Citizens’,Scientists’,and Policy Advisors’ Beliefs about Global Warming

Abstract

Numerous factors shape citizens’ beliefs about global warming, but there is very little research that compares the views of the public with key actors in the policymaking process. We analyze data from simultaneous and parallel surveys of (1) the U.S. public, (2) scientists who actively publish research on energy technologies in the United States, and (3) congressional policy advisors and find that beliefs about global warming vary markedly among them. Scientists and policy advisors are more likely than the public to express a belief in the existence and anthropogenic nature of global warming. We also find ideological polarization about global warming in all three groups, although scientists are less polarized than the public and policy advisors over whether global warming is actually occurring. Alarmingly, there is evidence that the ideological divide about global warming gets significantly larger according to respondents’ knowledge about politics, energy, and science.

Keywords

global warming scientists policy advisors politicization

Global warming is among the most complex and challenging problems facing humanity. Yet despite a consensus among climate scientists about the reality of human-induced global warming (Melillo, Richmond, and Yohe 2014), numerous studies find that segments of the public express doubt about whether this phenomenon is occurring and, if it is, whether this is the result of humans’ actions (anthropogenic) or the result of natural causes (Brulle, Carmichael, and Jenkins 2012; Hamilton 2011; Jones 2011; McCright and Dunlap 2011). It has also become apparent that “educating” the public about the “facts” associated with global warming rarely leads individuals to update their beliefs in a corrective fashion. This stems from the politicization of science that creates uncertainty about whether one can trust scientific information invoked in the context of political arguments (Bolsen, Druckman, and Cook 2014a).¹ It consequently has led to diminishing trust in science among some citizens (Akerlof et al. 2012; Gauchat 2012; Hmielowski et al. 2014; Lewandowsky, Gignac, and Vaughan 2013).

Extant work has identified a number of factors that shape citizens’ beliefs about global warming (Borick and Rabe 2010; Nisbet and Myers 2007; Nisbet 2009; Schuldt, Konrath, and Schwartz 2011), yet no studies that we have been able to find compare the views of the public with other key actors in the policymaking process. Currently most research on the drivers of beliefs about global warming relies on either individual-level cross-sectional analyses or state- and national-level time-series analyses of samples. No studies to date have assessed the degree to which scientists and policy advisors differ from the public in their beliefs about global warming and how the politicization of climate science shapes fundamental beliefs about global warming among these three distinct groups.

To address these issues, we draw on data from three simultaneous and parallel surveys of (1) the U.S. public, (2) scientists who work at universities in the United States and who actively publish research on energy technologies, and (3) congressional policy advisors. These three surveys enable us to pinpoint areas of agreement and disagreement regarding beliefs about anthropogenic global warming among these critical actors in the policymaking process. We also test the degree to which fundamental beliefs about climate science are associated with political ideology (Hamilton 2011; Hmielowski et al. 2014; Zia and Todd 2010), party identification (Dunlap and McCright 2008; Hart and Nisbet 2012; Weber and Stern 2011), values (Dietz 2013; Jones 2011; Kahan et al. 2012), and the interaction of factual knowledge with ideology and party identification (Hayes 2001; Malka, Krosnick, and Langer 2009; McCright 2010).

We find that beliefs about the existence of anthropogenic global warming diverge markedly in comparing the views of the public, scientists, and policy advisors. Compared with the public, scientists and policy advisors are more likely to express a belief in the existence of anthropogenic global warming. When it comes to ideological polarization, however, the public and policy advisors look the same whereas scientists differ, looking less polarized. The results highlight challenges that the politicization of science presents to achieving a consensus necessary for meaningful policy action.

Drivers of Beliefs about Anthropogenic Global Warming

A central departure point for any examination of public opinion on the issue of global warming involves measurement of the public’s belief that global warming is actually occurring … as the issue has evolved over the past two decades, there has been significant debate in the public forum regarding the reality of global warming. (Borick and Rabe 2010, 781)

In recent testimony to Congress on the issue of global warming, Professor Jon Krosnick (2013) provided an overview of beliefs labeled the fundamentals with respect to aspects of global warming attitudes: (1) the belief that global warming is happening and (2) that the observed warming trend is the result of humans’ actions. Any meaningful collective action when it comes to global warming requires coherence in the attitudes and beliefs among the key actors involved: policy advisors who create the laws, scientists who generate technologies, and the public who determines what laws and technologies survive in the political and economic marketplace. For example, policy-makers rarely take action without public support, and technologies that scientists believe are helpful will not survive in the marketplace without being endorsed by policy advisors and the public—i.e., there needs to be some agreement among these key actors for technologies and policies designed to combat global warming to take effect (Druckman 2013).

Although clear scientific consensus exists that anthropogenic global warming is occurring, the percentage of the U.S. public who reported a belief in its existence declined between 2008 and 2011 at a rate of about 5 percent per year (Krosnick and MacInnis 2012). Approximately two-thirds of Americans now believe that global warming is happening (63 percent) based on the results from a survey of a representative sample of the U.S. public in late November and early December 2013 (Leiserowitz et al. 2014).² Higher levels of education and knowledge tend to correlate with greater concern about anthropogenic global warming; however, the effect of knowledge depends on one’s ideology and party identification in the United States (Malka, Krosnick, and Langer 2009; Hamilton 2011). As McCright and Dunlap (2011, 161) have pointed out, “Citizens’ political orientations may lead them to perceive this politically contentious issue quite differently.” An abundance of survey data clearly shows that partisan and ideological divisions have emerged in recent years in the United States on the issue of global warming with liberals and Democrats more likely to accept the scientific consensus on this issue relative to conservatives and Republicans (McCright and Dunlap 2011; Mildenberger and Leiserowitz, n.d.; Schuldt, Konrath, and Schwarz 2011). Skepticism about whether anthropogenic global warming is happening is more common among Republicans, with 57 percent seeing “no solid evidence” compared with 17 percent of Democrats based on the results from a survey conducted in 2009 (Schuldt, Konrath, and Schwarz 2011).³ This leads us to offer the following prediction.

Individuals who identify as liberal or Democrat will be more likely, and individuals who identify as conservative or Republican will be less likely, to believe that (a) global warming is happening and that (b) humans are responsible for it among each of the three distinct samples we study (hypothesis 1).

Individuals’ values also play a key role in shaping beliefs about global warming. Values refer to concepts or beliefs about desirable end states or behaviors that transcend specific situations, guide evaluations, and are ordered in their relative importance (Davidov, Schmidt, and Schwartz 2008; Dietz 2013). Values play a central role in determining attitudes and behaviors toward environmental issues (Douglas and Wildavsky 1982; Stern, Dietz, and Guagnano 1995). One prominent set of values associated with beliefs about global warming stems from cultural cognition theory (Kahan et al. 2012). It posits that individuals form perceptions of societal risks that cohere with values characteristic of groups with which they identify (Kahan, Jenkins-Smith, and Braman 2011). According to Kahan et al. (2012, 732), individuals “who subscribe to a ‘hierarchical, individualistic’ worldview … tend to be skeptical of environmental risks [because] widespread acceptance of such risks would license restrictions on commerce and industry, forms of behavior that hierarchical individualists value. … Persons who hold an ‘egalitarian, communitarian’ worldview tend to be morally suspicious of commerce and industry, to which they attribute social inequity. They therefore find it congenial to believe those forms of behaviors are dangerous and worthy of restriction.” We expect that these values will not only shape the lay citizens’ beliefs about global warming but also the beliefs of scientists and policy advisors. Kahan et al. (2012) find that individuals with the highest levels of numeracy and scientific literacy were the ones among whom cultural polarization on global warming was the greatest. This leads us to our second hypothesis.

