You are on page 1of 7

Journal of Environmental Psychology 92 (2023) 102160

Contents lists available at ScienceDirect

Journal of Environmental Psychology


journal homepage: www.elsevier.com/locate/jep

Pro-environmental behavior in a common-resource dilemma: The role


of beliefs
Annika M. Wyss a, *, Sebastian Berger b, Daria Knoch a
a
Department of Social Neuroscience and Social Psychology, University of Bern, Switzerland
b
Department of Sociology, University of Bern, Switzerland

A R T I C L E I N F O A B S T R A C T

Handling Editor: W. Schultz Despite the overwhelming scientific evidence about the anthropogenic nature of climate change, a vocal minority
continues to spread skepticism. This inhibits pro-environmental action and fosters a false perception of social
reality, leading people to underestimate the pro-environmental intentions and actions of others required to
facilitate rapid and deep decarbonization. Previous efforts to address these beliefs through environmental in­
terventions have yielded inconsistent outcomes. In two studies, we jointly examine the role of first-order climate
change beliefs and beliefs about others’ behavior (Study 1) as well as second-order climate change beliefs (Study
2) in pro-environmental decision-making in controlled laboratory settings. The studies involve a common-
resource dilemma – in which a negative environmental externality is triggered depending on the group’s col­
lective decision-making. Our findings show that while first-order climate change beliefs weakly predict pro-
environmental behavior, second-order climate change beliefs do not correlate with participants’ choices when
accounting for first-order climate change beliefs. However, beliefs about others’ behavior strongly predict
people’s choices. We discuss the results in terms of the role different types of beliefs play in environmental
decision-making.

1. Introduction ignorance (Geiger & Swim, 2016; O’Gorman & J, 1986; Sargent &
Newman, 2021). Therefore, belief in climate change, belief about
Despite scientists’ overwhelming consensus about the anthropogenic others’ belief in climate change, as well as belief about others’ behavior
nature of climate change (IPCC, 2022), a considerable part of society have been particularly strong candidates for psychological in­
remains skeptic about the human responsibility for climate change (e.g., terventions, yielding mixed results (Allcott & Mullainathan, 2010;
27% in the United States, Leiserowitz et al., 2023). This climate skep­ Bergquist et al., 2023; Eyring et al., 2021; Goldberg et al., 2020;
ticism is insofar problematic as people’s climate change beliefs are Jachimowicz et al., 2018; van der Linden et al., 2019). How beliefs affect
routinely seen as a crucial variable influencing pro-environmental in­ behavior therefore remains a crucial research task of environmental
tentions, behavior, and policy support (Berger & Wyss, 2021b; Hornsey psychology.
et al., 2016; Steg, 2023; van Valkengoed et al., 2021). Furthermore, Here, we contribute to this broader research topic through a joint
climate skeptics are often a particularly vocal minority, which may assessment of first-order climate change beliefs, second-order climate
contribute to a false perception of the social reality, with some detri­ change beliefs, and beliefs about others’ behavior and their role in un­
mental behavioral consequences at the social level (Maibach et al., derstanding pro-environmental behavior. Following calls to more
2016). Exacerbating this problem, research has found that people vastly frequently use measures of actual behaviors rather than self-reports
overestimate their fellow citizens’ climate skepticism and strongly un­ (Lange et al., 2023), we rely on a behavioral experimental paradigm
derestimate their support for effective climate policy (Leviston et al., with personal and environmental consequences. In particular, we model
2013; Maibach et al., 2016; Mildenberger & Tingley, 2019; Sparkman pro-environmental behavior as a common-resource dilemma where
et al., 2022). This perception of a false social reality can inhibit one’s environmental consequences co-depend on a decision maker’s actions as
own willingness to combat climate change, consistent with pluralistic well as the decisions of other study participants. Thereby, we mimic an

* Corresponding author. Department of Social Neuroscience and Social Psychology, University of Bern, Bern, Switzerland.
E-mail address: annika.wyss@unibe.ch (A.M. Wyss).

https://doi.org/10.1016/j.jenvp.2023.102160
Received 27 April 2023; Received in revised form 5 September 2023; Accepted 15 October 2023
Available online 16 October 2023
0272-4944/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
A.M. Wyss et al. Journal of Environmental Psychology 92 (2023) 102160

