Social Justice Research

, Volume 26, Issue 3, pp 301–319 | Cite as

Making Sense of the Senseless: Identity, Justice, and the Framing of Environmental Crises

Article

Abstract

Responses to environmental crises will depend on the way in which these events are understood and characterized, perceptions that may be affected by media frames as well as by individual motivations. This paper reports on two studies looking at the role of justice and framing of environmental problems. In Study 1, 297 participants were asked to characterize the explosion of the Deepwater Horizon as an injustice, a crime, or a natural disaster following a description of the event that focused on one of several different types of harm. They also rated harm caused, responsibility for the harm, and their own affective response. In Study 2, 387 participants read a paragraph about climate change that focused on one of several targets of harm and then rated the threat of climate change, responsibility for addressing climate change, and affective response. In both studies, general belief in a just world was associated with weaker negative affect, whereas environmental identity and a liberal political orientation were associated with stronger responses. Business and industry were seen as primarily responsible for both causing and remediating the problems. Framing the issue had a limited influence. The results suggest that political differences in environmental concern are associated with different characterizations of environmental crises and that a desire for justice can both facilitate and hamper pro-environmental responses.

Keywords

Belief in a just world Environmental identity Political orientation Framing Climate change Environmental attitudes 

We live in challenging times. Human impacts on the natural environment present us with changes, the possible magnitude of which is so immense that some have suggested that the era we live in should be called the “anthropocene” (Crutzen & Stoermer, 2000) and the possible outcome of which is so devastating that people may be driven to escalating levels of denial (cf. Vess & Arndt, 2008). How can we reconcile this state of affairs with our desire to see the world as a fair place? And, how do beliefs about justice affect the way in which we respond to environmental crises?

Environmental issues present a compelling context for examining perceptions of justice (Clayton & Opotow, 2003; Kahn, 1997). Increasingly, environmental changes restrict access to environmental resources, some of which were previously considered for all practical purposes infinite. Thus, fossil fuels, clean water, the open ocean, and places to dump garbage are not just things to be taken for granted, but finite quantities that must be shared with other groups. This raises questions about the fairness of the distribution and the fairness of the process (Syme & Nancarrow, 2012). It also raises questions about inclusion: Whose needs must be considered in the allocation of natural resources? Future generations? Non-human animals? Ecosystems? And, it highlights questions of responsibility: Should those who caused the problem be responsible for fixing it? Or does everyone need to participate? In comparison to questions about the distribution of financial resources, these issues are relatively unexamined. In the absence of cultural consensus about what is “right” or “fair,” there is more room for individual differences in responses.

Perceptions of justice regarding environmental conflicts are important to investigate because they may have a practical impact on people’s behavior (Gross, 2008; Kals & Russell, 2001; Müller, Kals, & Maes, 2008). Many conservation measures are ineffective because local residents do not abide by rules of environmental protection, behavior that is encouraged when policies are perceived to be inequitable or unfair (Keane, Jones, Edwards-Jones, & Milner-Gulland, 2008). Protecting environmental resources requires consideration of perceived fairness. Perceived fairness, however, is affected by individual motivations and by media portrayals as well as by the actual attributes of a policy. The present studies were designed to investigate perceptions of justice in regard to a specific, concrete environmental problem—the explosion of the Deepwater Horizon in the Gulf of Mexico (Study 1)—as well as to a more abstract and diffuse problem, that of global climate change (Study 2), and to examine the impact of different possible frames.

The Deepwater Horizon

In April 2010, the Deepwater Horizon oil rig exploded. Eleven members of the crew were killed instantly and others injured. But, there were greater consequences from the massive amount (200 million gallons) of oil that eventually leaked into the Gulf of Mexico. Six hundred miles of coastline were oiled, and 37 % of the water in the Gulf was temporarily closed to fishing. As a consequence, 17,000 jobs were lost. Residents of the Gulf, even if they were not directly affected, tended to experience indirect economic effects, such as a loss in property value, as well as emotional and social consequences. Even if there was no clear threat to health, there were concerns about the possible impacts of both the oil and the dispersants used to clear it (Lehner & Deans, 2010). Comparisons with previous incidents, such as the Exxon Valdez, suggest that domestic abuse, depression, and post-traumatic stress disorder are likely to rise among those affected (Koger, 2010; Palinkas, Downs, Petterson, & Russell, 1993).

How do we make sense of events such as this? In particular, does it threaten a fundamental sense of justice? A common response is to look for someone to blame (Koger, 2010). The media, of course, do their part to help the public make sense. Unfortunately, media portrayals do not always enhance public understanding (Boykoff & Boykoff, 2004; Kim, 2010). The media make decisions about what information to include and how to frame it that can lead to inaccurate perceptions of environmental problems. In this case, there was a great deal of focus on possible shortcuts by BP that contributed to the disaster, but less focus on the socioeconomic context that led to the attempt to drill. Emphasizing different kinds of impacts is likely to have an effect on the way in which people think about an event. Kortenkamp and Moore (2001), for example, presented participants with ecological moral dilemmas and highlighted either human or environmental damage. They found that people were more likely to include environmental considerations in their moral reasoning when environmental damage was described than when human damage was the focus. Stressing a social conflict, alternatively, reduced reference to environmental considerations.