Among each of the three distinct groups that we study, individuals who possess hierarchical/individualist values will be less likely, relative to those who possess egalitarian/communitarian values, to believe that (a) global warming is happening and that (b) humans are responsible for it (hypothesis 2).

Politicization and ideological polarization

The politicization of climate science refers to a concerted effort that began in the 1990s to challenge the legitimacy of the scientific consensus emerging on this issue by organized interests within both the fossil fuel industry and conservative think tanks that were allied with business interests (Weber and Stern 2011). This has resulted in declining levels of trust toward scientists among conservatives and in ideological polarization on this issue (Gauchat 2012). In the past, research on public understanding of science emphasized the relationship between scientific literacy—i.e., knowledge of scientific facts—and support for science and scientists (Gauchat 2012, 169). However, factual knowledge and education do not always predict attitudes about global warming. A number of studies show that the marginal effect of an increase in education or knowledge depends on one’s partisanship and ideology (Hamilton 2011; Malka, Krosnick, and Langer 2009; Mildenberger and Leiserowitz 2013). This is likely because knowledgeable partisans and ideologues are more likely to engage in motivated reasoning when expressing their beliefs about global warming (Hart and Nisbet 2012; Taber and Lodge 2006).

Motivated reasoning refers to one’s goal in the process of forming a belief or opinion (Kunda 1999; Taber and Lodge 2006).⁴ Individuals can either pursue a directional goal, in which an evaluation or opinion expressed in a given context serves to bolster existing beliefs or identities; or an accuracy goal, in which one’s underlying motivation is to form and hold a correct or accurate belief. When people pursue a directional goal they tend to give more weight to evidence that is consistent with existing views, dismiss information that is inconsistent with existing views, and view evidence and arguments as stronger when they are consistent with one’s beliefs (e.g., see Druckman, Peterson, and Slothuus 2013; Kunda 1990, 1999). Past work demonstrates that the tendency to pursue a directional goal in forming an opinion in the face of new information increases as individuals become more knowledgeable and politically sophisticated (e.g., see Malka, Krosnick, and Langer 2009; Taber and Lodge 2006). Therefore, we expect higher levels of knowledge to lead to greater polarization among ideologues and partisans across the three distinct groups that we study due to a greater propensity to engage in motivated reasoning.⁵ In other words:

Among each of the three groups we study, partisanship and ideology will interact with knowledge such that liberals/Democrats who possess higher levels of knowledge will be more likely, and conservatives/Republicans who possess higher levels of knowledge will be less likely, to believe that (a) global warming is happening and that (b) humans are responsible for it (hypothesis 3).⁶

A number of other individual and contextual factors are associated with beliefs about global warming in the United States. We do not offer explicit hypotheses regarding these additional factors given that they largely serve as controls in the analyses below. Nonetheless, the literature shows that disbelief in anthropogenic global warming is more common among whites and males (McCright and Dunlap 2011; McCright 2010). Individuals who have higher levels of trust in science (Gauchat 2012), who experience extreme local weather (Egan and Mullin 2012; Hamilton and Stampone 2013; Hansen, Sato, and Ruedy 2012), who have higher levels of political awareness (Boykoff 2007; Feldman et al. 2011; Zhao et al. 2011), and who possess positive perceptions of scientists (Giddens 2009; Leiserowitz et al. 2014; Oreskes and Conway 2010) are more likely to believe in anthropogenic global warming. Thus, we account for these additional factors in the analyses reported below.

Surveys

As mentioned, existing work in the United States that focuses on beliefs about global warming: (1) has not simultaneously explored the attitudes of three key groups in the policymaking process (i.e., the public, scientists, policy advisors); (2) has not isolated the key factors driving opinions across these groups; and (3) has not explored the impact of the politicization of climate science on knowledgeable partisans and ideologues. To address these gaps, we draw on data from three simultaneous and parallel surveys conducted in August 2010 on samples of (1) the U.S. public, (2) scientists who actively publish research on energy technologies in the United States, and (3) congressional policy advisors.

Public sample

We contracted with a survey research company (Bovitz Inc.) to conduct a web-based survey of a representative sample of 1,600 citizens in the United States. The sample comes from a panel of respondents who have opted to complete online surveys. The panel was originally developed based on a random-digit-dial telephone survey, where to enter the panel a respondent needed to have access to the Internet.⁷ The panel has continued to grow based on ongoing random-digit-dial recruiting and referrals. From the panel, which has approximately one million members, a given sample is drawn using a matching algorithm (based on likely response rates) to ensure that those screened to qualify for the survey constitute a sample that demographically represents the United States.

Scientist sample

We surveyed a sample of scientists who publish research in areas related to energy technologies at universities in the United States to assess their beliefs about global warming. We focused on the population of scientists who work on energy technology and conduct potentially influential research. We identified our population based on a search of the Web of Science’s Science Citation Database to locate those who had published articles on energy technology between January 1, 2006, and October 17, 2009.⁸ We drew a random sample of 1,800 articles that met our search criteria over this period. We then identified the contact/lead author to ensure that the person was actively involved in a sustained research program. We removed scientists located outside of the United States and authors whose work had been cited less than five times. We recorded the contact information for each lead author and attempted to contact each person to invite him/her to participate in a survey about energy-related issues in the United States. We sent a $5 Starbucks gift card to each scientist invited to participate in the survey that they could keep regardless of whether they participated. We contacted a total of 827 scientists after removing emails and letters that were returned to sender. A total of 280 scientists completed the survey for a response rate of 34 percent. This is an exceptionally high response rate in comparison to similar types of surveys previously conducted on elite samples (Berkman and Plutzer 2011).⁹

Policy advisors

Our third survey focused on policy advisors who may affect U.S. energy policy. We initially defined our population of interest as legislative directors who work for members of Congress (N = 535); however, concerns about a low response rate prompted us to collect additional data that focused on lower-level staff members within each member of Congress’s office, including communication directors and legislative assistants (see Plutzer, Maney, and O’Connor 1998).¹⁰ We collected the names of up to three staff members from each office using the Congressional Staff Directory. As with our sample of scientists, we removed from the sample individuals whom we could not contact because of letters and emails that were returned to sender. We mailed a letter to each member of Congress’s office in advance of the survey that provided information about its purpose and provided a $5 Starbucks gift card that they could keep regardless of whether they completed the survey. In addition, staff members who completed the survey were offered a completion code that they could enter to receive $20 as compensation for their time. We contacted a total of 984 individuals. A total of fifty-five policy advisors completed the survey for a response rate of 5.56 percent. Although this is a disappointingly low response rate, of the fifty-five policy advisors who completed our survey, there is a good variation in party identification and ideological self-placement—49 percent identified as a Democrat, 35 percent as a Republican, and 16 percent as an independent; 38 percent identified as a liberal and 35 percent as a conservative. The total number of policy advisors we completed interviews with was comparable to other samples that have attempted to interview this population (Plutzer, Maney, and O’Conner 1998).

Measures

Our primary dependent variables include two questions that measure: (1) belief about whether global warming is happening (where 1 = definitely is not happening, 4 = not sure, and 7 = definitely is happening), and (2) if it is happening, whether the trend is a result of natural changes or humans’ actions (1–7 scale, where 1 = definitely naturally induced, 4 = not sure, and 7 = definitely human induced). The exact wording for each measure is included in the appendix.

