Scepticism in a changing climate: A cross-national study

Global Environmental Change 33 (2015) 154–164

Contents lists available at ScienceDirect

Global Environmental Change journal homepage: www.elsevier.com/locate/gloenvcha

Scepticism in a changing climate: A cross-national study Bruce Tranter *, Kate Booth University of Tasmania, Australia

A R T I C L E I N F O

A B S T R A C T

Article history: Received 4 October 2014 Received in revised form 29 April 2015 Accepted 11 May 2015 Available online 17 June 2015

Despite the findings of climate scientists, the proportions of climate sceptics appear to be increasing in many countries. We model social and political background, value orientations and the influence of CO2 emissions per capita and vulnerability to climate change upon climate scepticism, drawing upon data from the International Social Survey Programme. Substantial differences in the levels of climate scepticism are apparent between nations. Yet cross national data show that climate sceptics are not merely the mirror image of environmentalists. Typical predictors of environmental issue concern, such as education level, postmaterial value orientations and age are poor predictors of climate scepticism. Affiliation with conservative political parties, gender, being unconcerned about ‘the environment’ or having little trust in government are consistent predictors of scepticism. Climate change scepticism is also correlated positively with CO2 emissions and vulnerability to climate change. While high levels of scepticism have been documented among citizens of the United States, scepticism is as high or higher in countries such as Australia, Norway and New Zealand. ß 2015 Elsevier Ltd. All rights reserved.

Keywords: Climate change Climate change sceptics Partisans CO2 emission Cultural values Postmaterial values

1. Introduction A plethora of research outlines the characteristics of environmentally friendly attitudes and behaviours (e.g. Clements, 2012; Franzen and Vogl, 2013; Givens and Jorgenson, 2011; Tranter, 2010) and the political bases of environmental attitudes (e.g. Dunlap et al., 2001; Nawrotzki, 2012). This includes works that consider attitudes towards climate change, such as public opinion on the veracity of climate science (e.g. Brechin and Bhandari, 2011), political polarisation on climate change (e.g. McCright and Dunlap, 2011a) and the role of conservative political elites in influencing public opinion and determining climate change policy (e.g. McCright and Dunlap, 2010; Tranter, 2011). However, few scholars aim to specifically profile climate change sceptics. There are some exceptions such as Poortinga et al. (2011) who examined types of British sceptics and the association between scepticism and attitudinal constructs, and Engels et al. (2013) who assessed scepticism in Germany. Yet few studies examine associations between attitudinal and socio-demographic variables and a sceptical stance (Poortinga et al., 2011; Whitmarsh, 2011; McCright and Dunlap, 2011a). In addition, Crona et al. (2013) observe that most studies of climate change perceptions are ‘place-based’: situated within

* Corresponding author. Tel.: +61 362262362. E-mail address: [email protected] (B. Tranter). http://dx.doi.org/10.1016/j.gloenvcha.2015.05.003 0959-3780/ß 2015 Elsevier Ltd. All rights reserved.

particular locales, regions or nation states. While acknowledging the importance of these, they highlight the need for more comparative cross national studies. Engels et al. (2013) note that the ‘‘specific composition of factors correlating with climatechange scepticism depend strongly on the political and cultural context in which [a] model is tested’’ (2013: 1019). However, they also argue that accounting for such differences reduces the capacity for cross national comparisons that allow identification of differences and similarities between nation states (Engels et al., 2013). The strength of using the same model of scepticism for cross national analysis is that such differences can be gauged, compared and contrasted. According to Brechin and Bhandari (2011), there has been some consideration of cross-national data on public perceptions of climate change and limited determination of cross-national levels of scepticism. Yet systematic cross-national analyses that correlate scepticism with socio-demographic data and other mediating factors such as scientific knowledge, political party identification, levels of national CO2 emissions, climate vulnerability, and postmaterialist value orientations are lacking, a deficiency we begin to address here. We attempt to model levels of climate scepticism across 14 advanced industrialised nations using survey data from the 2010–2011 International Social Survey Program, Environment module (ISSP Research Group, 2012). Based upon several social and political correlates of environmental issue concern that we argue should be associated with acceptance of anthropogenic climate change, and the profiles of climate change

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

sceptics from relevant literature, we ask, to what extent are the background characteristics of climate sceptics the mirror image of environmentalists, and, to what extent are the socio-economic and political backgrounds of climate sceptics consistent cross-nationally? Specifically, we consider how well the following hypotheses, derived from existing research on climate scepticism, hold up within and between countries: 1. Conservative (right leaning) political party identifiers are more likely than progressive (left) identifiers to be climate sceptics. 2. Men are more likely than women to be climate sceptics. 3. Older people are more likely than younger people to be climate sceptics. 4. City dwellers are less sceptical than those living in other locations. 5. Those who believe that science can solve environmental problems are less sceptical. 6. Identifying with a religion increases levels of scepticism. 7. Understanding how to solve environmental problems is associated with less scepticism. 8. Those less trusting of government hold more sceptical positions on climate change. 9. Postmaterialists are less likely than materialists to be sceptical about climate change. 10. Concern over environmental issues is negatively associated with climate scepticism. 11. ‘Individualistic’ worldviews will be positively associated with scepticism. 12. The tertiary educated right will be more sceptical than the tertiary educated left (i.e. education by party left-right interaction effect is expected). 13. At the country level, CO2 emissions will be associated positively with climate change scepticism. 14. At the country level, climate vulnerability will be associated negatively with climate scepticism. A growing body of evidence suggests public concern over the risks of global warming is waning in advanced industrialised nations (e.g. McCright and Dunlap, 2011a; Leiserowitz et al., 2011; Ratter et al., 2012; Scruggs and Benegal, 2012). If this heralds a rise in actual climate scepticism, more research into climate scepticism is warranted and timely. We begin by reviewing empirical research on identified indicators and mediators of climate scepticism. After describing the data and methods and presenting the results of our analysis, we conclude with observations on the predictors of climate scepticism that apply across nations and a preliminary ranking of countries based on levels of climate scepticism. 2. Background Poortinga et al. (2011) note that ‘scepticism’ is an imprecise term that has multiple meanings in the context of complex debates surrounding climate change, with a significant degree of heterogeneity among those identified as sceptics. Hobson and Niemeyer (2012: 403–404) identify five discourses associated with climate change sceptics: empathic negation (we are not in the position to say whether it is real or not given the state of knowledge about climate change); unperturbed pragmatism (climate change policy rejection combined with belief that if there is a problem there is time to sort it out); proactive uncertainty (no need to do anything as rash as question that climate change is happening and what kind of impact it will have if it is. If something is to be done it should be focused on adaption); earnest acclimatisation (climate change is a natural phenomenon that we should be concerned about but as its natural the focus needs to be on adaptation, not, for example, on

155

emission reductions); and noncommittal consent (uncertainty about the knowledge to do with climate change but a sense that it is possibly happening and is anthropogenic – focus on managing impacts rather than causes). They argue that the complex nature of scepticism may result in an under-reporting of scepticism (Hobson and Niemeyer, 2012). Poortinga et al. (2011) found that climate scepticism is not widespread in Britain, although there is uncertainty about its potential effects. Citing a range of studies from the United Kingdom, Whitmarsh (2011: 691) similarly argues that climate change denial is fairly uncommon. However, she found that many people question the seriousness of climate change, and while complete rejection of human-induced climate change stands at between 10 and 20 per cent, public uncertainty is significantly higher. This is supported by previous UK studies, for example, Downing and Ballantyne (2007), who found 40 per cent agreed ‘climate change is too complex and uncertain for scientists to make useful forecasts’, while 56 per cent agreed that ‘many leading experts still question if human activity is contributing to climate change’. Country level research on attitudes towards climate change provides clues about the social and political background of climate sceptics. McCright and Dunlap (2011b, 2013) have shown conservative white males to be significantly more likely than others to endorse sceptical positions on climate change in the United States, while in the UK, sceptics tend to be aged over 65, are more likely to be male, right-of-centre in their political orientations and to hold individualistic worldviews (Whitmarsh, 2011). The most important predictors of scepticism in the UK are environmental values and political affiliation with demographic factors mediated by these (Whitmarsh, 2011). There are deep divides over global warming on the basis of political party identification in the United States (e.g. Hamilton, 2010; Dunlap and McCright, 2008; McCright and Dunlap, 2011a, 2011b), the United Kingdom (Poortinga et al., 2011; Whitmarsh, 2011) and Australia (Tranter, 2011, 2013). For example, Republicans and conservatives are far less likely than Democrats and liberals to accept climate change as a serious problem and are less willing to act to address climate change in the United States (e.g. Wood and Vedlitz, 2007; Jacques et al., 2008; McCright and Dunlap, 2011a, 2011b), while conservative Liberal and National party members exhibit similar sceptical attitudes in Australia (Tranter, 2011, 2013; Fielding et al., 2012). Political scientists have long argued that political leaders influence public attitudes and behaviour and provide ‘cues’ that partisans tend to follow (Campbell et al., 1960; Bartels, 2000; Green et al., 2002). By filtering complex political issues, party identification has a ‘simplifying function’, with political leaders providing cues that ‘guide the political thought and action of the party identifier’ (Miller, 1976: 23). Gilens and Murakawa (2002: 21) suggest partisans tend to follow the cues of political leaders more often than they engage with ‘substantive assessment of competing evidence and arguments’. Conservative politicians in many countries question and sometimes even reject climate science outright, with divisions among political elites leading to polarisation among party supporters (Sciarini et al., 2007). This process has been referred to as ‘party sorting’ among voters ‘wherein the more visible and active members of a party, especially its elected officials and party activists, sort first and provide cues to voters that party positions are evolving’ (Fiorina and Abrams, 2008: 581). Several scholars have pointed to the role of mass media in the formation of attitudes towards climate change and the dissemination of climate science. Brulle et al. (2012) found that media coverage of climate change in the US directly affects public concern about this issue, but public opinion in this area competes with

