Bachelor thesis Political Science – International Relations track
Cool Dudes in the Netherlands
Climate change denial among conservative Dutch men
January 2021
Name:
Sander Bernhard
Student number
11272139
Supervisor:
Joost Berkhout
Second reader:
Eelco Harteveld
Word count:
5616
Abstract
Studies in the United States and Norway have shown that conservative white males are more likely to deny climate change science than other individuals. This thesis aims to replicate these studies’ research design to examine whether this effect is also apparent in the Netherlands. Previous literature is used to describe the reasoning behind this observed effect. Data from the European Social Survey is used to analyse climate change beliefs in the Netherlands. This study found that contrasting to the US climate change denial is less common, and conservative white males had a moderate relation to only one denial indicator. The logistic regression analysis showed that the most important explanatory factors are political ideology, education and age. Most results in this study were not significant; however, there are suggested effects which could be studied in further research using a larger dataset.
Table of Contents
Abstract ... 2
1. Introduction ... 4
2. Literature review ... 5
2.1 The cool dude effect ... 52.2 The Dutch context ... 6
3. Theoretical framework ... 6
3.1 Forms of denial ... 73.2 Explanations for the cool dude effect by McCright & Dunlap ... 7
3.3 Hypothesis ... 8
4. Research design ... 8
4.1 Data ... 94.2 Operationalisation ... 9
Trend denial ... 10
Attribute denial ... 10
Impact denial ... 10
Worry about climate change ... 11
Party identification ... 11
Male ... 12
White ... 12
Cool dude ... 13
Control variables ... 13
4.3 Methods ... 15
5. Analysis and results ... 15
5.1 Conservative white males and climate change beliefs ... 155.2 Predicting factors ... 16
6. Conclusion and discussion ... 19
Biography ... 21
1. Introduction
A changing climate will lead to rising sea levels and more extreme weather events such as droughts and storms (IPCC, 2014). This will harm ecosystems, human health and economies (IPCC, 2014). In 2001, The Intergovernmental Panel on Climate Change (IPCC) stated that it is unequivocal that the world’s climate is changing (IPCC, 2001). Their assessment reports in 2007 and 2013 have reconfirmed this claim (IPCC, 2007, 2013). Furthermore, it is stated that it is virtually certain (a probability of >99%) that human activity is mainly responsible for this change (IPCC, 2001, 2007, 2013). These reports are composed by scientists from all over the world who asses thousands of published scientific papers on climate change (IPCC, n.d.). Despite the abundance of research and scientific consensus on climate change, some individuals are sceptic of these claims.
In the Paris Climate Agreement, it is recognised that climate change poses a significant risk, and therefore, action is needed (Paris Agreement, 2015). Most importantly, countries will take measures to reduce greenhouse gasses emission to limit the increase of the global temperature well below 2°C (Paris Agreement, 2015). Although rapid action is needed, some countries have scaled back their efforts in reducing climate change. This could be seen most notably in the US under the presidency of Donald Trump, who withdrew from the Paris Climate Agreement, but also in Poland, Hungary and Denmark (Lockwood, 2018). Delaying climate action means that emissions have to decrease faster in the future to reach 1,5°C as the risks significantly increase after this level (IPCC, 2018). Krange et al. (2019) state that if climate change is denied for ideological reasons, scientific communication cannot convince deniers otherwise. To adequately address climate change denial, it is crucial to find underlying motivations for this behaviour. Therefore it is essential to understand who are denying the science on climate change as a means to study why they choose to deny climate science. McCright & Dunlap (2011) have observed that in the United States, the most notable climate change deniers such as politicians, contrarian scientists, media personalities, and think tank representatives are conservative white males. McCright & Dunlap (2011) raise the question if this pattern can also be observed among the American public. The pattern in prominent climate change deniers can also be seen in the Netherlands, think of politicians such as Geert Wilders and Thierry Baudet, TV-personality Robert Jensen, and thinktank representative and scientist Marcel Crock. Therefore the question can be raised is also applicable in the Dutch public.
This research aims to replicate the research design of McCright & Dunlap (2011) and Krange et al. (2019) to study the ‘cool dude-effect’. The research question of this thesis will, therefore be:
Does the cool dude-effect exist in the Netherlands?
This question will be answered using data from the European Social Survey (ESS). First, an overview of relevant literature will be provided also relating to the Dutch context. After that, the theoretical framework is introduced, and relevant concepts will be outlined. Then research design used in this thesis is presented. From this follows the analysis in which the results of the logistic regression analysis will be discussed. Finally, the conclusions of this thesis are presented together with a discussion of the results.
2. Literature review
This section will provide an overview of earlier research into climate denial among certain social groups. First, McCright & Dunlap’s research will be discussed, followed by a review of a replication of their study in Norway by Krange et al. (2019). Hereafter literature related to the Dutch context will be reviewed. From this follows the identified gap in the literature and relevance of this study.