Individual Differences

In addition to the framing of the dilemma, Kortenkamp and Moore (2001) found that individual environmental attitudes also affected moral reasoning. A number of individual differences have been found to be important in understanding environmental attitudes. Studies looking at the impact of gender have found that women tend to demonstrate greater concern for the natural environment, though some studies find no gender difference and the effect tends to be small (Zelezny, Chua, & Aldrich, 2000). More recent research has focused on political ideology as a predictor of environmentalism. A 2008 national survey of over 2,000 Americans found that party affiliation was the strongest predictor of belief in climate change, for example, with a greater impact than gender, race, educational level, or age (Borick & Rabe, 2010). It seems to be the case that a belief in climate change is seen to imply a rejection of traditional American values among conservatives (Kahan et al., 2012). Feinberg and Willer (2013) found that the environment is an issue with more moral standing for liberals than for conservatives.

More directly associated with environmental concern, environmental identity (EID) refers to individuals’ sense of themselves as interdependent with the natural world (Clayton, 2003, 2012). Based on personal experiences as well as emotional responses, EID has been found to predict environmental attitudes in general and in regards to specific environmental issues (Clayton, 2012). It is argued that a high EID implies greater self-relevance for environmental issues and thus increases both attention to and concern about these issues. EID has also been associated with greater support for the attribution of moral standing to environmental entities (Clayton 2003, 2008).

When considering how justice relates to understandings of environmental problems, the belief in a just world (BJW; Lerner, 1977) is an important individual difference to examine. BJW, the belief that fundamentally people get what they deserve, is a worldview that allows a comfortable perception that the world is predictable. People high in BJW may be quicker to rectify injustice when they can, but when they cannot, they resolve their discomfort by minimizing the existence of injustice, for example, by blaming the victim of a negative event or by ignoring evidence of discrimination. Dalbert (1999; Dalbert, Montada, & Schmitt, 1987) has distinguished both conceptually and empirically between a personal and a general belief in a just world (GBJW): that is, a belief that one’s own life is fair, compared to a belief that life is fair in general. She argues that a personal belief in a just world (PBJW) is more strongly associated with psychological well-being. However, a GBJW may more strongly affect how one interprets events that do not have personal relevance. BJW for others has been more strongly associated with social attitudes than BJW for the self (Sutton & Douglas, 2005).

Maintaining the BJW sometimes requires a refusal to acknowledge reality. As applied to environmental issues, Feinberg and Willer (2011) found that BJW predicted denial of climate change information, and that priming just-world statements increased skepticism about climate change. The perception that the world is fair can be highly adaptive: Wu et al. (2011) found that survivors of the 2008 earthquake in Sichuan, China, maintained a high GBJW, which in turn predicted psychological resilience; others have found that high BJW was associated with less psychological distress after a natural disaster (Otto, Boos, Dalbert, Schops, & Hoyer, 2006; Xie, Liu, & Gan, 2011).

The Present Research

It is both interesting and important to understand perceptions of justice as part of people’s responses to environmental problems. In Study 1, we examined responses to a specific environmental disaster, the explosion of the Deepwater Horizon, and investigated whether or not the explosion was classified as an injustice. We mimicked the way in which media may frame an event by creating three preambles, each of which emphasized a different type of harm. Because perceived responsibility is relevant to constructions of justice, we looked at attributions of responsibility for the event. Finally, we examined individual differences in gender, political orientation, EID, and BJW. We anticipated that EID would be associated with stronger responses to the disaster, that conservatives would describe the event differently than would liberals, and that those high in GBJW would be less likely to characterize the event as unjust. Study 2 examined responses to a more general problem, global climate change.

Study 1

Method

Participants

Participants (N = 297: 126 males, 169 females, and 2 who declined to identify their sex) were recruited via Amazon’s Mechanical Turk to complete an on-line survey. MTurk has been shown to be a reliable way to recruit a diverse group of participants to on-line studies (Buhrmester, Kwang, & Gosling, 2011). Although they are not representative of the overall population, participants tend to be more representative than a typical in-person sample and their responses are similar (Berinsky, Huber, & Lenz, 2012). Participation was limited to people over 18 who were living in the United States. The participants were distributed among four experimental conditions: version A (N = 92), version B (N = 73), version C (N = 64), and version D (N = 68).

Materials

Experimental Condition

Based on a review of media coverage of the event, we created four versions of the survey: one control version (A) with no preamble and three experimental versions (B, C, and D). The purpose of the preamble was to see how the media’s portrayal of the Deepwater Horizon explosion could affect people’s perceptions of the disaster. Each preamble was one paragraph long and stressed a different type of harm caused by the oil spill: environmental (version B), social (version C), or economic harm (version D). The preambles may be seen in Appendix 1.

Dependent Variables

The first questions asked participants to rate their perceptions of the event: In separate items, they indicated the extent to which they saw it as an injustice, as a natural disaster, and as a crime. The survey then asked for ratings of responsibility, perceptions of harm, and emotional reactions. Respondents were asked “How much do you think each of the following was responsible for the oil spill?” with regard to five targets: BP, inadequate government regulation of drilling, government policies favoring drilling, insufficient governmental response, or American consumers who are too dependent on oil. We were interested in examining each item separately in order to understand who was seen as most responsible. However, to obtain an overall “blame” score, we also combined the five items to form a scale with an internal reliability, as assessed by Cronbach’s α, of .67.