156

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

concern over issues such as unemployment and economic prosperity. However, in line with the findings of McCright and Dunlap (2011a) regarding political polarisation in the US around issues such as climate change, elite driven partisanship over climate change was the most important factor influencing opinion on this issue. Hindman (2009) argues in relation to the US: . . .liberal Democrats, who get their news from National Public Radio (NPR), MSNBC, and the New York Times and follow the likes of Al Gore, are more likely to hear favourable messages about the reality and significance of climate change than are their conservative Republican counterparts, who get their news from conservative talk radio, FOX News, and the Wall Street Journal. . .. Hmielowski et al. (2013) confirm the alignment of particular news outlets with conservative and Republican interests suggesting they tend to run stories that question the existence of humaninduced climatic change as opposed to stories more accepting of climate science, and in the United States, increasing fragmentation of audiences is occurring across a diversity of news outlets. Similar conclusions were reached by Whitmarsh (2011) in the UK, while the political polarisation observed in Australia (Tranter, 2011, 2013; Fielding et al., 2012) reflects strong associations between particular publishers and the tone of articles on climate change (Bacon, 2013). Kahan et al. (2012: 732) tested what they refer to as the ‘scientific comprehensions thesis’ (SCT) which states ‘As members of the public do not know what scientists know, or think the way scientists think, they predictably fail to take climate change as seriously as scientists believe they should’. Contrary to the expectations of SCT, Kahan et al. (2012: 732) show that as scientific literacy/numeracy scores increase, perceptions of the risks associated with climate change decrease slightly. Hamilton et al. (2012), alternatively, found that higher basic science-literacy scores are associated with greater concern over the impacts of climate change.1 However, both Kahan et al. (2012) and Hamilton et al. (2012) detected interaction effects demonstrating that ideological polarisation over climate change increases as science literacy increases.2 The manner in which members of the public perceive the trustworthiness of climate change communicators is influenced strongly by their worldviews and values. Kahan et al. (2012) maintain cultural values have a mediating influence upon attitudes towards climate change in the US, with those holding individualistic worldviews more sceptical than collectivists. Corner et al. (2012: 463) found biased assimilation of information regarding climate change, with those less sceptical about climate change evaluating the reliability of newspaper editorials in a significantly different way to the sceptical. Australian evidence suggesting sceptical positions on climate change are quite entrenched (Hobson and Niemeyer, 2012), further supporting the culturally/ socially embedded nature of climate change beliefs. A substantial body of research links Inglehart’s postmaterial values thesis to various aspects of environmental support. Inglehart (1997) argues that since World War II, citizens of advanced industrialised countries have become increasingly likely to hold ‘postmaterial’ values, to prioritise free speech and seek greater say in political decisions. ‘Materialists’ tend to prioritise economic and national security related issues. Postmaterialists are

1 The opposite main effects for science literacy and climate change beliefs found by Kahan et al. (2012) and Hamilton et al. (2012) may be due to their different approaches to measuring scientific literacy. 2 For Kahan et al. (2012: 734) between egalitarian communitarians and hierarchical individualists; for Hamilton et al. (2012: 164) between liberals and conservatives.

more concerned than materialists about environmental issues, are more likely to join environmental groups, participate in protest actions and vote for Green candidates (Inglehart, 1990, 1997; Muller-Rommel, 1990; Papadakis, 1993; Crook and Pakulski, 1995; Dalton, 2005; Tranter, 2013). Examining national survey data from Australia, Tranter (2011) found postmaterialists to be more concerned than materialists that global warming will ‘pose a serious threat’ to Australians’ ‘way of life’, controlling for social and political background, while Kvaløy et al. (2012: 18) found postmaterialists to be more concerned about ‘global warming or the greenhouse effect’ based upon World Values data. Yet there is scant research on climate change scepticism viewed through this influential theoretical lens. Further, given the paucity of cross-national analysis of climate scepticism, few studies consider associations between national levels of CO2 emissions, climate vulnerability and scepticism towards climate change, a point we begin to address here. This deficiency remains despite some research demonstrating that forms of structural vulnerability do inform or influence perceptions and stances regarding climate change at the local level (e.g. Brody et al., 2008), and a small number of cross-national studies examining links between affluence and vulnerability (Kim and Wolinsky-Nahmias, 2014) and wealth and CO2 emissions (Kvaløy et al., 2012) on public opinion regarding climate change. 3. Data and method Data analysed here are from the 2010 International Social Survey Programme, Environment Module III (see ISSP Research Group, 2012). The following countries from the 2010 ISSP were selected: Australia (N = 1946), Austria (N = 1019), Canada (N = 985), Denmark (N = 1305), Finland (N = 1211), France (N = 2253), Germany (N = 1407), Great Britain (N = 928), New Zealand (N = 1172), Norway (N = 1382), Spain (N = 2560), Sweden (N = 1181), Switzerland (N = 1212), and the United States (N = 1430). Surveys were administered in 2010 and 2011. While sampling strategies and modes of survey administration varied somewhat between countries and are beyond the scope of detailed description here, samples were designed to be nationally representative of adults aged 18 and over (see Gendal, 2012 for response rates, survey administration methods and the follow up strategies applied). We model three dependent variables (see Appendix table for descriptive statistics). Two variables are operationalised to allow us to (indirectly) contrast the profile of climate sceptics with those of general concern about the environment, across 14 advanced industrialised countries in the 2010 ISSP. The first variable measures responses to the statement: ‘How concerned are you about environmental issues?’ (1 to 5 scale: not at all concerned to very concerned). Environmental concern is modelled using ordinal logistic regression (Agresti and Finlay, 1999: 599) in Table 1. Known correlates of environmental concern are expected to show opposite signed effects when predicting climate sceptics. The premise here is that those who are concerned about environmental issues will be very unlikely to be climate sceptics (Corbett and Durfee, 2004), but very likely to ‘believe’ climate scientists in relation to climate change. For example, men are generally less supportive of environment issues than women are, so men should be more likely than women to be climate sceptics, older people should be more sceptical than younger people, the highly educated less sceptical than the less educated, urban cosmopolitans less sceptical than those living outside of large cities and materialists more sceptical than postmaterialists. We also construct two dichotomous dependent variables to measure climate change scepticism by combining responses to ISSP questions. The first question (‘climate sceptics 10 ) is based on

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

157

Table 1 ‘How concerned are you about environmental issues?’ (Ordered logit models with odds ratios shown.). Australia ***

Austria ***

Canada **

Denmark

Finland ***

France ***

Germany ***

NZ

Norway ***

***

Spain ***

Sweden

Switz.