We also included measures of the previously discussed demographic and political characteristics posited to be associated with beliefs about anthropogenic global warming.¹¹ This included measures of cultural cognition theory’s worldview variables—hierarchical (as opposed to egalitarianism) and individualism (as opposed to communitarianism)—measured on 7-point scales with higher values indicating hierarchical tendencies or individualism.¹² We measured energy, science, and political knowledge and created a global measure of knowledge based on the number of correct responses across these items (see Table A1 in the appendix for a list of the individual items and responses). Finally, we included standard measures that asked for respondents’ gender, ethnicity,¹³ education,¹⁴ age,¹⁵ media exposure,¹⁶ trust in government,¹⁷ trust in science,¹⁸ local weather experience,¹⁹ ideology,²⁰ and party identification.²¹

Results

We begin, in Figure 1, by presenting the percentage in each sample that believes global warming is happening and that it is human induced as opposed to a result of natural changes. The majority of respondents in all three samples say they believe global warming is happening. However, there are striking differences in the beliefs about global warming in comparing responses across the three samples.²² Not surprisingly, scientists working at universities in the United States and who publish research in energy-related areas are far more likely than policy advisors or members of the public to say that global warming is happening (89 percent of scientists, 64 percent of the public, and 71 percent of policy advisors). The strong endorsement of global warming among our sample of scientists comports with data recently released from a panel evaluating levels of scientific consensus regarding climate change among climate scientists for the American Association for the Advancement of Science (AAAS), which reports that 97 percent of climate scientists believe in anthropogenic global warming (AAAS 2014). The slightly lower percent of energy scientists who endorse global warming in our sample—compared with climate scientists’ views—is likely due to the fact that the scientists in our sample are not climate scientists but rather scholars working on energy-related research at universities in the United States. The percentage of the public in our sample that believes global warming is happening is nearly identical to what other surveys of the U.S. public have reported (Leiserowitz et al. 2014). Similarly, the energy scientists who completed our survey overwhelmingly attribute the rise in Earth’s temperature to humans’ actions (81 percent of those who believe it is happening). Policy advisors were largely in agreement with scientists that global warming is a result of human actions (70 percent of those who believe it is happening). The U.S. public was more divided on this issue, with 57 percent of those who believe global warming is happening seeing it as human induced. Thus, the data suggest that the views of these critical actors in the policymaking process diverge markedly about the fundamentals related to global warming. We next tested the aforementioned hypotheses about the effect of party identification and ideology, values, and the politicization of climate science on beliefs about whether global warming is happening and whether it is human induced. In particular, we wanted to learn the extent to which the same factors or different ones shape the responses of the public, energy scientists, and policy advisors.

Figure 1

Citizens’, Scientists’, and Policy Advisors’ Beliefs about Global Warming

In Table 1, we report the results from a series of ordered probit models that estimate factors that increase or decrease the likelihood of expressing a belief that global warming is happening among members of the U.S. public, energy scientists, and policy advisors for members of Congress. The first model for each sample (model 1) allows a test of hypothesis 1 about the impact of party identification and political ideology on the likelihood that one expresses a belief that global warming is happening. In strong support of hypothesis 1, and in line with the results of numerous recent surveys (Brulle, Carmichael, and Jenkins 2012; Malka, Krosnick, and Langer 2009; McCright and Dunlap 2011; Hamilton 2011), we find that party identification and political ideology play a powerful role in determining individuals’ beliefs about global warming. Among members of the public, Democrats are significantly more likely to express a belief that global warming is happening (72 percent predicted probability of expressing a belief global warming is happening), while Republicans are significantly less likely to express this view (57 percent predicted probability of expressing a belief that global warming is happening).²³ Similarly, among members of the public, liberals are significantly more likely than conservatives to believe global warming is happening (73 percent versus 60 percent predicted probability of expressing a belief that global warming is happening, respectively). This result is true not only for members of the public but also for policy advisors and scientists where ideology plays a significant role in determining these groups’ views (but not party identification for scientists and policy advisors; McCright and Dunlap 2011). The predicted probability of a liberal scientist expressing a belief that global warming is happening is 94 percent, whereas the probability that a conservative scientist expresses a similar view is 84 percent. The same pattern is observed among the sample of policy advisors; however, the sample is too small to accurately estimate changes in predicted probabilities among this group. In short, ideology clearly plays a central role in driving beliefs about whether global warming is happening among all three of our samples.

Table 1

Determinants of the Belief that Global Warming Is Happening

	Public Sample				Scientist Sample				Policy Advisor Sample
	Model 1		Model 2		Model 1		Model 2		Model 1		Model 2
Demographics
Education	−.00	(.03)	−.00	(.03)	−.58	(.57)	−.55	(.56)	−.49	(.81)	−.21	(.92)
Age	−.00	(.00)	.00	(.00)	.01	(.00)	.01	(.00)	.16**	(.08)	.13	(.09)
Female	.06	(.05)	.04	(.05)	.63*	(.33)	.62*	(.34)	−.65	(1.2)	−2.42	(2.20)
Income	.04	(.03)	.04*	(.02)	.04	(.10)	.04	(.10)	.38	(.43)	.64	(.66)
Nonwhite	.14**	(.06)	.16***	(.06)	.10	(.25)	.08	(.25)	7.25**	(2.85)	.64	(.66)
Knowledge	−.06	(.12)	.41**	(.20)	.25	(.74)	1.83*	(.97)	−2.95	(3.91)	6.84	(5.39)
Local Weather	.10***	(.03)	.10***	(.03)	−.09	(.12)	−.09	(.13)	2.97***	(1.06)	−.15**	(7.36)
Media	.32***	(.11)	.31***	(.11)	−.05	(.4)	.05	(.4)	−10.2***	(3.01)	2.77***	(1.40)
Trust Gov.	.15***	(.04)	.14***	(.04)	−.09	(.19)	−.10	(.19)	−.81	(.59)	−8.74***	(3.25)
Trust Science	.18***	(.03)	.18***	(.03)	.01	(.11)	.00	(.11)	−.18	.62	−.18	.62
Values
Hierarchy	−.06***	(.01)	−.06***	(.01)	−.06	(.06)	−.05	(.06)	−1.64***	(.46)	−1.52***	(.54)
Individualism	−.09***	(.01)	−.08***	(.01)	−.13*	(.07)	−.11	(.07)	.41	(.35)	.06	(.54)
Political Characteristics
Republican	−.28***	(.07)	.16	(.20)	−.03	(.34)	1.01	(3.24)	15.24	(17.06)	25.09***	(15.09)
Democrat	.25***	(.07)	.27*	(.16)	.26	(.29)	1.99	(2.05)	16.37	(17.09)	22.49***	(12.26)
Liberal	.32***	(.07)	.19	(.19)	.22	(.27)	.81	(1.99)	6.19***	(2.32)	47.41	(41.09)
Conservative	−.20***	(.07)	.33*	(.18)	−.68**	(.33)	3.14	(2.77)	−4.33***	(1.60)	6.02	(8.42)
Liberal × Knowledge	—	—	.17	(.30)	—	—	−.77	(2.3)	—	—	−43.77	(41.45)
Conservative × Knowledge	—	—	−.90***	(.20)	—	—	−4.53*	(3.05)	—	—	−12.34	(10.24)
Republican × Knowledge	—	—	−.68**	(.30)	—	—	−1.17	(3.50)	—	—	−2.82	(9.12)
Democrat × Knowledge	—	—	−.05	(.27)	—	—	−1.9	(2.36)	—	—	—	—
Pseudo R²	0.09		0.10		0.15		0.16		0.69		0.72

