Mapping Racial and Ethnic Variation in Climate Belief Networks

Abstract

Past research observing differences in environmental attitudes across racial and ethnic groups often mischaracterized those differences as deficits, casting environmental concern as a predominately White issue. Our study contributes to current work correcting this account by mapping substantive differences in the structure of climate attitudes across racial and ethnic groups. We use belief network analysis on 14 years of survey data from the Climate Change in the American Mind (CCAM) survey (n = 20,396), and we find substantive differences in the climate belief networks of White, Black, and Hispanic survey respondents. These differences are not primarily about the strength or weakness of associations between attitudes, as theorized by deficit accounts. Instead, we find different attitudes are most central to these respective belief networks. We argue research on the social construction of race and ethnicity can better measure substantive variation in climate attitudes among Black, Indigenous, and People of Color (BIPOC) respondents by attending to how racialized experiences with climate change may produce aggregated belief networks with different profiles of salient issues and different interpretive frameworks.

Keywords

race/ethnicity climate change public opinion

Introduction

To better address the climate crisis, research needs to understand the different interpretive frameworks people use to make sense of climate change. People draw on a variety of moral (Graham, Haidt, and Nosek 2009) and cultural frameworks (Martin and Desmond 2010) to make sense of political issues, and so a “one-size-fits-all” messaging strategy is unlikely to overcome political polarization and accelerate support for effective mitigation policies. The first step in addressing this issue is to better measure and explain variation in climate attitudes across sociodemographic groups, which can guide future research on opinion formation at the individual level.

One area where research can do this more effectively is race and ethnicity. Early research on climate and environmental attitudes often perpetuated the notion that White respondents were more concerned about the environment than respondents from other groups racialized as non-White, primarily because it relied on very specific attitudinal measures (Hershey and Hill 1978). Later work correcting this account emphasized more varied measurement strategies, finding the interpretive frames people bring to climate and environmental policy vary across racialized communities (Goldberg et al. 2021; Pearson et al. 2021) because of racialized experiences with environmental injustice (Bhardwaj 2023).

Recent findings in the literature on climate attitudes and the sociology of race and ethnicity motivate our work in this area. First, given the volume of research on the association between partisanship and climate attitudes (McCright 2011; McCright and Dunlap 2008, 2011), it is easy to assume that partisan-motivated reasoning is a core driver of these attitudes. Recent work by Matthew H. Goldberg et al. (2021), however, finds that other factors matter as well, namely worry about climate change, risk perceptions of climate change, certainty that change is happening, and belief that change is human-caused. This work suggests that more attitudinal factors are at play in climate belief systems, and therefore, there is more variation to map across different social groups. Second, Salil Benegal, Flávio Azevedo, and Mirya R. Holman (2022) find that conventional assumptions about the association between partisanship and climate views only hold as expected for White Americans, and that climate attitudes vary across racial and ethnic groups in meaningful ways. Third, theories of racialized subjectivity and cognition give us reason to extend these findings with additional attitudes about climate change. A potential source of differences in climate attitudes could be the racialized exposure to environmental risks (Bhardwaj 2023; Brekhus et al. 2010; Taylor 2014). Therefore, there may be additional variance in climate attitudes, which can be explained as one of many facets in a composite understanding of race (Sen and Wasow 2016).

In this study, we propose a way to validate these findings and extend them with a method that can better map variation in climate attitudes across racial and ethnic groups in the face of the climate crisis. In short, we conceptualize climate attitudes as belief networks (Boutyline and Vaisey 2017; Brandt and Sleegers 2021; DellaPosta 2020; Fishman and Davis 2022; Lee et al. 2024), and we ask, is there substantive variation in the aggregated climate belief networks across different racial and ethnic groups? Our analysis of public, nationally representative survey data from the Climate Change in the American Mind (CCAM) survey (2008–2022, n = 20,396) finds that the strength and network centrality between attitudes substantively varies across groups. We find some minor differences in the strength of partisanship ties to climate attitudes, but we also find differences in network centrality that emphasize different aspects of climate beliefs. Our results demonstrate substantive variation not only in which issues matter to key stakeholders in the climate crisis but also how those issues are mutually constituted in belief systems. To close, we discuss how these findings will help future research and advocacy better conceptualize and operationalize opinions about the climate crisis to study movements in different racialized communities.

Racialized Climate Attitudes

The literature examining environmental and climate attitudes across race and ethnicity can be characterized by three waves dating back to the 1970s with the emergence of the international environmental movement and scientific consensus on global warming (Dunlap and Catton 1994; Lazri and Konisky 2019). The first wave focused on differences in environmental interest and concern among Black and White respondents and implicitly framed environmental issues as a predominately White concern due to imprecise measurement strategies (Hershey and Hill 1978; Hohm 1976; Kellert 1984; Kreger 1973). Specifically, this research framed Black respondents’ lower levels of support for the environmental movement in a way that presumed Black indifference to general environmental matters (Hershey and Hill 1978; Theodore Washington 1976). Like other “deficit models” of conceptualizing both racial inequality (Cabrera 2019; Valencia 2002) and climate attitudes (Suldovsky 2017), these findings reify assumptions about the climate and environment, which further perpetuate inaccurate distinctions between racial groups.

In a second wave, researchers used a wider range of measures, including the perceived seriousness of various environmental issues, pollution (e.g., air, water, and general pollution), nature preservation (e.g., loss of animal habitats and acid rain), and global environmental problems (e.g., ozone layer depletion and global warming). This research found Black respondents expressing equal, even higher, levels of climate and environmental concern relative to White respondents (Caron 1989; Mohai 1990; Mohai and Bryant 1998; Uyeki and Holland 2000). Past research showing Black respondents’ lower levels of participation in climate and environmental matters was associated with feelings of powerlessness (Parker and McDonough 1999) contradicting the idea that economically vulnerable groups express climate and environmental apathy.

