ستایش عمومی در مقابل پرداخت خصوصی: اثرات پاداش ها در حفاظت از انرژی در محل کار

Any solution to rising levels of CO2 depends on human behavior. One common approach to changing human behavior is rewarding desired behavior. Because financial incentives often have side effects that diminish efficacy, we predict that social rewards are more effective, because they invoke adherence to descriptive and injunctive social norms. We investigated this by measuring electricity use for 13 weeks at a Dutch firm. Each week, employees were rewarded for conserving energy. They either received monetary rewards (€0–€5) or social rewards (grade points with a descriptive comment). Rewards were either private or public. In both the short and long term, public rewards outperformed private rewards, and social rewards outperformed monetary rewards. This suggests that private monetary rewards, although popular, may be ineffective. Instead, public social rewards may be a more promising approach to stimulating energy conservation. We argue that this approach should be considered more frequently by policy-makers.

Any solution to rising levels of CO2 in the atmosphere critically depends on changing human behavior. Although technological solutions (electric cars, energy-efficient appliances, CFL bulbs) can help reduce CO2-emissions, behavioral changes are necessary to achieve sufficient reductions (Attari et al., 2010, Dietz et al., 2009 and Pacala and Socolow, 2004). Here we focus on changing energy conservation behavior (also termed curtailment behavior, or usage related behavior), and not on the adoption of technology (or efficiency-related change; see Barr et al., 2005 and Gardner and Stern, 2008). Studies investigating how to motivate behavior changes in the environmental domain have focused on household energy use (Schultz et al., 2007) and car use (Graham et al., 2011), frequently using monetary rewards to encourage energy conservation (Abrahamse et al., 2005). Indeed, policymakers seem to prefer an economic cost–benefit analysis approach to energy conservation. However, such financial incentives often have side effects that diminish efficacy (Ariely et al., 2009, Bowles, 2008 and Fehr and Falk, 2002). Moreover, many of the behaviors under scrutiny here are habitual, and such routine behaviors are not easily changed by using financial incentives, which require at least some conscious trade-off to be effective (Barr et al., 2005). We discuss these concerns and investigate whether more psychologically-motivated interventions are more effective (Crompton, 2011). In a field-experiment, we tested whether social rewards lead to better results than a strictly economic approach (Allcott and Mullainathan, 2010). 1.1. Monetary Incentives Despite its popularity, there are several problems with the traditional economic approach of providing monetary rewards to incentivize energy conservation. First, recent evidence strongly suggests that monetary rewards can have detrimental effects (Ariely et al., 2009, Bowles, 2008 and Fehr and Falk, 2002). Monetary incentives may “crowd out” and replace other motivations (Deci et al., 1999 and Frey and Jegen, 2001), such as ethical or moral goals. This is particularly problematic when cost–benefit analyses favor the wrong behavior, as is often the case with energy conservation at the individual level, where small amounts of money are at stake. To illustrate, compare the social request “Please recycle paper” to the monetarily-incentivized request “Please recycle paper, for every pound recycled you will receive a dime”. From an economic perspective, ten cents is better than zero cents, and recycling should be more likely given the second request. However, the monetary reward changes the request from a social/normative request to an economic trade-off, and ten cents may seem insufficient to justify the effort required to recycle. Thus, the monetary incentive may actually decrease the likelihood of recycling. Similarly, offering monetary incentives for energy conservation may change what was initially a moral or social issue (i.e., acting for the greater good) into an economic trade-off with small monetary gains. People may be less likely to conserve energy than if no monetary incentive was offered. Although offering larger monetary incentives may solve some of these problems, they may still lead to, and possibly even exacerbate, the crowding out of intrinsic motivation (Ariely et al., 2009). Of course one could argue that, as long as the incentives are large enough, people will conserve energy. Although this may be true, it represents a costly approach since monitoring behavior and administering rewards involve additional costs. Moreover, there is the question of who will bear these costs. Punishments are another way to encourage conservation. Although punishment may potentially be less costly (even though monitoring and administration costs would be similar), it, too, may crowd out intrinsic motivation (Mulder et al., 2006). Additionally, from a policy standpoint, punishments involve unpopular measures such as taxes or fines. Another problem with monetary incentives is that they can be interpreted as information about other people's intrinsic motivation. People may reason that, if an incentive is necessary, it must be because others would not cooperate without incentives. Such inferences strengthen the Norm of Self-Interest; the belief that other people are mainly motivated by economic goals ( Miller, 