NOTE: The columns contain ordered probit coefficient estimates predicting experimental treatment effects relative to the control (baseline) condition. The coefficients and standard errors for cut points 1 through 6 for model 1 (Public) are −1.54 (0.22), −1.07 (0.22), −0.65 (0.22), −0.11 (0.22), 0.48 (0.22), and 1.09 (0.22). The coefficients and standard errors for cut points 1 through 5 for model 1 (Scientist) are −5.02 (3.03), −4.35 (3.01), −3.72 (3.03), −2.87 (3.03), and −2.07 (3.03). The coefficients and standard errors for cut points 1 through 6 for model 1 (Policy Advisors) are −.533 (17.3), .40 (17.3), 3.4 (17.2), 4.7 (17.2), 7.9 (17.3), and 15.2 (17.7). The coefficients and standard errors for cut points 1 through 6 for model 2 (Public) are −1.26 (0.24), −0.788 (0.23), −0.36 (0.23), 0.17 (0.23), 0.79 (0.23), and 1.4 (0.23). The coefficients and standard errors for cut points 1 through 5 for model 2 (Scientist) are −3.73 (3.07), −3.01 (3.05), −2.33 (3.04), −1.42 (3.04), and −0.60 (3.04). The coefficients and standard errors for cut points 1 through 6 for model 2 (Policy Advisors) are 3.95 (11.05), 4.96 (11.00), 8.35 (10.88), 9.72 (10.9), 13.2 (11.32), and 18.9 (12.85).

***

p ≤ .01; **p ≤ .05; *p ≤ .10 (one-tailed tests).

We also find strong support for hypothesis 2 regarding the role that values play in shaping individuals’ beliefs about whether global warming is happening. Cultural cognition theory posits that the values of hierarchy (as opposed to egalitarianism) and individualism (as opposed to communitarianism) reduce the likelihood that an individual will express a belief in global warming because it would lead to unwanted restrictions on businesses and individuals (Kahan, Jenkins-Smith, and Braman 2011; Kahan et al. 2012). As predicted, hierarchical individualists are significantly less likely to accept the scientific consensus that global warming is happening relative to egalitarian communitarians (52 percent versus 85 percent predicted probability of expressing a belief that global warming is happening, respectively). Values not only shape the public’s views but also shape the views of scientists and policy advisors in the expected direction. The predicted probability of expressing a belief that global warming is happening for a scientist who values hierarchy and individualism is 81 % the probability of expressing a similar belief for a scientist who values egalitarianism and communitarianism is 97 percent. Thus, in line with a growing literature, we find that values play a central role in shaping views about the fundamentals of global warming (Dietz 2013; Kahan et al. 2012; Weber and Stern 2011).²⁴

We report the results from a second model (Table 1, model 2) for each sample to test hypothesis 3 about the politicization of climate science and ideological and partisan polarization on this issue rooted in the theory of motivated reasoning (Kunda 1990; Taber and Lodge 2006). We test this hypothesis by creating dichotomous measures for individuals who identified as a Democrat or Republican, and as a liberal or conservative, and interacting these variables with a measure of knowledge based on the number of correct responses to factual questions about politics, energy, and science (see Table A1). The results are striking and offer strong support for hypothesis 3. Members of the public who identify as Republican and as conservative and who are relatively more knowledgeable about politics, energy, and science are significantly less likely to say that global warming is happening relative to less knowledgeable conservatives. Among scientists and policy advisors, we do not find significant effects when party identification is interacted with knowledge; however, the coefficients for conservative and Republican are all signed in the expected negative direction, and knowledgeable, conservative scientists are significantly less likely to believe global warming is happening (p = .06, one-tailed test). Taken together, the results presented in Table 1 clearly demonstrate that the politicization of climate science has resulted in a partisan and ideological divide among members of the public, and to a lesser extent among scientists and policy advisors, over whether global warming is actually occurring, and this divide gets significantly wider as individuals become more knowledgeable about politics, energy, and science.

The results reported in Table 1 also show that minorities are significantly more likely than whites to express a belief in global warming among members of the public and policy advisors, but race does not play a role in determining the views of scientists. Among members of the public, people who perceive that they experience extreme local weather are significantly more likely to express a belief that global warming is happening. This is true both for members of the public and for policy advisors but not for scientists. Finally, media use, trust in government, and trust in science are associated with acceptance of the scientific consensus that global warming is happening among members of the public. Interestingly, among members of the public (see Table 1, model 1, public sample), education, knowledge, gender, and age appear to have no impact on the beliefs about whether global warming is happening; however, female scientists (see Table 1, model 1, scientist sample), minority policy advisors, and older policy advisors (see Table 1, model 1, policy advisors sample) are significantly more likely to express a belief that global warming is happening.

In Table 2, we report the results from a second series of ordered probit models that estimate factors that increase or decrease the likelihood of expressing a belief among members of the U.S. public, energy scientists, and policy advisors that global warming is happening due to humans’ actions as opposed to natural changes. In support of hypothesis 1, we find that members of the public who identify as liberal or as a Democrat are significantly more likely (64 percent predicted probability), and conservatives are significantly less likely (50 percent predicted probability), to express a belief that global warming is occurring as a result of humans’ actions. Similarly, the predicted probability of a liberal scientist expressing a belief that global warming is human induced is 92 percent, whereas for a conservative scientist, the predicted probability drops significantly to 63 percent. Thus, ideology and partisanship also play a central role in shaping each sample’s views that the warming trend is due to anthropogenic forces.

Table 2

Determinants of Belief that Global Warming Is Anthropogenic

	Public Sample				Scientist Sample				Policy Advisor Sample
	Model 1		Model 2		Model 1		Model 2		Model 1		Model 2
Demographics
Education	−.06**	(.03)	−.06**	(.03)	−.54	(.53)	−.58	(.54)	−.41	(.57)	−.07	(.67)
Age	−.00	(.00)	−.00	(.00)	.01*	(.00)	.01	(.00)	.02	(.05)	.00	(.05)
Female	.17***	(.05)	.17***	(.05)	.18	(.27)	.24	(.28)	.43	(1.17)	−.15	(1.35)
Income	−.01	(.02)	−.01	(.02)	−.13	(.09)	−.11	(.10)	−.21	(.36)	−.00	(.39)
Nonwhite	.10*	(.06)	.10*	(.06)	.61**	(.25)	.65**	(.25)	−.13	(1.21)	.19	(1.27)
Knowledge	.43***	(.12)	.72***	(.20)	1.25*	(.70)	2.60***	(.95)	.33	(1.75)	−5.41	(5.20)
Local Weather	.08***	(.03)	.08**	(.03)	.00	(.11)	.03	(.12)	.18	(.71)	.14	(.75)
Media	−.28**	(.11)	−.29**	(.11)	−.02	(.37)	.08	(.37)	−1.39	(1.18)	−1.51	(1.20)
Trust Gov.	−.13***	(.04)	−.12***	(.04)	−.10	(.18)	−.10	(.18)	.12	(.36)	−.21	(.56)
Trust Science	.04	(.03)	.04	(.03)	−.09	(.10)	−.12	(.11)	−.18	(.28)	−.34	(.32)
Values
Hierarchy	−.07***	(.01)	−.07***	(.01)	−.02	(.06)	−.02	(.06)	−.24	(.22)	−.44*	(.25)
Individualism	−.04***	(.01)	−.04**	(.01)	−.28***	(.07)	−.27***	(.07)	.23	(.28)	.27	(.29)
Political Characteristics
Republican	−.03	(.08)	.17	(.21)	.57*	(.34)	5.52*	(3.22)	−1.15	(2.57)	−3.12	(6.79)
Democrat	.14**	(.07)	.24	(.16)	−.39	(.27)	.24	(1.97)	1.04	(1.79)	1.83	(1.91)
Liberal	.26***	(.07)	.13	(.19)	−.41	(.26)	1.72	(1.96)	.02	(.91)	−5.05	(5.02)
Conservative	−.18**	(.07)	.14	(.18)	−.99***	(.32)	−1.35	(2.27)	.39	(1.06)	8.45	(6.45)
Liberal × Knowledge	—	—	.20	(.29)	—	—	−1.53	(2.20)	—	—	6.12	(5.67)
Conservative × Knowledge	—	—	−.56**	(.27)	—	—	.07	(2.90)	—	—	−9.44	(7.66)
Republican × Knowledge	—	—	−.31	(.30)	—	—	−5.40	(3.40)	—	—	3.78	(6.83)
Democrat × Knowledge	—	—	−.19	(.26)	—	—	−.83	(2.20)	—	—	—
Pseudo R²	0.04		0.04		0.14		0.15		0.18		0.22