***

*

UK

USA *

Men Aged 65+ Degree Live in large city Non-religious Postmaterial (+post) Party left-right (+right) Distrusts government

0.7 1.1 1.4** 1.1 1.3* 1.1 0.6*** 0.8*

0.7 1.7** 1.2 1.4** 1.5* 1.4** 0.9 1.4**

0.7 1.4* 1.6* 0.9 0.7 1.4** 0.8** 1.1

0.97 1.8*** 1.2 0.9 1.3 1.5*** 0.9* 0.99

0.6 1.1 1.4* 1.2 1.2 0.96 0.9 1.1

0.8 0.8* 1.4** 0.9 1.2 1.3*** 0.8** 1.05

0.7 0.8* 1.1 0.9 0.9 1.5*** 0.8*** 1.3*

0.6 0.8 1.2 1.2 0.9 1.2 0.6*** 0.98

0.7 1.03 1.1 1.1 1.02 1.5*** 0.8*** 0.8*

0.8 0.6*** 1.1 1.3** 1.2 1.2*** 0.9 1.1

0.7 0.9 1.1 1.1 0.8 1.2 0.8** 1.1

0.8 1.2 1.05 0.8 1.1 1.2 0.8** 1.2

0.7 1.0 1.5** 1.1 0.9 1.1 0.8* 0.8

0.9 0.9 1.07 1.2 1.2 1.0 0.7*** 0.97

Science solves environmental problems Knowledge of solutions to environmental problems Private enterprise Reduce income inequality

0.9

0.8*

0.8

1.2

0.8

0.7***

1.05

1.2

0.6***

1.1

0.7*

0.6***

1.3

1.00

1.3

2.0***

1.9***

3.5***

1.9***

2.1***

1.6***

1.9***

1.9***

2.5***

2.1***

1.6**

2.2***

2.0***

0.8** 1.5***

0.9 1.1

0.7** 1.4**

0.98 1.1

0.9 1.7***

0.9 1.2

0.9 1.2

0.9 1.5***

0.9 1.4**

0.99 1.4***

0.9 1.2

0.9 1.3*

1.3 1.7***

0.9 1.7***

Nagelkerke R2 N

0.13 (1929)

0.10 (1019)

0.11 (958)

0.12 (1264)

0.08 (1179)

0.10 (2173)

0.07 (1380)

0.15 (1151)

0.11 (1351)

0.08 (2545)

0.07 (1169)

0.06 (1204)

0.09 (898)

0.08 (1404)

Source: ISSP (2010). Dependent variable: ‘How concerned are you about environmental issues?’ (1 not at all concerned to 5 very concerned). * p < 0.05. ** p < 0.01. *** p < 0.001.

two questions. The first states ‘Many of the claims about environmental issues are exaggerated?’ (responses: strongly agree, agree, neither agree nor disagree, disagree, strongly disagree). The second question asks ‘In general, do you think that a rise in the world’s temperature caused by climate change is. . . extremely dangerous for the environment, very dangerous, somewhat dangerous, not very dangerous, or, not dangerous at all for the environment? Respondents who agree or strongly agree that ‘many of the claims about environmental issues are exaggerated’ are combined with those who also select climate change as not very dangerous, or not dangerous at all for the environment. Responses to the climate sceptics 1 variable (Table 2) can be compared to the environmental concern variable (Table 1) as an indirect ‘test’ of differences between environmental supporters and climate sceptics. If climate change sceptics are mainly the mirror image of environmentalists, most of those who are not concerned about environmental issues should also tend to be climate change

sceptics. The second measure of scepticism (‘climate sceptics 20 ) comprises an attempt to directly assess the research question ‘to what extent are sceptics the mirror image of environmentalists?’ This variable also uses responses from the ‘rise in the world’s temperature caused by climate change’ question (not very dangerous, or not dangerous at all = 1; otherwise = 0), but restricts these responses to only those who are not very concerned, or not at all concerned about the environment (from the question ‘How concerned are you about environmental issues?’). There are limitations with this approach. The question ‘. . .a rise in the world’s temperature caused by climate change. . .’ clearly refers to the concept of climate change, but does not distinguish anthropogenic and non-anthropogenic causes. A change in question wording from the 2000 to the 2010 ISSP Environment module surveys also saw the substitution of the phrase ‘greenhouse effect’ in the 2000 survey (i.e. ‘. . .a rise in the world’s temperature caused by the greenhouse effect . . .’) with the phrase

Table 2 Profile of climate change ‘sceptics’ (binomial logit models with odds ratios shown). Australia

Austria

Canada

Denmark Finland France

Germany NZ

Norway Spain

Sweden

Switz.

UK

USA

Men Aged 65+ Degree Live in large city Non-religious Postmaterial (+Post) Party left-right (+right) Distrusts government

1.5** 1.3 0.6** 0.7* 0.7* 1.1 1.7*** 1.7***

1.3 1.4 1.1 0.8 0.7 0.97 1.1 0.7

1.8* 1.04 0.8 1.03 1.6 0.9 1.5* 1.5

2.3*** 1.7* 0.6 0.7 1.1 0.9 2.2*** 1.2

1.8** 1.8* 0.6 1.05 1.1 0.7 1.3 1.7*

1.8*** 1.2 0.8 1.1 1.1 1.03 1.2 1.6*

1.7 1.3 0.9 0.99 1.04 1.3 1.3 1.5

1.7** 2.2*** 0.6 1.3 1.3 0.97 1.5** 2.0***

1.7** 1.1 0.6** 0.7 1.2 0.95 1.3* 1.5*

1.8* 1.6 1.6 0.8 1.1 1.7* 1.3 2.1*

2.4*** 1.3 0.6 0.4** 0.9 0.9 1.3 1.6

1.9 1.6 1.4 0.4 1.1 1.04 1.5 1.5

1.6 1.3 0.1* 0.6 1.2 1.04 1.6** 2.1**

1.3 0.9 0.97 1.2 0.6 2.0*** 1.8*** 2.1***

Science solves environmental problems Unconcerned about environment Knowledge of solutions to environmental problems Private enterprise Reduce income inequality Degree * left–right

1.1

0.6

1.4

1.8**

1.7

1.5*

1.05

1.3

1.3

1.8*

2.1*

1.04

0.7

1.4

4.4*** 1.8*

4.0*** 1.6

5.3*** 0.8

1.7* 0.6

7.4*** 2.1**

2.9*** 1.04

2.4** 0.8

4.7*** 1.5*

5.0*** 0.9

2.3* 0.4

8.2*** 1.3

4.3*** 0.8

6.3*** 1.2

5.6*** 1.3

1.6*** 0.6** 1.7*

1.7 1.2 1.6

1.2 0.5* 2.9

1.4 0.8 2.1

1.8** 0.99 1.9

1.4 0.6** 1.1

1.6 0.7 1.3

2.0*** 0.6* 1.3

1.5* 0.9 1.6*

1.9* 0.7 0.8

0.96 0.9 1.5

1.6 1.2 0.4

1.5 1.4 6.6

3.7*** 0.3*** 1.4

0.26 (1946)

0.11 (1019)

0.17 (985)

0.18 (1305)

0.24 (1211)

0.09 (2253)

0.06 (1407)

0.21 (1172)

0.24 (1382)

0.09 0.29 (2560) (1181)

0.09 (1212)

0.26 (928)

0.36 (1430)

Nagelkerke R2 N

Source: ISSP (2010). Dependent variable ‘climate change sceptics’ = 0; or 1 if they agree or strongly agree that ‘Many of the claims about environmental threats are exaggerated’ and that ‘a rise in the world’s temperature caused by climate change is. . .not dangerous or not dangerous at all for the environment’. * p < 0.05. ** p < 0.01. *** p < 0.001.

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

158

‘climate change’. Further, the item ‘. . .claims about environmental issues are exaggerated’ does not directly measure climate change. However, the ‘climate change’ question does serve as a useful proxy measure for our purposes. In lay usage the term ‘climate change’ is increasingly associated with human-induced climate change. Further, while ‘climate change’ is an outcome of global warming rather than a cause (as the wording of the 2010 ISSP question seems to imply), according to Whitmarsh (2009), researchers sometimes combine the two concepts as ‘climate change/global warming’ in survey items (e.g. MORI, 2005; Poortinga et al., 2006). Whitmarsh (2009: 403) also claims, citing Lorenzoni and Pidgeon (2006) as an example, that ‘there is typically an implicit assumption that the two terms ‘‘global warming’’ and ‘‘climate change’’—refer to the same thing’ while confusion among the lay public may be heightened because ‘media coverage of the issue often uses the two terms indiscriminately or interchangeably’. even within the international policy community there is inconsistency in how the term ‘‘climate change’’ is used. . .Given Given this variation amongst decision-makers and communicators in the use and meaning of these terms, this would suggest that the public too are unlikely to have a clear understanding of key terminology (Whitmarsh, 2009: 404). Schuldt et al. (2011: 116) also suggest ‘the terms ‘‘global warming’’ and ‘‘climate change’’ are used interchangeably in public discourse and opinion polls on this issue’. Using survey data from the United States they found that variations in question wording elicit different responses according to one’s political affiliation,