2.1 The cool dude effect
McCright & Dunlap (2011) combine race, gender and political ideology in order to study whether conservative white males are more likely to be climate change deniers than other adults in the US In order to answer this question the researchers use a multivariate logistics regression using public opinion data collected by Gallup Inc. Gallup is a private company specialised in opinion polls (Gallup Inc., n.d.). They find that the relationship between the created cool dude dummy and climate change denial indicators is moderate to very strong and significant with a p < 0.001 (McCright & Dunlap, 2011). This stud also reported a positive relationship between conservative white males with a self-reported high understanding of climate change and climate change denial. In contrast, this relationship was weaker among all other adults for one of the indicators of climate change denial, no correlation for the second indicator and a negative relationship for the last two. McCright & Dunlap (2011) suggest further research to determine whether the cool dude effect exists outside of the UUS
This suggestion has resulted in a similar study carried out in Norway. This study by Krange et al. (2019) uses similar variables to replicate the research design by McCright & Dunlap as
closely as possible. Their findings are that conservative Norwegian males more often have climate change denial beliefs than all other adults. This resembles the results of the study by McCright & Dunlap. However, they conclude that being ethnically Norwegian has a weak relationship with climate change denial (Krange et al., 2019). This study also added a variable to test whether xenosceptic views relate to climate change denial; they found a robust significant relation. This means conservative Norwegian males with xenosceptic views are more likely to be climate change deniers than all other adults, as well as conservative Norwegian males who do not have xenosceptic views (Krange et al., 2019).
2.2 The Dutch context
Most research on climate change denial is focused on Anglo-American countries (Björnberg et al., 2017). Especially in the US climate change has become a highly politicised issue (Dunlap & McCright, 2015). In the case of the US, with two dominant parties in the political system, denialist views are more common among the Republican Party. In the Netherlands, climate denial can be found mostly with right-wing populist parties such as Forum voor Democratie (FvD) and Partij Voor de Vrijheid (PVV). These political parties explicitly deny the science on climate change (Haagsma, 2020).
Researchers have recently started to explore the links between far-right and right-wing populist (RWP) parties in Europe, their supporters, and climate change denial (Forchtner, 2019; Lockwood, 2018). A couple of suggested explanations have been given on the association between the far-right/RWP and climate change denial. The first possible explanation is that the far-right/RWP tend to focus on local issues with relevance for the ordinary man and behave hostile towards the cosmopolitan elite and global abstract problems such as climate change (Forchtner, 2019; Lockwood, 2018). A second explanation might be that parties and supporters at the edges of the political spectrum, both left and right, are more likely to believe in conspiracy theories, which can include conspiracy theories stating climate change is a hoax (van Prooijen et al., 2015). Furthermore, lastly, it is suggested that the RWP appeal to voters who feel left behind by globalisation and technological progress resulting in economic insecurity for this group (Lockwood, 2018).
3. Theoretical framework
This part will give an overview of the essential concepts used in this thesis and provide a theoretical background on the observed white male effect. This will lead to the formulation of the hypothesis, which will later be tested in the analysis.
3.1 Forms of denial
Individuals who doubt the evidence presented by climate scientist are called climate change sceptics or climate change deniers. Within this thesis, the concept of climate change denier will be used. However, climate change deniers are not a homogeneous group. There are different forms of climate change denial to consider. The following three are the most notable. The first form of denial is described as trend scepticism. These deniers reject that there is a significant change in the climate at all (Hobson & Niemeyer, 2013; Rahmstorf, 2004). Trend denial is slowly fading as the scientific evidence for climate change is abundantly clear (Rahmstorf, 2004). The second form of denial recognises that the climate is changing but doubt that humans are responsible for this change (Hobson & Niemeyer, 2013; Rahmstorf, 2004). Most of these so-called attribute deniers acknowledge that humans are responsible for the rise in CO2 levels. However, they argue that this is not the cause of the changing climate (Hobson & Niemeyer, 2013; Rahmstorf, 2004). Lastly, there are deniers who that claim the impacts of climate change are not necessarily harmful (Hobson & Niemeyer, 2013; Rahmstorf, 2004). These impact deniers highlight positive impacts such as agricultural potential in higher latitudes but downplay the adverse effects of a rapidly changing climate (Hobson & Niemeyer, 2013; Rahmstorf, 2004). These three forms of denial will be tested separately by means of different indicators. This will be discussed in more detail in the operationalisation.
3.2 Explanations for the cool dude effect by McCright & Dunlap
McCright & Dunlap (2011) base their theory of the cool dude effect on earlier research in the United States on perceptions of (environmental) risks. This research has uncovered that white males accept higher risks than other adults (Finucane et al., 2010; Flynn et al., 1994; Kahan et al., 2007; Kalof et al., 2002; Satterfield et al., 2004). This is described as a so-called white male effect. A biological explanation is rejected in favour of three socio-political explanations. The theory is based on a difference in experienced vulnerability between white males and females and non-whites. This theory states that gender is a potent risk indicator, as well as race to a lesser extent, but the effect of these two demographic variables (especially race) is partly explained by vulnerability indicators (Satterfield et al., 2004). Vulnerability has been found one of the strongest predictors of environmental risk perception (Marshall, 2004). because of their dominant position in the social system, white males are more tolerant of threats as they feel less vulnerable (Marshall et al., 2006). A second theory claims that cultural worldviews such as hierarchicalism, egalitarianism, and individualism shape risk perception (Kahan et al., 2007). It is argued that because white males tend to create, manage, control and benefit of the current system they place more trust in authorities, are more positive towards hierarchy and oppose democratisation of risk management (Flynn et al., 1994).