Five objects were listed after “who or what do you think was harmed by the oil spill?”: people’s physical health, people’s economic well-being, communities who live on the Gulf, wildlife, or the natural environment in general. In order to assess the type of harm that was seen as dominant, each item was examined separately, but we were also interested in obtaining an overall “harm” score so we combined the six items to form a scale with a reliability of .76. Ratings from 1 to 7 were used to measure attributions of responsibility (ranging from 1 as “no responsibility” to 7 as “very high responsibility”) and harm (ranging from 1 as “not harmed” to 7 as “very much harmed”).

Participants were asked to indicate whether any groups were particularly harmed: children, men, minorities, women, or Whites. They could check none, some, or all of these groups. All of these items also had opportunities for free responses.

Participants also rated their emotional responses to climate change by agreeing or disagreeing with the statements “Thinking about the oil spill makes me (sad, angry, ashamed),” ranging from 1 (totally disagree) to 7 (totally agree). Each item was examined separately, but they were also combined for an overall negative affect score with a reliability of .76.

Subject Variables

Participants were asked to identify their gender and political orientation (ranging from 1 “very liberal” to 5 “very conservative”). PBJW and GBJW scales from Dalbert et al. (1987) were used to measure the extent to which raters believe that their own life is fair (PBJW) and that the world is a fair place (GBJW). Ratings were made on seven-item PBJW and 6-item GBJW scales ranging from 1 (strongly disagree) to 6 (strongly agree). The two sets of items were combined for the participants, but separated for analysis. In the present study, the reliability was .78 for the GBJW scale and .89 for the PBJW. The scales were highly correlated at r = .71.

A short version of Clayton’s (2003) EID scale was used to assess the degree to which people have a sense of themselves as interdependent with the natural world. Ratings were made on a 11-item scale where responses ranged from 1 (not true of me at all) to 7 (completely true of me). Internal reliability of the scale was .89.

Procedure

Participants were recruited during Fall 2010 using Amazon.com’s MTurk on-line subject recruitment; they were compensated a moderate amount for their participation (20 cents per survey). Participants were told the survey was about their reactions to the Gulf oil spill. All responses were anonymous and participants gave consent to participate. The different versions were presented at different times, by putting up one version (on MTurk, linked to Surveymonkey.com) at a time and switching them twice per day until the target sample number was reached. Participants were asked to read the preamble before completing the survey. All items on all scales were randomized.

Results

Descriptive Results

Respondents were most likely to describe this event as a “crime” (M = 4.83) rather than an injustice (M = 4.57) or a natural disaster (M = 3.34). A repeated measures ANOVA found a significant difference among these ratings: F (2, 2361) = 44.5, p < .001, η2 = .13. “Crime” and “injustice” were highly correlated, r = .67, p < .001, and both were negatively correlated with “natural disaster” (r = −.36, −.20, respectively; both p’s < .001). High ratings for harm were given to many things, but particularly to wildlife (M = 6.7) and the natural environment (M = 6.7). People’s physical health was rated as being least harmed by the event (M = 5.3). A repeated measures ANOVA found a significant effect of type of harm: F (4, 1132) = 145.7, p < .001, η2 = .34.

Among the different groups affected, people were most likely to indicate that children had been harmed (52 %) and least likely to identify Whites (30 %). Men (46 %), minorities (41 %), and women (40 %) fell in the middle. Interestingly, there was some opposition to identifying specific groups as having been victimized. Thirty percent of respondents did not select any groups, and this question received more open-ended responses (49) than any other, with many people writing “none more than others.” One respondent wrote, “There is no reason to say one specific group is threatened. That’s sick. The people who depend on that oil and who live on that coast are threatened, whether they’re black, white, purple, green, or polka dotted.”

For the responsibility ratings, BP was clearly identified as the villain (M = 6.3), while American consumers were let off the hook (M = 3.8). Government responsibility fell in the middle, ranging from mean ratings of 4.5 for policies and responses to 5.0 for regulations. A repeated measures ANOVA found a significant effect of target of responsibility: F (4, 1116) = 137.4, p < .001, η2 = .33.

Affectively, sadness was the predominant response (M = 5.8), closely followed by anger (5.6). Feeling ashamed was a distant third, though still above the midpoint of the scale (4.1). A repeated measures ANOVA found a significant effect of type of emotion: F (2, 580) = 179.6, p < .001, η2 = .38.

Effects of Preamble

The introductory paragraph had an effect on perceived harm; a significant effect of condition on overall harm rating (F [3, 283] = 3.4, p = .02, η2 = .04) was qualified by an interaction between condition and type of harm (F [12, 1132 = 4.0, p < .001, η2 = .04). “Economic well-being” and “physical health” received lower ratings of harm in the condition emphasizing environmental harm and received their highest ratings in the condition emphasizing social harm. Mean ratings of harm to wildlife, the natural environment, and communities remained fairly similar (a difference of .1–.2) across experimental conditions. Chi square analyses showed that children were significantly more likely to be identified as having been particularly harmed in the condition stressing social harm; χ2 (3) = 12.3, p = .007. There were no other effects of condition on identification of groups that were harmed.

