and it shares with most other proenvironmental behaviours that it typically entails extra costs; notably the premium-price that is usually charged for such products. It also represents a class of behaviours where money is naturally involved and which naturally involves considerations about money, also without a monetary inducement. For six weeks, participants were asked to record their purchases within a specified set of product categories in a shopping diary, specifying cases where they chose an environment-friendly version (e.g: organic milk vs traditional milk, etc). Appendix 2 reports the complete list of product categories as well as the average number of items for each category purchased by participants in the monetary and in the praise condition, and the share of the “green” alternative. Receipts had to be kept to allow crosschecking at the end of the experiment. Participants in the monetary condition were informed that they would receive a monetary compensation for the premium price they had to sustain to purchase environment-friendly products, with no further specification on how the amount of the compensation was to be calculated. They were also informed about the lottery, with the specification that the more “green” products one purchased, the more lottery slots one would receive. Participants in the non-monetary experimental condition instead received messages of praise (e.g. “Your participation in the project contributes to building knowledge that will in turn be useful for developing specific and effective policies addressing environmental problems”, “We can all make a difference … it might be a drop in the ocean, but isn't the ocean made up of drops?”, and so on). Halfway through the experiment (that is, three weeks after the kick-off of the manipulations) participants received a reminderemail. The message for those in the monetary condition was a simple reminder of both the activities required for the project and of the compensation and incentive they would receive at the end.

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

5. Results 5.1. Direct effects of the experimental interventions We analyse the direct effects of the interventions on the change in “green” buying from before to after the intervention (i.e., the analysed behaviour at t2 minus the same behaviour at t1), as well as possible spillover effects to other pro-environmental behaviours, by means of multiple regression analysis (SPSS 21). In all analyses, we control for the person's behaviour at baseline (i.e., how much room there is for improvement). The results regarding the direct effects of the intervention are reported in Table 2. As expected, the amount of “green” buying at baseline had a significant negative impact on the change in this behaviour over time. The more a person bought “green” at baseline, the more difficult or costly it would be to increase “green” buying even further. This causes the well-known phenomenon of regression towards the mean (Cohen et al., 2003). Also as expected, both of the two interventions had a positive impact on “green” buying, but only the monetary condition produced a statistically significant impact. According to Cohen (1988), a standardized regression coefficient of .20 represents a small effect, though. Still, this analysis confirms Hypothesis 1.2 and partially supports Hypothesis 1.1. The analysis reported in Table 2 also reveals a significant negative interaction between the monetary condition dummy and behaviour at baseline. For this and all the following analyses of interaction effects, all continuous independent variables were standardized before calculating product terms (Cohen et al., 2003). For ease of interpretation, the significant interaction effect in Table 2 is plotted in Fig. 1. The plot is based on the unstandardised regression coefficients (including the constant) and produced by means of Jeremy Dawson's online Excel worksheets (http://www. jeremydawson.co.uk/slopes.htm). As clear from Fig. 1, this analysis confirms Hypothesis 1.3, that the intervention had a stronger

Table 2 The direct effects of the interventions on the change in “green” buying (N ¼ 194).

(Constant) GreenBuying1 Money (Dummy) Praise (Dummy) Money*GreenBuying(t1) Praise*GreenBuying(t1) Note: R2-adj. ¼ .21.

B

Std. Error

.079 .403 .516 .223 .369 .080

.113 .124 .181 .192 .185 .193

Beta

t

Sig.

.335 .200 .081 .174 .035

.692 3.247 2.852 1.164 1.995 .414

.490 .001 .005 .246 .047 .679

1.5

Increase in Green Buying

Participants in the praise condition received instead further praising messages and links to documents on the importance of shifting towards green purchasing patterns (e.g. links to webpages of official national and European certification schemes for food and nonfood products). After six weeks, the link to a second online questionnaire (exact replication of the first) was e-mailed to participants. This allows us to gain insights into the effects of the intervention and possible spillover between behaviours. Praise and monetary inducements were terminated at this point, shopping diaries and receipts were collected from participants in the monetary condition, and participants were de-briefed thanking them for their participation in the study and informing them that they might be contacted in the upcoming weeks for some follow-up questions. Participants in the monetary condition were also informed about where and when to collect the compensation for their green purchases and the lottery prize, after the calculations and the checks on the receipts were completed.