The third and current wave is characterized by a transition to focusing on climate attitudes in the face of climate change, disaggregation of racial and ethnic categories, attention to an even wider range of attitudinal measures, and new insights about the mechanisms that produce variation in climate attitudes across racial and ethnic groups. Along with larger survey samples, research during this period disaggregates the Black/White racial dichotomy by pursuing more representative samples including Latinos, Asians, Native Americans, and Other historically underrepresented groups (Burger et al. 2004). In the larger context of work reevaluating classic measures of political ideology among Black respondents specifically, researchers have good reason to question the conventional wisdom of oft-cited mechanisms for attitude formation, such as political polarization, in different racialized groups (Jefferson 2020). For example, Benegal, Azevedo, and Holman (2022) found that conventional assumptions about the association between partisanship and climate views only hold as expected for White Americans, and that climate attitudes vary across racial and ethnic groups in meaningful ways (Hegtvedt, Parris, and Johnson 2019). Extant literature shows that researchers can map meaningful variation in how people in different social locations conceptualize issues and coordinate opinions (Ford and Norgaard 2020; Martin and Desmond 2010), and we leverage both these insights and recent methodological developments to better understand variation in climate attitudes among racial groups.

Racial Subjectivity

This third wave of research in racialized climate attitudes shows us how continuing this work requires theorizing race in a framework of both social stratification and meaning-making processes. This kind of cognitive approach to the sociology of race (Brekhus et al. 2010) focuses on how social structures produce variation in the interpretive frames that people bring to bear on their lives. Rather than treating race as an immutable characteristic, a cognitive approach encourages researchers to consider race as a “bundle of sticks” (Sen and Wasow 2016:506)—a composite of other social indicators in the way that social class is composed of similar sets of wealth, income, and socialization.

In this sense, we treat the composition of climate and environmental attitudes as one “stick” among many in the bundle that we can study as a manifestation of racialized experiences in the context of the United States. Following a Du Boisian approach (Itzigsohn and Brown 2020), we expect existing racialized social structures to “map onto our cognitions” that guide our lives (Brekhus et al. 2010:64). Disparate exposure to environmental hazards through residential segregation is an essential example of this process, with racially privileged communities benefiting from being proximally distant from climate and environmental hazards (Carrillo 2021:646–47). Similarly, poor and marginalized communities are more exposed and vulnerable to the adverse impacts of climate change (Blaikie et al. 2014; Collins 2010; Grineski et al. 2012; Maldonado et al. 2016; Zahran et al. 2008).

The development of the environmental state shows how social stratification, unequal access to environmental resources, and disproportionate exposure to climate risk may lead to racial variation in climate attitudes. Ian Carrillo (2021:652) describes a “treadmill of production” in the environmental state whereby the state supports large firms in their pursuit of economic growth and consumption while adopting a “racial fix to bend social forces to its interests.” This better accounts for explaining how communities of color are disproportionately exposed to environmental risks (Ard 2015; Bell and Ebisu 2012; Crowder and Downey 2010; Mohai et al. 2009; Taylor 2014). This approach keeps racialized politics at the center for theories of the environmental state and not along the periphery, as the dominant theories, such as the Treadmill of Production Theory and Ecological Modernization Theory, have suggested (Carrillo 2021).

Moreover, these environmental inequalities foster what Ankit Bhardwaj (2023) identifies in W.E.B. Du Bois’ work as “environmental racialization”—the process through which environmental conflict encourages and fosters the social construction of race. As a result, we expect climate attitudes to vary racially along a “color line,” though not necessarily in terms of simply stronger or weaker attitudes among specific groups. Instead, we expect that racialized subjectivity maps onto climate attitudes in distinct, measurable ways across groups, and that more empirical work is needed to map the full extent of this variation and to understand how race and climate are coconstitutive in contemporary belief systems. In the study that follows, we leverage a new and growing method in public opinion research to do this.

The Current Study: Racial and Ethnic Variation in Climate Belief Networks

Studies investigating changes in public opinion have traditionally focused on polarization, where attitudes become either more extreme among different social groups or more clearly sorted into different ideological camps (Bail et al. 2018; Baldassarri and Gelman 2008; DiMaggio, Evans, and Bryson 1996; Iyengar and Westwood 2015). But an emerging body of work understands polarization differently—not necessarily as changes in ideological extremity, but changes in the extent to which additional beliefs become more tightly integrated into a collection of partisan views (DellaPosta 2020; DellaPosta, Shi, and Macy 2015; Quintana 2023). This work understands attitudes as “belief networks” (Boutyline and Vaisey 2017; Brandt and Sleegers 2021; Epskamp, Borsboom, and Fried 2018; Lee et al. 2024). A belief network is a series of interdependent attitudes that demonstrate the tight or loose integration of a person’s or a group’s ideological characteristics. While work in political psychology focuses on networks at the individual level, belief network analysis (BNA) is perhaps most useful for inferring attitudinal patterns that emerge as characteristics of the social structure at the population level (Fishman and Davis 2022:661), rather than inferring any single person’s attitudinal structure.

In a BNA, researchers treat survey items as nodes and associations between survey items as indicators of the strength of a network tie between them. The advantage of this approach is that this strength measure is not the only characteristic of the network. BNA allows researchers to estimate other elements of network structure including centrality and closeness. As a result, researchers can observe more characteristics of ideological constraint than conventional correlation matrices and other similar measurement strategies typically allow.