NOTE: The columns contain ordered probit coefficient estimates predicting experimental treatment effects relative to the control (baseline) condition. The coefficients and standard errors for cut points 1 through 6 for model 1 (Public) are −0.80 (0.22), −0.01 (0.22), 0.67 (0.22), 0.99 (0.22), 1.54 (0.22), and 2.29 (0.23). The coefficients and standard errors for cut points 1 through 6 for model 1 (Scientist) are 1.68 (2.7), 2.63 (2.7), 3.52 (2.7), 3.70 (2.7), 4.2 (2.7), and 5.1 (2.8). The coefficients and standard errors for cut points 1 through 5 for model 1 (Policy Advisors) are −0.04 (3.2), 1.14 (3.2), 2.01 (3.3), 3.05 (3.3), and 3.5 (3.3). The coefficients and standard errors for cut points 1 through 6 for model 2 (Public) are −0.97 (0.24), −0.18 (0.23), 0.50 (0.23), 0.82 (0.23), 1.38 (0.24), and 2.13 (0.24). The coefficients and standard errors for cut points 1 through 6 for model 2 (Scientist) are 0.66 (2.9), 1.62 (2.9), 2.53 (2.9), 2.72 (2.9), 3.31 (2.9), and 4.24 (2.9). The coefficients and standard errors for cut points 1 through 5 for model 2 (Policy Advisors) are 3.33 (4.6), 4.6 (4.7), 5.5 (4.7), 6.6 (4.7), and 7.11 (4.7).

***

p ≤ .01; **p ≤ .05; *p ≤ .10 (one-tailed tests).

We again find strong support for hypothesis 2. Individuals who subscribe to a hierarchical and individualistic worldviews are significantly less likely to view global warming as happening due to humans’ actions. This is consistent with the cultural cognition thesis that the acceptance that global warming is happening threatens these individuals’ worldviews by leading to inevitable efforts to regulate businesses and place constraints on individuals’ freedoms (Kahan, Jenkins-Smith, and Braman 2011). These values not only shape the U.S. public’s views but also influence the way that scientists and policy advisors form opinions about the issue of global warming. In short, people who value individualism and hierarchy are less likely to believe in anthropogenic global warming compared with those who value egalitarianism and communitarianism.

To test hypothesis 3, we once again interact measures of party identification and ideology with our knowledge measure. The results, presented in Table 2 (model 2), offer strong support for the ideological polarization thesis rooted in the theory of motivated reasoning (see Gauchat 2012). Among members of the public, knowledgeable conservatives are significantly more likely than conservatives who lack factual knowledge about politics, energy, and science to express a belief that global warming is happening because of natural changes (as opposed to humans’ actions). There is less support for hypothesis 3 in looking at the effects of knowledge interacted with party identification and ideology among our sample of scientists and policy advisors. Although most of the coefficients are signed in the expected direction for scientists and policy advisors, they are not statistically significant.

To better illustrate the magnitude of the impact of knowledge among ideological subgroups within the public, we plot, in Figure 2, the predicted probability of expressing a belief among liberals and conservatives that global warming is happening and human induced, as knowledge increases from its minimum to its maximum value, holding all other covariates at their means.²⁵ The predicted value for the belief global warming is happening on a 7-point scale among the least knowledgeable conservative in the public sample is 5.32 (where 1 = definitely not happening and 7 = definitely happening); however, the predicted value for the most knowledgeable conservative on this question drops to 3.72. In contrast, there is a marginal increase in the belief among liberals that global warming is happening, in moving from the least (4.98) to most knowledgeable (5.25) in the sample. Policy advisors also display a tendency to engage in motivated reasoning on this issue to a greater extent as their levels of knowledge increase. The least knowledgeable conservative policy advisor’s predicted score for the belief that global warming is happening is a remarkably high 6.61 but drops significantly to 5.35 for the most knowledgeable conservative policy advisor. The reverse pattern—i.e., significant knowledge increases—is not detected for liberal policy advisors likely due to ceiling effects in terms of support for the scientific consensus (e.g., the predicted score for the least knowledgeable liberal in the policy-maker sample is 6.87). The predicted value that the belief that global warming is happening among the least knowledgeable conservative scientist is 5.49, and this increases to 5.70 for the most knowledgeable conservative scientist. The predicted increase for a liberal scientist in moving from the least to most knowledgeable is a modest 6.09 to 6.30 on the 7-point scale.

Figure 2

Predicted Values for Beliefs about Global Warming for Liberals and Conservatives among the Public

The right side of Figure 2 plots predicted values for the least and most knowledgeable liberals and conservatives in our public sample on whether global warming is human induced. The results follow the same pattern as the slope of the lines on the left side of Figure 2. Low-knowledge liberals and conservatives do not possess significantly different views about the fundamental cause of global warming; however, as conservatives become more knowledgeable, they become less likely to accept the scientific consensus regarding human-induced global warming. Conversely, knowledge marginally (but not significantly, see Table 2) increases the likelihood that liberals express a view that is consistent with the scientific consensus. These relationships are not apparent among the most and least knowledgeable conservative and liberal scientists and policy advisors; however, conservative scientists who believe global warming is happening are significantly more likely to report that it is the result of natural changes to Earth as opposed to human induced (see Table 2, model 1, scientist sample).

Although we did not offer explicit hypotheses about the impact of the control variables in Table 2, we find that females, minorities, individuals with greater factual knowledge, and those who experience extreme local weather are significantly more likely to see global warming as happening as a result of humans’ actions; however, the relationships are weaker or nonexistent in many cases among scientists and policy advisors.

Conclusion

Three sets of findings emerge from our analyses. First, we find significant differences in comparing the views of the public, scientists, and policy advisors on the issue of global warming and its fundamental cause. Similar to the results of recent surveys of nationally representative samples of the U.S. public, we find that 64 percent of the public believes that global warming is happening (Leiserowitz et al. 2014). This compares to nearly 90 percent of the energy scientists in our sample who express this belief and 71 percent of the policy advisors who completed our survey. Moreover, significant differences were observed in evaluating the belief that global warming is human caused across samples. Although the majority of the public (57 percent) sees global warming as caused by human action, widespread misperceptions of the scientific consensus still persist on this issue. Nearly one in five energy scientists (19 percent) and about one in three policy advisors (30 percent) in our sample express doubt or uncertainty about whether global warming is human induced (see Figure 1).