385

1 0.5

Low Green Buying at Baseline

0 Control -0.5

Monetary Inducement

High Green Buying at Baseline

-1 Fig. 1. Plot of the significant interaction effect in Table 2, based on the unstandardised regression coefficients (including the constant).

effect for people who bought relatively little “green” products at baseline, compared to those that already bought them a lot. Since we already controlled for regression towards the mean by including the direct effect of behaviour at baseline, it can be inferred that this interaction represents more than that. 5.2. Spillover from an increase in “green” buying to other proenvironmental behaviours In the next two steps, we investigate spillover from the induced change in “green” buying behaviour to other pro-environmental behaviours. First, we look at whether the change in “green” buying spilled over to other green behaviours. We use in principle the same statistical procedure as in Table 2, the dependent variable in each regression analysis being the change in a proenvironmental behaviour from before to after the intervention and, as before, controlling for the same behaviour at baseline. In each analysis, the change in “green” buying behaviour is included as a second independent variable and we also included the interaction between the two independent variables. The results are reported in Table 3 for the six out of nine possible cases where the regression analysis revealed a significant spillover effect. As expected, the more the person performed a proenvironmental behaviour at baseline, the less that behaviour increased over time. This is the same as we found with regard to “green” buying, and the assumed reason is the same: the more a person did the behaviour at baseline, the more difficult or costly it would be to increase it even further. Further, as hypothesized, changes in “green” buying behaviour spilled over and was echoed in changes in the same direction in a range of other proenvironmental behaviours: recycling, use of public transport, carpooling, printing on both sides of the paper, switching off the light when leaving a room, and saving water.2 All the spillover effects are small, though, according to Cohen (1988). For three other behaviours included in the survey, no spillover effects were detected: cycling, volunteering and reading documents on the screen rather than printing them.3 In one case, there is also a significant interaction effect. For ease of interpretation, this interaction effect is plotted in Fig. 2. As clear from the figure, this analysis shows that there is a stronger spillover effect on switching off light among people who did this to a relatively low extent at baseline.

2 We merged the two water saving items for this analysis, although they are not very strongly correlated, because calculations based on the individual items pointed in the same direction, but the results were not significant at the 5 pct. level. Hence, we infer that the two items are formative indicators of the same behaviour category. 3 In order to save space, non-significant results are not shown, but they can be acquired from the authors.

386

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

Table 3 A change in “green” buying spilling over to other pro-environmental behaviours. B Recycling (Constant) Behaviour(t1) Change in green buying Public transport (Constant) Behaviour(t1) Change in green buying Car-pooling (Constant) Behaviour(t1) Change in green buying Printing on both sides (Constant) Behaviour(t1) Change in green buying Saving water (Constant) Behaviour(t1) Change in green buying Switching off light (Constant) Behaviour(t1) Change in green buying ChangeGreenBuying*Behaviour(t1)

Std. Beta Error

t

Sig.

R2-adj

B

.313 .083 .490 .085 .170 .082

3.784 .000 .16 .384 5.797 .000 .138 2.078 .039

.185 .126 .438 .130 .311 .126

1.463 .145 .08 .247 3.474 .001 .170 2.389 .018

.137 .110 .401 .108 .262 .110

1.250 .213 .11 .284 3.702 .000 .183 2.382 .018

.079 .094 .225 .046 .264 .079

.841 .401 .16 .332 4.915 .000 .226 3.344 .001

.085 .072 .163 .047 .128 .061

1.173 .242 .07 .244 3.466 .001 .147 2.089 .038

.064 .647 .190 .212

1.004 .317 .39 .568 9.806 .000 .176 3.026 .003 .199 3.427 .001

.064 .066 .063 .062

Table 4 Interventions promoting “green” buying spilling over to other pro-environmental behaviours.

5.3. Effects of experimental interventions on other behaviours Next, we investigate whether the spillover effects identified above can be traced directly back to the interventions. If this is the case, there should be a more positive change in other proenvironmental behaviours in the intervention groups than in the control group, and specifically a more positive change in the group receiving monetary inducement than in the group getting praise and the control group (according to Table 2). In principle, we repeated the statistical analyses in Table 3, with the only exception that the intervention dummies included in Table 2 replaced the change in “green” buying behaviour. The results are reported in Table 4 for the five out of nine possible cases where the regression analysis revealed a significant spillover effect. It appears that participants receiving experimental treatment behaved more consistently (i.e., acted in a more pro-environmental way) also in other domains than the one targeted by the intervention at the end of the six-week period. For one of the five behaviours where a spillover effect was identified in Table 3 (i.e., recycling), the behaviour change could be traced back to the experimental treatment that had significantly influenced the “source” behaviour according to Table 2. In addition, in one case