1999). People do not want to cooperate if they think that they are the only one cooperating (the ‘Sucker Effect’; Kerr, 1983) or if they think that the effect of their cooperation is too small to matter (the ‘Drop in Bucket Effect’; Larrick and Soll, 2008). Yet another problem with economic approaches is that monetary incentives are commonly part of short-term initiatives or programs that eventually run out of funds. Once people have adapted to behaving correctly because of external reinforcement, ending or even merely reducing that reinforcement can spell the end of the desired behavior. A recent example is the purchase of hybrid cars, the sales of which are strongly correlated with oil prices (Beresteanu and Li, 2011). In some cases, the cessation of external reinforcement can even lead to lower than baseline levels of the desired behavior (Mulder et al., 2006). Moreover, many people do not pay directly for much of the energy they use. In fact, the highest levels of energy use are at the workplace (Kempton et al., 1992), where the employer pays the bill. Even at home, many renters pay a flat rent that includes energy usage. Thus, for many segments of energy consumption, monetary savings are irrelevant and a different approach to encouraging energy conservation is required. In this respect, the situation of energy conservation at work represents a ‘delayed social fence’, in that the costs for the individual (i.e., changing routine behavior) are immediate and salient, whereas the potential benefits (a decrease in CO2 emissions) are very distant in both time and relevance (Kollock, 1998 and Platt, 1973). This constitutes a situation that is difficult to solve, because the temptation to defect (i.e., not changing behavior) in such a social dilemma situation instead of cooperating (i.e., conserving energy) is very strong. Finally, the effectiveness of monetary incentives relies strongly on the involvement of conscious reasoning (or System 2 thinking; see Kahneman, 2003) for making tradeoffs between the effort invested and the money earned by doing so. As has been argued many times before, such extensive conscious reasoning is rare (e.g. Busenitz and Barney, 1997), especially when routine behaviors are involved, we argue that changes in energy conservation behavior are to a large extent governed by bounded rationality (or System 1 thinking). For such System 1 decision-making, information about the behavior of peers may provide important guidelines for behavior. We therefore anticipate that social norms, which provide such guidelines, will have a strong influence on energy conservation behavior. 1.2. Social Norms In contrast to the problems of the economic approach, the social norms approach seems to bear promise (Nolan et al., 2008). In many studies, simply telling people what other people do (providing descriptive social norms) or what is commonly approved or disapproved (providing injunctive social norms) has relatively strong and lasting positive effects on behavior (Abrahamse et al., 2005). In addition to alerting them to what other people do, providing a descriptive social norm tells people what is ‘normal’. An unintended side effect of this can be that people who perform better than the norm may decrease their effort (a boomerang effect). However, this can be overcome by adding an injunctive message indicating that the desired behavior is approved (Schultz et al., 2007). Thus, social norms are a promising approach for incentivizing energy conservation. Performance feedback may be more effective if it is acknowledged by others, or if praise is given when positive goals are reached. Acknowledgement and praise are examples of types of social recognition that may increase intrinsic motivation when provided as information rather than as an attempt to control behavior (Deci et al., 1999). Social recognition communicates an injunctive message; a perception of what is approved or disapproved of within a given culture (Reno et al., 1993). This may also be adopted as a standard from which people do not want to deviate, partly because social recognition may serve as a predictor of desired future rewards (e.g., by fostering a good reputation). As a result, people will engage in behaviors that receive approval and avoid behaviors that lead to disapproval (Bandura, 1997). While social approval may be valued positively because it sometimes generates future benefits, it is believed that most people value social recognition positively (and disapproval negatively) for its own sake (Fehr and Falk, 2002). The provision of social recognition has led to success in behavioral management studies: it led to increased improvements in work performance in the fast-food industry (Peterson and Luthans, 2006) and in an operations division of a credit card company (Stajkovic and Luthans, 2001). Surprisingly, social recognition has rarely been studied in the domain of environmental behavior. One exception is provided by Schultz et al. (2007), who provided feedback on household energy conservation. Social recognition was given with a positively valenced emoticon (☺) or a negatively valenced emoticon (☹). People received a positive emoticon if they consumed less than the average consumption in their neighborhood and a negative emoticon if they consumed more than the average consumption of the neighborhood. Households that received a negatively valenced emoticon tried to obtain a positively valenced emoticon and therefore decreased their consumptions; households that received a positively valenced emoticon tried to maintain that emoticon by keeping their consumption levels low. Schultz et al. (2007) argued that these emoticons were effective because they improved people's compliance with what was considered appropriate and increased or maintained people's motivation to conserve. If people are sensitive to how appropriate their behavior is, their sensitivity to appropriate action may be activated even more strongly if their behavior is made known to relevant others. By providing public feedback, relevant others receive information about each other's behavior. This may yield advantages that are not present when rewards are provided in a private context, as explained below. 