Second, we find strong support for the argument that values shape people’s beliefs about the fundamentals of global warming (Kahan et al. 2012). Scientists and policy advisors also appear to form beliefs about global warming that are consistent with their underlying values. People who value individualism and hierarchy are less supportive of action on global warming, whereas those who value egalitarianism and communitarianism are more supportive of action on global warming. This result holds true across all three of our samples.

Third, we demonstrate that the propensity to engage in motivated reasoning related to global warming increases as individuals become more knowledgeable about politics, energy, and science. Among the public and policy advisors, Democrats and liberals are more likely to accept the scientific consensus regarding the reality of global warming, while conservatives and Republicans are significantly less likely to express a belief that is consistent with the scientific consensus on this issue. Unfortunately, the tendency to pursue a directional goal rather than an accuracy goal on this issue becomes more pronounced as individuals become more knowledgeable (Kahan et al. 2012; McCright and Dunlap 2011; Malka, Krosnick, and Langer 2009), at least among members of the public.

The divisions we pinpoint across the public, energy scientists, and policy advisors—all key actors in the policymaking process—highlight major challenges to meaningful policy action to address the problem of global warming. They raise questions about the long-term implications of the politicization of science given the tendency for individuals to engage in directional motivated reasoning—the process whereby they process information with the goal of arriving at an opinion that is consistent with their partisan and ideological identities rather than the goal of holding an accurate opinion. Conservatives have become increasingly distrustful of science due to the rise of a “new Right” skeptical of organized science and the intellectual establishment. As Gauchat (2012) argues, “the relationship between public trust in science and political orientations also poses larger questions about the unevenness of the cultural authority of science and the potential for deep sociocultural divisions in the public sphere” (p. 168).

Understanding the opinions of key groups in the policymaking process including the public, scientists, and policy advisors is a crucial first step if political science is to contribute to developing responses to the vexing problem of global warming (McCright and Dunlap 2011; Javeline 2014). In moving forward, it is important to consider what, if anything, can be done to communicate climate science more effectively to create greater consensus necessary for meaningful policy action. Unfortunately, the problem is deeper than informing the public about the correct facts in the “debate” over global warming. Weber and Stern (2011, 323) explain that the problem in understanding global warming is not one of “illiteracy”—in comparison to the rest of the world, the U.S. public has average levels of knowledge—but rather that some individuals’ “mental models” of the world conflict with scientific consensus. In such cases, science communicators may find it difficult to motivate citizens to accurately process information in a way necessary to form a belief that is congruent with a consensus of scientists (Lupia 2013; Pidgeon and Fischoff 2011). Science can play an important role in informing citizens about what they should believe about how the world works, but it cannot tell us what people should care about: as Dietz (2013, 14085) has pointed out, “Science has no privilege with regard to values. However, continuing research on how values influence and are influenced by decision-making processes can help us hone better processes for identifying and coping with the diversity of values engaged around complex societal decisions.”

Given that citizens tend to conform their beliefs to those of their peers, “communicators should endeavor to create a deliberative climate in which accepting the best available science does not threaten any group’s values” (Kahan et al. 2012, 734). Thus, there is a need to find more effective frames and ways to counter the politicization of science so that citizens are open to new information that may lead to support for scientific adaptations that would benefit society (Bolsen, Druckman, and Cook 2014a; Emanuel 2013; Nisbet 2009; Stern 2011). One promising avenue for overcoming politicalization is to motivate citizens to form accurate beliefs when there is a clear scientific consensus surrounding a given issue (Bolsen, Druckman, and Cook 2014b). Future research must find additional ways to counter the politicization of science if we hope to combat the challenging and complex problems that we collectively face.

Footnotes

Appendix

Table A1

Descriptive Statistics across Samples

Variables	Question / Distribution	Public^a	Scientist^b	Policy Adv.^c
Variables	Question / Distribution	Mean (SD)	Mean (SD)	Mean (SD)
Female	Are you male (50%) or female (50%)	N/A	N/A	N/A
Nonwhite	Which of the following do you consider to be your primary racial or ethnic group? (white = 73%; minority = 27%)	N/A	N/A	N/A
Age	What is your age?	44.75 (16.43)	47.81 (12.04)	36.57 (8.58)
Education	What is the highest level of education you have completed? 1 = less than high school (1%^a); 2 = high school (16%^a); 3 = some college (39%^a; 0.41%^b); 4 = four year college degree (31%^a; 0.81%^b; 54%^c); 5 = advanced degree (13%^a; 99%^b; 46%^c)	3.37 (0.94)	4.98 (0.16)	4.45 (0.50)
Income	Estimate of family income (before taxes)	2.24 (1.02)	3.81 (0.97)	3.64 (0.98)
	1 = < $30,000 (24%^a; 3%^b; 2%^c); 2 = $30,000−$69,999 (42%^a; 7%^b; 16%^c); 3 = $70,000−$99,999 (19%^a; 18%^b; 11%^c); 4 = $100,000−$200,000 (13%^a; 50%^b; 58%^c); 5 = > 200,000 (2%^a; 22%^b; 13%^c)
Media	How often do you obtain energy information from … newspapers, TV, online (0–1 scale, alpha = .54)	N/A	0.51 (0.24)	0.79 (0.21)
PIDrep	Generally speaking, do you consider yourself a Democrat, independent, or Republican? 1 = strong Democrat (16%^a; 27%^b; 47%^c); 2 = weak Democrat (9%^a; 15%^b; 2%^c); 3 = lean Democrat (14%^a; 13%^b; 9%^c); 4 = independent (33%^a; 33%^b; 2%^c); 5 = lean Republican (12%^a; 7%^b; 6%^c); 6 = weak Republican (6%^a; 3%^b; 2%^c); 7 = strong Republican (12%^a; 3%^b; 32%^c).	3.81 (1.81)	3.00 (1.62)	3.53 (2.71)
Ideology	Which point on this scale best describes your political views? 1 = very liberal (6%^a; 12%^b; 4%^c); 2 = mostly liberal (10%^a; 28%^b; 32%^c); 3 = somewhat liberal (11%^a; 17%^b; 9%^c); 4 = moderate (36%^a; 22%^b; 15%^c); 5 = somewhat conservative (14%^a; 12%^b; 9%^c); 6 = mostly conservative (13%^a; 7%^b; 21%^c); 7 = very conservative (9%^a; 3%^b; 11%^c).	4.19 (1.59)	3.23 (1.55)	3.98 (1.93)
Hierarchy	Agreement with “We have gone too far in pushing equal rights in this country.” 1 = strongly disagree (21%^a; 35%^b; 33%^c); 2 = moderately disagree (9%^a; 17%^b; 16%^c); 3 = slightly disagree (9%^a; 9%^b; 11%^c); 4 = neither disagree nor agree (19%^a; 23%^b; 18%^c); 5 = slightly agree (16%^a; 7%^b; 4%^c); 6 = moderately agree (11%^a; 5%^b; 9%^c); 7 = strongly agree (16%^a; 4%^b; 9%^c)	3.95 (2.07)	2.83 (1.79)	3.07 (2.04)
Individualism	Agreement with “If the government spent less time trying to fix everyone’s problem, we’d all be a lot better off.” 1 = strongly disagree (6%^a; 24%^b; 26%^c); 2 = moderately disagree (6%^a; 24%^b; 13%^c); 3 = slightly disagree (7%^a; 12%^b; 6%^c); 4 = neither disagree nor agree (19%^a; 16%^b; 9%^c); 5 = slightly agree (17%^a; 9%^b; 13%^c); 6 = moderately agree (16%^a, 10%^b; 13%^c); 7 = strongly agree (28%^a, 7%^b; 21%^c)	4.96 (1.82)	3.17 (1.89)	3.94 (2.35)
TrustGov	How much of the time do you think you can trust the government in Washington to do what is right? 1 = never (20%^a; 5%^b; 9%^c); 2 = only some of the time (61%^a; 72%^b; 54%^c); 3 = most of the time (18%^a; 23%^b; 37%^c); 4 = just about always (2%^a)	2.02 (0.67)	2.18 (0.50)	2.28 (0.62)
TrustScience	To what extent do you think science enables us to overcome almost any problem?	3.71 (0.81)	3.72 (0.85)	3.41 (0.858)
	1 = not at all (0.5%^a; 0%^b; 2.17%^c) 2 = to a small extent (4.75%^a; 7.45%^b; 10.87%^c) 3 = to a moderate extent (34%^a; 31.77%^b; 36.96%^c) 4 = to a great extent (44.06%^a; 41.96%^b; 43.48%^c) 5 = to a very great extent (16.68%^a; 18.82%^b; 6.52%^c)
Political knowledge	Know majority required to override veto (56% correct)	0.68 (.33)	0.77 (0.36)	0.84 (0.35)
	Know which party has majority in U.S. House = (72% correct)
	Know whose responsibility it is to declare law unconstitutional = (76% correct)
	Know current U.S. Secretary of State = (67% correct)
Energy knowledge	Know the world’s largest exporter of oil = (63% correct)	0.50 (.30)	0.67 (0.32)	0.66 (0.38)
Energy knowledge	Know renewable energy sources = (63% correct)
	Know most U.S. oil not imported from Middle East = (24%)
Science knowledge	Is it true or false that lasers work by focusing sound waves?	0.64 (0.35)	0.85 (0.33)	0.61 (0.38)
	Which travels faster: light or sound?
	0 correct = 14%; 1 correct = 44%; 2 correct = 42%
Local Weather	Self-report relative climate where live where 1 = moderate (42%^a; 39%^b; 72%^c); 2 = relatively cold or warm (31%^a; 45%^b; 28%^c); 3 = extreme cold or warm (27%^a; 16%^b)	1.85 (0 .81)	1.76 (0.70)	1.28 (0.46)
Global Warming Happening	Global warming refers to the idea that the world’s average temperature has been increasing over the past 150 years and may be increasing more in the future. What do you think? Do you think that global warming is happening?	4.92 (1.80)	6.12 (1.17)	5.45 (1.67)
Global Warming Anthropogenic	If global warming is happening, to what extent do you think it is caused by human activities, as opposed to natural changes in the environment? (If you are “extremely sure global warming is not happening,” you can leave this answer blank.)	3.51 (1.71)	2.55 (1.60)	3.06 (1.77)