Reading documents on screen (Constant) .100 Behaviour(t1) .254 Money (Dummy) .599 Praise (Dummy) .052 Biking (Constant) .539 .481 Behaviour(t1) Money (Dummy) .192 Praise (Dummy) .564 Public transport (Constant) .369 Behaviour(t1) .448 Money (Dummy) .153 Praise (Dummy) .586 Recycling (Constant) .152 Behaviour t1 .411 Money (Dummy) .344 Praise (Dummy) .169 Money*Behaviour(t1) .545 Praise*Behaviour(t1) .110 Switching off light (Constant) .112 Behaviour(t1) .329 Money (Dummy) .226 Praise (Dummy) .379 Money*Behaviour(t1) .297 Praise*Behaviour(t1) .645

Std. Error

Beta

t

Sig.

R2-adj

.160 .158 .259 .277

.622 1.610 2.315 .187

.535 .109 .022 .852

.04

.166 .178 .014

.206 .140 .323 .339

2.611 3.432 .594 1.662

.010 .001 .553 .098

.06

.248 .046 .130

.195 .127 .300 .313

1.891 3.523 .510 1.876

.060 .001 .611 .062

.06

.253 .040 .147

.121 .123 .192 .205 .213 .195

1.261 3.339 1.790 .827 2.561 .565

.209 .001 .075 .409 .011 .573

.21

.322 .126 .059 .205 .048

.097 .133 .154 .159 .164 .186

1.161 2.479 1.471 2.388 1.804 3.470

.247 .014 .143 .018 .073 .001

.37

.289 .094 .152 .180 .290

where no behavioural spillover was detected in Table 3, the analysis in Table 4 revealed an impact from the successful behavioural intervention on “another” pro-environmental behaviour (reading texts on the screen rather than printing them out before reading). Furthermore, in one case there was a significant and in two cases a marginally significant impact on another pro-environmental behaviour (switching off light when leaving a room, biking and using public transport) from the behavioural intervention that did not have a significant impact on the source behaviour (praise). Again, all the detected spillover effects are small, according to Cohen (1988). Notice that the (only marginally significant) spillover effects for the two travel behaviours have opposite signs. It seems from these analyses that the praise intervention made some participants substitute some of their use of public transport with biking. In two cases, recycling and switching off light when leaving the room, significant interaction effects are registered. For space reasons, these interaction effects are not plotted, but since the pattern is similar to what was shown in Figs. 1 and 2, it is clear that these analyses show that there is a stronger effect of the interventions on these two behaviours among people who performed them to a relatively low extent at baseline.

Increase in Switching off Light

1.5

6. Discussion and conclusions

1

0.5 Low Switching off at Baseline

0

Low Increase Green Buying High Increase Green Buying -0.5

High Switching off at Baseline

-1

Fig. 2. Plot of the significant interaction effect in Table 3 based on the unstandardised regression coefficients (including the constant).

This study investigated the possible spillover from one proenvironmental behaviour to another by means of experimental interventions (verbal and monetary inducement in combination with keeping record of green purchasing behaviour) in a field setting. The interventions aimed at stimulating “green” purchases and this and a range of other pro-environmental behaviours were measured before and after. The evidence emerging from this study corroborates earlier findings suggesting that pro-environmental behaviours are connected in the sense that undertaking a pro-environmental behaviour in one domain can, under given circumstances, spill over and increase