Different theories of interpretation and social behavior may posit that different beliefs are most central to their respective belief networks because those central items are more strongly correlated with a wider range of other attitudes. For example, Andrei Boutyline and Stephen Vaisey (2017) use this approach to formally test George Lakoff’s (2002) theory of moral politics by evaluating belief networks of political attitudes. Lakoff’s theory posits that parenting styles are a central element of the construction of different moral and political frameworks. Their network analysis finds that left/right ideology is more central to the belief networks of survey respondents relative to parenting styles, which emerged in the periphery of the network. Methodologically, this shows why belief networks are so useful; they let us observe and adjudicate claims about the “core” and the “periphery” of ideological systems in a more comprehensive way than other measures of association.

Network approaches allow us to capture additional variation in the structure of attitudes among different groups of people. It is possible that variation in climate attitudes and partisan association observed by Salil Benegal et al. (2022) is due to substantive variation in climate attitudes across racialized groups. Because the literature is moving to correct the implicit assumption that specific attitudinal measures are implicitly racialized as White (Jefferson 2020), variation in the belief networks of different racialized groups can yield additional insights into substantive, interpretative variation in support for environmental protection than previously understood. We can benefit from understanding which attitudes are most central to the interrelated network of beliefs and opinions about climate change and whether those central attitudes have changed over time. It may be that different attitudes are central to these different belief system networks, or that different attitudes serve as “bridges” between other attitudes.

If climate beliefs are racialized, that is, if people in different racialized groups have substantively different experiences that lead to varied conceptualizations of those climate beliefs (Bhardwaj 2023), we expect to find variation in the structural composition of belief networks. We focus on climate attitudes—opinions about the risks, concerns, and causes of climate change—as opposed to environmental attitudes more broadly. We do this because the stakes of climate change stand to impose even stronger racialized environmental inequities and because there can be substantive differences in how people view environmental and climate issues.

The literature suggests that we should consider five major dimensions of climate beliefs in our analysis. The first dimension is partisan-motivated reasoning which expects climate beliefs to be shaped by partisan predispositions, cues, and leaders (Bisgaard 2015; Bolin and Hamilton 2018; Bolsen, Druckman, and Cook 2014; Hart and Nisbet 2012; Knight 2016; Lenz 2012; McCright and Dunlap 2008, 2011) meaning that partisan predispositions ought to have a high centrality score in the belief network. The second dimension is beliefs about human-caused climate change. These may be differentially integrated into different belief networks whereby some people, for example, may not necessarily need to accept the factual explanation of the problem to be concerned about it, while others may accept the facts without expressing substantive concern due to partisanship. The third dimension is support for specific policies designed to mitigate climate change. The fourth is risk perceptions or the extent of harm that climate change would cause on oneself and others. Finally, the fifth dimension is worry about climate change or general concern about the issue. Using relative weight analysis, Goldberg et al. (2021) found that the fifth dimension, affect and worry, had the largest effect of the dimensions on climate policy support.

People may not express equal levels of support across these different measures of climate attitudes. Our study, however, focuses more on whether these attitudes are integrated in different ways. Varied levels of exposure and perceived climate risks (O’Brien et al. 2020) may motivate different kinds of urgency—even independent of partisanship—toward different forms of climate and environmental reforms (Dacey and Stewart 2023; Kyselá, Tvinnereim, and Ivarsflaten 2019).

Our null hypothesis to this theory is that climate belief networks among all racial groups will show little to no differences in the structural composition of the relationships between climate attitude dimensions. Thus, we expect the following:

Null Hypothesis: Estimates of network centrality will be equal for similar climate attitudes across racial and ethnic groups.

But if theories of racial subjectivity and environmental racialization hold in this case, we expect, in contrast, the following:

Hypothesis 1: Estimates of network centrality for similar climate attitudes will vary across racial and ethnic groups.

Method

Data

We use data from the CCAM survey (Ballew et al. 2019). CCAM is a publicly available, nationally representative cross-sectional data set of 26 waves collected from 2008 through 2022.¹ Surveys were completed once in 2008, then twice each year from 2010 to 2022 (no survey was administered in 2009 and only one wave exists for 2022). Respondents were recruited using Ipsos KnowledgePanel with a probability sampling design including random digit dialing and address-based sampling strategies to generate a nationally representative sample of U.S. residents aged 18 years and older. Sampling weights were computed to the Current Population Survey and are included in all analyses.

All waves of the survey included our core measures of interest. The pooled sample of 14 years of data totals 30,136 observations. After dropping respondents who answered “don’t know” on each attitudinal item and using listwise deletion to account for missing data, our analytic sample consists of 20,396 observations.

Measures

Core measures of interest include risk perception toward climate change, attitudes about environmental policy, worry about climate change, and belief about the cause of climate change. We present descriptive statistics for our primary variables in Table 1.

Table 1.

Descriptive Statistics for Core Measures.

Measure	M	SD	Min	Max
Party x Ideology Scale (5-Liberal Democrat)	3.04	1.58	1	5
Belief in Human-Caused Climate Change	3.12	1.05	1	4
Worry about Climate Change	2.66	0.99	1	4
Perceived Harm (Personal)	2.3	1.01	1	4
Perceived Harm (To the US)	2.73	1.06	1	4
Perceived Harm (To Developing Countries)	2.94	1.11	1	4
Perceived Harm (To Future Generations)	3.17	1.08	1	4
Perceived Harm (To Plants & Animals)	3.15	1.08	1	4
Support for Regulating CO₂ Pollutants	2.97	0.95	1	4
Support for Funding Research on Climate Change	3.22	0.87	1	4
	N	%
Black	1531	7.5
Hispanic	1916	9.4
Other	1353	6.6
White	15,596	76.5

Source. Climate Change in the American Mind survey, Aggregated Sample (2008–2022).

To measure partisanship, we used the CCAM’s included “ideology x partisanship” scale, which aggregates political party and political ideology identification to generate a 5-point scale ranging from conservative Republican to progressive Democrat.