with Republicans more sceptical when responding to questions that contain the term ‘global warming’ than those using ‘climate change’ (Schuldt et al., 2011: 120). If similar bias regarding ‘global warming’ holds among political conservatives cross nationally, our findings may offer a conservative estimate of the differences between the left and the right, given that the ISSP question refers to ‘climate change’. The concept ‘danger’ in relation to climate change is also complex, Dessai et al. (2004) maintain, because it includes both internal and external perspectives on risk perceptions. While some may respond due to their knowledge of ‘external’ definitions of danger, on the basis of ‘scientific risk analysis, performed by experts, of system characteristics of the physical or social world’ (2004: 11), the 2010 ISSP climate change question is more likely to tap the ‘perceived insecurity arising from realised (experienced) or anticipated (perceived) impacts associated with changing extreme weather events, and often immediate threats to life and livelihood, which are of greatest concern to individuals or, collectively, to society’ (2004: 14). Finally, our first operational definition of sceptics is strengthened by restricting them to those who question the veracity of environmental claims more broadly. The binary dependent variable ‘climate sceptics 10 contrasts all other responses on these survey items and is modelled using binary logistic regression (Agresti and Finlay, 1999: 575) in Table 2 and multilevel binary regression in Table 3. The dependent variable ‘climate sceptics 20 is based on a smaller sample as it is restricted to those who are not concerned about ‘environmental issues’. It is also modelled with multilevel binary regression in Table 3. The signs for our independent variables should differ (i.e. if effects are opposite) between Table 1 (country comparisons of environmental support)

Table 3 Profile of sceptics 1 and sceptics 2 with multilevel binary regression (binomial logit models). Sceptics 1 Model 1 Country level error variance Level 2 Intercept (estimate) Fixed effects Intercept (estimate)

Model 3

0.34**

0.20**

0.20**

2.41***

3.51***

3.50***

1.7*** 1.4*** 0.8** 0.8** 0.99 1.1* 1.6*** 1.7*** 1.3*** 4.2*** 1.1 1.7*** 0.7*** 1.08** 1.006 –

1.7*** 1.4*** 0.7*** 0.8** 1.00 1.1* 1.5*** 1.7*** 1.3*** 4.2*** 1.1 1.7*** 0.7*** 1.08** 1.006 1.4***

Fixed effects (odds ratios) Men Aged 65+ Degree Live in large city Non-religious Postmaterial (+post) Party left-right (+right) Distrusts government Science solves environmental problems Not concerned about environmental issues Knowledge of solutions to environmental problems Private enterprise solves economic problems Government should reduce income inequality CO2 emissions per capita (2010) Global climate risk index Degree * left–right Model fit 2LL N

Sceptics 2 Model 2

11,714.4 19,991

10,208.5 19,991

10,195.08 19,991

Model 4 0.66** 0.63*

2603.6 2194

Model 5

Model 6

0.22*

0.22*

1.30**

1.29**

1.4** 1.4** 1.1 0.8 1.1 1.3** 1.2** 1.5*** 0.9 – 1.5** 1.5*** 0.7** 1.11** 1.01**

1.4** 1.3* 1.02 0.8 1.2 1.3** 1.2 1.5*** 0.9 – 1.5** 1.5*** 0.7** 1.11** 1.01** 1.4

2461.9 2194

2458.1 2194

Source: ISSP (2010). ‘Sceptics 10 ICC = 0.09; ‘Sceptics 20 ICC = 0.17. Values based on SAS PROC GLIMMIX; Estimation Method = Laplace. Dependent variables: ‘Climate Change Sceptics 10 = 0; or 1 if respondents agree or strongly agree that ‘Many of the claims about environmental threats are exaggerated’ and that ‘a rise in the world’s temperature caused by climate change is. . .not dangerous or not dangerous at all for the environment’. ‘Climate Sceptics 20 = 0; or 1 if respondents are not very concerned, or not at all concerned about ‘environmental issues’ and consider ‘a rise in the world’s temperature caused by climate change is. . .not dangerous or not dangerous at all for the environment’. Model 1 is an unconditional model with no predictors; Model 2 adds the independent variables; Model 3 adds an interaction term for degree by party left-right scale. * p < 0.05. ** p < 0.01. *** p < 001.

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

and Table 2 (country comparisons of climate scepticism). As suggested above, our independent variables include a range of socio-demographic variables (i.e. sex, age, education level, location in a large city) that are known correlates of environmental concerns. In addition, we model Inglehart’s (1997) short (4 item) postmaterial values measure. Inglehart’s short values index is constructed from selections of two national priorities from the following list 1. 2. 3. 4.

Maintain order in the nation Give people more say in government decisions Fight rising prices Protect freedom of speech

Those choosing options 1 and 3 in any order are deemed ‘materialists’ (scored 1) while options 2 and 4 are ‘postmaterialists’ (scored 1). All other combinations are scored 0 and classified as ‘mixed’ values (Inglehart, 1997: 389). Postmaterialists tend to prioritise quality of life issues such as the environment to a greater extent than materialists who are more likely to be concerned with economic growth and national security (Inglehart and Welzel, 2005). While Inglehart’s values change hypothesis has been challenged (e.g. Clarke et al., 1997, 1999; Davis and Davenport, 1999; Clarke, 2000), it has also been defended rigorously by Inglehart and his collaborators (see Abramson, 2011 for a summary of responses). Nevertheless, the short values index predicts variations in support for environmental issues (Crook and Pakulski, 1995; Pakulski et al., 1998) and environmental behaviour (Inglehart, 1990, 1997; Dalton, 2005; Tranter, 2010) based upon survey data from a range of countries. We measure scientific solutions to environmental problems with an item that taps agreement or strong agreement with the statement ‘Modern science will solve our environmental problems with little change to our way of life’. Those who believe that scientific solutions to climate change are possible are also expected to believe in the claims made by scientists per se, including the predictions of climate scientists. Ceteris paribus, they should also be less likely to be climate sceptics. Self-assessed knowledge of environmental problems is measured with the item ‘how much do you feel you know about solutions to these sorts of environmental problems?’ The question ranges from 1 ‘know nothing at all’ to 5 ‘know a great deal’ represent high self-assessed knowledge (scored 4 + 5 = 1; 1–3 = 0). The expectation here is that those who believe they understand how to solve environmental problems should be less concerned about climate change. General environmental concern is also entered into the regression equation as a control variable when modelling climate sceptics 1, measured with the question ‘Generally speaking, how concerned are you about environmental issues?’ (responses: 1 not at all concerned to 5 very concerned). Low concern is modelled by contrasting options 1 and 2 on this scale (scored 1) with other responses (3 to 5 scored 0). Low concern for environmental issues is a control variable in Tables 2 and 3 as ‘environmental awareness’ is a known predictor of environmental scepticism (Engels et al., 2013). Citing Cvetkovich and Lofstedt (1999) and Rohrmann and Renn (2000), Lorenzoni and Pidgeon (2006: 85) argue ‘[T]rust in institutional performance is another major influence on people’s responses to risk’. Trust in political institutions is potentially important in relation to public views on climate change as it ‘reflects people’s confidence in both the expertise and actions of agencies and institutions that initiate and control risk’ (2006: 85). We model lack of trust in government as a potential predictor of scepticism; those less trusting are hypothesised to hold more sceptical positions on climate change. The question wording is ‘[M]ost of the time we can trust people in government to do what is