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

the likelihood of performing other pro-environmental behaviours, even in different domains (Berger, 1997; De Young, 2000; Kals, Schumacher, & Montada, 1999; Maiteny, 2002). Because most everyday behaviours tend to be pretty stable over time (Thøgersen, € 2006; Thøgersen & Olander, 2003), we did not expect major changes in behaviour to develop over the six weeks time-span of the experiment, and this expectation was confirmed. However, among those increasing “green” purchasing, significant increases were observed in a range of other pro-environmental behaviours as well. Hence, it seems that getting people to adopt or increase an everyday pro-environmental behaviour can indeed act as a wedge and trigger a virtuous circle leading to broader impacts on the consumption pattern. However, it should be noted that the obtained spillover effects are not very strong. It is a limitation of this study that the sample consists of students only. Besides an obvious over-representation of a narrow age group, students have peculiar features making them scarcely representative of the overall population. They usually have very tight budgets (which might hinder the adoption of costly proenvironmental behaviours) and very specific behavioural patterns heavily influenced by the situational conditions typical of that period of their lives, like living at campuses or eating often in canteens and cafeterias. Hence, replications should be made with samples representative of the overall adult population. The limited time-span of the project is a further limitation. Although a pretty long period in experimental research, six weeks is still a very limited time period to change long established behavioural routines. Hence, future research should cover longer time periods, to study whether even bigger leaps and spillover effects might gradually evolve, or the effects rather dissipate over time. Perhaps the biggest limitation of this study is that behaviours were registered by means of self-reports, which are subject to a number of well-known errors and biases. There is even the risk that the first questionnaire and perhaps even more so the participation in an experiment might have produced a bias in responding to the second questionnaire. In that case, participants might have changed the way they reported their behaviour, rather than the behaviour itself. However, the fact that we did not find effects on all behaviours speaks against biased responding driving our main findings, and so does the consistency of the evidence emerging from this study with other studies mentioned earlier not suffering from this drawback. However, future research on this topic should search for ways of registering behaviour that are less error prone. Our results are broadly consistent with the earlier finding that individuals are more likely to adopt new pro-environmental behaviours that are not costly, and that spillover is more likely to impact low-cost rather than high-cost behaviour (Thøgersen & Crompton, 2009). Most of the reviewed theories predicting spillover also have in common that they predict rather weak effects only. Consistent with this being the case, no spillover effects could be registered on the most taxing or difficult of the included behaviours: volunteering. This is consistent with the proposition that behavioural spillover is limited to activities that are not too difficult and € rfer, 2003). costly to carry out (cf. also Diekmann & Preisendo Similarly, for example, Stern (1992, p. 285) reports that “psychological variables such as attitudes and personal norms appear to have more effect on relatively inexpensive, easy-to-perform energysaving actions.” Hence, the strength of effects that can be obtained via the spillover mechanism is probably similar in size to that which can be obtained by other means aimed at increasing people's environmental awareness (mostly information means), which are rarely strong enough to induce them to adopt more costly behaviours. This implies that there is a risk that when people have performed a few easy tasks towards an environmental goal, they might believe that they have done their fair share (Guagnano, Dietz, &

387

Stern, 1994; Kahneman, Ritov, & Jacowitz, 1993), as suggested by the contribution ethics hypothesis (Thøgersen & Crompton, 2009). Still, it cannot be ruled out that spillover from previously performed pro-environmental behaviours might eventually push people to adopt even more relevant behaviour changes, towards the broad goal of preserving the environment. Once the first small steps sediment and get internalized, becoming the new baseline rather than something more the person does for the environment, they might be ready to make a further step up the ladder. Hence, future research should address this issue by following individuals over a longer time-span. Spillover processes have been investigated and explained with theories focussing on a range of different psychological mechanisms: the activation of broad pro-environmental goals, self-perception inferred from past behaviour, the quest for consistency and learning-by-doing processes. These theories all suggest that there are good reasons to expect spillover between pro-environmental behaviours and that such spillover is mediated through psychological mechanisms making environmental concern or goals more salient and/or self-relevant. This study finds that a campaign that successfully promotes “green” buying behaviour can have additional spillover effects on other pro-environmental behaviours, such as switching off the lights, turning off water or recycling, which is consistent with all of the reviewed theories. However, the presented evidence is mute about the actual mechanism responsible for the spillover effects (i.e., goal priming, self-perception, and/or learning). Hence, this is a question for future research. A possible avenue of research might be to refine predictions about the direction of spillover (positive vs negative) and the target behaviours that are more likely to be affected, from the different theories. For instance, one might speculate that consistency based spillover is most likely to affect behaviours sharing a symbolic similarity with the behaviour from which the spillover originates, whereas learning based spillover is most likely to affect behaviours sharing a functional similarity with the origin behaviour.4 Further, we might speculate that spillover stemming from goal priming is most likely to affect behaviours that are either perceived as very similar to the source behaviour or enjoy a strong generic association to the primed goal in the context (e.g., recycling seems to be perceived as a generic pro-environmental behaviour in Europe and North America). This study is one of the first attempts to empirically investigate the effectiveness of monetary inducements in generating proenvironmental behavioural spillover in a field setting. Contrary to popular beliefs (e.g., Thøgersen & Crompton, 2009), our results suggest that financial inducements are not only more effective than verbal encouragement at promoting the desired behaviour, but also at least no less likely to produce spillover effects on other green behaviours, perhaps the contrary. As regards verbal encouragement, Swim and Bloodhart (2013) also found that praising participants for their pro-environmental behaviour (taking the stairs instead of the elevator) only produced a modest spillover on other, not related “green” behaviours (turning off lights and monitor at the end of the experiment). The authors suggest that the modest result “may be attributable to a lack of internalization of admonishment and praise, as participants may not have felt pride for taking the stairs because their decision was not based upon its environmental or social implications” (Swim & Bloodhart, 2013, p.33). In one of the studies questioning whether behavioural spillover can be expected from campaigns relying on monetary incentives, Evans et al. (2013) found no spillover from carpooling to recycling among participants that were primed with financial benefits of carpooling; only among those primed with environmental benefits.