Lastly, Kahan et al. (2007) also present the identity-protective theory. According to the authors, individuals prefer to accept views held by members of influential in-groups, frequently avoiding revision of those beliefs when faced with contradictory evidence from presumed out-groups. Environmental risks will most likely be downplayed or ignored by white males with a hierarchical cultural worldview as they challenge the existing social, political and economic hierarchy (Kahan et al., 2007).
Next to gender and race, US studies found that conservatives are more likely to deny climate change (Hamilton, 2008; McCright, 2010; Wood & Vedlitz, 2007). On explanation for this relation is that conservatives are more likely to maintain the societal status quo and oppose efforts to change it (Jost et al., 2008). Adding to this, system justification is associated with the denial of problems that threaten system function, such as climate change (Feygina et al., 2010).
McCright & Dunlap (2011) provide two additional arguments for why it is expected that conservative white males are more likely to deny climate change science. First, they state that prominent conservative white males have consistently sent the message that climate change is not real through media for the last 20 years. Conservative white males in the general public perceive these prominents as part of their in-group and are therefore more likely to resonate with them and dismiss climate scientist and other groups. A second explanation given argues that conservative white males have significantly benefitted from the industrial capitalist system and are therefore more likely to defend it. Climate change and mitigation to prevent climate change pose significant challenges to this system and is therefore rejected.
3.3 Hypothesis
Considering the theoretical substantiation of McCright & Dunlap (2011), it could be speculated whether such an effect would also be observable in a context outside of the US, as much of the literature is focussed on the US. Considering the results from the study by Krange et al. (2019) showed a similar effect in a Western-European context the hypothesis of this thesis will be that the effects of political ideology, gender and race will also be observable in the Netherlands.
4. Research design
This section, the research design of this thesis will be discussed which will be used to answer the hypothesis. The research design is based on the research design of McCright & Dunlap
(2011). This is done to study whether the cool dude effect observed in the US can also be seen in the Netherlands. First, the data that is used will be deliberated on. Subsequently, the variables from this dataset that are used in the analysis will be operationalised. Finally, the methods of analysis will be explained.
4.1 Data
This research aims to replicate the research done by McCright & Dunlap (2011) in the context of the Netherlands. The data used in this research is derived from the eight-round of the European Social Survey (ESS). The ESS is a Pan-European research infrastructure which provides data on social attitudes and behaviour for academics, policymakers, civil society and the wider public (European Social Survey, 2013). A survey is organised every two years in member states, which are mostly European countries (European Social Survey, 2013). Respondents are selected by strict random probability, and samples must represent residents aged 15 or above (European Social Survey, 2013).
The ESS is suitable for this research because it combines views on climate change with demographic information and political orientation which is needed to differentiate between conservative white males and others of the population. In addition, this dataset also allows to include demographic control variables such as income and education, among others. The 8th round of the ESS (ESS8), surveyed in 2016/2017 is chosen because in this round the rotating module is on climate change and energy.
Only data gathered in the Netherlands will be used (n = 1681).
4.2 Operationalisation
The ESS8 which is discussed above has similarities to the data used by McCright & Dunlap (2011). The variables that are used in the statistical model will be discussed in more detail, as well as the coding that has been used. Not all variables needed are directly available within the ESS8. Therefore proxies are created to resemble the variables used McCright & Dunlap (2011). Not all variables used in McCright & Dunlap (2011) and Krange et al. (2019) will be studied in this research. In addition, next to studying the cool dude-effect, both studies study a second independent variable in the statistical model. McCright & Dunlap (2011) study white, conservative males with a self-reported high understanding of climate change; this will not be replicated in this study as a result of a limitation in the data. Krange et al. (2019) examine if xenoscepticism adds to the white male effect, this will not be replicated in this study because it does not directly relate to the original study of McCright & Dunlap (2011).
Trend denial
Climate change denial can be manifested in three different types, namely trend denial, attribute denial and impact denial (Hobson & Niemeyer, 2013; Rahmstorf, 2004). These three are all measured separately in the ESS8. Trend denial is operationalised as follows: question D19 is used “Do you think the world’s climate is changing?”. The coding of the answer option can be seen in the table below and is in line with the coding that McCright & Dunlap (2011) used.
Table 1: operationalisation of question D19
Original coding
Answer options Coding used in this thesis
1 Definitely changing 0
2 Probably changing 0
3 Probably not changing 1
4 Definitely not changing 1
6 Not applicable
7 Refusal
8 Do not know
9 No Answer
Attribute denial
For this concept question, D22 is used “Do you think that climate change is caused by natural processes, human activity, or both? The coding of the answer option can be seen in the table below and is in line with the coding that McCright & Dunlap (2011) used.