To measure belief about the cause of climate change, we used CCAM’s measure asking respondents to Assume global warming is happening do you think it is . . . caused mostly by natural changes in the environment, caused mostly by human activities and natural changes caused mostly by human activities, neither because global warming isn’t happening, don’t know, and other, please specify. We recoded this measure into a 4-point ordinal scale for strength of belief in human-caused climate change.

Attitudes about environmental policy were measured with two survey questions. These included whether respondents felt strongly opposed, somewhat opposed, somewhat supported, or strongly supported regulating carbon dioxide (the primary greenhouse gas) as a pollutant and fund[ing] more research into renewable energy sources, such as solar and wind power. The mean average for regulating carbon dioxide emission was 2.97, while support for funding research was 3.22, suggesting that support for one kind of environmental regulation does not mirror support for others.

Risk perception variables were measured by a series of questions asking respondents to what extent they thought climate change will harm [them] personally, harm plants and animals, harm developing countries, or harm future generations. For each risk perception question, respondents indicated if they did not know, not at all, only a little, a moderate amount, or a great deal. Respondents on average were least concerned about climate change harming them personally and were most concerned about climate change harming plants and animals.

Finally, worry about climate change was measured by asking respondents if they were not at all worried, not very worried, somewhat worried, or very worried about climate change. The mean for worry was 2.66, suggesting respondents were only moderately worried about climate change, which reflects a national trend extending back to Anthony A. Leiserowitz’s (2005) study of climate concern.

To measure racial and ethnic identities, CCAM survey used two questions, beginning with Hispanic self-identification and followed by identification as White, Black or African American, American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, or Some Other race. The public release of the survey data includes a four-category classification based on these two items which includes White, non-Hispanic; Black, non-Hispanic; Other, non-Hispanic; and Hispanic. We used this item to create four dichotomous indicators for Black, Hispanic, White, and “other” groups to compare attitudes across groups which, unfortunately, require the aggregation of smaller racial and ethnic groups into an “other” category. As we note below, we include this “other” category because we do find useful comparisons between the belief network for this group and the other groups included in the study. However, we also advise strong caution in the interpretation of findings about this group, and we encourage future replication in different survey data sets to better evaluate climate attitudes in other ethnic communities.

Analytic Approach

Following previous work examining political ideology and polarization, we use a BNA approach to understand the structure of the relationships between these five sets of attitudes. BNA treats each measure as a “node” in a network and uses the partial correlation coefficients between measures to construct weighted ties between those nodes.

Our analysis uses the bootnet package in R (Epskamp et al. 2018) with the estimateNetwork() command to generate our belief networks. Because many of our indicators are ordinal measures, we use the command’s cor auto option for polychoric and polyserial correlations between the measures, and we use the command’s default estimation approach with Gaussian Markov random field estimation.

First, we estimate the full sample belief network. To address our core research expectation, we aggregate the survey waves to maximize our sample size of Black and Hispanic respondents and estimate separate belief networks for each racial and ethnic identification option in the survey data. After a visual comparison of the networks, we present estimates of key network centrality indices for each attitudinal measure in each network: the strength, betweenness, and closeness for each survey item. Strength is based on the direct correlation of each node to all other nodes in the network, closeness is based on the indirect correlation of each item to all other nodes in the network, and betweenness is an indicator of the importance of each node in terms of its “average path between two other nodes” (Epskamp et al. 2018:196). We compare estimates and bootstrapped 95-percent confidence intervals (CIs) for each of these indicators from 1,000 resamples from each network data set. By comparing these measures, we can identify substantively different characteristics of the belief networks from each racialized subsample of respondents.

Finally, it may be the case that racialized variation in belief networks fluctuates over time. To evaluate this, we also conduct separate network estimates by year for key indicators identified in our aggregated analysis and visualize changes over time.

Results

Full Sample

First, we visualize the belief network for the entire aggregated CCAM survey in Figure 1. Thicker edges in the network indicate stronger correlations, and we see unsurprisingly strong correlations between conceptually similar measures. For example, support for regulating carbon dioxide pollution and support for funding research are positively correlated, as are many of the measures of perceived harm due to climate change.

Figure 1.

Climate belief network for full aggregated sample, 2008 to 2022.

Figure 2 presents centrality indices for each item in the full network—closeness, betweenness, and strength—using z scores to interpret results in terms of standard deviations across each measure. These illustrate the importance of using separate measures for different kinds of centrality. In terms of strength—direct ties between measures—harm indicators are among the strongest elements of the network due to their close interrelationship with each other. Harm to the United States (1.12) and harm to future generations (1.77) each had a strength estimate over one standard deviation higher than the other indicators. While worry about climate change has a comparatively lower strength indicator to these factors, it is among the highest in closeness (indirect ties between measures: z = 1.67) and betweenness (relative importance to paths between measures: z = 1.60). Visually, the summary of closeness, betweenness, and strength measures for the network presented in Figure 2 corresponds to the visual patterns in Figure 1—we can see how worry about climate change is the “bridge” node that sits between clusters of perceived harm items and other ideological considerations such as partisanship and support for regulation. This corroborates previous findings that climate concern explains a high proportion of variance in climate attitudes and prior analyses of aggregated climate belief networks.²

Figure 2.

Centrality plot for full aggregated sample.

Contrary to expectations from partisan-motivated reasoning literature, the strength of partisanship occupies the periphery of the belief network with below-average scores on each measure (closeness z = −1.21, betweenness z = −1.02, strength z = −1.51). This is the advantage of a belief network approach—it reveals different contributions of different items relative to their overall association to the network structure, and not just items that exhibit the strongest correlations.