actual commons problem, as climate mitigation hinges on our ability to 5. Measures and procedures
successfully coordinate collective action (S. Chen et al., 2022), as
frequently done in related disciplines such as ecological economics (see 5.1. Assessment of pro-environmental behavior and belief about others’
Lange, 2023, for a review on validated and ad-hoc behavioral behavior
paradigms).
As a central consequential measure of pro-environmental behavior,
2. The present research we relied on a common-resource dilemma, a threshold game that trig­
gers a negative environmental externality conditional on excessive
In the present research, we report on two behavioral studies group consumption. Although such stylized games under controlled
involving an ad-hoc behavioral paradigm to assess pro-environmental conditions tend to oversimplify the complexity of real-world climate
behavior in an intergroup context. Due to the structural similarity be­ behavior (Brekke & Johansson-Stenman, 2008; Tavoni et al., 2011),
tween both studies, we report both studies jointly. Separate analyses for they enable testing critical contingencies of interest (Lange, 2023),
each study – showing qualitatively identical results – are presented in avoid common method bias of purely self-report research (Lange et al.,
the Supplementary Material. The behavioral assessment of pro- 2023), and allow interdisciplinary engagement between social scientific
environmental behavior was accompanied by self-report assessments disciplines (Berger & Wyss, 2021a).
of first- and second-order climate change beliefs, alongside items Upon enrolling in the study via Prolific, participants formed quasi-
assessing support for climate policy and demographics. The overarching random groups of four. Participants were given the opportunity to
research questions tested the relative importance of first-order climate anonymously extract between one and six points out of a common pool
change beliefs, beliefs about others’ behavior, and second-order climate in ten rounds of play, with a monetary reward of 0.10 GBP attached to
change beliefs in predicting pro-environmental behavior. each point taken. Thus, an additional monetary payoff of between 1 GBP
and 6 GBP was possible. Collectively extracting more than 80 points
3. Open science and ethical statement after ten rounds, however, resulted in a detrimental environmental
consequence. If a group stayed under the critical threshold, we pur­
Both studies were implemented using the free software oTree (D. L. chased and retired a certificate of 200 lbs of carbon dioxide from the
Chen et al., 2016), and all data was analyzed using the open source European Union Emission Trading System (EU-ETS). If a group exceeded
software R (R Core Team, 2017). All original experimental code to the threshold, no certificate was purchased and retired. Thus, excessive
replicate the studies, all raw data, and all statistical code to computa­ consumption had an actual environmental consequence of 200 lbs of
tionally reproduce all reported results are available in the associated carbon dioxide emitted to the atmosphere.
OSF project (https://osf.io/6ymwz). For both studies, the data inclusion The dependent variable was the total extraction per individual after
protocol was pre-registered using the As Predicted template within the the first five rounds of play. After round five, the game continued with a
OSF, but the specific correlational pattern was not. The studies received behavior-contingent warning (as in Wyss et al., 2021; and the in­
ethical approval from the University of Bern (protocol number structions in the supplementary material). After the warning or control
2020-08-00006] and were conducted according to the principles message respectively, participants were asked to indicate their current
expressed in the Declaration of Helsinki. All participants were recruited emotional state in terms of five emotions (guilty, worried, hopeful,
via the platform Prolific (Palan & Schitter, 2018), provided written angry, powerless) on a scale from 1 (not at all) to 7 (very much). The
informed consent, and were financially compensated for their partici­ attention check was embedded into this assessment of emotional states,
pation, in part using behavior-contingent payoffs (see below). Neither asking participants to select the option “not at all” in one survey item.
study involved deception. In order to be eligible to participate, pro­ Behavior after the warning, the emotions, and belief updates after the
spective participants had to be at least 18 years of age and fluent in warning are not relevant in the present study and merely served as a
English. means to gather data for potential replication of prior work.
In order to assess participants’ beliefs about others’ behavior, we
4. Pre-registered participant exclusion protocol and statistical asked them to indicate their beliefs about the average extraction rate of
power the other three group members (following their own decisions) after
round two. At the end of the game, participants received feedback about
Across both studies, sample sizes were determined so that samples how much their group had collectively extracted as well as their indi­
exceed the sample size at which correlations stabilize (i.e., 250) vidual extractions and the corresponding bonus payment. They were
(Schönbrodt & Perugini, 2013), alongside budgetary constraints also informed whether their group had surpassed the critical threshold
(Lakens, 2021). In each study, we posted 800 participant slots on the of 80 points, which caused the actual carbon emission. Although par­
platform Prolific. We followed an identical, pre-registered exclusion ticipants interacted in groups, all analyses are performed on an indi­
protocol in both studies. Participants were excluded from the analysis if vidual level, as group members made anonymous and independent
they failed to correctly pass two bot checks, if they failed to correctly decisions, not receiving any feedback whatsoever until round 5.
answer three control questions, or an additional attention check. Finally,
participants who timed out on more than 20% of the extraction decisions 5.2. Questionnaires
(see below) were also excluded from the analysis. Following these
pre-registered exclusion criteria resulted in samples sizes of n = 500 Following the common-resource dilemma, participants continued to
(41.20% female, age: M = 26.58, SD = 9.47) for Study 1, and n = 397 complete a battery of self-report scales. In Study 1, three constructs were
(42.57% female, age: M = 26.36, SD = 8.56) for Study 2. Consequently, administered. First, participants completed the Social Value Scale (SVS;
Study 1 yields 80% power to detect zero-order correlations of r ≥ 0.12 Steg & de Groot, 2012), including egoistic, hedonic, altruistic, and
and 95% power to detect correlations of r ≥ 0.16. Study 2 yields 80% biospheric values. Items were rated on a scale ranging from − 1 (opposed
power to detect zero-order correlations of r ≥ 0.14 and 95% power to to my guiding principles) to 7 (extremely important). Next, we assessed
detect correlations of r ≥ 0.18. All power analyses relied on the “pwr” participants’ belief in climate change with five items (Lewandowsky
package within R. et al., 2013) using a scale ranging from 1 (strongly disagree) to 5
(strongly agree). To measure participants’ willingness to accept climate
policies, we included and adapted items from Nisbet et al. (2013). All
eight items were rated on a scale from 1 (strongly disagree) to 7
(strongly agree). Finally, participants completed a series of demographic