Table 2: operationalisation of question D22
Original coding
Answer options Coding used in this
thesis
1 Entirely by natural processes 1
2 Mainly by natural processes 1
3 About equally by natural processes and human activity
1
4 Mainly by human activity 0
5 Entirely by human activity 0
55 I do not think climate change is happening
66 Not applicable
77 Refusal
88 Do not know
99 No answer
Impact denial
This question was not included in the research by McCright & Dunlap (2011) or Krange et al. (2019). However, it is one of the three forms of climate change denial described by (Hobson & Niemeyer, 2013; Rahmstorf, 2004). Therefore this question will be used because it is a relevant form of denial. This question also serves as an alternative for the variables’ seriousness of global warming’ which measures whether respondents think climate change is
generally exaggerated in the news and ‘scientific consensus on global warming’ which measures if respondents think there is a general consensus among scientists or not, used by both McCright & Dunlap (2011) and Krange et al. (2019). These variables could not be included due to limitations in the available data. The variable impact denial is based on question D25 ‘How good or bad do you think the impact of climate change will be on people across the world?
Table 3: operationalisation of question 25
Original
coding Answer options Coding used in this thesis
0 Extremely bad 0 1 0 2 0 3 0 4 0 5 0 6 1 7 1 8 1 9 1 10 Extremely good 1 66 Not applicable 77 Refusal 88 Do not know 99 No answer
Worry about climate change
The variable impact denial is based on question D24 ‘How worried are you about climate change’. This coding is equal to the coding applied by McCright & Dunlap (2011).
Table 4: operationalisation of question D24
Original coding
Answer options Coding used in this thesis
1 Not at all worried 1
2 Not very worried 0
3 Somewhat worried 0 4 Very worried 0 5 Extremely worried 0 6 Not applicable 7 Refusal 8 Do not know 9 No answer
Party identification
The Netherlands have a multi-party system. Therefore party identification such as in McCright & Dunlap (2011) is not very suitable. Norway has a multiple party system as well, Krange et al. (2019) categorised parties considered as conservative a considered not conservative. Respondents were asked which party they voted for in the last parliamentary election.
Respondents who voted for the parties to be considered conservative were labelled ‘confirmed conservative’, all others were labelled not (confirmed) conservatives (Krange et al., 2019). As the Dutch system is similar within this research, the same approach will be followed. Appendix 3 of the ESS8 is used to classify parties as conservative and coded with 1 or non-conservative coded with 0, Christian parties have also been labelled as conservative. The data of question B14 will be used. It is important to note that 29,8% of the answers are missing, meaning that for 501 out of 1681 respondents cannot be confirmed they are a conservative voter and were therefore coded 0 this is equivalent to the operationalisation of Krange et al. (2019).
Table: operationalisation of question B14
Original
coding
Answer options
Coding used in this thesis
1 People’s Party for Freedom and Democracy 0
2 Labour Party 0
3 Party for Freedom 1
4 Socialist Party 0
5 Christian Democratic Appeals 1
6 Democrats ’66 0
7 Christian Union 1
8 Green Left 0
9 Reformed Political Party 1
10 Party for the Animals 0
11 50PLUS 0 16 Other 0 17 Blanc 0 66 Not applicable 0 77 Refusal 0 88 Do not know 0 99 No answer 0
Male
The ESS8 uses a binary indicator of gender and has only coded male and female. This is done in question F21. Male is coded as 1 female as 0.
White
Race is not directly included within ESS8; therefore, another variable will be used as a proxy. Krange et al. (2019) describe that white is not used commonly as a social category in Norway. They argue that ethnic Norwegian comes closest to correspond to the concept of white. They choose to operationalise the concept of ethnic Norwegian as a person with two Norwegian born parents (Krange et al., 2019). This results in some severe limitations because people that are not white but have two parents born in Norway will now be labelled as ‘white’, and the other way around, people who are white but with one parent, for example, born in Sweden will now be considered ‘non-white’. McCright & Dunlap (2011) use white rather than, for example, immigrant because of the privileges attached to this race compared to minorities. In this
research, white will be operationalised as not belonging to an ethnic minority. In the Dutch questionnaire of ESS8 question C26 was formulated as “Do you belong to an ethnic minority in the Netherlands? By this we mean ethnic groups such as Turks, Moroccans, Surinamese and Antilleans who did not originally come from the Netherlands.” (translated) (European Social Survey, 2017, p. 25). This operationalisation gives a clearer view of whether someone is part of a racial minority or not. Belonging to an ethnic minority has been coded as 0 not belonging to an ethnic minority as 1.
Cool dude
To study if the combination of political ideology, race and gender has a unique effect in conservative white males McCright and Dunlap (2019) and Krange et al. (2019) both create a dummy variable. They both combine the gender, race and political ideology. Krange et al. (2019) use the confirmed conservative vote variable to determine political ideology.