The effect on affective responses approached significance at p = .055: Negative affect was significantly higher when social harm was stressed (M = 5.44) than in the control condition (M = 4.83), with the economic and environmental harm conditions falling in the middle. There was no effect of condition on the way in which the event was characterized (crime, injustice, natural disaster) or on ratings of responsibility. ANOVAs found no significant interaction between EID (dichotomized at the mean) and condition, between GBJW (dichotomized at the mean) and condition, or between political orientation (similarly dichotomized) and condition. In total, these results suggest that the preamble was read and attended to, and influenced perceptions of harm, but did not have a wide-ranging effect in reframing the way people characterized the justice or injustice of the event.

Individual Differences

There was a small effect of gender on affect (F [1, 292] = 4.9, p = .03, η2 = .017) and on perceptions of harm (F [1, 283] = 4.7, p = .03, η2 = .016). In both cases, women gave higher scores. There were no effects of gender on characterization of the event as a crime, injustice, or natural disaster; on responsibility ratings; or on identification of some groups as particularly harmed. There were no significant gender differences in EID, GBJW, PBJW, or political orientation.

GBJW and PBJW were uncorrelated with EID, but were significantly correlated with a conservative political orientation (r = .29, p < .001; r = .19, p < .01, respectively). EID was negatively correlated with conservative political orientation (r = −.17, p < .01).

Zero-order correlations showed that GBJW, but not PBJW, tended to be associated with lower ratings of responsibility (r = −.10, p = .08) and negative affect (r = −.11, p = .05).

Regression Analyses

Regression analyses were run to predict characterization of the event based on personality variables. Gender, EID, political orientation, GBJW, and PBJW were entered as predictors. GBJW, political orientation, and EID were significant predictors of perceptions of the event as a natural disaster; only EID and political orientation predicted perception as an injustice or a crime. See Table 1. Those high in EID or more politically liberal were more likely to consider it an injustice or crime and less likely to consider it a natural disaster; those high in GBJW were more likely to consider it as like a natural disaster. (Zero-order correlations showed a negative correlation between GBJW and ratings of injustice, r = −.18, p < .01, or crime, r = −.15, p < .05, but this relationship was suppressed when the shared variance due to political orientation was included.)
Table 1

Beta coefficients for predictors of event characterization, Study 1

Predictor

Criterion

Natural disaster

Injustice

Crime

Gender (male)

−.05

.01

.02

EID

−.16**

.28***

.46***

Conservative political orientation

.20**

−.19**

−.18**

GBJW

.27**

−.01

.06

Total R2

.37

.39

.53

** p < .01

*** p < .001

A second set of analyses used hierarchical regressions, in which the above predictors were entered in the first stage and characterization of the event was entered in the second stage, to predict ratings of harm, attributions of responsibility, and affective responses. EID was the only significant personality predictor of overall harm, but perceptions of the event as an injustice or a crime added significant variance (F change [3, 250] = 16.6, p < .001). EID and liberal political orientation predicted overall ratings of responsibility; perceptions of the event added significant predictive ability (F change [3, 257] = 3.6, p = .01), although none of the items individually was a significant predictor. Finally, both EID and liberal political orientation predicted negative affect, with perceptions of the event as an injustice or crime adding significant variance (F change [3, 258] = 39.4, p < .001). See Table 2. The effects of EID and of political orientation were partly, but not completely, reduced by including characterizations as predictors.
Table 2

Beta coefficients for predictors of dependent variables, Study 1

Predictor

Criterion

Harm

Responsibility

Negative affect

Step 1

 Gender (male)

−.08

.01

−.09

 EID

.38***

.33***

.47***

 Conservative political orientation

.03

−.17**

−.18**

 GBJW

−.03

−.03

−.05

 Total R2

.16

.16

.30

Step 2

 Gender (male)

−.10

.01

−.10

 EID

.20**

.24***

.25***

 Conservative political orientation

.12*

−.13*

−.06

 GBJW

−.02

−.03

−.05

 Like a natural disaster

−.07

−.02

−.04

 An injustice

.19**

.13

.29***

 A crime

.26**

.09

.29***

 Total R2

.30

.20

.52

** p < .01

*** p < .001

Discussion

At the time these data were collected (Fall 2010), the explosion of the Deepwater Horizon was still fresh in people’s minds. Participants’ responses to the survey indicated a strong affective response and a high rating of harm and responsibility. There was also a dominant narrative constructed about the event: It was a crime, for which BP was largely responsible, and it was an environmental disaster more than a social one. Although the preamble did serve to highlight certain types of harm more than others, it did not have a significant general effect on the way in which the event was perceived, as reflected in responsibility attributions and in characterizations of the explosion as an injustice, crime, or natural disaster.