159

right’ (in some countries the phrase ‘of any party’ was added after ‘government’ to broaden the question beyond the current government). Agree and strongly agree responses are coded 1; either disagree, strongly disagree or neither agree nor disagree responses coded 0. We operationalise two attitudinal questions from the ISSP as proxies for capturing the different worldviews Kahan et al. (2012) argue have a mediating influence on attitudes towards climate change. These are ‘Private enterprise is the best way to solve [country’s] economic problems’ and ‘It is the responsibility of the government to reduce the differences in income between people with high incomes and those with low incomes’ (5 point scales from strongly agree to strongly disagree). Those who strongly agree or agree are coded 1, otherwise 0. The expectation here is that those holding ‘individualistic’ worldviews (i.e. disagree with governments redistributing income) and believe in the importance of the free market (i.e. emphasise the importance of private enterprise) should be more sceptical than others. As noted above, there is mounting evidence from a range of countries of the political divide over climate change, with supporters of conservative political parties far more likely than those who identify with progressive parties to reject global warming influenced by human behaviour. The political partisan phenomenon maps onto the left-right political dimension, with the ‘left’ more supportive of action on climate change than the ‘right’. We therefore include a variable from the ISSP that measures political party identification aligned on a left-right scale. The ISSP investigators present a range of scales for parties in different countries. For some countries the parties are placed on a scale that ranges from far left to far right (e.g. 1 = far left to 5 = far right), while in others, such as the USA, the scale has three positions (i.e. 1 = Democrat, 2 = other, 3 = Republican). To consistently measure political party left-right position across countries, this variable is coded as a three point scale; parties of the left or far left are coded 1, the right or far right coded 1 and other responses coded 0. As the postmaterial values dimension is claimed to be a relatively consistent predictor of environmental support in Western countries (Inglehart and Welzel, 2005) and is hypothesised be an important measure of climate scepticism, the analyses are limited to Nordic countries, Western Europe, North America and the antipodes, as these have relatively high levels of GDP per capita, which is linked with higher levels of economic security and postmaterial values (Inglehart, 1997: 150). Estimates for each independent variable control for other independent variables with odds ratios presented in the regression tables. 4. Results If climate sceptics approximate those who are not concerned about environmental issues, then while we may expect to find cross national variations in levels of scepticism and environmental concerns, the patterns between the two measures should be relatively consistent within countries. That is to say, where scepticism in a given country is high, environmental concern should be low. The results in Fig. 1 indicate considerable variation across countries in the proportions of those unconcerned about the environment and climate sceptics. They also suggest that these two measures are not strongly correlated. While we expect that the environmentally unconcerned may well be sceptics, the Pearson’s correlation r = 22 between these variables for the pooled sample of 14 countries indicates a modest association, suggesting climate sceptics are not merely those who care little about the environment. Those least concerned about the environment are found in Denmark, Sweden, Austria and the United Kingdom, with the

160

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

Fig. 1. Not concerned or not very concerned about the environment and climate sceptics (%).

highest environmental concern apparent in Canada, New Zealand and Switzerland. Australia has the highest proportion of sceptics at 17%, followed by Norway (15%), New Zealand (13%) and the USA (12%). At the other extreme, only 2% of the Spanish and 4% of Germans and Swiss are climate sceptics, while the proportion of sceptics in all other countries is only 10% or less. Based upon research findings on climate sceptics from the United States (e.g. McCright and Dunlap, 2011a, 2011b) and the failure of that country to adopt the Kyoto Protocol, the USA may be expected to be home to the largest proportion of sceptics. Although these findings suggest citizens of the USA are indeed sceptical of climate change in comparative terms, levels of scepticism for the United States are somewhat lower than Australia and Norway. At the other extreme, our findings support Engels et al. (2013: 1023), who found that in Germany ‘sceptics form a relatively small group’. Given the high levels of postmaterial value orientations in Germany (Inglehart and Welzel, 2005: 90), we expected to find low levels of scepticism, with scepticism expected to be associated with materialist rather than postmaterial values. There is also a potential association here for countries with high levels of carbon dioxide emissions and climate scepticism. For example, the USA, Australia, Canada and Norway appear at the top of the list of CO2 emitting countries (i.e. among the 14 ISSP countries analysed here) based upon World Bank figures for 2010 (World Bank, 2015). Perceived vulnerability to the risks associated with climate change are also expected to be associated with lower levels of climate scepticism, although Kvaløy et al. (2012: 17) found that the ‘risk of being affected by climate-related natural disasters is negatively correlated with level of concern for global warming’.3 We return to analyse this potential association below when we model CO2 emissions (Kim and Wolinsky-Nahmias, 2014) and climate vulnerability measured with the Global Climate Risk Index (Kreft et al., 2015) with multilevel regression analysis.

3 We thank an anonymous reviewer for highlighting these points and inspiring us to examine the association between CO2 emissions, climate vulnerability and scepticism empirically.

4.1. Environmental concern Turning now to the regression analyses, to what extent are the hypotheses regarding environmentalists and sceptics having approximately opposite social and political profiles supported empirically? To examine this question, in Table 1 the dependent variable measuring the statement ‘How concerned are you about environmental issues?’ (1–5 not at all concerned to very concerned) is considered. Disagreement with the statement is modelled using ordinal logistic regression and the independent variables outlined above. Statistical significance is reported at p < 05 or better, although given the relatively large country samples, the 0.01 and 0.001 levels of statistical significance indicate more reliable estimates. The odds ratios suggest men are less likely than women to be concerned about environmental issues in all countries apart from Denmark and the USA. Age is an inconsistent predictor, with older citizens more environmentally concerned in Austria, Canada and Denmark, but younger people more concerned in France, Germany and Spain. Tertiary education is somewhat more consistent as a predictor, although positive effects are only apparent in five countries. As Hamilton (2010) argues, ‘the social bases of environmental concern could be changing. Education no longer has a simple positive effect on concern.’ Postmaterialist effects are apparent in half of the countries examined (along with several other weaker positive, but non-significant effects). Believing that science will solve environmental problems tends to be associated with lower environmental concern, although in less than half of the countries analysed here. The association between political parties on the left of the political spectrum is a far more consistent indicator of environmental concern, while in all countries with the exception of Australia, self-assessed knowledge of the solutions to environmental problems is associated with increased concern over environmental issues. These results are largely in line with previous studies of environmental issues, although in some instances, such as age and education, the magnitudes of effects are modest. Finally, environmental concern is positively associated with believing that governments should be involved in the redistribution of income, although estimates for the ‘private enterprise’ variable are generally weak and non-significant.

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

4.2. Climate sceptics The results in Table 2 examine the social and political background of climate sceptics. These binary logistic regression estimates indicate some relatively consistent predictors of climate scepticism. Respondent sex shows statistically significant estimates for nine of the fourteen countries, with men more likely than women to be sceptical about the danger of climate change. Age is not a particularly consistent predictor of climate scepticism, nor it appears is education, city location, religious orientation or postmaterial values. In the case of postmaterial values, the only statistically significant findings emerge for the USA and Spain. Yet contrary to theoretical expectations, the odds ratios for American and Spanish postmaterialist value orientations are larger than unity, indicating a potential postmaterial association with scepticism in these countries. This is explored further in Table 3. Politically, those who identify with parties of the left are more likely than supporters of right aligned parties to hold that climate change is dangerous for the environment, while those who place little trust in governments are more sceptical in half of these countries. Having little concern for environmental issues produces large odds ratios that are statistically significant at 0.01 or better for all but two of the 14 countries. There is some support also for the hypothesis that individualistic values are associated with higher levels of scepticism, although in less than half of the countries analysed. The discrepancies between the results in Tables 1 and 2 suggest climate sceptics are not merely anti- or non-environmentalists. While sex and political orientations show opposite patterns (i.e. men and the right tend to be sceptics but are not environmentally supportive), age, education and value orientations do not. So far the results suggest that while similar to McCright and Dunlap’s (2011b) findings on the political conservatism and gender among US sceptics (the ISSP does not measure race), climate sceptics, are not necessarily older in cross national perspective, nor is there evidence that materialists are more likely than postmaterialists to be climate sceptics. Lack of trust in governments, but positive attitudes towards private enterprise both tend to increase scepticism in several countries. Tertiary education, in most countries has a negative association with scepticism but only reaches statistical significance for Australia, Norway and the UK. We also test for interactions between education and political party identification, for individual countries (hypothesis 12). While all countries apart from Spain and Switzerland have odds ratios for the interaction term that are larger than one, only Australia and Norway reached statistical significance (p < 0.05). The interaction term is also included in the multilevel models in Table 3, where it does show a significant interaction effect for tertiary education*party. 4.3. Multilevel analysis of scepticism While in Tables 1 and 2 we attempted to illustrate how the independent variables are associated with scepticism within countries, in Table 3 we combine data from 14 countries (N 19, 991) and estimate multilevel binary regression models (using SAS PROC GLIMMIX). A multilevel approach is appropriate for these analyses as individuals are nested within countries in the ISSP data, while scepticism appears to vary at both the individual and country level. If significant country level variation in scepticism is evident, this should be accounted for when considering how our independent variables impact upon climate scepticism as this generates ‘statistically efficient estimates of regression coefficients’ and ‘correct standard errors, confidence intervals and significance tests’ (Goldstein, 2003: 3). The multilevel approach