2
A.M. Wyss et al. Journal of Environmental Psychology 92 (2023) 102160

questions, which included information about their gender, age, level of beliefs about others behavior (r = − 0.15, p < .001). Income and edu­
highest education, employment, household income, as well as their cation correlated with each other (r = .18, p < .001), but not with any
political orientation on the liberal-conservative spectrum, ranging from other measure. Political views were further correlated with personal
1 (very liberal) to 7 (very conservative). climate change beliefs (r = − 0.25, p < .001), policy support (r = − 0.23,
In Study 2, we additionally included a scale to measure second-order p < .001), and weakly with pro-environmental behavior (r = 0.09, p <
climate change beliefs by adapting the first-order climate change beliefs .05), but not with any other measure.
questionnaire. Instead of reporting their own beliefs, prompts were re-
phrased to target second-order climate change beliefs. The items were 7. The relative importance of beliefs in predicting pro-
preceded by the sentence: “To the best of your knowledge, how much on environmental behavior
average do participants in this study agree with the statement: (…)”.
Note that in Study 2, first and second-order climate change beliefs as As a next step, we aimed to assess the (relative) importance of both
well as climate policy acceptance were assessed before the common- types of beliefs in predicting participants’ pro-environmental behavior.
resource dilemma in randomized order. To do so, we included our predictors of interest in a multiple regression
model and assessed their relative importance. Table 2 (Models 1 to 3)
6. Results shows that belief in climate change and beliefs about others’ behavior
were both uniquely linked to participants’ extraction decisions. Second-
6.1. Behavioral results and zero order correlations order climate change beliefs, on the other hand, did not predict pro-
environmental choices above and beyond beliefs about others’
The results are based on a total of 897 participants (Study 1: n = 500; behavior and first-order climate change beliefs (Models 2 to 3). In order
Study 2: n = 397), following the pre-registered exclusion protocol dis­ to estimate the variance explained by each individual predictor variable,
played above. During the first five rounds of play, participants extracted we applied the “lmg” method (Lindeman et al., 1980) using the statis­
on average 2.44 (SD = 1.4) points per round (see Fig. 1). They expected tical package “relaimpo” (Grömping, 2006) in R (R Core Team, 2017).
the average extraction behavior of their group members to be 2.92 (SD “Lmg” calculates the relative contribution of each individual predictor
= 1.2) points per round. Thus, people extracted less than they expected to the R2 taking into account the order in which they appear in the
others to do (exploratory paired t-test: t(897) = 12.37, p < .001). Ulti­ regression model.
mately, 496 participants were part of a group that collectively exceeded As displayed in Fig. 2A, the relative contribution of belief about
the critical threshold of 80 points. Thus, behavior in both studies others’ behavior was 91.80% (95% CI [85.09%, 96.62%]) compared to
resulted in a total carbon emission of 11.2 tons. 8.20% (95% CI [3.38%, 14.91%]) for first-order climate change beliefs
Table 1 displays the zero-order correlations. Participants’ beliefs (total R2 of all predictors = 36.04%). This difference was statistically
about others’ behavior and their first and second-order climate change significant (Δbeliefs = 83.60%, 95% CI [70.57%, 93.95%]). Fig. 2B
beliefs significantly correlated with their extraction behavior. More shows that when including second-order climate change beliefs, the
specifically, beliefs about others’ behavior were strongly correlated with relative contribution of belief about others’ behavior was 86.74% (95%
pro-environmental behavior (r = 0.59, p < .001), whereas the correla­ CI [72.27%, 95.95%]) compared to 10.03% (95% CI [3.25%, 21.76%])
tions between first and second-order climate change beliefs and pro- for first-order and 3.23% (95% CI [0.64%, 9.57%]) for second-order
environmental behavior were weaker (first-order: r = − 0.21, p < .001; climate change beliefs (total R2 of all predictors = 31.5%). The differ­
second-order: r = − 0.16, p < .01). Second-order climate change beliefs ence in contributions between first-order climate change beliefs and
correlated with beliefs about others’ behavior (r = − 0.14, p < .01). belief about others’ behavior (Δbeliefs = 76.71%, 95% CI [49.55%,
Extraction behavior was also correlated with biospheric values (r = 92.57%]), as well as the difference between second-order climate
− 0.30, p < .001), which were significantly linked to all types of beliefs change beliefs and beliefs about others’ behavior (Δbeliefs = 83.52%,
(about others’ behavior: r = − 0.18, p < .001; first-order: r = 0.31, p < 95% CI [64.47%, 94.57%]) were statistically significant. However, there
.001; second-order: r = 0.18, p < .001) as well as policy support (r = was no statistically significant difference between first- and second-
0.43, p < .001) and political views (r = − 0.10, p < .01). order climate (Δbeliefs = 6.80%, 95% CI [− 2.25%, 18.59%]). Thus,
Policy support was further correlated with first-order climate change participants’ beliefs about the behavior of other group members
beliefs (r = 0.60, p < .001), own behavior (r = − 0.19, p < .001), and explained more variance in pro-environmental behavior than their first-
and second-order climate change beliefs.