4

We are grateful to an anonymous reviewer for proposing this prediction.

388

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

However, in their case participants were not actually carpooling, only thinking about it. The difference in results between the two studies suggests that when people actually perform the proenvironmental behaviour over an extended period of time, induced by the monetary compensation and incentive, their attention to the monetary inducement dwindles whereas their attention to the pro-environmental goal of the behaviour increases. Replications of our findings are definitely needed and future research should also investigate more systematically if participants' attention really shifts from the monetary inducement to the goal of the behaviour, as suggested. However, for the time being, our results suggest that the common reservations about potentially negative impacts of monetary inducements on spillover may be exaggerated. It is important to remember that participants in the monetary condition were induced to buy “green” by means of both an incentive (the lottery) and monetary compensation (for the premium price paid for “green” products).5 Notice that the compensation they received to cover the extra costs of buying “green” products was really a compensation only for individuals purchasing no or few “green” products at baseline, whereas for individuals purchasing a lot of “green” products already at baseline it was to a large extent a monetary reward for doing what they already did. This study was not designed to reveal specifically how an incentive and a compensation differ in affecting participants' behaviour, which should be addressed in future research. However, the significant negative interaction between the monetary condition dummy and behaviour at baseline (Table 2) gives an indication of what to expect in terms of relative impacts of the two types of

Green purchasing Organic Milk Organic Potatoes Organic Beef Eco-labelled paper Eco-labelled soap Recycling Glass Paper Plastic Batteries Public transport Bus/train university Bus/train shopping Biking Bike university Bike shopping Carpooling Switching off lights when exiting room as last person Turn off water when brushing teeth Turn off water in shower Printing documents on both sidesa Reading documents on screen Volunteeringa Nature conservation Environmental education International aid Animal care

inducement. Considering that the analysis had already controlled for the regression towards the mean (i.e., the direct effect of behaviour at baseline), the significant interaction shows that those buying relatively little at baseline, and who therefore received a compensation for additional costs, were influenced more than those buying a lot at baseline, and for whom the compensation was mostly a reward for their good behaviour. Although there might be other reasons for this effect, the interaction does suggest that a compensation for the incurred costs from changing behaviour is a more powerful tool than a reward for good behaviour. An interesting question for future research in this connection is whether a compensation framework makes environmental goals more salient than an incentive framework. Previous research suggests that a monetary incentive can activate self-enhancement values that compete with self-transcending goals like promoting the environment (Bolderdijk, Steg, Geller, Lehman, & Postmes, 2013). However, because compensation is about facilitating fairness and equality, perhaps making a compensation salient might promote self-transcending goals (Schwartz, 1994).6 Acknowledgements We are grateful to Greg Maio and two anonymous reviewers for valuable comments to an earlier version of the manuscript. Appendix 1. Self-reported pro-environmental behaviours (mean scores)

Control (time 1)

Control (time 2)

Monetary (time 1)

Monetary (time 2)

Praise (time 1)

Praise (time 2)

2.83 2.79 2.52 1.87 2.73

2.80 2.71 2.56 2.04 2.75

3.39 3.14 2.95 2.42 3.33

4.37 3.25 2.84 2.89 3.42

3.56 2.88 2.85 2.58 3.58

3.27 3.04 2.85 2.60 3.46

4.91 4.47 3.13 4.87

4.87 4.83 3.26 4.79

5.30 4.49 3.28 4.88

5.37 4.81 3.93 5.47

5.56 4.71 3.40 5.08

5.50 4.92 3.79 5.29

3.09 2.45

3.55 2.52

3.46 2.39

3.30 2.07

2.69 1.83

2.81 2.08

5.27 4.53 4.75 6.35

5.07 4.43 4.42 6.17

4.21 4.47 5.16 5.95

4.56 4.77 4.89 6.25

5.17 4.83 4.58 6.10

5.54 4.67 5.17 6.40

6.19 3.35 3.99 3.46

6.21 3.48 3.80 3.49

5.96 3.51 4.12 3.70

5.96 3.77 4.12 3.88

6.42 3.54 4.04 3.25

6.42 3.60 4.29 3.50

1.82 1.65 2.08 1.45

1.70 1.60 2.10 1.52

1.79 1.70 2.14 1.54

1.84 1.67 2.18 1.64

2.02 1.68 2.21 1.58

2.04 1.85 2.15 1.71

Note: a Scale from 1 ¼ I never did it to 5 ¼ I do it regularly, each year. All others: scale from 1 ¼ never to 7 ¼ always.