Perceptions of justice were influential in two ways. First, pre-existing GBJW affected the characterization of the event: Those high in GBJW were more likely to describe it as “like a natural disaster” and less likely to describe it as a crime or injustice. This is consistent with a large body of research showing that BJW is associated with a tendency to interpret events in a way that minimizes the existence of injustice; in the present study, GBJW was the strongest predictor of the way the event was characterized. High GBJW was also slightly associated with a weaker affective response and lower ratings of blame, particularly for American consumers. Blaming consumers would imply not only that someone has done wrong but also that the respondent him- or herself may have done wrong, so it is understandable that people with high GBJW would reject that attribution.

Second, characterization of the event, as a crime, injustice, or natural disaster, affected reactions to the event. In general, perceiving it as a crime or injustice was associated with more negative affect and greater ratings of harm. This was true even after accounting for pre-existing individual differences in EID, BJW, and political orientation. People care very much about justice, so that defining an event as unjust endows it with greater significance.

Individual differences were important. The strongest effects were found for EID, which was associated with characterization of the event as a crime or injustice and also directly related to affective response, responsibility ratings, and harm ratings. EID can affect the perception of justice by helping to define a “moral universe”—a range of entities that deserve moral consideration. Previous research has found that people high in EID are more likely to endow animals with rights (Clayton, 2008) and more likely to consider the natural environment and future generations as morally relevant (Clayton, 2003). Unsurprisingly, political orientation was also important; people who rated themselves as more liberal responded similarly to those high in EID. Political conservatism was also associated with a BJW, particularly GBJW, and the inclusion of political orientation as a variable may partly mask the impact of GBJW on responses. Gender differences, meanwhile, were consistent with the stereotype that women are more willing to express emotion (Brody & Hall, 2008).

Study 2 was designed to explore how this pattern of findings would generalize to an environmental problem that is more abstract and where no clear villain can be identified. How do these individual differences, and particularly the belief in justice, affect responses to climate change? We constructed a study to examine the impact of EID, BJW, and political orientation on perceived harm, ratings of responsibility, and negative affect. In addition to changing the topic from the Gulf oil spill to climate change, we made two significant changes in Study 2. First, the preamble focused more directly on harm to humans versus harm to animals. Because the human costs of climate change are often unrecognized, we wanted to focus on the impact of describing human consequences in general and not separate it into social versus economic. Second, instead of asking who was responsible for causing climate change, we asked who was responsible for addressing it. Because individuals can change their behavior in ways that will mitigate climate change, we wanted to see where individuals lay the responsibility for taking action and particularly whether they took personal responsibility.

Study 2

Method

Participants

Participants (N = 378, 171 males, 204 females) were recruited by Amazon’s Mechanical Turk requester program. Data for Study 2 were collected about 12 months after the data for Study 1; so given the large number of potential participants (over 500,000), it is unlikely that any of the same participants were included. Participation was limited to people over 18 who were living in the United States. The participants were distributed among four experimental conditions: version A (n = 90), version B (n = 90), version C (n = 106), and version D (n = 92).

Materials

Four versions of the survey were created: one control version (A) with no preamble and three experimental versions (B, C, and D). Participants were primed with 1 of 3 introductory paragraphs. Each paragraph described a different type of harm caused by climate change: harm to humans (version B), animals (version C), or a specific zoo animal (version D). All articles were crafted specifically for this study. The paragraphs can be seen in Appendix 2.

Dependent Variables

The survey asked for ratings of responsibility, perceptions of harm, and emotional reactions. Consistent with research by Krosnick, Holbrook, and Visser (2000), respondents were asked “How much do you think each of the following should take responsibility for solving problems resulting from climate change?” with regard to four targets: the average person, the U.S. government, foreign countries, and business and industry. Although the four items were examined separately, they were also combined for an overall responsibility scale with a reliability of .83. Five objects were listed after “who or what do you think may be harmed by climate change?”: people’s physical health, people’s economic well-being, some species of wildlife, all wildlife, or the natural environment in general. Each was examined separately, but they were also combined for an overall harm measure with a reliability of .86. Ratings from 1 to 7 were used to measure attributions of responsibility (ranging from 1 as “no responsibility” to 7 as “very high responsibility”) and harm (ranging from 1 as “not harmed” to 7 as “very much harmed”). These scales also had opportunities for free responses. Participants were also asked to rate their emotional responses to climate change (from sad to happy, calm to angry, guilty to proud, hopeful to anxious), ranging from 1 (sad) to 7 (happy). Items were reverse-coded, as appropriate, so that a higher number indicated a more negative response. Each item was examined separately, but they were also combined for an overall negative affect score with a reliability of .78.

Subject Variables

As in Study 1, participants were asked to identify their gender and political orientation (ranging from 1 “very liberal” to 5 “very conservative”), and to complete PBJW and GBJW scales and the short EID scale. PBJW items and GBJW items were presented together to the participants, but separated upon analysis. In the present study, the reliability was .81 for the GBJW scale and .82 for the PBJW. The scales were highly correlated at r = .75. Internal reliability of the EID scale was .91.

Procedure

Participants were recruited using Amazon.com’s Mechanical Turk and compensated a moderate amount for their participation (20 cents per survey). Participants were told that the survey was about their responses to climate change. All responses were anonymous and participants gave consent to participate. The different experimental versions were presented at different times, by putting up one version (on MTurk, linked to Surveymonkey.com) at a time and switching them three times a day until the target sample number was reached. The participants were asked to read the preamble before completing the survey. They then completed the questions on responsibility and harm, rated their emotional responses to climate change, completed the BJW and EID scales, and identified their gender and political orientation. All items on all scales were randomized.