8. General discussion

In the present research, we investigated the role and relative


importance of first-order climate change beliefs, second-order climate
change beliefs, and beliefs about others’ behavior on pro-environmental
behavior in a common-resource dilemma game. In terms of zero-order
correlations, all three forms of beliefs correlated with pro-
environmental (i.e., individual consumption) behavior. Regarding the
relative importance of the three forms of beliefs, beliefs about others
behavior were the variable that most strongly correlated with behavior
and hence explained most of the variance. Second-order climate change
beliefs, on the other hand, did not predict pro-environmental decisions
above and beyond first-order climate change beliefs as well as beliefs
about others’ behavior.

8.1. Contribution to environmental psychology

Taken together, the results contribute to a better understanding of


the factors that relate to pro-environmental behavior. First, the fact that
Fig. 1. Mean extraction rate (collapsed data). beliefs about others’ behavior correlate quite strongly with pro-

3
A.M. Wyss et al. Journal of Environmental Psychology 92 (2023) 102160

Table 1
Zero order correlations between main variables (collapsed data).
Variable M SD 1 2 3 41 5 6 7 8 ω2
1. Extraction behavior 2.44 1.40

2. Beliefs about others’ behavior 2.93 1.20 .59***


[.54, .63]

3. 1st order climate change 4.21 0.60 − .21*** − .14*** 0.77/


belief 0.76
[-.27, [-.21,
− .14] − .08]

4. 2nd Order Climate change 3.90 0.49 − .16** − .14** .49*** 0.68
belief1
[-.25, [-.24, [.41, .56]
− .06] − .04]

5. Biospheric values 5.41 1.26 − .30*** − .18*** .31*** .18*** 0.92/


0.92
[-.36, [-.25, [.25, .37] [.08, .27]
− .24] − .12]

6. Policy acceptance 5.17 0.90 − .19*** − .15*** .60*** .27*** .43*** 0.86/
0.83
[-.25, [-.21, [.56, .64] [.18, .36] [.37, .48]
− .12] − .09]

7. Education 1.75 0.93 − .05 − .02 .01 − .02 .07* .08*


[-.12, .01] [-.09, .04] [-.05, .08] [-.12, [.01, .14] [.02, .15]
.08]

8. Income 2.54 2.34 .00 .02 − .04 .10 − .08* − .02 .18***
[-.06, .07] [-.05, .08] [-.10, .03] [.00, .20] [-.14, [-.09, .04] [.11, .24]
− .01]

9. Political views 2.30 1.87 .09* .03 − .25*** − .08 − .10** − .23*** − .03 .04
[.02, .15] [-.03, .10] [-.32, [-.18, [-.17, [-.29, [-.09, [-.03,
− .19] .02] − .04] − .17] .04] .10]

Note. Correlation coefficients represent Pearson’s r. Education and political views are treated as numerical variables, with higher values representing higher levels of
education or more conservative views, respectively. Values in square brackets indicate the 95% confidence interval. All correlations based on n = 897 except from
correlations with second order climate change beliefs, where n1 = 397. * indicates p < .05. ** indicates p < .01. *** indicates p < .001. 2First value refers to Study 1, and
second number to Study 2.