Control variables
McCright & Dunlap (2011) employ eight control variables that have been used as control variables in previous studies on the conservative white male effect (Finucane et al., 2010; Flynn et al., 1994; Kahan et al., 2007; Marshall, 2004; Palmer, 2010; Satterfield et al., 2004). This research will use seven similar control variables or use proxies to emulate the same effect. This research will not control for environmental movement identity because the data within the ESS8 does not allow for this; a proxy is also not possible. The control variables that will be used are:
Political ideology
In McCright & Dunlap’s (2011) research, a scale was used to determine whether people identified themselves to be more conservative or liberal. A conservative-liberal scale is not included in the ESS8. Mccright et al. (2015) wanted to examine whether the divide in public opinion between conservatives and liberals could also be observed in countries outside the UUS As a means, they used a left-right scale as they argue this is similar to the conservative-liberal divide in the UUS Therefore for the sake of this thesis question B26 will be used in which respondents were asked where they would place themselves on a left-right scale. This scale ranges from 0 (left) to 10 (right).
Age
For this, the calculated age variable will be used based on a respondent’s year of birth question F31b. This variable is similar to the age variable used in McCright & Dunlap (2011).
Education:
The ESS8 uses a version of the ISCED, an international standard to classify education levels, question F15. It comprises a scale with 7 categories ranging from 1 (low) to 7 (high).
Household income
The ESS8 uses a system where respondents indicate that decile best describes their total household income after taxes and deductions with the national median income as the 5th category (European Social Survey, 2016). Question F41 is used for this variable.
Fulltime employment
Respondents are not directly asked in the ESS8 whether they have full-time employment. However, they are asked in question F30 for the total number of contracted hours per week in their primary job. This will be used as a proxy. The Dutch Central Bureau of Statistics (2020), has described a full-time job in the Netherlands as working 35 hours or more. Therefore the proxy is created as 0 to 34 hours not full-time coded as 0, and 35 hours or more as full-time coded as 1. Not applicable has been coded as 0 as this indicates people have no fulltime employment.
Parenthood
For this control variable, a proxy was needed, and the ESS8 does not ask respondents whether they are parents. As an alternative, the question F12 on whether the respondent lives with children is used. This is somewhat similar to the control variable used by Krange et al. (2019) which uses children under the age of 15 as a proxy for parenthood.
Religiosity
McCright & Dunlap (2011) use the frequency people visit a church to determine how strongly religious a person is. The ESS8 has a more direct alternative which lets respondents put themselves on a scale not religious at all (0) to very religious (10) in question C15.
Table 5: coding, mean and standard deviation for the variables in this thesis
Variable Coding Mean SD
Trend denial 0 (all else) to 1 (trend denial) 0,04 0,19 Attribute denial 0 (all else) to 1 (attribute denial) 0,43 0,50 Impact denial 0 (all else) to 1 (impact denial) 0,20 0,40 No worry about global
warming
0 (all else) to 1 (not worried) 0,03 0,18 Political ideology 0 (right-wing) 10 (left-wing) 4,71 2,29 Confirmed conservative
voter
0 (no) to 1 (yes) 0,26 0,44
Ethnically Dutch 0 (belongs to an ethnic minority) 1 (does not belong to an ethnic minority)
0,94 0,23
Gender 0 (female) to 1 (male) 0,45 0,50
Conservative Dutch male 0 (no) to 1 (yes) 0,10
Age 15 – 97 (numbers in years) 51,22 18,70
Educational 1 (less than lower secondary) – 7
(Higher tertiary education >= MA level) 3,89 3,21 Household income 1 (lowest 10%) – 10 (highest 10%) 5,99 2,73
Full time employment 0 (no) to 1 (yes) 0,50 0,50
Parenthood 0 (no) to 1 (yes) 0,31 0,46
4.3 Methods
Using the data discussed in the operationalisation, the analysis will test the stated hypotheses. First climate change beliefs of conservative white males will be examined and compared to beliefs of all other adults to assess whether conservative white males report higher rates of climate change denial. Subsequently, a binary logistics regression will be used. The reason for choosing a logistics regression rather than a linear regression is that linear regression requires a continuous dependent variable. The variables of interest in this thesis are dichotomous as they have been coded to 0 no climate change denial 1 climate change denial. A binary logistics regression provides a conditional probability that the independent variable will have a value of one based on the dependent variable.
5. Analysis and results
This section contains the analysis of the data using the methods as described above. Data from the ESS8 will be used to study if there is a discrepancy between the beliefs of conservative white males and all other adults. In addition, the variables will be tested separately together with control variables in a binary logistics regression. The results will be discussed and compared to the earlier studies in the US and Norway.
5.1 Conservative white males and climate change beliefs
Table 6 shows the total sample’s climate change views, the conservative white males, and all other adults. As shown in the table, impact denial is the only variable with a significant relationship with conservative white males. The association of,259 can be described as moderate. Conservative white males in the Netherlands are more likely than other adults to have denial believes around the impact of climate change. For the other variables, there was no significant relationship found. This means that being a conservative white male is not an explanatory factor for denialist beliefs. When compared to the study by McCright & Dunlap (2011), there are some notable differences when looking at climate change beliefs. Looking at the trend denial and worry about climate change variables McCright & Dunlap (2011) reported 11,6% of the sample denied that global warming was happening, among conservative white males, this was 29,6%. 19% of the American sample was not worried about climate change. 39,1% of conservative white male reported to have no concern. In the case of the Netherlands, these values are much lower, only a minimal amount of the sample reported denialist views for these variables. Looking at the study in Norway, the values reported by Krange et al. (2019) for trend denial are even lower in Norway, with only 1,3% of the sample denying the trend of climate change. Both Krange et al. (2019) and McCright & Dunlap (2011) reported scores around 40% for attribute denial. This is somewhat higher in the
Netherlands. The other studies did not measure impact denial. Therefore it is not possible to compare.