161

also enables us to assess the potential impact of two country level variables, CO2 emissions per capita and climate vulnerability.4 Both climate change scepticism variables are modelled in Table 3.5 We begin by estimating the amount of variance in scepticism at the country level (i.e. an unconditional model with no predictors). Intra-class correlations (ICC) indicate that nine per cent of the variation in the first measure of climate scepticism is accounted for at the country level while seventeen per cent of the variation in sceptics 2 is ‘explained’ at the country level.6 The statistically significant intercept for Models 1 and 4 suggests that a multilevel model is appropriate to analyse these data for each dependent variable. We next examine a random intercept model with individual level independent variables and the two country level variables added as fixed effects in Models 2 and 5. Models 3 and 6 then include an interaction term for degree*party. Several of the predictors of scepticism show similar effects for both climate scepticism dependent variables. Gender is a statistically significant predictor of climate scepticism after adjusting for country level variation, as is age, party affiliation, trust in government and the proxy worldviews measures – ‘private enterprise solves economic problems’ and ‘reducing income inequality’. The associations of these independent variables with scepticism are in the expected direction for both dependent variables, supporting hypotheses 1, 2, 3, 8 and 11. Hypothesis 6 regarding the impact of (no) religion is not supported, while other independent variables show statistically significant odds ratios for one dependent variable only. For sceptics 1, high educational achievement exhibits the expected negative association with climate scepticism, as does living in a large city (hypothesis 4). Being unconcerned about environmental issues (as sceptics 2 is constructed using the environmental issues variable, the latter is omitted from Models 5 and 6) shows a strong positive association with scepticism. Believing that science can solve environmental problems (Models 2 and 3) and self-assessed knowledge of solving environmental problems (in Models 5 and 6) are associated with higher levels of scepticism, supporting hypotheses 6 and 7. The odds ratios for postmaterial values scale are positively signed, while weak for sceptics 1, but show stronger and statistically significant associations with the second dependent variable in Models 5 and 6. This finding is contra to our expectations (hypothesis 9) that materialists should be more sceptical than are postmaterialists. The reason for this positive association between postmaterialism and scepticism is not entirely clear. However, analysis of USA ISSP data (not shown), suggests that republican leaning independents and strong republican identifiers are much more likely than any other party identifiers to be postmaterialists. This effect in the US may reflect the way some republicans (i.e. perhaps Tea Party in particular) are attracted to the ‘more say in government decision making’ and ‘freedom of speech’ items that comprise the postmaterial values index, and subsequently classified as postmaterialists. Our findings suggest that postmaterial values, at least as they are measured by the short values index (Inglehart, 1997: 389), contribute little to explanations of climate scepticism. The positive associations between postmaterialism and scepticism question the efficacy of the postmaterial thesis as an account of climate scepticism. Unlike Kvaløy et al. (2012: 18) who also employed a random intercept multilevel model to examine concern for global warming, 4

CO2 and the Global Climate Risk Index are centred variables. Given its construction, the ‘Sceptics 2’ variable is based upon a smaller sample than ‘Sceptics 1’ and is therefore only analysed with the combined sample of 14 countries in the multilevel analyses. 6 By contrast, twenty per cent and twenty four per cent of the variation for sceptics 1 and 2 respectively was accounted for by country level variation using the full sample of 33 ISSP countries. 5

162

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

we found the country level independent variable CO2 emissions per capita to be a significant predictor of climate scepticism. A measure of climate vulnerability measured with the Global Climate Risk Index (Kreft et al., 2015) is also significant for the second sceptics variable only, suggesting that hypothesis 13 is supported while 14 is not. The latter finding, however, is in line with Kvaløy et al. (2012: 17–18) who found that ‘people living in countries that are highly exposed to. . .natural disasters are less concerned about global warming than those living in less disaster prone countries’. Finally, the multilevel regression results indicate that having a degree reduces the odds of climate scepticism, however, the interaction term indicates that tertiary educated right wing party identifiers are more likely to be climate sceptics (supporting hypothesis 12).7 5. Conclusion Our research based on analysis of 14 advanced industrialised nations from the International Social Survey Programme shows substantial cross-national differences in the levels of climate scepticism and the social and political background of climate sceptics. As Engels et al. (2013) observe, the factors that correlate with climate scepticism appear to vary according to the political and cultural context of each country. Contrary to expectations, climate sceptics are not merely the mirror image of environmentalists. Some predictors of environmental issue concern, such as postmaterial value orientations, prove to be poor predictors of climate change scepticism within countries, while lack of concern over environmental issues and low trust in governments are associated with higher levels of scepticism. The social and political background of environmental issue concern follows previous studies. For example, across 12 of the 14 nations analysed, men were less likely than women to be concerned about environmental issues. In addition, political party affiliation and postmaterial value orientations were relatively consistent predictors of environmental issue concern (see also Inglehart, 1990; Dalton, 2005; Tranter, 2013), as was high selfassessed knowledge of the solutions to environmental problems. Yet only political orientations (conservative), gender (male) and being unconcerned about environmental issues are relatively consistent predictors of climate scepticism on a country by country basis. McCright and Dunlap have shown gender and political party allegiance to be important correlates of attitudes towards climate change in the USA (Dunlap and McCright, 2008; McCright and Dunlap, 2010, 2011a, 2011b; McCright, 2010). Our findings show they are also consistent predictors of climate scepticism cross nationally. Believing science will solve environmental problems, selfreported knowledge of the solutions to environmental problems, and tertiary level education per se, are all inconsistent predictors of climate scepticism in the country by country analyses. Hamilton’s (2010) country level results regarding the interaction of tertiary educated supporters of conservative parties as climate sceptics are also supported by our multilevel findings (although not within individual countries), indicating the rejection of climate science by (presumably) quite well informed, conservatively oriented citizens is apparent cross-nationally, albeit relatively modest in magnitude. Yet this does not necessarily indicate a lack of understanding of climate science among educated conservatives. As Kahan (2010: 296) suggest, those who 7 We also tested an interaction term for men by political party affiliation to consider the multiplicative association of right wing males in the pooled dataset (not presented here). However McCright and Dunlap’s (2011b) finding for the USA was not replicated in the multi-level analyses (non-significant effect p > 0.05).

hold individualistic as opposed to collectivist worldviews are highly likely to reject the proposition of anthropogenic climate change ‘because the widespread acceptance of such evidence would lead to restrictions on commerce and industry, activities they admire’. A related account advanced by Dunlap and others as the ‘anti reflexivity thesis’ (Dunlap, 2014; McCright et al., 2013) argues that ‘an ‘anti-reflexive’ coalition of corporate interests and political conservatives tends to undermine the development and employment of scientific evidence documenting environmental problems in order to defend the current economic system of production’ (Dunlap, 2014). Indeed, we found some evidence to support such claims cross-nationally. Low levels of trust in ‘the people in government’ is associated with climate scepticism, as is valuing private enterprise to solve economic problems. Rejecting government intervention in the reduction of income inequality also reflects laissez-faire attitudes towards ‘big government’ among climate sceptics. Coupled with findings that climate scepticism is higher among the ideologically conservative, our research buttresses that of Heath and Gifford (2006: 64–65), who found that ‘those who value the free market system over environmental quality tend to believe that global climate change is not occurring, that the causes of global climate change are more natural than human caused, and that its consequences will not be negative’. There is evidence in our analyses too of substantial differences in levels of climate scepticism across nations. Following a study of climate scepticism in Germany, Engels et al. (2013: 1025) speculated: ‘[S]kepticism might thus be a phenomenon of the Anglo-American cultural sphere rather than a worldwide trend’. While there are very low levels of scepticism in Spain, Germany and Switzerland, and high levels in Australia, New Zealand, and the USA, countries with relatively high levels of postmaterial value orientations such as Sweden and Finland (Inglehart, 1997: 151; Inglehart and Welzel, 2005) whose citizens tend to espouse pro-environmental attitudes were more sceptical than expected. Norway was a particular exception here as the second most sceptical of all countries studied. These countries are also high on the list of CO2 emitting countries, with higher levels of CO2 emissions per capita associated with higher levels of climate scepticism. Perceptions of the risks associated with climate change are also correlated positively with increased climate scepticism (O’Brien et al., 2007), suggesting the need for further research in this area. At present cross-national research regarding public opinion on climate change (as distinct from public scepticism) and structural issues utilises varying measures and appears inconclusive. For example, Kvaløy et al. (2012) found no significant relationship across nations between wealth and CO2 emissions, and Kim and Wolinsky-Nahmias (2014) demonstrated concern about climate change to be higher in developing countries with levels of concern to be little effected by climate vulnerability. However, we did find significant associations between scepticism and the Global Climate Risk Index (Kreft et al., 2015). Our findings on cross national variations in climate scepticism largely support the patterns found by Toke and Vezirgiannidou’s (2013) observations of cross-national differences in associations between energy security and climate change. There are also substantial differences between ‘green’ European countries such as Germany, and the USA, where climate change and climate protection are typically given very low priority. Vezirgiannidou (2013: 605) argues that in the USA ‘the message is no longer that climate change needs to be addressed, but that economic growth and energy security can be addressed in more climate-friendly ways. Thus, the message of climate change loses its potency’. Such observations have led to an assumption that the USA is leading the charge of rising climate scepticism. Yet based upon ISSP data, the