Table 2
Multiple regression models predicting mean extraction behavior.
Predictors Model 1 Model 2 Model 3

B CI p B CI p B CI p

(Intercept) 2.13 1.50–2.75 <0.001 1.94 0.91–2.96 <0.001 2.01 0.93–3.10 <0.001
1st order climate change beliefs − 0.31 − 0.43–− 0.18 <0.001 − 0.25 − 0.45–− 0.04 0.017 − 0.24 − 0.45–− 0.03 0.025
2nd order climate change beliefs − 0.08 − 0.32–0.17 0.548 − 0.09 − 0.34–0.15 0.464
Belief about others’ behavior 0.66 0.60–0.72 <0.001 0.58 0.48–0.67 <0.001 0.57 0.48–0.67 <0.001
Study − 0.24 − 0.39–− 0.09 0.002
Income − 0.03 − 0.07–0.02 0.240
Political views 0.02 − 0.04–0.08 0.487

Observations 897 397 396


R2/R2 adjusted 0.367/0.365 0.315/0.310 0.323/0.315

Note. The dependent variable is participants’ extraction behavior in the threshold public goods game during the first five rounds of play, with higher values reflecting
less pro-environmental behavior. Model 1 is based on collapsed data whereas Model 2 and Model 3 are based on data from Study 2 only, as second order climate change
beliefs were not assessed in Study 1.

environmental behavior may inform research about the relevant ante­ 2002; Nielsen et al., 2022; Wyss et al., 2022; Zhang et al., 2021). The
cedents of pro-environmental behavior. Research has found that indi­ strong correlation between belief about others’ behavior and
vidual values, climate change beliefs, or other individual factors pro-environmental choices may contribute to the fact that individual
sometimes stand only in a modest relationship with actual pro- motivation and behavior are often misaligned. If people systematically
environmental behavior (see, e.g., the broad literature on the environ­ underestimate other people’s support for climate mitigation, as is re­
mental attitude-behavior gap, Farjam et al., 2019; Kollmuss & Agyeman, ported in various research (Sparkman et al., 2022), this could explain an

4
A.M. Wyss et al. Journal of Environmental Psychology 92 (2023) 102160

Fig. 2. Relative Importance (Percentage in R2) for Extraction Behavior


Note. A) Based on collapsed data (n = 896), Total R2 = 36.04%. B) Based on data from Study 2 (n = 396), Total R2 = 31.5%. LMG Method was applied, vertical lines
indicate 95% bootstrap confidence intervals. Metrices are normalized to sum 100%.

own reluctance to engage in costly pro-environmental behavior despite e., self-reported intentions or past behaviors) or in field settings without
one’s own climate change beliefs or environmental values – in addition strict control about behavior and contexts under which the behavior
to variables such as working memory (Langenbach et al., 2020), occurs (Jachimowicz et al., 2018). Contributing direct behavioral evi­
self-control (Wyss et al., 2022), or behavioral costs (Berger & Wyss, dence through stylized behavioral paradigms supplements the existing
2021c; Kaiser & Lange, 2021). Second, and quite surprisingly, work in a meaningful way (Lange et al., 2023). For example, in a
second-order climate change beliefs did not predict pro-environmental discipline that routinely relies on self-reports (Lange & Dewitte, 2019),
choices above and beyond first-order climate change beliefs and be­ adding behavioral data about antecedents of pro-environmental
liefs about others’ behavior, inconsistent with recent theorizing (e.g., behavior may help to provide a more multifaceted evidence base that
Eyring et al., 2021; Goldberg et al., 2020) and empirical findings (e.g., can increase our confidence in the robustness of correlational patterns.
Jachimowicz et al., 2018; Schuldt et al., 2019; van der Linden et al., Furthermore, adopting methods that are used by related disciplines (e.
2019). However, research suggests that second-order climate change g., economics, political science) may foster interdisciplinary exchange
beliefs may have a stronger effect on peoples’ willingness to engage in (Berger & Wyss, 2021a; Lange et al., 2023).
pro-environmental action if the belief is held by closer and more influ­
ential others (Goldberg et al., 2020; Schuldt et al., 2019). Thus, it is
8.2. Limitations
possible that second-order climate change beliefs are in fact not as
relevant as previously thought, but also that our contradictory finding
Turning to some limitations, two distinctive elements emerge. First,
may be explained by the fact that we assessed second-order climate
studying critical contingencies of pro-environmental behavior through
change beliefs of an unknown reference group (i.e., anonymous partic­
highly abstract behavioral games necessarily comes at the cost of losing
ipants on Prolific). Third, the fact that belief in climate change is only
some of the realism that is typically present in naturally occurring be­
moderately linked to private pro-environmental behavior in prior
haviors (Tavoni et al., 2011). This artificiality may lower the external
research (Hornsey et al., 2016) may underestimate the true behavioral
validity if the chosen paradigm misses some of the crucial functional
consequences of climate skepticism. If climate skepticism also operates
drivers of real-world behavior (Lange et al., 2023). A second relevant
via belief about others’ behavior, this would necessarily imply lower
limitation is the correlational nature of present study. Here, we
correlation between one’s own beliefs and behavior (Eyring et al.,
measured three forms of beliefs and pro-environmental behavior, and as
2021).
a result provide merely correlational patterns between the four assessed
Another contribution of the present work is methodological in na­
variables. Although theoretical models (e.g., Battigalli & Dufwenberg,
ture. Research on second-order climate change beliefs has thus far been
2009) and related empirical results on belief-dependent behavior (All­
largely conducted through a collection of verbal behaviors in surveys (i.
cott, 2011; Andre et al., 2022; Berger et al., 2019; Fischbacher et al.,