Results

Descriptive Results

Similar to Study 1, ratings of harm were high in general, but climate change was seen as most likely to harm nature and least likely to affect people’s economic well-being. Harm to people’s physical health fell in the middle. A repeated measures ANOVA found a significant effect of type of harm: F (4, 1472) = 71.9, p < .001, η2 = .16.

Also similar to Study 1, responsibility attributions were the lowest to the average citizen (M = 4.8) and the highest to business and industry (M = 5.7). The US government and foreign countries received intermediate ratings. A repeated measures ANOVA found a significant effect of target of responsibility: F (3, 1110) = 51.13, p < .001, η2 = .12.

Negative affect ratings were above the mean. As in Study 1, the strongest response was sadness (M = 5.3). Anger was less common, getting similar ratings (M = 4.9) to guilt (5.0) and anxiety (4.9). A repeated measures ANOVA showed a significant effect of type of emotion: F (3, 1098) = 14.2, p < .001, η2 = .04.

Effects of Preamble

In general, the introductory paragraph had little impact. It did not have a significant effect on the overall ratings of responsibility, harm, or affect. When individual items were examined, there was only one significant difference: The preamble affected ratings of how responsible the average person was to mitigate climate change, F (3, 370) = 4.03, p = .008, η2 = .032. Ratings were the lowest when participants read about a specific zoo animal (4.4) and the highest when they read about environmental impact (5.1). Post-hoc tests showed that only these two means were significantly different, with the control condition (4.6) and harm to humans (5.0) falling in between. As for Study 1, ANOVAs found no significant interaction between EID (dichotomized at the mean) and condition, between GBJW (dichotomized at the mean) and condition, or between political orientation (similarly dichotomized) and condition.

Individual Differences

There were main effects of gender on overall harm, overall responsibility, and negative affect; women had higher average scores on each. There were no interactions between gender and the within-subject variables of type of harm, negative emotion, or target of responsibility. Women were significantly less conservative and higher in EID than men. See Table 3. There were no gender differences in GBJW or PBJW.
Table 3

Gender differences, Study 2

Dependent variable

Men Mean (SD)

Women Mean (SD)

F

df

p

eta2

Overall harm

5.3 (1.3)

5.9 (1.0)

28.0

1, 361

<.001

.072

Overall responsibility

5.1 (1.3)

5.5 (1.2)

9.89

1, 363

.002

.027

Negative affect

4.7 (1.1)

5.3 (1.0)

24.8

1, 359

<.001

.065

Conservative political orientation

2.9 (1.1)

2.6 (1.0)

5.49

1, 371

.02

.015

EID

4.9 (1.1)

5.2 (1.1)

5.86

1, 360

.02

.016

GBJW and PBJW were significantly correlated with EID (.18, p < .01, .21, p < .001, respectively) and with political orientation (.22, .19, p’s < .001). EID was not correlated with political orientation.

Zero-order correlations showed that GBJW, but not PBJW, tended to be associated with lower ratings of negative affect (r = −.14, p = .01). An examination of the individual items showed that the buffering effect on affect was significant only for sadness (r = −.19, p < .001). GBJW had no relationship with overall responsibility, but it was significantly associated with a greater attribution of responsibility to the average individual (r = .13, p = .01).

Regression Analyses

Regression analyses were run to predict ratings of harm, attributions of responsibility, and affective responses based on personality variables. Gender, EID, political orientation, and GBJW were entered as predictors. EID (high) and political orientation (more liberal) were significant predictors of attributed responsibility. These, along with gender (female), also predicted perceived harm. Finally, EID (high), gender (female), and GBJW (low) predicted negative affect. See Table 4.
Table 4

Beta coefficients for predictors of dependent variables, Study 2

Predictor

Criterion

Harm

Responsibility

Negative affect

Gender (male)

−.19***

−.08

−.22***

EID

.40***

.35***

.38***

Conservative political orientation

−.20***

−.28***

−.08

GBJW

−.01

.09

−.24***

Total R2

.28

.24

.27

*** p < .001

Discussion

The effect of the preamble suggests that describing the harm of climate change is more effective than not doing so and that attempting to personalize it by focusing on a specific animal is ineffective. However, the effects of this preamble were limited. Because there has been a high level of media attention to the topic, participants probably had already formed some opinions and it is not surprising that a single paragraph had little impact on their responses.

Once again, GBJW appeared to protect people from the negative affect, particularly sadness, associated with climate change. Political orientation predicted perceived harm and attributions of responsibility. EID was a predictor of all the dependent variables. While GBJW predicted less individual blame, in Study 1, here it predicted according greater responsibility to individuals for responding to climate change. This pattern of responses is quite consistent with the idea that a high BJW will encourage people to remedy injustice when they feel that they can. Because responding to climate change is an event in the future, whereas the explosion of the Deepwater Horizon was in the past, a high BJW seems to lead individuals to feel greater responsibility for action.