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

citizens of Australia, New Zealand and Norway are at least as sceptical, if not more so, than Americans, which warrants further investigation. Climate scepticism persists despite overwhelming scientific evidence that anthropogenic climate change is occurring (IPPC, 2013). The reasons for this are varied and complex. Understanding why climate scepticism endures or is even on the rise, and why levels of scepticism vary across countries, requires accounts that recognise that biased information assimilation is mediated by cross-cultural and intra-national differences in values and worldviews. Fruitful explanations of scepticism must also account for the way in which partisans are influenced by their political leaders. Integrating such accounts may provide a way to both understand and address the social problem of climate scepticism. Acknowledgements This research received funding from the Institute for the Study of Social Change at the University of Tasmania. We thank the journal editors and the anonymous reviewers for their helpful comments and suggestions. Appendix. Descriptive statistics Mean Dependent variables Climate change sceptics 1 Climate change sceptics 2

0.09 0.35

Standard Min deviation 0.29 0.48

Country level independent variable CO2 emissions per 3.71  108 4.26 capita (2010) 2.49  108 35.73 Global climate risk index Individual level independent variables Men 0.48 Aged 65+ 0.23 Degree 0.21 Live in large city 0.21 Non-religious 0.25 Postmaterial values 0.08 Party left–right 0.03 Distrusts government 0.43 Science solves 0.21 environmental problems Not concerned about 0.12 environmental issues 0.21 Knowledge of solutions to environmental problems 0.41 Private enterprise solves economic problems 0.57 Government should reduce income inequality

0 0

4.52

Max

1 1

N

19,991 2194

8.08 19,991

31.13 77.19 19,991

0.50 0.42 0.41 0.41 0.43 0.57 0.75 0.49 0.41

0 0 0 0 0 1 1 0 0

1 1 1 1 1 1 1 1 1

19,991 19,991 19,991 19,991 19,991 19,991 19,991 19,991 19,991

0.33

0

1

19,991

0.40

0

1

19,991

0.49

0

1

19,991

0.50

0

1

19,991

References Abramson, P., 2011. Critiques and Counter-Critiques of the Postmaterialism Thesis: Thirty-four Years of Debate. CSD Working Papers. Center for the Study of Democracy, UC Irvine. , http://escholarship.org/uc/item/3f72v9q4 (accessed 09.04.14). Agresti, A., Finlay, B., 1999. Statistical Methods for the Social Sciences. Prentice Hall, New Jersey. Bacon, W., 2013. Sceptical Climate. Part 2: Climate Science in Australian Newspapers. Australian Centre for Independent Journalism, http://investigate.org. au/wp-content/uploads/2013/10/Sceptical-Climate-Part-2-Climate-Sciencein-Australian-Newspapers.pdf.

163

Bartels, L., 2000. Partisanship and voting behavior 1952–1996. Am. J. Polit. Sci. 44 (1), 35–50. Brechin, S., Bhandari, M., 2011. Perceptions of climate change worldwide. WIREs Clim. Change 2, 871–885. Brody, S.D., Zahran, S., Vedlitz, A., Grover, H., 2008. Examining the relationship between physical vulnerability and public perceptions of global climate change in the United States. Environ. Behav. 40 (1), 72–95. Brulle, R., Carmichael, J., Jenkins, J., 2012. Shifting public opinion on climate change: an empirical assessment of factors influencing concern over climate change in the U.S. 2002–2010. Clim. Change 114, 169–188. Campbell, A., Converse, P., Miller, W., Stokes, D., 1960. The American Voter. University of Chicago Press, Chicago. Clarke, H., 2000. Refutations affirmed: conversations concerning the euro-barometer values battery. Polit. Res. Q. 53 (3), 477–494. Clarke, H., Kornberg, A., McIntyre, C., Bauer-Kaase, P., Kaase, M., 1999. The effect of economic priorities on the measurement of value change: new experimental evidence. Am. Polit. Sci. Rev. 93 (3), 637–647. Clarke, H., Dutt, N., Rapkin, J., 1997. Conversations in context: the (mis)measurement of value change in advanced industrialised societies. Polit. Behav. 19, 19–40. Clements, B., 2012. The sociological and attitudinal bases of environmentallyrelated beliefs and behaviour in Britain. Environ. Polit. 21 (6), 901–921. Corbett, J., Durfee, J., 2004. Testing public (un)certainty of science: media representations of global warming. Sci. Commun. 26, 129–151. Corner, A., Whitmarsh, L., Xenias, D., 2012. Uncertainty, scepticism and attitudes towards climate change: biased assimilation and attitude polarisation. Clim. Change 114, 463–478. Crona, B., Wutich, A., Brewis, A., Garint, M., 2013. Perceptions of climate change: linking local and global perceptions through a cultural knowledge approach. Clim. Change 119, 519–531. Crook, S., Pakulski, J., 1995. Shades of green: public opinion on environmental issues in Australia. Aust. J. Polit. Sci. 30 (1), 39–55. Dalton, R., 2005. The greening of the globe? Cross national levels of environmental group membership. Environ. Polit. 44 (4), 441–459. Davis, D., Davenport, C., 1999. Assessing the validity of the postmaterialism index. Am. Polit. Sci. Rev. 93 (3), 649–664. Dessai, S., Adger, W., Hulme, M., Turnpenny, J., Ko¨hler, J., Warren, R., 2004. Defining and experiencing dangerous climate change: and editorial essay. Clim. Change 64 (1), 11–25. Downing, P., Ballantyne, J., 2007. Tipping Point or Turning Point: Social Marketing and Climate Change. Ipsos-MORI Social Research Institute, http://www. ipsos-mori.com/DownloadPublication/ 1174_sri_tipping_point_or_turning_point_climate_change.pdf. Dunlap, R., 2014. Clarifying anti-reflexivity: conservative opposition to impact science and scientific evidence. Environ. Res. Lett., http://dx.doi.org/10.1088/ 1748-9326/9/2/021001. Dunlap, R., McCright, A., 2008. A widening gap: republican and democratic views on climate change. Environment 50 (5), 26–35. Dunlap, R., Xiao, C., McCright, A., 2001. Politics and environment in America: partisan and ideological cleavages in public support for environmentalism. Environ. Polit. 10 (4), 23–48. Engels, A., Hu¨ther, O., Scha¨fer, M., Held, H., 2013. Public climate-change scepticism, energy preferences and political participation. Glob. Environ. Change 23, 1018– 1027. Fielding, K., Head, B., Laffan, W., Western, M., Hoegh-Guldberg, O., 2012. Australian politicians’ beliefs about climate change: political partisanship and political ideology. Environ. Polit. 21 (5), 712–733. Fiorina, M., Abrams, S., 2008. Political polarization in the American public. Annu. Rev. Polit. Sci. 11, 563–588. Franzen, A., Vogl, D., 2013. Two decades of measuring environmental attitudes: a comparative analysis of 33 countries. Glob. Environ. Change 23 (5), 1001– 1008. Gendal, P., 2012. International Social Survey Programme Study Monitoring 2010 Environment III, http://www.gesis.org/issp/issp-modules-profiles/environment/ 2010/ (accessed 27.08.14). Gilens, M., Murakawa, N., 2002. Elite cues and political decision making. In: DelliCarpini, M., Huddy, L., Shapiro, R. (Eds.), Political Decision-Making, Deliberation and Participation. JAI, Oxford. Givens, J., Jorgenson, A., 2011. The effects of affluence, economic development, and environmental degradation on environmental concern: a multilevel analysis. Organ. Environ. 24 (1), 74–91. Goldstein, H., 2003. Multilevel Statistical Models, 4th ed. John Wiley and Sons, West Sussex. Green, D., Palquist, B., Schickler, E., 2002. Partisan Hearts and Minds. Yale University Press, New Haven. Hamilton, L., Cutler, M., Schaefer, A., 2012. Public knowledge and concern about polar-regions warming. Polar Geogr. 35 (2), 155–168. Hamilton, L., 2010. Education, politics and opinions about climate change: evidence for interaction effects. Clim. Change, http://dx.doi.org/10.1007/s10584-0109957-8. Heath, Y., Gifford, R., 2006. Free-market ideology and environmental degradation: the case of belief in global climate change. Environment and Behaviour 38 (1), 48–71. Hindman, D., 2009. Mass media flow and differential distribution of politically disputed beliefs: the belief gap hypothesis. Journal. Mass Commun. Q. 86, 790– 808.