5
A.M. Wyss et al. Journal of Environmental Psychology 92 (2023) 102160

2001) support a causal path from climate change beliefs or beliefs about Chen, D. L., Schonger, M., & Wickens, C. (2016). oTree—an open-source platform for
laboratory, online, and field experiments. Journal of Behavioral and Experimental
others behavior to one’s own behavior, other results raise some caution
Finance, 9, 88–97. https://doi.org/10.1016/j.jbef.2015.12.001
(see Bellmare & Sebald, 2022, and the references therein). For example, Chen, S., Wan, F., & Yang, S. (2022). Normative misperceptions regarding pro-
cognitive dissonance theory (Festinger, 1957) or the effects surrounding environmental behavior: Mediating roles of outcome efficacy and problem
the false consensus effect (Ross et al., 1977) provide conceptual awareness. Journal of Environmental Psychology, 84, Article 101917. https://doi.org/
10.1016/j.jenvp.2022.101917
frameworks that allow beliefs to follow from behavior, rather than the Eyring, V., Mishra, V., Griffith, G. P., Chen, L., Keenan, T., Turetsky, M. R., Brown, S.,
other way around. Jotzo, F., Moore, F. C., & van der Linden, S. (2021). Reflections and projections on a
decade of climate science. Nature Climate Change, 11(4), 279–285. https://doi.org/
10.1038/s41558-021-01020-x
9. Conclusion Farjam, M., Nikolaychuk, O., & Bravo, G. (2019). Experimental evidence of an
environmental attitude-behavior gap in high-cost situations. Ecological Economics,
As the urgent transformation of service provisioning systems requires 166, Article 106434. https://doi.org/10.1016/j.ecolecon.2019.106434
Fischbacher, U., Gächter, S., & Fehr, E. (2001). Are people conditionally cooperative?
a profound understanding of the complex interaction of individual, so­ Evidence from a public goods experiment. Economics Letters, 71(3), 397–404.
cial, and structural factors surrounding choice (Creutzig et al., 2022), an https://doi.org/10.1016/S0165-1765(01)00394-9
improved understanding of the factors that promote or inhibit individ­ Geiger, N., & Swim, J. K. (2016). Climate of silence: Pluralistic ignorance as a barrier to
climate change discussion. Journal of Environmental Psychology, 47, 79–90. https://
ual willingness to engage in pro-environmental behavior remains a doi.org/10.1016/j.jenvp.2016.05.002
crucial research topic in environmental psychology. In the present Goldberg, M. H., Gustafson, A., & van der Linden, S. (2020). Leveraging social science to
research, we show that beliefs about others’ behavior are strongly generate lasting engagement with climate change solutions. One Earth, 3(3),
314–324. https://doi.org/10.1016/j.oneear.2020.08.011
correlated with pro-environmental choices, whereas first- and second-
Grömping, U. (2006). Relative importance for linear regression in R: The package
order climate change beliefs are not. This may critically augment prior relaimpo. Journal of Statistical Software, 17(1). https://doi.org/10.18637/jss.v017.
work and inspire new work to disentangle how beliefs about behavior i01
and beliefs about beliefs are interrelated and how they functionally Hornsey, M. J., Harris, E. A., Bain, P. G., & Fielding, K. S. (2016). Meta-analyses of the
determinants and outcomes of belief in climate change. Nature Climate Change, 6(6),
affect pro-environmental behavior. 622–626. https://doi.org/10.1038/nclimate2943
IPCC. (2022). Global Warming of 1.5◦ C: IPCC special Report on Impacts of global Warming of
Declaration of competing interest 1.5◦ C above pre-industrial Levels in Context of strengthening Response to climate change,
sustainable development, and Efforts to eradicate poverty (1. Aufl.). Cambridge
University Press. https://doi.org/10.1017/9781009157940
We have no known conflict of interest to disclose. Daria Knoch Jachimowicz, J. M., Hauser, O. P., O’Brien, J. D., Sherman, E., & Galinsky, A. D. (2018).
gratefully acknowledges the financial support from the Typhaine The critical role of second-order normative beliefs in predicting energy conservation.
Nature Human Behaviour, 2(10), 757–764. https://doi.org/10.1038/s41562-018-
Foundation. 0434-0
Kaiser, F. G., & Lange, F. (2021). Offsetting behavioral costs with personal attitude:
Author statement Identifying the psychological essence of an environmental attitude measure. Journal
of Environmental Psychology, 75, Article 101619. https://doi.org/10.1016/j.
jenvp.2021.101619
Annika M. Wyss: Conceptualization, Methodology, Analysis, Kollmuss, A., & Agyeman, J. (2002). Mind the Gap: Why do people act environmentally
Writing, Reviewing and Editing. Sebastian Berger: Conceptualization, and what are the barriers to pro-environmental behavior? Environmental Education
Research, 8(3), 239–260. https://doi.org/10.1080/13504620220145401
Methodology, Writing, Reviewing and Editing. Daria Knoch: Supervi­
Lakens, D. (2021). Sample size justification [preprint]. PsyArXiv. https://doi.org/
sion, Conceptualization, Methodology, Reviewing and Editing, Funding 10.31234/osf.io/9d3yf