Network Variation by Race and Ethnicity

If the structure of climate attitudes is uniform across race and ethnicity, we would not expect to see much variation when estimating belief networks separately among Black, Hispanic, White, and “Other” respondents identified in the survey sample. But as the networks show visually in Figure 3, these groups exhibit distinct differences in network composition. Most notable is the varied placement of worry about climate change. Worry appears most centrally in the network of White respondents only. For Black respondents, the item assessing belief in human-caused climate change is more central. For Hispanic respondents, and to a lesser extent respondents in the “other” group, harm measures appear more central in the belief network. This descriptive pattern provides some preliminary evidence that these groups have structurally different belief networks.

Figure 3.

Climate belief networks for racial and ethnic groups, 2008 to 2022.

Network composition is sensitive to researcher design choices. While we selected identical layout rules for each network (the “spring” layout in bootnet’s plotting rules), these visual differences may not necessarily indicate substantive differences in the underlying network structure. To evaluate differences in betweenness, closeness, and strength across groups, we used a bootstrapping estimation strategy on 1,000 sampled iterations of each group to produce average estimates and 95-percent CIs. Figure 4 summarizes these results and highlights some important insights about racial and ethnic variation in climate beliefs.

Figure 4.

Bootstrapped centrality plot for racial and ethnic groups.

Three key differences in network centrality emerge from Figure 4. First, we find that the betweenness score for worry about climate change for White respondents (M: 29.11, CI: 25.00–42.99) is substantively higher than the betweenness scores for worry among Black respondents (M: 14.1, CI: 6.17–29.80), Hispanic respondents (M: 13.1, CI: 4.71–31.30), and Other respondents (M: 15.1, CI: 5.91–30.10). Second, we find that Black respondents had substantively higher betweenness for belief in human-caused climate change (M: 12.2, CI: 6.00–38.00) than White respondents (M: .80, CI: −3.12 to 3.12), Hispanic respondents (M: 11.6, CI: −0.15 to 20.1), and Other respondents (M: 5.63, CI: −4.07 to 12.1). Third, Hispanic respondents had higher betweenness scores for perceived harm to developing countries (M: 4.65, CI: 2.71–21.30), than White (M: 2.34, CI: −5.96 to 5.96), Black (M: 7.09, CI: −4.37 to 12.40), or Other respondents (M: 3.12, CI: −2.24 to 10.20). We focus on these three differences because they are the largest in magnitude for each group.

Other suggestive differences are present in Figure 4 as well, such as similar betweenness scores for perceived harm to future generations among White and Hispanic respondents. Other similarities also emerge. For example, our bootstrapped estimates for strength of ties across measures are relatively equal across groups, with slightly higher strength of ties for partisanship among White respondents. It is important to emphasize that our comparisons do not find strong, clearly structured attitudes among White respondents and weak, diffuse, or apathetic attitudes among other racial and ethnic groups.

One limitation of CCAM data is that the public release of the data set aggregated all other racial and ethnic identifications into a composite “other” group. We advise caution in the interpretation of results for this group, but we do find some notable differences here as well. Respondents in this group exhibited substantively lower estimates for the betweenness of regulating carbon dioxide pollutants (M: 3.36, CI: −4.58 to 8.58) and substantively higher betweenness estimates for concern about harm to plants and animals (M: 10.2, CI: 2.28–21.7) and support for funding research (M: 5.79, CI:1.77–18.2). This group also produced some of the highest closeness estimates for many measures in the sample, indicating that they had a stronger average of indirect correlations across all measures.

While belief networks are primarily characteristics of groups, one important implication of our theory of racial subjectivities is that differences are not simply reducible to a single belief network for each racial and ethnic group. It is also important to examine whether and how these network characteristics might vary within groups. To illustrate this concept, we examine two key correlates of ideological change and constraint—time and educational attainment—with respect to worry about climate change, belief in human-caused change, and concern about harm to developing countries.

Recall that betweenness is a measure of the relative importance of each node in terms of its placement in the average path between other nodes in the network. In Figure 5, we examine changes in betweenness over time, estimated in separate year samples for Black, Hispanic, and White respondents. Disaggregating by survey year requires sacrifices in sample size. We use 2016 to 2021 as the period to present a five-year time span with consistently high subsample sizes of Black and Hispanic respondents. Here we see that estimates for White respondents are relatively stable over time, especially the elevated betweenness scores for worry. In contrast, estimates of betweenness scores are more comparable and trend downward over time for Black and Hispanic respondents, suggesting that the relative importance of any single measure has declined over time. Because cohort change is an important component of large-scale attitudinal change (Kiley and Vaisey 2020), these results suggest that over time, patterns in cohort replacement might also vary by race and ethnicity, suggesting unique, potentially racialized trajectories in opinion change for future research to examine.

Figure 5.

Change in network betweenness of key climate attitudes over time.

In Figure 6, we use CCAM’s four-category education measure (less than high school, high school, some college, and bachelor’s degree or higher) to stratify our full sample and estimate four belief networks for each racial and ethnic group at each level of educational attainment. Here we see differences within racial and ethnic groups. For example, Black respondents show a unique divide in the betweenness of belief in human-caused climate change; this measure is more central to Black respondents with a high school diploma or less, and less central to Black respondents with college exposure. Worry is also more central to belief networks among more highly educated White respondents, but less central for the most highly educated Black and Hispanic respondents. These descriptive results further demonstrate the stakes of our core argument about racial subjectivities: The relationship between demographic predictors and attitudes may vary both within and between racial and ethnic groups.

Figure 6.

Variation in network betweenness of key climate attitudes by education.

Discussion

Using BNA, this study sought to test variation in the network structure of climate attitudes among respondents from different racial and ethnic groups in the United States. Theories of racialized subjectivity and opinion formation suggested that we might find evidence for different attitudinal formation around the climate crisis, and our findings support these theories.