1.3. Social Comparison Social comparison theory claims that people have a fundamental desire to evaluate their opinions and abilities and that they strive to have stable, accurate appraisals of themselves (Festinger, 1954). According to Festinger, objective non-social criteria for self-evaluations of ability are hard to come by and, as a consequence, people often compare themselves with others to assess their abilities. Public feedback may thus give more meaning to individual scores and a thorough understanding of one's personal score may activate a desire to improve. A field experiment by Schultz (1999) about the effects of individual versus individual plus group feedback on curbside recycling provides some suggestive evidence for this expectation. Both feedback groups increased recycling. However, the people that also received group feedback (i.e., information about the recycling behavior of their neighbors) continued to increase their behavior from the intervention period to the post-intervention period, whereas behavior change in the individual feedback condition (i.e., information about own recycling behavior only) leveled off. Schultz explained his findings as follows: When feedback was withheld, people that had received comparison information about others' behavior still had a standard against which to compare their behavior, whereas people that received private feedback quickly assumed that they were no longer being watched and reverted to their original behavior patterns. Thus, public feedback can help to establish and maintain normative behavior after feedback is withheld, thereby producing longer-lasting changes than private feedback. 1.4. Goal Framing Theory Another reason to expect that social incentives will be most effective is based on goal-framing theory (Lindenberg and Steg, 2007). The central idea of goal-framing theory is that people have multiple goals that lead them to act: a hedonic goal ‘to feel better now,’ a material goal ‘to guard and improve one's resources,’ and a normative goal ‘to act appropriately’ (Lindenberg, 2001). These goals ‘frame’ what people attend to, what knowledge and attitudes become cognitively most accessible, how people evaluate various aspects of situations, and which alternatives are considered. Goal-framing theory proposes that although behavior is driven by multiple goals that act together, one goal dominates the framing process. Goal-framing theory postulates that, to stimulate pro-social behavior, a normative goal frame should dominate because a normative goal frame often implies acting pro-socially, whereas material and hedonic goal frames often lead people to act in line with individual interests, which in many cases are not in line with collective interest (Lindenberg and Steg, 2007). 1.5. The Current Experiment To compare economic and social/normative rewards we ran a field experiment investigating energy usage in the workplace. Specifically, we compared the effects on energy conservation of monetary rewards and social rewards (a grade point accompanied by an explanatory comment) when the rewards were either private (participants knew only how they themselves performed) or public (participants knew how others in their group performed). We hypothesized that social rewards would outperform monetary rewards, and that public feedback would outperform private feedback. Psychologically, monetary rewards appeal to self-interest (Bowles, 2008), which can be a threat to intrinsic motivation for pro-social behavior (Deci et al., 1999), whereas social rewards can increase motivation to act pro-socially through the activation of normative standards. Thus, we expected that social rewards would have a stronger positive effect on both motivation and actual energy conservation. Similarly, through the activation of social norms (Schultz et al., 2007), we expected that public feedback would have a stronger effect on both motivation and the conservation of energy than private feedback. We expected that social and monetary rewards differentially affect positive emotions. Social rewards can be experienced as more informational and less controlling compared to monetary rewards (Deci et al., 1999). Thus, it is likely that social rewards affirm a feeling of being self-determined and competent to a greater extent than monetary rewards do. We expected that this would make feel people proud and satisfied with their own behavior (Fehr and Falk, 2002). We therefore predicted that social rewards would lead to more positive emotions. Finally, we predicted that public feedback would stimulate people to interact about their scores. Thus, we expected that public (vs. private) feedback would lead to more social interaction about energy conservation.