acquisition. Lange, F. (2023). Behavioral paradigms for studying pro-environmental behavior: A
systematic review. Behavior Research Methods. https://doi.org/10.3758/s13428-022-
01825-4
Appendix A. Supplementary data Lange, F., Berger, S., Byrka, K., Brügger, A., Henn, L., Sparks, A. C., Nielsen, K. S., &
Urban, J. (2023). Beyond self-reports: A call for more behavior in environmental
Supplementary data to this article can be found online at https://doi. psychology. Journal of Environmental Psychology, 86, Article 101965. https://doi.
org/10.1016/j.jenvp.2023.101965
org/10.1016/j.jenvp.2023.102160. Lange, F., & Dewitte, S. (2019). Measuring pro-environmental behavior: Review and
recommendations. Journal of Environmental Psychology, 63, 92–100. https://doi.org/
References 10.1016/j.jenvp.2019.04.009
Langenbach, B. P., Berger, S., Baumgartner, T., & Knoch, D. (2020). Cognitive resources
moderate the relationship between pro-environmental attitudes and green behavior.
Allcott, H. (2011). Social norms and energy conservation. Journal of Public Economics, 95
Environment and Behavior, 52(9), 979–995. https://doi.org/10.1177/
(9–10), 1082–1095. https://doi.org/10.1016/j.jpubeco.2011.03.003
0013916519843127
Allcott, H., & Mullainathan, S. (2010). Behavior and energy policy. Science, 327(5970),
Leiserowitz, A., Maibach, E., Rosenthal, S. A., Kotcher, J. E., Carman, J., Verner, M.,
1204–1205. https://doi.org/10.1126/science.1180775
Lee, S., Ballew, M. T., Uppalapati, S., Campbell, E., Myers, T., Goldberg, M. H., &
Andre, P., Boneva, T., Chopra, F., & Falk, A. (2022). Misperceived social norms and
Marlon, J. (2023). Climate change in the American mind: Beliefs & attitudes, December
willingness to act against climate change. Econtribute Discussion Paper.
2022. Yale University and George Mason University.
Battigalli, P., & Dufwenberg, M. (2009). Dynamic psychological games. Journal of
Leviston, Z., Walker, I., & Morwinski, S. (2013). Your opinion on climate change might
Economic Theory, 144(1), 1–35. https://doi.org/10.1016/j.jet.2008.01.004
not be as common as you think. Nature Climate Change, 3(4), 334–337. https://doi.
Berger, S., Christandl, F., Bitterlin, D., & Wyss, A. M. (2019). The social insectivore: Peer
org/10.1038/nclimate1743
and expert influence affect consumer evaluations of insects as food. Appetite, 141,
Lewandowsky, S., Gignac, G. E., & Oberauer, K. (2013). The role of conspiracist ideation
Article 104338. https://doi.org/10.1016/j.appet.2019.104338
and worldviews in predicting rejection of science. PLoS One, 8(10), Article e75637.
Berger, S., & Wyss, A. M. (2021a). Measuring pro-environmental behavior using the
https://doi.org/10.1371/journal.pone.0075637
carbon emission task. Journal of Environmental Psychology, 75, Article 101613.
Lindeman, R. H., Merenda, P. F., & Gold, R. Z. (1980). Introduction to bivariate and
https://doi.org/10.1016/j.jenvp.2021.101613
multivariate analysis. Foresman: Scott.
Berger, S., & Wyss, A. M. (2021b). Climate change denial is associated with diminished
van der Linden, S., Leiserowitz, A., & Maibach, E. (2019). The gateway belief model: A
sensitivity in internalizing environmental externalities. Environmental Research
large-scale replication. Journal of Environmental Psychology, 62, 49–58. https://doi.
Letters, 16(7), Article 074018. https://doi.org/10.1088/1748-9326/ac08c0
org/10.1016/j.jenvp.2019.01.009
Berger, S., & Wyss, A. M. (2021c). Climate change denial is associated with diminished
Maibach, E., Woods Placky, B., Witte, J., Seitter, K., Gardiner, N., Myers, T., Sublette, S.,
sensitivity in internalizing environmental externalities. Environmental Research
& Cullen, H. (2016). Oxford research encyclopedia of climate science. Oxford University
Letters, 16(7), Article 074018. https://doi.org/10.1088/1748-9326/ac08c0
Press. https://doi.org/10.1093/acrefore/9780190228620.013.505
Bergquist, M., Thiel, M., Goldberg, M. H., & van der Linden, S. (2023). Field
Mildenberger, M., & Tingley, D. (2019). Beliefs about climate change beliefs: The
interventions for climate change mitigation behaviors: A second-order meta-
importance of second-order opinions for climate politics. British Journal of Political
analysis. Proceedings of the National Academy of Sciences, 120(13), Article
Science, 49(4), 1279–1307. https://doi.org/10.1017/S0007123417000321
e2214851120. https://doi.org/10.1073/pnas.2214851120
Nielsen, K. S., Brick, C., Hofmann, W., Joanes, T., Lange, F., & Gwozdz, W. (2022). The
Brekke, K. A., & Johansson-Stenman, O. (2008). The behavioural economics of climate
motivation–impact gap in pro-environmental clothing consumption. Nature
change. Oxford Review of Economic Policy, 24(2), 280–297. https://doi.org/10.1093/
Sustainability, 5(8), 665–668. https://doi.org/10.1038/s41893-022-00888-7
oxrep/grn012
Nisbet, E. C., Hart, P. S., Myers, T., & Ellithorpe, M. (2013). Attitude change in
competitive framing environments? Open-/Closed-Mindedness, framing effects, and