It is interesting that BJW was positively correlated with EID, given that they tended to have opposite relationships to some of the dependent variables. The conceptual linkage between an EID and BJW is plausible, in that one might feel the world is more just if one also feels more interconnected with elements of the natural world. However, particularly given the lack of relationship between these two variables in Study 1, the relationship requires further investigation.

General Discussion

The studies presented here give us a sense of general themes in response to environmental problems and specifically how motivated reasoning can explain some responses. For both a specific incident and a more general problem, people report a negative emotional response in which sadness—a passive emotion, rather than one like anger that motivates action—predominates. They think about harm to the natural environment more than to humans and human society. They do not see individuals as highly responsible for either causing or mitigating the problem.

Framing the story, as specific media accounts might do, did not have a strong general effect in either case. However, this should be seen in the light of typical media coverage. Our respondents had no doubt seen other stories about both crises, which may have already created an interpretive frame. If people encounter multiple stories about environmental problems that are similarly framed, those effects are likely to be stronger. Indeed, the trends we found in terms of who is harmed and who is responsible probably reflect the media accounts present in the wider society.

Individual identities mattered. As expected, a high level of EID was consistently associated with a stronger reaction to environmental problems: greater perceived harm and more negative affect. Expected gender differences were seen in responses to climate change, but not to the explosion of the Deepwater Horizon. There may have been a more compelling narrative for the latter which reduced the gender differences in responses. Alternatively, the relative newness of the event may have meant that stereotypes about appropriately gendered responses had not had time to develop.

Political identity was also important. In general, a conservative political orientation was associated with a weaker response to both problems. This is to be expected given previous research showing the relationship between political ideology and environmental attitudes (Borick & Rabe, 2010; Kahan et al., 2012). The present Study 1 adds to our understanding of this relationship by suggesting that political conservatives define environmental crises differently than do liberals: as unfortunate events, but not as crimes or acts of injustice. Thus, ideology is important in constructing an understanding of environmental threats.

Overall, it seems clear that justice is important in the way people interpret environmental problems. (See Kahn, 1997, for evidence that even children consider principles of justice in such events.) Justice can constitute its own ideology, suggesting that bad things do not happen without reason. In the present studies, this ideological bent led people high in GBJW to search for a way of characterizing the explosion of the Deepwater Horizon other than as an injustice and to let themselves (i.e., American consumers) off the hook for the event, but also to assign more responsibility to individuals for addressing climate change. Such responses, which presumably enhance their sense of control over environmental crises, had the effect of reducing their level of anger or sadness. Meanwhile, those who did characterize an event as unjust (in Study 1) perceived greater harm from that event and experienced more negative affect. These studies also provided more evidence that GBJW and PBJW should be conceptually distinguished: Although the two were highly correlated, GBJW showed more significant relationships with the responses to both specific and general environmental issues. GBJW was also more strongly correlated with political orientation.

Although the desire for justice can make it more difficult to perceive environmental problems, this desire can also be channeled to encourage more pro-environmental action. Future research should explore further the ways in which a narrative of injustice can be used to motivate environmental concern and a sense of individual responsibility.