164

B. Tranter, K. Booth / Global Environmental Change 33 (2015) 154–164

Hmielowski, J., Feldman, L., Myers, T., Leiserowitz, A., Maibach, E., 2013. An attack on science? Media use, trust in scientists, and perceptions of global warming. Public Underst. Sci., http://dx.doi.org/10.1177/0963662513480091. Hobson, K., Niemeyer, S., 2012. What sceptics believe: the effects of information and deliberation on climate change scepticism. Public Underst. Sci. 22 (4), 396–441. Inglehart, R., 1997. Modernization and Postmodernization: Cultural, Economic and Political Change in 43 Societies. Princeton University Press, New Jersey. Inglehart, R., 1990. Values, ideology and cognitive mobilization in new social movements. In: Dalton, R., Kuechler, M. (Eds.), Challenging the Political Order. Polity Press, Cambridge, pp. 43–66. Inglehart, R., Welzel, C., 2005. Modernization, Cultural Change and Democracy, The Human Development Sequence. Cambridge University Press, New York. International Panel on Climate Change (IPPC), 2013. Climate Change 2013: The Physical Science Basis, http://www.ipcc.ch/report/ar5/wg1/ (accessed 15.04.14). ISSP Research Group, 2012. International Social Survey Programme: Environment III – ISSP 2010. GESIS Data Archive, Cologne. ZA5500 Data file Version 2.0.0, http:// dx.doi.org/10.4232/1.11418. Jacques, P., Dunlap, R., Freeman, M., 2008. The organization of denial: conservative think tanks and environmental scepticism. Environ. Polit. 17, 349–385. Kahan, D., Peters, E., Wittlin, M., Slovic, P., Ouellette, L., Braman, D., Mandel, G., 2012. The polarizing impact of science literacy and numeracy on perceived climate change risks. Nat. Clim. Change, http://dx.doi.org/10.1038/NCLIMATE1547. Kahan, D., 2010. Fixing the communications failure. Nature 463, 296–297. Kim, S.Y., Wolinsky-Nahmias, Y., 2014. Cross-national public opinion on climate change: the effects of affluences and vulnerability. Glob. Environ. Polit. 14 (1), 79–106. Kreft, S., Eckstein, D., Junghans, L., Kerestan, C., Hagen, U., 2015. Global Climate Risk Index, http://germanwatch.org/en/download/10333.pdf (accessed 13.02.15). Kvaløy, B., Finseraas, H., Listhaug, O., 2012. The publics’ concern for global warming: a cross-national study of 47 countries. J. Peace Res. 49 (1), 11–22. Leiserowitz, A., Maibach, E., Roser-Renouf, C., Smith, N., 2011. Climate Change in the American mind: Americans’ Global Warming Beliefs and Attitudes in May 2011. Yale Project on Climate Change Communications, Yale University and George Mason University, New Haven, CT. http://environment.yale.edu/climatecommunication/files/ClimateBeliefsMay2011.pdf. Lorenzoni, I., Pidgeon, N., 2006. Public views on climate change: European and USA perspectives. Clim. Change 77 (1–2), 73–95. McCright, A., 2010. The effects of gender on climate change knowledge and concern in the American public. Popul. Environ. 32 (1), 66–87. McCright, A., Dunlap, R., 2010. Anti-reflexivity: the American conservative movement’s success in undermining climate science and policy. Theory Cult. Soc. 27 (2–3), 100–133. McCright, A., Dunlap, R., 2011a. The politicization of climate change and polarization in the American public’s views of global warming, 2001–2010. Sociol. Q. 52, 155–194. McCright, A., Dunlap, R., 2011b. Cool dudes: the denial of climate change among conservative white males in the United States. Glob. Environ. Change 21 (4), 1163–1172. McCright, A., Dunlap, R., 2013. Bringing ideology in: the conservative white male effect on worry about environmental problems in the USA. J. Risk Res. 16 (2), 211–226. McCright, A., Dentzman, K., Charters, M., Dietz, T., 2013. The influence of political ideology on trust in science. Environ. Res. Lett. 8 (4) , http://iopscience.iop.org/ 1748-9326/8/4/044029/pdf/1748-9326_8_4_044029.pdf (accessed 16.04.14).

MORI, 2005. Science in Society: Findings From Qualitative and Quantitative Research. MORI, London. Miller, W., 1976. The cross-national use of party identification as a stimulus to political inquiry. In: Budge, I., Crewe, I., Farlie, D. (Eds.), Party Identification and Beyond. Wiley and Sons, London, pp. 21–31. Muller-Rommel, F., 1990. New political movements and ‘new politics’ parties in Western Europe. In: Dalton, R., Kuechler, M. (Eds.), Challenging the Political Order: New Social and Political Movements in Western Democracies. Oxford University Press, New York, pp. 209–231. Nawrotzki, R., 2012. The politics of environmental concern: a cross-national analysis. Organ. Environ. 25 (3), 286–307. O’Brien, K., Eriksen, S., Nygaard, L.P., Schjolden, A., 2007. Why different interpretations of vulnerability matter in climate change discourses. Clim. Policy 7 (1), 73–88. Pakulski, J., Tranter, B., Crook, S., 1998. Dynamics of environmental issues in Australia: concerns, clusters and carriers. Aust. J. Polit. Sci. 33 (2), 235–253. Papadakis, E., 1993. Politics and the Environment: The Australian Experience. Allen and Unwin, Sydney. Poortinga, W., Spence, A., Whitmarsh, L., Capstick, S., Pidgeon, N., 2011. Uncertain climate: an investigation into public scepticism about anthropocentric climate change. Glob. Environ. Change 21, 1015–1024. Poortinga, W., Pidgeon, N., Lorenzoni, I., 2006. Public Perceptions of Nuclear Power, Climate Change and Energy Options in Britain: Summary Findings of a Survey Conducted during October and November 2005. Understanding Risk Working Paper 06-02. School of Environmental Sciences, University of East Anglia, Norwich. Ratter, B., Philipp, K., von Storch, H., 2012. Between hype and decline: recent trends in public perception of climate change. Environ. Sci. Policy 18, 3–8. Sciarini, P., Bornstein, N., Lanz, B., 2007. The determinants of voting choices on environmental issues: a two-level analysis. In: de Vreese, C. (Ed.), The Dynamics of Referendum Campaigns: An International Perspective. Palgrave Macmillan, New York. Schuldt, J.P., Konrath, S.H., Schwarz, N., 2011. ‘‘Global warming’’ or ‘‘climate change’’? Whether the planet is warming depends of question wording. Public Opin. Q. 75 (1), 115–124. Scruggs, L., Benegal, S., 2012. Declining public concern about climate change: can we blame the great recession? Glob. Environ. Change 22, 505–515. Toke, D., Vezirgiannidou, S., 2013. The relationship between climate change and energy security: key issues and conclusions. Environ. Polit. 22 (4), 537–552. Tranter, B., 2010. Environmental activists and non-active environmentalists in Australia. Environ. Polit. 19 (3), 413–429. Tranter, B., 2011. Political divisions over climate change. Environ. Polit. 20 (1), 78–96. Tranter, B., 2013. The great divide: political candidate and voter polarisation over global warming in Australia. Aust. J. Polit. Hist. 59 (3), 397–413. Vezirgiannidou, S., 2013. Climate and energy policy in the United States: the battle of ideas. Environ. Polit. 22 (4), 593–609. Whitmarsh, L., 2009. What’s in a name? Commonalities and difference in public understanding of ‘‘climate change’’ and ‘‘global warming’’. Public Underst. Sci. 18, 401–420. Whitmarsh, L., 2011. Scepticism and uncertainty about climate change: dimensions, determinants and change over time. Glob. Environ. Change 21, 690–700. Wood, D., Vedlitz, A., 2007. Issue definition, information processing and the politics of global warming. Am. J. Polit. Sci. 51 (3), 552–568. World Bank, 2015. CO2 emissions (metric tons per capita), http://data.worldbank. org/indicator/EN.ATM.CO2E.PC (accessed 09.02.15).