The results of this study supported our main hypotheses. As predicted, social rewards had a stronger positive effect on energy conservation than monetary rewards, and public feedback had a stronger positive effect on energy conservation than private feedback. The effects of social rewards on energy conservation were strong: eight weeks after we stopped providing feedback and rewards, participants in the social reward conditions still conserved energy. Public social rewards were most effective: during all intervention and post-intervention periods, the public social condition showed the highest energy conservation of all conditions. On average, overall energy conservation in these conditions was 6.4%. The effects of private monetary rewards were also remarkable: From the first manipulation week on, private monetary rewards worked counterproductively. This condition showed more energy consumption than the control condition during the intervention and post-intervention periods, although the difference with baseline consumption was non-significant. Interestingly, there seemed to be an initial positive effect of public monetary rewards on energy conservation: the first week of intervention shows a significant interaction effect between reward type and feedback type (F(1,62) = 5.268, p = .025, η2 = .078) such that the public monetary condition shows a similar increase in energy savings to both social conditions. However, this initially positive effect soon declines (see Fig. 1). It is possible that the decline of the initially positive effect of public monetary effect is caused by negative feedback from colleagues who consider it ‘not done’ to save energy for money. However, since this finding is based on only one of our 11 weeks of measurement, and the effect quickly disappears, caution is warranted to avoid overinterpretation. Future research could investigate this possibility. Our results confirm the predictions of goal-framing theory (Lindenberg and Steg, 2007), which posits that strategies focusing on normative aspects of pro-social behavior may be more effective in stimulating such behavior than strategies focusing on materialistic or hedonic aspects. In our study, not only did social rewards lead to more conservation, conservation also continued after the social rewards were no longer provided. Interestingly, our results are in contrast to people's lay intuitions about what will motivate them. Others have shown that, when asked to rate which factors (e.g., environmental protection, self-interest, etc.) would influence their energy consumption, people estimated that ‘what their neighbors did’ (i.e., the social norm), would be the least influential, whereas it actually was the most influential motivation (Nolan et al., 2008). There is some evidence in the literature that indicates that it is difficult to maintain positive effects of these types of interventions (De Young, 1993; see for a review Abrahamse et al., 2005). However, this picture is not very clear, since, as Abrahamse et al. mention, there is a serious lack of studies that look at long-term effects of these interventions. We consider our study to be a positive addition to the literature in this respect, because we did measure the effects over time of our measure, up until 8 weeks after we stopped our intervention. Our results show that although the expected decline in conservation is there, the effect of reward type remains significant until the final week of our measurements. Even though we cannot know for certain that these changes are permanent, we do consider it likely that for this effect to remain for 8 weeks after the intervention, some change in routine behavior is likely. It should be noted that in the current study, despite the fact that the monetary incentives were relatively minor in absolute terms (participants could earn up to approximately $7.35 a week), this actually represented a hugely inflated reward, since the actual monetary equivalent of their savings by participants were about a factor 100 smaller.2 It is likely that differences between conditions might have been more pronounced (i.e., that the monetary conditions would have scored even lower) had we rewarded people according to the actual energy savings. Even though in this study money saved would have gone to the employer, this indicates that even transferring such savings in their entirety in monetary form to the employees will not lead to positive results, since we found hardly any positive effects when amounts were multiplied by 100. This is important since many (behavioral) energy savings measures lead to only relatively minor monetary savings, and therefore a focus on monetary gains is extra risky in terms of effectiveness of intervention. The results of our study show that social rewards have stronger, more lasting effects on energy conservation than monetary rewards. Further, public rewards led to more conservation than private rewards. Overall, public social rewards were most successful, both in the short run and the long run. Interestingly, despite their popularity with policymakers, private monetary rewards were least successful. 