Two important patterns inform our conclusions. First, focusing only on the strength of association between measures yields many commonalities across racial and ethnic groups, though White respondents do exhibit some of the strongest associations between partisanship and other climate attitudes as observed in previous research. Second, focusing on other measures of network centrality—especially betweenness—finds substantive differences in which measures appear most central to climate belief networks among different racial and ethnic groups. This variation is not primarily in terms of strong or weak attitudes among different groups, but rather which attitudes are most tightly integrated into the network and therefore might be more important to emphasize in different stakeholder communities affected by climate change.

In interpreting these findings and applying them to the literature on race, environmental issues, and political attitudes, it is important to emphasize that network centrality is a feature of population-level issue salience more so than individual-level beliefs that produce changes in opinions (Fishman and Davis 2022; Kiley and Vaisey 2020). Our goal is not to say that people who belong to any particular racial or ethnic group will necessarily view climate change differently, but rather to note how different compositions of belief networks indicate broader patterns in issue salience that could be a result of different racialized experiences, socialization, and interpretive frames in response to environmental inequality (Bhardwaj 2023; Brekhus et al. 2010; Sen and Wasow 2016).

For example, our measure of belief in human-caused climate change was most central to Black respondents’ belief network, suggesting that it is more salient and occupies a stronger “bridging” position in relation to other attitudes than the position occupied by climate concern in the White network and the full sample. This suggests a wider variety of attitudes are indirectly related to climate beliefs, rather than affective worry, across our aggregated sample of Black respondents. This conclusion makes sense in light of previous research showing greater support for climate change mitigation policies among Black respondents (Dietz, Dan, and Shwom 2007), and that Black respondents’ pro-environmental behaviors are associated with both environmentalist and environmental justice-oriented frames (Hegtvedt et al. 2019). In other words, respondents here do not necessarily have to be as worried about climate change to link perceived harms of climate change to support for mitigation policies.

Another possible explanation for our finding is what Salil D. Benegal (2018) shows is a spillover of race and racial attitudes into public opinion about climate change. The election of Barack Obama included more progressive policy reforms such as in health care and climate change policy. The view that human activity was causing climate change was central to Obama’s climate initiatives, and Benegal (2018) argues that this attitude has spilled over and remained stable among Black American public opinion and conversely for mostly White conservative Americans. Both explanations could fit our finding that belief in anthropogenic climate change is more central to Black Americans’ climate belief network. We also found, however, that estimates of this betweenness vary with education—belief is much less central in networks estimated with a subsample of Black respondents who attended college. Although further research can elaborate on and test these explanations, our key observation here is that these observed Black/White differences in attitudinal structure are important, and they do not necessarily indicate a deficit of concern or support among Black respondents, relative to White respondents.

Harm measures are central to the Hispanic respondent belief network, specifically perceived harm to developing countries and perceived harm to future generations. Climate change beliefs and risk perceptions have been tightly associated with cultural values and worldviews such as egalitarianism, and individualism (Kahan 2007). Our findings confirm previous research on Latinos’ high perception of climate risks compared with Whites and suggest that Latinos’ alignment with egalitarian values shapes their climate belief network (Ballew et al. 2019; Goldberg et al. 2020). Two potential explanations for the centrality of harm in the Latino community are the immigration experience and the role of the cultural value of familism. Past research has shown that familism—a value representing the prioritization of the family—is a stronger predictor for Latinos’ concern for cross-generational harms of climate change compared with Whites, for which education and partisanship were stronger predictors (Pearson et al. 2021).

In contrast to belief and perceived harm to future generations and developing countries, the measure for worry about climate change was most central to the climate belief network for White respondents. Here, views about harm and regulation are bridged by the central position of affective concern about climate change. We also show that estimates of the network centrality of worry are stable over time for White respondents, while they decline over time for Black and Hispanic respondents, and higher for more highly educated White respondents.

Past research shows that White respondents, and to a larger extent, White men, are more likely to align with individualistic worldviews that are associated with weaker climate concern and pro-environmental attitudes (McCright et al. 2016). Davin L. Phoenix (2019) shows that White people in the United States evoke emotions, particularly anger, within political systems because it tends to produce positive political outcomes especially during presidential elections, while for Black Americans, evoking emotions tends to have negative political outcomes. Our findings show that personal harm and negative affect, understood as individualist frames of climate impacts, are most central for White respondents. This is an additional reason why worry might not occupy the core of Black respondents’ belief networks as well, because worry as an affective state may not be perceived as an avenue to address climate issues in the same way as it is for White respondents.

These findings contribute to three key bodies of literature. First, they help us to better conceptualize racial and ethnic variation in climate attitudes and experiences. We know that the impacts of climate change are not equally distributed—Black, Indigenous, and People of Color (BIPOC) communities are much more likely to be exposed to elevated levels of pollution (Li, Konisky, and Zirogiannis 2019), extreme heat (O’Brien et al. 2020), and flooding (Grineski, Chakraborty, and Montgomery 2015; Maldonado et al. 2016). More classic deficit models might expect that this inequality would produce resignation in the face of the climate crisis, as there is evidence that experiences with exposure to climate/environmental risk and other adverse events can produce diminished political efficacy and even apathy as a coping mechanism (Desmond and Travis 2018; Eliasoph 1997). While our findings do show some weaker associations between partisanship and climate attitudes among non-White respondents—as seen in past research (Schuldt and Pearson 2016)—they also show that the climate belief networks of Black and Hispanic respondents are equally—if not more—integrated than White respondents; they are simply integrated around different salient attitudes. Therefore, these results help to challenge theories that conceptualize racialized differences in climate attitudes as deficits (Cabrera 2019; Suldovsky 2017; Valencia 2002). Instead, they support environmental subjectivities, a concept that considers the interpretive frames and situated knowledges of climate issues (Martin and Desmond 2010) that emerge in relation to differences in cultural worldviews, race, class, and overall racialized experiences in the context of the United States (Bhardwaj 2023; Ford and Norgaard 2020). Climate and environmental attitudes are two sticks—within the bundle of sticks of race—that deserve additional attention in research.