Table 6: per cent’s reported by the total sample, conservative white males and all other adults on selected climate change views.
Climate change believes
Total (N) Conservative white males (N)
All other adults (N)
γ
aTrend denial 3,7% (55) 7,2% (10) 2,4% (45) ,380 Attribute denial 57,9% (762) 54,1% (73) 57,6% (835) -,078 Impact denial 19,8% (283) 28,4% (38) 18,9% (245) ,259** No worry about climate change 3,0% (44) 4,4% (6) 2,9% (38) ,220
a Gamma for the relationship between the conservative white male dummy variable and the denial variables.
* P<0,05 ** P<0,025 ** P<0,005
5.2 Predicting factors
As discussed, there is no significant relationship between the climate change believes variables and conservative white males, for impact denial the relationship was only moderate. As in previous research, other factors that might explain climate change denial are analysed. This is done for all four aspects of climate change denial using eight logistic regression models which can be seen in table 7 and 8. To understand to what degree the variance in the observed data can be explained, the Nagelkerke R2 is used. The value of the R2 can be interpreted as the measure of success in predicting the dependent variable from the independent variable (Nagelkerke, 1991). As can be seen in the tables, the degree to which the models can predict whether an individual has denialist believes ranges from 5,3% to 11,1%. Therefore it can already be stated that the variables included in the model explain climate believes to a limited degree. Nevertheless, there are still notable findings which will be discussed.
For each denial variable, two models were created. First, a base model which studies the effects of political ideology, race, and gender separately, while seven other relevant, social, demographical, and political variables are used as control variables. The second model ads the conservative white male dummy variable to the base model. First, the three key variables will be reviewed. Subsequently, the control variables will be discussed.
Out of the three key variables, only political ideology can be considered significant with the exemption of the model on-trend denial in which none of the key variables is significant. In the other model’s political ideology shows it has a positive explanatory attribute. This means that the more someone identifies himself as right-wing, the more likely they will deny climate
science. This correlates with McCright & Dunlap (2011) findings and Krange et al. (2019). This is not surprising, as Mccright et al. (2015) have demonstrated that Europeans on the left of the political spectrum are more likely to believe in climate change confirmed by this model. When looking at race, it can be observed that it is suggested that this variable has an explanatory factor, being ethnically Dutch results being less likely to hold climate sceptic views. This would be contrary to the results of McCright & Dunlap (2011) and Krange et al. (2019), however, these results are not significant. In order to state whether this is the case, more research is needed. Gender is also not a significant explanatory factor. The values also do not show a consistent image. For some, the model suggests that gender explains a positive relationship; in others, a negative relationship is shown.
Table 7: logistic regression models predicting climate change beliefs about trend denial and attribute denial
Independent variables Trend denial Attribute denial
Base Conservative white male Base Conservative white male Political ideology ,155 (,081) ,142 (,087) ,161*** (,032) ,176*** (,034) Race -1,053 (,587) -1,141 (,602) -,274 (,286) -,282 (,291) Gender ,575 (,345) ,207 (,409) -,282 (,132) -,181 (,158) Age ,007 (,009) ,003 (,010) ,023*** (,004) 0,025*** (0,004) Educational Attainment -,202** (,082) -,166 (,087) -,028 (,024) -,042 (,033) Annual income ,090 (,064) ,121 (,071) -,042 (,023) -,029 (,026) Full-time employment ,118 (,333) ,062 (,360) ,196 (,132) ,195 (,142) Parenthood -,001 (,341) -,143 (,368) -,035 (,131) -,043 (,139) Religiosity -,050 (,049) -,043 (,053) ,027 (,019) ,304 (,020) Conservative voter -,072 (,382) -1,484 (1,044) -,200 (,158) ,007 (,214) Conservative white male 2,018
(1,116) -,533 (,298) Constant 3,438*** (,917) -3,225*** (,966) -1,127*** (,383) -1,365*** (,412) -2log likelihood 422,543 359,038 1768,619 1570,376 Nagelkerke R2 ,056 ,066 ,097 ,111 Sample size 1392 1249 1370 1230 * P<0,05 ** P<0,025 ** P<0,005
Moving over to the control variables, it can be seen that most do not display a significant relationship. For some of these variables, this was to be expected. Both McCright & Dunlap (2011) and Krange et al. (2019) found that income, employment and parenthood had no
significant explanatory factor. McCright & Dunlap (2011) found that education has no significant impact on most of the denialist variables whereas Krange et al. (2019) did find that people with higher education are more likely to believe climate change science. This is also found in the model for impact denial in the Dutch context and the base model of trend denial. Also in line with Krange et al. (2019) findings, the models of attribute and impact denial and the base model for lack of worry about climate change show a significant positive relationship between age and climate change denial. However, these relationships can be described as very weak. McCright & Dunlap (2011) only found a weak significant positive relationship between age and denial indicators. Lastly, the impact denial model shows for impact denial. There is a significant positive relationship between religiosity and impact denial. However, this relationship is also weak.