4.1. Limitations One limitation of the current study is that the effects of rewards and feedback were combined in one intervention. In all four experimental conditions, participants received feedback and a reward to stimulate energy conservation. Therefore, we cannot make a comparison with a ‘feedback only condition’. However, since all participants in all experimental conditions received feedback, the differential effects we found of our manipulation of social vs. monetary and public vs. private feedback are not caused by feedback per se. It should be noted that we consciously decided to give priority to a ‘no-feedback’ control condition instead of a ‘feedback only’ condition, since we wanted to make sure we could compare conservation rates to what would have happened without any intervention. A second limitation of the current setup is that we cannot be sure whether the positive effect of public feedback is caused by the fact that these participants knew other people's behavior, or that they knew that others could observe their behavior. Future studies should aim to disentangle these two effects. Another limitation of this study is the quasi-experimental set-up that we used. We could not randomly assign participants to conditions because it was of importance that participants did not communicate with participants in other conditions about manipulation differences. This was a potential threat to internal validity. However, we randomly assigned groups (i.e., departments) to conditions and used a pre-test questionnaire to control for possible differences between participants. Therefore, we ruled out as much of this potential threat as possible. 4.2. Strengths Field studies also have an important advantage: the results can more easily be translated into a useful policy. We showed in a ‘real life’ situation that stimulating employees to conserve energy at work can be effective when social rewards and public feedback are used. Although it may take some time and effort for managers to organize these incentives, we showed that it is an effective and relatively cheap strategy. Providing social rewards for employees' energy conservation does not cost any money, and because employees will likely conserve energy, the provision of public and social rewards reduces a company's energy bill. This is a much more cost-effective strategy than the use of monetary rewards. Monetary rewards are more expensive: the reward itself is more costly and their use is also inefficient because, as we have shown, it is likely that these rewards will not result in an improvement in energy conservation. Other strengths of this experiment are related to the innovative measurement method that we used. The measurement devices that we used gave us the possibility to unobtrusively collect extremely detailed information about peoples' energy consumption. In contrast with most other energy conservation studies that report short-term effects or self-reports (Abrahamse et al., 2005 and Lehman and Geller, 2004), our method enabled us to precisely record actual energy conservation behavior unobtrusively over a prolonged period. Moreover, participants indicated that they believed the feedback that we sent them was reliable and accurate, which indicates that people trusted this new measurement system. 4.3. Generalizability and Implications of This Research The success of an incentive program also depends on the socio-psychological make-up of the population (Samuelson, 1990). The people who participated in this study worked for a company that designed products and processes for a sustainable energy supply. It is likely that they had a relatively positive attitude towards energy conservation compared with employees in other companies. This has implications for the generalizability of the results. It could be argued that the counterintuitive effect of a monetary reward is mainly present in such a group of participants, and that in other groups monetary incentives are more motivating. There are examples abound of monetary incentives as very successful drivers of behavioral change in the environmental domain (Abrahamse et al., 2005 and Stern, 1999). An example of a study that showed positive effects of monetary incentives in certain groups comes from Bashir et al. (2011). They compared Canadian immigrants with materialistic goal frames to non-immigrants who had a relatively less materialistic goal frame about their intentions to engage in pro-environmental behaviors. For immigrants, highlighting the financial aspect of pro-environmental messages had a positive effect on pro-environmental behaviors. This can be seen as an indication of weak generalizability of our findings. However, the non-immigrants in this study were not sensitive to such financial emphasis, which is in line with our reasoning. Future research may give stronger clues as to the boundaries of our findings. Our results confirm that social rewards may be a better approach to changing environmental behavior than financial incentives. Social rewards do not crowd out intrinsic motivation, they have less need for large-scale institutions or exogenous funding, and they work regardless of who is paying the energy bill. We show here that the social reward approach is fairly successful in stimulating cooperation (i.e., energy conservation) in this notoriously difficult delayed social fence-type situation, where costs to the individual are immediate and salient, and potential positive outcomes are delayed and fuzzy. In the environmental domain, many situations resemble such a difficult social dilemma. Thus, we argue that the social norms approach should be considered more frequently as a valuable tool in the intervention tool-kit, especially when focusing on low-cost environmental behavior.