Note:

The authors thank Elizabeth Karampelas and Adnan Rasool for research assistance. They are also grateful to the Initiative for Sustainability and Energy at Northwestern for research funding.

Notes

Toby Bolsen is an assistant professor of political science at Georgia State University.

James N. Druckman is the Payson S. Wild Professor of Political Science, and a faculty fellow at the Institute for Policy Research at Northwestern University.

Fay Lomax Cook is on leave from Northwestern University as assistant director of the National Science Foundation and director of Social, Behavioral, and Economic Sciences at the National Science Foundation.

References

Akerlof

Karen

Rowan

Katherine E.

Fitzgerald

Dennis

Cedeno

Andrew Y.

2012. Communication of climate projections in U.S. media amid politicization of model science. Nature Climate Change 2 (9): 648–54.

American Association for the Advancement of Science (AAAS). 2014. What we know: The reality, risks, and response to climate change. Climate Science Panel. Washington, DC: AAAS. Available from http://whatweknow.aaas.org/get-the-facts.

Berkman

Michael B.

Plutzer

Eric

. 2011. Defeating creationism in the courtroom, but not in the classroom. Science 331 (28): 404–405.

Bolsen

Toby

. 2011. The construction of news: Energy crises, advocacy messages, and frames toward conservation. International Journal of Press / Politics 16 (2): 143–62.

Bolsen

Toby

Druckman

James N.

Cook

Fay Lomax

. 2014a. How frames can stunt support for scientific adaptations: Politicization and the status quo bias. Public Opinion Quarterly 78 (1): 1–26.

Bolsen

Toby

Druckman

James N.

Cook

Fay Lomax

. 2014b. The influence of partisan motivated reasoning on public opinion. Political Behavior 36 (2): 235–62.

Borick

Christopher P.

Rabe

Barry G.

2010. A reason to believe: Examining the factors that determine individual views on global warming. Social Science Quarterly 91 (3): 777–800.

Boykoff

Maxwell T.

2007. Flogging a dead norm? Newspaper coverage of anthropogenic climate change in the United States and United Kingdom from 2003 to 2006. Area 39:470–81.

Brulle

Robert J.

Carmichael

Jason

Jenkins

J. Craig

. 2012. Shifting public opinion on climate change: An empirical assessment of factors influencing concern over climate change in the U.S., 2002–2010. Climatic Change 114 (2): 169–88.

10.

Davidov

Eldad

Schmidt

Peter

Schwartz

Shalom H.

2008. Bringing the values back in the adequacy of the European Social Survey to measure values in 20 countries. Public Opinion Quarterly 72 (3): 420–45.

11.

Dietz

Thomas

. 2013. Bringing values and deliberation to science communication. Proceedings of the National Academy of Sciences 110 (3): 14081–14087.

12.

Douglas

Mary

Wildavsky

Aaron

. 1982. Risk and culture: An essay on the selection of technological and environmental dangers. Berkely, CA: University of California Press.

13.

Druckman

James N.

Peterson

Erik

Slothuus

Rune

. 2013. How elite partisan polarization affects public opinion formation. American Political Science Review 107 (1): 57–79.

14.

Druckman

James N.

2013. Stunted policy support. Nature Climate Change 3 (7): 617.

15.

Dunlap

Riley E.

McCright

Aaron M.

2008. A widening gap: Republican and Democratic views on climate change. Environment 50 (5): 26–35.

16.

Egan

Patrick J.

Mullin

Megan

. 2012. Turning personal experience into political attitudes: The effect of local weather on American’s perceptions about global warming. Journal of Politics 74 (3): 796–809.

17.

Emanuel

Ezekiel J.

2013. The future of biomedical research. Journal of the American Medical Association 309 (15): 1589–90.

18.

Feldman

Lauren

Maibach

E. W.

Roser-Renouf

Leiserowitz

Anthony

. 2011. Climate on cable: The nature and impact of global warming coverage on Fox News, CNN, and MSNBC. The International Journal of Press/Politics 17 (1): 3–31.

19.

Gauchat

Gordon

. 2012. Politicization of science in the public sphere: A study of public trust in the United States, 1974–2010. American Sociological Review 77 (2): 167–87.

20.

Giddens

Anthony

. 2009. The politics of climate change. Cambridge: Polity Press.

21.

Hamilton

Lawrence C.

2011. Education, politics, and opinions about climate change: Evidence for interaction effects. Climatic Change 104 (2): 231–42.

22.

Hamilton

Lawrence C.

Stampone

Mary D.