Second, our study also contributes to emerging work on racialized variation in public opinion more generally. Across many fields, research is moving to more deeply evaluate how core aspects of public opinion like political ideology (Jefferson 2020), concern in the face of social problems (Grace and García 2024), and beliefs and practices (Wilde and Glassman 2016; Yukich and Edgell 2020) work differently across different racialized contexts. Recent research comparing climate opinions across race and ethnicity shows that predictors of climate attitudes can vary across groups. Once-blanket predictors of climate attitudes such as partisanship, ideology, education, age, gender, and income manifest in different ways for different racial and ethnic groups (Benegal et al. 2022). Our use of BNA approach provides a more holistic methodology that can help to measure additional racial and ethnic variation on other issues in the future.

Third, our findings provide an important contribution to the study of racialized emotions in public opinion. Emotions have varying potency among racial groups and motivate political will and action that shapes public opinion (Valentino et al. 2011; Valentino, Gregorowicz, and Groenendyk 2009). Still, political will and action is conditioned on how well emotions engender a shared sense of community, group consciousness, and legitimacy that actively engages with the greater public to address social issues (Teeger 2023). Research shows that anger, for example, has a stronger effect on aggregate public opinion among White respondents compared with Black respondents. Our study shows how this emotional valence is not uniform along racial and ethnic lines for addressing the climate crisis, as results showed that group consciousness better predicted BIPOCs’ environmental attitudes relative to Whites’ (Li 2021). Our study could only measure worry due to data limitations, though future research should further disaggregate which kinds of emotions (e.g., anger, happiness, joy, etc.) structure group consciousness among BIPOCs’ climate belief networks. It is possible, for example, that the elevated salience of the perceived harm measures indicates a different kind of concern or fear at play, and specific survey questions can evaluate this in future research.

Conclusion

Every study also has limitations. One advantage of the CCAM data in our study is that the aggregation of multiple annual cross-sectional survey samples allowed us to mitigate typical concerns about sample size when surveying BIPOC populations. However, the public release of the aggregated data we used included coarse and rather limited measures of racial and ethnic identification. Future work using oversamples of specific groups, including Asian Americans and indigenous populations, can further extend and validate the results presented here, especially by considering additional measures of ethnicity including languages spoken and country of origin. Additional qualitative interviews, focus groups, and ethnographic work can also help to explain the variation in these descriptive patterns identified here, along with work in global, comparative contexts outside the United States. Much more remains to be understood about the cultural conditions of racialization, racialized institutions, and community mobilization that create and change the belief networks of respondents (Ray 2019).

In addition, work should be careful not to conflate attitudes about the environment with attitudes about the climate. The belief network measures we used focused on climate attitudes to specifically evaluate whether variation in public opinion might map onto climate inequality. In contrast, much of the early work highlighted in the literature review focused on different environmental attitudes, such as concern about local air pollution, littering, toxic waste disposal, and water pollution, among others. Although attitudes about the environment and climate may be very similar and related at times, we suggest future research to disaggregate these concepts from one another.

But in the face of stark racial inequality, in the United States and abroad, we know that the worst effects of the climate crisis will not be evenly distributed. Our work contributes to studies of the impact of the climate crisis on BIPOC communities by highlighting differences in the formation of the interconnected climate attitude networks among different racial and ethnic groups. Because not all communities are affected by the climate crisis equally, understanding these differences can guide researchers to expand the study of BIPOC climate perceptions and lead to more effective support for communities most at risk to climate change. Understanding the extent to which generalized theories about climate attitudes are based on mostly White perceptions can lead to more inclusive communication strategies for change and just climate policies in the future.

Footnotes

ORCID iDs

Evan Stewart

Katsyris Rivera-Kientz

Notes

Author Biographies

Evan Stewart is an assistant professor of sociology at University of Massachusetts Boston. His research on belonging in public life, religious change, and public opinion has appeared in American Sociological Review, Social Forces, Social Problems, and Sociological Theory, among other outlets. He earned a PhD in sociology from the University of Minnesota and has held research fellowships with the Public Religion Research Institute and the Social Science Research Council.

Katsyris Rivera-Kientz is a doctoral candidate in sociology at the University of Massachusetts Boston. Her research examines climate change attitudes and behaviors, particularly how different communities understand and act on the current climate crisis. She earned an MS in Critical Ethnic and Community Studies from the University of Massachusetts and worked as a research fellow with the Sustainable Solutions Lab.

Timothy Dacey is an assistant professor of sociology at SUNY-Oswego. His research examines the social side of the climate crisis, specifically how climate related disasters shape public opinion towards climate change. He earned his PhD in sociology from the University of Massachusetts, Boston and has worked as a research fellow at the Center for Survey Research at UMass Boston.

References

Ard

Kerry

. 2015. “Trends in Exposure to Industrial Air Toxins for Different Racial and Socioeconomic Groups: A Spatial and Temporal Examination of Environmental Inequality in the U.S. from 1995 to 2004.” Social Science Research 53:375–90. doi:10.1016/j.ssresearch.2015.06.019.

Bail

Christopher A.

Argyle

Lisa P.

Brown

Taylor W.

Bumpus

John P.

Chen

Haohan

Hunzaker

M. B. Fallin

Lee

Jaemin

Mann

Marcus

Merhout

Friedolin

Volfovsky

Alexander

. 2018. “Exposure to Opposing Views on Social Media Can Increase Political Polarization.” Proceedings of the National Academy of Sciences 115(37):9216–21. doi:10.1073/pnas.1804840115.

Baldassarri

Delia

Gelman

Andrew

. 2008. “Partisans without Constraint: Political Polarization and Trends in American Public Opinion.” American Journal of Sociology 114(2):408–46. doi:10.2139/ssrn.1010098.