5 We are grateful to Greg Maio for pointing out this distinction between two types of monetary inducement.

6

Greg Maio suggested this excellent idea for future research.

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

Appendix 2. Shopping diary evidence: average number of items (“green” items) purchased

Milk Yoghurt Eggs Meat Vegetables Fruit Detergents for the house Personal hygiene soaps Kitchen paper Toilet paper

Monetary group

Praise group

Total

8.57 2.57 2.09 6.33 12.52 9.35 1.09 2.02 0.35 0.93

6.18 1.94 1.94 4.88 10.18 7.74 0.65 1.59 0.35 0.85

7.55 2.30 2.03 5.71 11.53 8.66 0.90 1.84 0.35 0.90

(5.70) (1.39) (0.89) (1.22) (4.17) (2.52) (0.39) (0.63) (0.22) (0.50)

(2.09) (0.68) (0.82) (0.44) (1.76) (1.26) (0.26) (0.62) (0.03) (0.24)

(4.16) (1.09) (0.86) (0.89) (3.15) (1.99) (0.34) (0.63) (0.14) (0.39)

References Abelson, R. P. (1968). Psychological implication. In R. P. Abelson, E. Aronson, W. J. McGuire, T. M. Newcomb, M. J. Rosenberg, & P. H. Tannenbaum (Eds.), Theories of cognitive consistency: A sourcebook (pp. 112e139). Chicago: Rand McNally. Aronson, E., & Carlsmith, J. M. (1962). Performance expectancy as a determinant of actual performance. Journal of Abnormal and Social Psychology, 65, 178e182. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Bem, D. J. (1972). Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (pp. 1e62). New York: Academic Press. Berger, I. E. (1997). The demographics of recycling and the structure of environmental behavior. Environment and Behavior, 29, 515e531. € sler, H., & Aiking, H. (2014). “Meatless days” or “less but better”? de Boer, J., Scho Exploring strategies to adapt Western meat consumption to health and sustainability challenges. Appetite, 76, 120e128. Bolderdijk, J. W., Steg, L., Geller, E., Lehman, P., & Postmes, T. (2013). Comparing the effectiveness of monetary versus moral motives in environmental campaigning. Nature Climate Change, 3, 413e416. Burger, J. M. (1999). The foot-in-the-door compliance procedure: A multipleprocess analysis and review. Personality and Social Psychology Review, 3, 303e325. Cann, A., Sherman, S. J., & Elkes, R. (1975). Effects of initial request size and timing of a second request on compliance: The foot-in-the-door and the door-in-the-face. Journal of Personality and Social Psychology, 32, 774e782. Chaiken, S., & Baldwin, M. W. (1981). Affective-cognitive consistency and the effect of salient behavioral information on the self-perception of attitudes. Journal of Personality and Social Psychology, 41, 1e12. Cialdini, R. B., Trost, M. R., & Newsom, J. T. (1995). Preference for consistency: The development of a valid measure and the discovery of surprising behavioral implications. Journal of Personality and Social Psychology, 69, 318e328. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Mahwah, NJ: Lawrence Erlbaum. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/ correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum. Collins, B., & Hoyt, M. (1972). Personal responsibility-for-consequences: An integration and extension of the ‘forced compliance’ literature. Journal of Experimental Social Psychology, 8, 558e593. Cornelissen, G., Pandelaere, M., Warlop, L., & Dewitte, S. (2008). Positive cueing: Promoting sustainable consumer behavior by cueing common environmental behaviors as environmental. International Journal of Research in Marketing, 25(1), 46e55. Defra. (2008). A framework for pro-environmental behaviours. London: DEFRA. De Young, R. (2000). Expanding and evaluating motives for environmentally responsible behavior. Journal of Social Issues, 56(3), 509e526. Dhar, R., & Simonson, I. (1999). Making complementary choices in consumption episodes: Highlighting versus balancing. Journal of Marketing Research, 36, 29e44. Dickerson, C., Thibodeau, R., Aronson, E., & Miller, D. (1992). Using cognitive dissonance to encourage water conservation. Journal of Applied Social Psychology, 22, 841e854. €rfer, P. (1998). Environmental behavior: Discrepancies Diekmann, A., & Preisendo between aspirations and reality. Rationality and Society, 10(1), 79e102. € rfer, P. (2003). Green and greenback: The behavioral Diekmann, A., & Preisendo effects of environmental attitudes in low-cost and high-cost situations. Rationality and Society, 15, 441e472. Evans, L., Maio, G. R., Corner, A., Hodgetts, C. J., Ahmed, S., & Hahn, U. (2013). Selfinterest and pro-environmental behaviour. Nature Climate Change, 3(2), 122e125. http://dx.doi.org/10.1038/nclimate1662. Festinger, L. (1957). A theory of cognitive dissonance. Evanston: Row Peterson.