Table 8: logistics regression models predicting climate change beliefs about impact denial and no concern about climate change
Independent variables Impact denial No worry about climate change Base Conservative white male Base Conservative white male Political ideology ,115*** (,040) ,121** (,042) ,204* (,091) ,230** (,102) Race -,393 (,330) -,445 (,336) -,990 (,618) -1,180 (,691) Gender ,104 (,164) ,033 (,202) ,316 (,385) -,198 (,477) Age ,013*** (,005) ,014** (,005) ,011 (,010) ,034** (,014) Educational Attainment -,172*** (,041) -,156*** (,043) ,051 (,028) ,128 (,082) Annual income -,058 (,031) -,061 (,032) -,064 (,064) -,037 (,081) Full-time employment ,189 (,163) ,070 (,174) ,228 (,377) ,494 (,424) Parenthood ,046 (,175) -,028 (,186) -,037 (,410) ,311 (,484) Religiosity ,071*** (,025) ,066** (,026) -,046 (,064) -,044 (,062) Conservative voter ,027 (,181) -,108 (,246) -,055 (,444) -,175 (,684) Conservative white male ,311
(,340) ,412 (,858) Constant -1,840*** (,476) -1,780*** (,501) -4,250*** (,979) -6,161*** (1,277) -2log likelihood 1265,962 1130,572 338,272 271,476 Nagelkerke R2 ,101 ,102 ,053 ,096 Sample size 1360 1224 1378 1237 * P<0,05 ** P<0,025 ** P<0,005
6. Conclusion and discussion
This thesis attempted to replicate the research design of McCright & Dunlap (2011) to test whether the conservative white male effect could also be observed in the Netherlands. It has been shown that for most of the climate change indicators, no significant relationship has been found. The indicator of impact denial showed a significant but moderate relationship. Considering these results, there cannot be made a conclusive claim that there is a conservative white male effect observed in the Netherlands.
These findings contrast to the findings by McCright & Dunlap (2011) and Krange et al. (2019). Who found significant relationships between climate change indicators and conservative white males in the US and Norway, respectively. The logistic regression showed that for most climate belief indicators, political ideology was a predicting factor, although the probability was low. Considering the research on RWP and far-right beliefs on climate change is probability is lower than might have been predicted.
A serious implication of this research is the limited amount of data. Although the ESS8 counts 1681 respondents form the Netherlands when testing the for the climate belief indicators the problem arose that for trend denial and lack of concern on climate change only a minimal number of respondents could be marked as climate change deniers. This applies to both the conservative white men as to the entire population. Although differences between these groups can be observed, the results are not significant. Before the start of this research, it was not expected that the N for these indicators would be so low. On the other hand, this result can be perceived as good news. An overwhelming majority of Dutch individuals are aware of the fact that climate change is happening, the results will have negative impacts about which they worry. However, more than half of the respondents are not convinced that climate change is primarily caused by human behaviour rather than natural processes.
In order to get more insight into the explanatory factors for climate change denial in the Netherlands, it would be suggestible to gather a more extensive data sample. In addition, other factors might be included such as views on whether the effects of climate change are exaggerated in the media as has been done by McCright & Dunlap (2011) and Krange et al. (2019).
It is clear that this thesis could not answer the research question: Does the cool dude-effect
research is needed to determine why individuals in the Netherlands endorse climate change denial beliefs.
Biography
Björnberg, K. E., Karlsson, M., Gilek, M., & Hansson, S. O. (2017). Climate and
environmental science denial: A review of the scientific literature published in 1990– 2015. In Journal of Cleaner Production (Vol. 167, pp. 229–241). Elsevier Ltd. https://doi.org/10.1016/j.jclepro.2017.08.066
CBS. (2020). Meer dan de helft werkt voltijds. 20-2. https://www.cbs.nl/nl-nl/nieuws/2020/08/meer-dan-de-helft-werkt-voltijds
Dunlap, R. E., & McCright, A. M. (2015). Challenging climate change. Climate Change and
Society: Sociological Perspectives, 300.
European Social Survey. (2013). About the European Social Survey European Research
Infrastructure. http://www.europeansocialsurvey.org/about/index.html
European Social Survey. (2016). Version Notes , ESS8 Appendix A2 INCOME. https://www.europeansocialsurvey.org/data/download.html?r=8
European Social Survey. (2017). Dutch questionnaire.
Feygina, I., Jost, J. T., & Goldsmith, R. E. (2010). System justification, the denial of global warming, and the possibility of “system-sanctioned change.” Personality and Social
Psychology Bulletin, 36(3), 326–338.
Finucane, M. L., Slovic, P., Mertz, C. K., Flynn, J., & Satterfield, T. A. (2010). Gender, race,
and perceived risk: The “white male” effect. https://doi.org/10.1080/713670162
Flynn, J., Slovic, P., & Mertz1, C. K. (1994). Gender, Race, and Perception of Environmental
Health Risks.