6
A.M. Wyss et al. Journal of Environmental Psychology 92 (2023) 102160

climate change: Framing and open-mindedness. Journal of Communication, 63(4), Steg, L. (2023). Psychology of climate change. Annual Review of Psychology, 74(1),
766–785. https://doi.org/10.1111/jcom.12040 391–421. https://doi.org/10.1146/annurev-psych-032720-042905
O’Gorman, H., & J. (1986). The discovery of pluralistic ignorance: An ironic lesson. Steg, L., & de Groot, J. I. M. (2012). Environmental values. Oxford University Press.
Journal of the History of the Behavioral Sciences, 22(4), 333–347. https://doi.org/10.1093/oxfordhb/9780199733026.013.0005
Palan, S., & Schitter, C. (2018). Prolific.ac—a subject pool for online experiments. Tavoni, A., Dannenberg, A., Kallis, G., & Loschel, A. (2011). Inequality, communication,
Journal of Behavioral and Experimental Finance, 17, 22–27. https://doi.org/10.1016/j. and the avoidance of disastrous climate change in a public goods game. Proceedings
jbef.2017.12.004 of the National Academy of Sciences, 108(29), 11825–11829. https://doi.org/
R Core Team. (2017). R: A language and environment for statistical com-puting. 10.1073/pnas.1102493108
Ross, L., Greene, D., & House, P. (1977). The “false consensus effect”: An egocentric bias van Valkengoed, A. M., Steg, L., & Perlaviciute, G. (2021). Development and validation
in social perception and attribution processes. Journal of Experimental Social of a climate change perceptions scale. Journal of Environmental Psychology, 76, Article
Psychology, 13(3), 279–301. https://doi.org/10.1016/0022-1031(77)90049-X 101652. https://doi.org/10.1016/j.jenvp.2021.101652
Sargent, R. H., & Newman, L. S. (2021). Pluralistic ignorance research in psychology: A Wyss, A. M., Berger, S., Baumgartner, T., & Knoch, D. (2021). Reactions to warnings in
scoping review of topic and method variation and directions for future research. the climate commons. Journal of Environmental Psychology, 78, Article 101689.
Review of General Psychology, 25(2), 163–184. https://doi.org/10.1177/ https://doi.org/10.1016/j.jenvp.2021.101689
1089268021995168 Wyss, A. M., Knoch, D., & Berger, S. (2022). When and how pro-environmental attitudes
Schönbrodt, F. D., & Perugini, M. (2013). At what sample size do correlations stabilize? turn into behavior: The role of costs, benefits, and self-control. Journal of
Journal of Research in Personality, 47(5), 609–612. https://doi.org/10.1016/j. Environmental Psychology, 79, Article 101748. https://doi.org/10.1016/j.
jrp.2013.05.009 jenvp.2021.101748
Schuldt, J. P., Yuan, Y. C., Song, Y., & Liu, K. (2019). Beliefs about whose beliefs? Zhang, Y., Bai, X., Mills, F. P., & Pezzey, J. C. V. (2021). Examining the attitude-behavior
Second-Order beliefs and support for China’s coal-to-gas policy. Journal of gap in residential energy use: Empirical evidence from a large-scale survey in
Environmental Psychology, 66, Article 101367. https://doi.org/10.1016/j. Beijing, China. Journal of Cleaner Production, 295, Article 126510. https://doi.org/
jenvp.2019.101367 10.1016/j.jclepro.2021.126510
Sparkman, G., Geiger, N., & Weber, E. U. (2022). Americans experience a false social
reality by underestimating popular climate policy support by nearly half. Nature
Communications, 13(1), 4779. https://doi.org/10.1038/s41467-022-32412-y

You might also like