389

Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58, 203e210. Freedman, J., & Fraser, S. (1966). Compliance without pressure: The foot-in-the-door technique. Journal of Personality and Social Psychology, 4, 195e202. Frey, B. S. (1993). Motivation as a limit to pricing. Journal of Economic Psychology, 14, 635e664. Gneezy, A., Imas, A., Brown, A., Nelson, L. D., & Norton, M. I. (2012). Paying to be nice: Consistency and costly prosocial behavior. Management Science, 58(1), 179e187. http://dx.doi.org/10.1287/mnsc.1110.1437. Guadagno, R. E., Demaine, L. J., & Cialdini, R. B. (2001). When saying yes leads to saying no: Preference for consistency and the reverse foot-in-the-door effect. Personality and Social Psychology Bulletin, 27(7), 859e867. Guagnano, G. A., Dietz, T., & Stern, P. C. (1994). Willingness to pay for public goods: A test of the contribution model. Psychological Science, 5, 411e415. Guagnano, G. A., Stern, P. C., & Dietz, T. (1995). Influences on attitude-behavior relationships. A natural experiment with curbside recycling. Environment and Behavior, 27, 699e718. Holland, R. W., Verplanken, B., & Van Knippenberg, A. (2002). On the nature of attitude-behaviour relations: The strong guide, the weak follow. European Journal of Social Psychology, 32, 869e876. Hughner, R. S., McDonagh, P., Prothero, A., Shultz, C., & Stanton, J. (2007). Who are organic food consumers? A compilation and review of why people purchase organic food. Journal of Consumer Behaviour, 6, 94e110. Kahneman, D., Ritov, I., & Jacowitz, K. E. (1993). Stated willingness to pay for public goods: A psychological perspective. Psychological Science, 4, 310e316. Kaiser, F. G., & Wilson, M. (2004). Goal-directed conservation behavior: The specific composition of a general performance. Personality and Individual Differences, 36, 1531e1544. Kals, E., Schumacher, D., & Montada, L. (1999). Emotional affinity toward nature as a motivational basis to protect nature. Environment & Behavior, 31, 178e202. Kollmuss, A., & Agyeman, J. (2002). Mind the Gap: Why do people act environmentally and what are the barriers to pro-environmental behavior? Environmental Education Research, 8, 239e260. Ludwig, T. D., & Geller, E. S. (1991). Improving the driving practices of pizza deliverers: Response generalization and moderating effects of driving history. Journal of Applied Behavior Analysis, 24, 31e44. http://dx.doi.org/10.1901/ jaba.1991.24-31. Maiteny, P. T. (2002). Mind in the Gap: Summary of research exploring ‘inner’ influences on pro-sustainability learning and behaviour. Environmental Education Research, 8, 299e306. Marian, L., Chrysochou, P., Krystallis, A., & Thøgersen, J. (2014). The role of price as a product attribute in the organic food context: An exploration based on actual purchase data. Food Quality and Preference, 37, 52e60. Mazar, N., & Zhong, C.-B. (2010). Do green products make us better people? Psychological Science, 21, 494e498. Monin, B., & Miller, D. T. (2001). Moral credentials and the expression of prejudice. Journal of Personality and Social Psychology, 81, 33. € Olander, F., & Thøgersen, J. (1995). Understanding of consumer behaviour as a prerequisite for environmental protection. Journal of Consumer Policy, 18, 317e357. € Olander, F., & Thøgersen, J. (2014). Informing versus nudging in environmental policy. Journal of Consumer Policy, 1e16. http://dx.doi.org/10.1007/s10603-0149256-2. Pieters, R. G. M., Bijmolt, T., van Raaij, W. F., & de Kruijk, M. D. (1998). Consumers' attribution of pro-environmental behavior, motivation and ability to self and others. Journal of Public Policy and Marketing, 17, 215e225. Pliner, P., Hart, H., Kohl, J., & Saari, D. (1974). Compliance without pressure: Some further data on the foot-in-the-door technique. Journal of Experimental Social Psychology, 10, 17e22. Poortinga, W., Whitmarsh, L., & Suffolk, C. (2013). The introduction of a single-use carrier bag charge in Wales: Attitude change and behavioural spillover effects. Journal of Environmental Psychology, 36, 240e247. Schwartz, S. H. (1994). Are there universal aspects in the structure and content of human values? Journal of Social Issues, 50(4), 19e45. Scott, C. A. (1977). Modifying socially-conscious behavior: The foot in the door technique. Journal of Consumer Research, 4, 156e163. Snyder, M., & Cunningham, M. R. (1975). To comply or not comply: Testing the selfperception explanation of the “foot-in-the-door” phenomenon. Journal of Personality and Social Psychology, 31, 64e67. Souchet, L., & Girandola, F. (2013). Double foot-in-the-door, social representations, and environment: Application for energy savings. Journal of Applied Social Psychology, 43, 306e315. Spence, A., Leygue, C., Bedwell, B., & O'Malley, C. (2014). Engaging with energy reduction: Does a climate change frame have the potential for achieving broader sustainable behaviour? Journal of Environmental Psychology, 38, 17e28. Staats, H. J., Wit, A. P., & Midden, C. Y. H. (1996). Communicating the greenhouse effect to the public: Evaluation of a mass media campaign from a social dilemma perspective. Journal of Environmental Management, 45, 189e203. Stall-Meadows, C., & Davey, A. (2013). Green marketing of apparel: Consumers' price sensitivity to environmental marketing claims. Journal of Global Fashion Marketing, 4, 33e43. Stern, P. C. (1992). Psychological dimensions of global environmental change. Annual Review of Psychology, 43, 269e302. Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis,10, 349e367. http://dx.doi.org/10.1901/jaba.1977.10-349.