Forchtner, B. (2019). Climate change and the far right. Wiley Interdisciplinary Reviews:
Climate Change, 10(5), e604. https://doi.org/10.1002/wcc.604
Gallup Inc. (n.d.). See the World in Data | Gallup. Retrieved January 23, 2021, from https://www.gallup.com/analytics/318875/global-research.aspx
Haagsma, B. (2020). Het eerlijke verhaal? [Universiteit Utrecht]. https://doi.org/10.1007/s12428-016-0009-7
Hamilton, L. C. (2008). Who cares about polar regions? Results from a survey of U.S. public opinion. In Arctic, Antarctic, and Alpine Research (Vol. 40, Issue 4, pp. 671–678). https://doi.org/10.1657/1523-0430(07-105)[HAMILTON]2.0.CO;2
Hobson, K., & Niemeyer, S. (2013). “What sceptics believe”: The effects of information and deliberation on climate change scepticism. Public Understanding of Science, 22(4), 396–412. https://doi.org/10.1177/0963662511430459
IPCC. (n.d.). About — IPCC. Retrieved January 21, 2021, from https://www.ipcc.ch/about/ IPCC. (2001). Climate Change 2001: The Scientific Basis. Contribution of Working Group I
to the Third Assessment Report of the Intergovernmental Panel on Climate Change.
https://doi.org/10.1016/S1058-2746(02)86826-4
IPCC. (2007). Climate Change 2007: The Physical Science Basis. Contribution of Working
Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. https://doi.org/10.1256/wea.58.04
IPCC. (2013). Climate Change 2013: The Physical Science Basis. Contribution of Working
Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change.
IPCC. (2018). Global Warming of 1.5°C. An IPCC Special Report on the impacts of global
warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, (Issue October).
Jost, J. T., Nosek, B. A., & Gosling, S. D. (2008). Ideology Its Resurgence in Social,
Personality, and Political Psychology. https://implicit.harvard.edu/
Kahan, D. M., Braman, D., Gastil, J., Slovic, P., & Mertz, C. K. (2007). Culture and Identity-Protective Cognition: Explaining the White-Male Effect in Risk Perception. Journal of
Empirical Legal Studies, 4(3), 465–505.
Kalof, L., Dietz, T., Guagnano, G., & Stern, P. C. (2002). Race, gender and
environmentalism: The atypical values and beliefs of white men. Race, Gender & Class, 112–130.
Krange, O., Kaltenborn, B. P., & Hultman, M. (2019). Cool dudes in Norway: climate change denial among conservative Norwegian men. Environmental Sociology, 5(1), 1–11. https://doi.org/10.1080/23251042.2018.1488516
Lockwood, M. (2018). Environmental Politics Right-wing populism and the climate change
agenda: exploring the linkages. https://doi.org/10.1080/09644016.2018.1458411
Marshall, B. K. (2004). Gender, race, and perceived environmental risk: The" white male" effect in cancer alley, La. Sociological Spectrum, 24(4), 453–478.
Marshall, B. K., Picou, J. S., Formichella, C., & Nicholls, K. (2006). Environmental Risk Perceptions and the White Male Effect: Pollution Concerns Among Deep-South Coastal Residents. Journal OfApplied Sociology, 23(2), 31–49.
McCright, A. M. (2010). The effects of gender on climate change knowledge and concern in the American public. Population and Environment, 32(1), 66–87.
McCright, A. M., & Dunlap, R. E. (2011). Cool dudes: The denial of climate change among conservative white males in the United States. Global Environmental Change, 21(4), 1163–1172. https://doi.org/10.1016/j.gloenvcha.2011.06.003
Mccright, A. M., Dunlap, R. E., & Marquart-Pyatt, S. T. (2015). Environmental Politics
Political ideology and views about climate change in the European Union.
https://doi.org/10.1080/09644016.2015.1090371
Nagelkerke, N. J. D. (1991). A note on a general definition of the coefficient of determination.
Biometrika, 78(3), 691–692. https://doi.org/10.1093/biomet/78.3.691
Palmer, C. (2010). Risk perception: Another look at the “white male” effect. https://doi.org/10.1080/1369857031000066014
Rahmstorf, S. (2004). The climate sceptics. Potsdam Institute for Climate Impact Research, 76–82.
http://www.pik-potsdam.de/stefan/Publications/Other/rahmstorf_climate_sceptics_2004.pdf
Satterfield, T. A., Mertz, C. K., & Slovic, P. (2004). Discrimination, Vulnerability, and Justice in the Face of Risk. Risk Analysis, 24(1), 115–129. https://doi.org/10.1111/j.0272-4332.2004.00416.x
Paris Agreement, (2015).
van Prooijen, J.-W., Krouwel, A. P. M., & Pollet, T. V. (2015). Political Extremism Predicts Belief in Conspiracy Theories. Social Psychological and Personality Science, 6(5), 570– 578. https://doi.org/10.1177/1948550614567356
Wood, B. D., & Vedlitz, A. (2007). Issue Definition, Information Processing, and the Politics of Global Warming. American Journal of Political Science, 51(3), 552–568.