References

  1. Berinsky, A., Huber, G., & Lenz, G. (2012). Evaluating online labor markets for experimental research: Amazon.com’s Mechanical Turk. Political Analysis, 20, 351–368.CrossRefGoogle Scholar
  2. Borick, C., & Rabe, B. (2010). A reason to believe: Examining the factors that determine individual views on global warming. Social Science Quarterly, 91, 777–800.CrossRefGoogle Scholar
  3. Boykoff, M. T., & Boykoff, J. M. (2004). Balance as bias: Global warming and the U.S. prestige press. Global Environmental Change, 14, 125–136.CrossRefGoogle Scholar
  4. Brody, L., & Hall, J. (2008). Gender and emotion in context. In M. Lewis, J. Haviland-Jones, & L. Barrett (Eds.), Handbook of emotion (3rd ed., pp. 395–408). New York: Guilford.Google Scholar
  5. Buhrmester, M., Kwang, T., & Gosling, S. (2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high quality, data? Perspectives on Psychological Science, 6, 3–5.CrossRefGoogle Scholar
  6. Clayton, S. (2003). Environmental identity: A conceptual and an operational definition. In S. Clayton & S. Opotow (Eds.), Identity and the natural environment (pp. 45–65). Cambridge, MA: MIT Press.Google Scholar
  7. Clayton, S. (2008). Attending to identity: Ideology, group membership, and perceptions of justice. In K. Hegtvedt & J. Clay-Warner (Eds.), Advances in group processes: Justice (pp. 241–266). Bingley, UK: Emerald.Google Scholar
  8. Clayton, S. (2012). Environment and identity. In S. Clayton (Ed.), Handbook of environmental and conservation psychology (pp. 164–180). New York: Oxford.CrossRefGoogle Scholar
  9. Clayton, S., & Opotow, S. (2003). Justice and identity: Changing perspectives on what is fair. Personality and Social Psychology Review, 7, 298–310.PubMedCrossRefGoogle Scholar
  10. Crutzen, P. J., & Stoermer, E. F. (2000). The ‘anthropocene’. Global Change Newsletter, 41, 17–18.Google Scholar
  11. Dalbert, C. (1999). The world is more just for me than generally: About the personal belief in a just world scale’s validity. Social Justice Research, 12, 79–98.CrossRefGoogle Scholar
  12. Dalbert, C., Montada, L., & Schmitt, M. (1987). Belief in a just world: Validity correlates of two scales. Psychologische Beitrage, 29, 596–615.Google Scholar
  13. Feinberg, M., & Willer, R. (2011). Apocalypse soon?: Dire messages reduce belief in global warming by contradicting just-world beliefs. Psychological Science, 22, 34–38.PubMedCrossRefGoogle Scholar
  14. Feinberg, M., & Willer, R. (2013). The moral roots of environmental attitudes. Psychological Science, 24, 56–62.PubMedCrossRefGoogle Scholar
  15. Gross, C. (2008). A measure of fairness: An investigative framework to explore perceptions of fairness and justice in a real-life social conflict. Human Ecology Review, 15, 130–140.Google Scholar
  16. Kahan, D. M., Peters, E., Wittlin, M., Slovic, P., Ouellette, L. L., Braman, D., et al. (2012). The polarizing impact of science literacy and numeracy on perceived climate change risks. Nature Climate Change. doi:10.1038/nclimate1547.Google Scholar
  17. Kahn, P. H. (1997). Children’s moral and ecological reasoning about the Prince William Sound oil spill. Developmental Psychology, 33(6), 1091–1096.PubMedCrossRefGoogle Scholar
  18. Kals, E., & Russell, Y. (2001). Individual conceptions of justice and their potential for explaining proenvironmental decision making. Social Justice Research, 14(4), 367–385.CrossRefGoogle Scholar
  19. Keane, A., Jones, J. P. G., Edwards-Jones, G., & Milner-Gulland, E. J. (2008). The sleeping policeman: Understanding issues of enforcement and compliance in conservation. Animal Conservation, 11, 75–82.CrossRefGoogle Scholar
  20. Kim, K. S. (2010). Public understanding of the politics of global warming in the news media: The hostile media approach. Public Understanding of Science, 20, 690–705.CrossRefGoogle Scholar
  21. Koger, S. (2010). Coping with the deepwater horizon disaster: An ecopsychology interview with Deborah DuNann Winter. Ecopsychology, 2, 205–209.CrossRefGoogle Scholar
  22. Kortenkamp, K., & Moore, C. (2001). Ecocentrism and anthropocentrism: Moral reasoning about ecological commons dilemmas. Journal of Environmental Psychology, 21, 261–272.CrossRefGoogle Scholar
  23. Krosnick, J., Holbrook, A., & Visser, P. (2000). The impact of the fall 1997 debate about global warming on American public opinion. Public Understanding of Science, 9, 239–260.CrossRefGoogle Scholar
  24. Lehner, P., & Deans, R. (2010). In deep water: The anatomy of a disaster, the fate of the gulf, and ending our oil addiction. New York: The Experiment Publishing Company.Google Scholar
  25. Lerner, M. J. (1977). The belief in a just world: A fundamental delusion. New York: Plenum.Google Scholar
  26. Müller, M., Kals, E., & Maes, J. (2008). Fairness, self-interest, and cooperation in a real-life conflict. Journal of Applied Social Psychology, 38, 684–704.CrossRefGoogle Scholar
  27. Otto, K., Boos, A., Dalbert, C., Schops, D., & Hoyer, J. (2006). Posttraumatic symptoms, depression, and anxiety of flood victims: The impact of the belief in a just world. Personality and Individual Differences, 40, 1075–1084.CrossRefGoogle Scholar
  28. Palinkas, L., Downs, M., Petterson, J., & Russell, J. (1993). Social, cultural, and psychological impacts of the Exxon Valdez oil spill. Human Organization, 52, 1–13.Google Scholar
  29. Sutton, R. M., & Douglas, K. M. (2005). Justice for all, or just for me? More evidence of the importance of the self-other distinction in just-world beliefs. Personality and Individual Differences, 39(3), 637–645.CrossRefGoogle Scholar
  30. Syme, G., & Nancarrow, B. (2012). Justice and the allocation of natural resources: Current concepts and future directions. In S. Clayton (Ed.), Handbook of environmental and conservation psychology (pp. 93–112). New York: Oxford.Google Scholar
  31. Vess, M., & Arndt, J. (2008). The nature of death and the death of nature: The impact of mortality salience on environmental concern. Journal of Research in Personality, 42(5), 1376–1380.CrossRefGoogle Scholar
  32. Wu, M., Yan, X., Zhou, C., Chen, Y., Li, J., Zhu, Z., et al. (2011). General belief in a just world and resilience: Evidence from a collectivistic culture. European Journal of Personality, 25, 431–442.CrossRefGoogle Scholar
  33. Xie, X., Liu, H., & Gan, Y. (2011). Belief in a just world when encountering the 5/12 Wenchuan earthquake. Environment and Behavior, 43, 566–586.CrossRefGoogle Scholar
  34. Zelezny, L., Chua, P., & Aldrich, C. (2000). Elaborating on gender differences in environmentalism. Journal of Social Issues, 56, 443–457.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  1. 1.The College of WoosterWoosterUSA

Personalised recommendations