390

P. Lanzini, J. Thøgersen / Journal of Environmental Psychology 40 (2014) 381e390

Susewind, M., & Hoelzl, E. (2014). A matter of perspective: Why past moral behavior can sometimes encourage and other times discourage future moral striving. Journal of Applied Social Psychology, 44, 201e209. Swim, J. K., & Bloodhart, B. (2013). Admonishment and praise: Interpersonal mechanisms for promoting proenvironmental behavior. Ecopsychology, 5, 24e35. Thibodeau, R., & Aronson, E. (1992). Taking a closer look: Reasserting the role of the self-concept in dissonance theory. Personality and Social Psychology Bulletin, 18, 591e602. Thøgersen, J. (1999). Spillover processes in the development of a sustainable consumption pattern. Journal of Economic Psychology, 20, 53e81. Thøgersen, J. (2004). A cognitive dissonance interpretation of consistencies and inconsistencies in environmentally responsible behavior. Journal of Environmental Psychology, 24, 93e103. Thøgersen, J. (2006). Understanding repetitive travel mode choices in a stable context: A panel study approach. Transportation Research Part A: Policy and Practice, 40, 621e638.

Thøgersen, J., & Crompton, T. (2009). Simple and painless? The limitations of spillover in environmental campaigning. Journal of Consumer Policy, 32, 141e163. Thøgersen, J., Haugaard, P., & Olesen, A. (2010). Understanding consumer responses to ecolabels. European Journal of Marketing, 44(11/12), 1787e1810. € Thøgersen, J., & Olander, F. (2003). Spillover of environment-friendly consumer behavior. Journal of Environmental Psychology, 23, 225e236. € Thøgersen, J., & Olander, F. (2006). To what degree are environmentally beneficial choices reflective of a general conservation stance? Environment and Behavior, 38, 550e569. Van der Werff, E., Steg, L., & Keizer, K. (2014). I am what I am, by looking past the present: The influence of biospheric values and past behavior on environmental self-identity. Environment and Behavior, 46, 626e657. Whitmarsh, L., & O'Neill, S. (2010). Green identity, green living? The role of proenvironmental self-identity in determining consistency across diverse proenvironmental behaviours. Journal of Environmental Psychology, 30, 305e314.