2013. Blowin’ in the wind: Short-term weather and belief in anthropogenic climate change. Weather, Climate, and Society 5 (2): 112–19.

23.

Hansen

James

Sato

Makiko

Ruedy

Reto

. 2012. Perception of climate change. Proceedings of the National Academy of Sciences 109 (37): E2415–E2423.

24.

Hart

P. Sol

Nisbet

Erik C.

2012. Boomerang effects in science communication: How motivated reasoning and identity cues amplify opinion polarization about climate mitigation policies. Communication Research 39 (6): 701–23.

25.

Hayes

Bernadette C.

2001. Gender, scientific knowledge, and attitudes toward the environment. Political Research Quarterly 54 (3): 657–71.

26.

Hmielowski

Jay D.

Feldman

Lauren

Myers

Teresa A.

Leiserowitz

Anthony

Maibach

Edward

. 2014. An attack on science? Media use, trust in scientists, and perceptions of global warming. Public Understanding of Science 23 (7): 866–83.

27.

Javeline

Debra

. 2014. The most important topic political scientists are not studying: Adapting to climate change. Perspectives on Politics 12 (2): 420–34.

28.

Jones

Michael D.

2011. Leading the way to compromise? Cultural theory and climate change opinion. PS (October):720–25.

29.

Kahan

Dan M.

Jenkins-Smith

Hank

Braman

Donald

. 2011. Cultural cognition of scientific consensus. Journal of Risk Research 14 (2): 147–74.

30.

Kahan

Dan M.

Peters

Ellen

Wittlin

Maggie

Slovic

Paul

Ouellette

Lisa L.

Braman

Donald

Mandel

Gregory

. 2012. The polarizing impact of scientific literacy and numeracy on perceived climate change risks. Nature Climate Change 2 (10): 732–35.

31.

Krosnick

Jon A.

2013. Red states, blue states agree on climate change. Testimony to the Bicameral Task Force on Climate Change. December 12, 2013. Available from http://www.casbs.org/casbs-fellow-jon-krosnick-red-states-blue-states-agree-climate-change (accessed 3 March 2014).

32.

Krosnick

Jon A.

MacInnis

. 2012. Trends in American public opinion on global warming policies between 2010 and 2012. Unpublished manuscript.

33.

Kunda

Ziva

. 1990. The case for motivated reasoning. Psychological Bulletin 108 (3): 480–98.

34.

Kunda

Ziva

. 1999. Social cognition: Making sense of people. Cambridge, MA: The MIT Press.

35.

Leiserowitz

Anthony

Maibach

Edward

Roser-Renouf

Connie

Feinberg

Geoff

Rosenthal

Seth

. 2014. Americans’ actions to limit global warming. Yale University and George Mason University. New Haven, CT: Yale Project on Climate Change Communication.

36.

Lewandowsky

Stephan

Gignac

Gilles E.

Vaughan

Samuel

. 2013. The pivotal role of perceived scientific consensus in acceptance of science. Nature Climate Change 3 (4) 399–404.

37.

Lupia

Arthur

. 2013. Communicating science in politicized environments. Proceedings of the National Academy of Sciences 110 (Supplement 3): 14048–14054.

38.

Malka

Ariel

Krosnick

Jon A.

Langer

Gary

. 2009. The association of knowledge with concern about global warming: Trusted information sources shape public thinking. Risk Analysis: An Official Publication of The Society for Risk Analysis 29 (5): 633–47.

39.

McCright

Aaron

. 2010. The effects of gender on climate change knowledge and concern in the American public. Population & Environment 32 (1): 66–87.

40.

McCright

Aaron M.

Dunlap

Riley E.

2011. The politicization of climate change and polarization in the American public’s views of global warming, 2001–2010. Sociological Quarterly 52 (2): 155–94.

41.

Melillo

Jerry M.

Richmond

Terese (T. C.)

Yohe

Gary W.

, eds. 2014. Highlights of climate change impacts in the United States: The third national climate assessment. Washington, DC: U.S. Global Change Research Program. Available from www.nca2014.globalchange.gov.

42.

Mildenberger

Matto

Leiserowitz

Anthony

. n.d. Why did public concern for climate change decline? Evidence from an opinion panel. Unpublished manuscript.

43.

Nisbet

Matthew C.

2009. Communicating climate change: Why frames matter for public engagement. Environment: Science and Policy for Sustainable Development 51 (2): 12–23.

44.

Nisbet

Matthew C.

Myers

Teresa

. 2007. The polls—trends: Twenty years of public opinion about global warming. Public Opinion Quarterly 71 (3): 444–70.

45.

Oreskes

Naomi

Conway

Erik M.

2010. Merchants of doubt: How a handful of scientists obscured the truth on issues from tobacco smoke to global warming. London: Bloomsbury Publishing.

46.

Petersen

Michael Bang

Skøv

Martin

Serritzlew

Søren

Ramsøy

Thomas

. 2013. Motivated reasoning and political parties: Evidence for increased processing in the face of party cues. Political Behavior 35 (4): 831–54.

47.

Pidgeon

Nick

Fischoff

Baruch

. 2011. The role of social and decision sciences in communicating uncertain climate risk. Nature Climate Change 1 (1): 35–41.

48.

Plutzer

Eric

Maney

Ardith

O’Conner

Robert E.

1998. Ideology and elites’ perceptions of the safety of new technologies. American Journal of Political Science 42 (1): 190–209.

49.

Rothman

Stanley

Lichter

Robert S.

1987. Elite ideology and risk perception in nuclear energy policy. American Political Science Review 81 (2): 383–404.

50.

Schuldt

Jonathon P.

Konrath

Sarah H.

Schwarz

Norbert

. 2011. “Global warming” or “climate change”? Whether the planet is warm depends on question wording. Public Opinion Quarterly 75 (1): 115–24.

51.

Schuldt

Jonathon P.

Roh

Sungjong

Schwarz

Norbert

. 2015. Questionnaire design effects in climate change surveys: Implications for the partisan divide. The ANNALS of the American Academy of Political and Social Science (this volume).

52.

Stern

Paul

. 2011. Psychological contributions to limiting climate change. American Psychologist 66 (4): 303–14.

53.

Stern

Paul C.

Dietz

Thomas

Guagnano

Gregory A.

1995. The new ecological paradigm in social-psychological context. Environment and Behavior 27 (6): 723–43.

54.

Taber

Charles S.

Lodge

Milton

. 2006. Motivated skepticism in the evaluation of political beliefs. American Journal of Political Science 50 (3): 755–69.

55.

Tomz

Michael

Wittenberg

Jason

King

Gary

. 2003. CLARIFY: Software for interpreting and presenting statistical results. Version 2.1. Stanford University, University of Wisconsin, and Harvard University. Available from http://gking.harvard.edu/.

56.

Weber

Elke U.

Stern

Paul C.

2011. Public understanding of climate change in the United States. American Psychologist 66 (4): 315–28.

57.

Zaller

John

. 1992. The nature and origins of mass opinion. New York, NY: Cambridge University Press.

58.

Zhao

Xiaoquan

Leiserowitz

Anthony A.

Maibach

Edward W.

Roser-Renouf

Connie

. 2011. Attention to science/environment news positively predicts and attention to political news negatively predicts global warming risk perceptions and policy support. Journal of Communication 61 (4): 713–31.

59.

Zia

Asim

Todd

Anne Marie

. 2010. Evaluating the effects of ideology on public understanding of climate change science: How to improve communication across ideological divides. Public Understanding of Science 19 (6): 743–61.