Ballew

Matthew T.

Leiserowitz

Anthony

Roser-Renouf

Connie

Rosenthal

Seth A.

Kotcher

John E.

Marlon

Jennifer R.

Lyon

Erik

Goldberg

Matthew H.

Maibach

Edward W.

2019. “Climate Change in the American Mind: Data, Tools, and Trends.” Environment: Science and Policy for Sustainable Development 61(3):4–18. doi:10.1080/00139157.2019.1589300.

Bell

Michelle L.

Ebisu

Keita

. 2012. “Environmental Inequality in Exposures to Airborne Particulate Matter Components in the United States.” Environmental Health Perspectives 120(12):1699–704. doi:10.1289/ehp.1205201.

Benegal

Salil D.

2018. “The Spillover of Race and Racial Attitudes into Public Opinion about Climate Change.” Environmental Politics 27(4):733–56. doi:10.1080/09644016.2018.1457287.

Benegal

Salil D.

Azevedo

Flávio

Holman

Mirya R.

2022. “Race, Ethnicity, and Support for Climate Policy.” Environmental Research Letters 17(11):114060. doi:10.1088/1748-9326/aca0ac.

Bhardwaj

Ankit

. 2023. “The Soils of Black Folk: W.E.B. Du Bois’s Theories of Environmental Racialization.” Sociological Theory 41(2):105–28. doi:10.1177/07352751231164999.

Bisgaard

Martin

. 2015. “Bias Will Find a Way: Economic Perceptions, Attributions of Blame, and Partisan-motivated Reasoning during Crisis.” The Journal of Politics 77(3):849–60. doi:10.1086/681591.

10.

Blaikie

Piers

Cannon

Terry

Davis

Ian

Wisner

Ben

. 2014. At Risk. New York: Routledge.

11.

Bolin

Jessica L.

Hamilton

Lawrence C.

2018. “The News You Choose: News Media Preferences Amplify Views on Climate Change.” Environmental Politics 27(3):455–76. doi:10.1080/09644016.2018.1423909.

12.

Bolsen

Toby

Druckman

James N.

Cook

Fay Lomax

. 2014. “The Influence of Partisan Motivated Reasoning on Public Opinion.” Political Behavior 36(2):235–62. doi:10.1007/s11109-013-9238-0.

13.

Boutyline

Andrei

Vaisey

Stephen

. 2017. “Belief Network Analysis: A Relational Approach to Understanding the Structure of Attitudes.” American Journal of Sociology 122(5):1371–447. doi:10.1086/691274.

14.

Brandt

Mark J.

Sleegers

Willem W. A

. 2021. “Evaluating Belief System Networks as a Theory of Political Belief System Dynamics.” Personality and Social Psychology Review 25(2):159–85. doi:10.1177/1088868321993751.

15.

Brekhus

Wayne H.

Brunsma

David L.

Platts

Todd

Dua

Priya

. 2010. “On the Contributions of Cognitive Sociology to the Sociological Study of Race.” Sociology Compass 4(1):61–76. doi:10.1111/j.1751-9020.2009.00259.x.

16.

Burger

Joanna

Myers

Boring

C. S.

Dixon

Lord

Ramos

Shukla

Gochfeld

2004. “Perceptions of General Environmental Problems, Willingness to Expend Federal Funds on These Problems, and Concerns Regarding the Los Alamos National Laboratory: Hispanics Are More Concerned than Whites.” Environmental Research 95:174–83.

17.

Cabrera

Nolan L.

2019. “Critical Race Theory V. Deficit Models.” Equity & Excellence in Education 52(1):47–54. doi:10.1080/10665684.2019.1630342.

18.

Caron

Anne C.

1989. “Environmental Perspectives of Blacks: Acceptance of the ‘New Environmental Paradigm. .’” The Journal of Environmental Education 20(3):21–26.

19.

Carrillo

Ian

. 2021. “The Racial Fix and Environmental State Formation.” Current Sociology 69(5):641–59. doi:10.1177/0011392120913099.

20.

Collins

Timothy W.

2010. “Marginalization, Facilitation, and the Production of Unequal Risk: The 2006 Paso Del Norte Floods.” Antipode 42(2):258–88. doi:10.1111/j.1467-8330.2009.00755.x.

21.

Crowder

Kyle

Downey

Liam

. 2010. “Interneighborhood Migration, Race, and Environmental Hazards: Modeling Microlevel Processes of Environmental Inequality.” American Journal of Sociology 115(4):1110–49. doi:10.1086/649576.

22.

Dacey

Timothy

Stewart

Evan

. 2023. “The Green Elephants in the Room: Perceived Environmental Harm and Support for Regulation Among Republicans.” Sociological Inquiry 93(2):250–72. doi:10.1111/soin.12511.

23.

DellaPosta

Daniel

. 2020. “Pluralistic Collapse: The ‘Oil Spill’ Model of Mass Opinion Polarization.” American Sociological Review 85(3):507–36. doi:10.1177/0003122420922989.

24.

DellaPosta

Daniel

Shi

Yongren

Macy

Michael

. 2015. “Why Do Liberals Drink Lattes?” American Journal of Sociology 120(5):1473–511. doi:10.1086/681254.

25.

Desmond

Matthew

Travis

Adam

. 2018. “Political Consequences of Survival Strategies among the Urban Poor.” American Sociological Review 83(5):869–96. doi:10.1177/0003122418792836.

26.

Dietz

Thomas

Dan

Amy

Shwom

Rachael

. 2007. “Support for Climate Change Policy: Social Psychological and Social Structural Influences.” Rural Sociology 72(2):185–214. doi:10.1526/003601107781170026.

27.

DiMaggio

Paul

Evans

John

Bryson

Bethany

. 1996. “Have American’s Social Attitudes Become More Polarized?.” American Journal of Sociology 102(3):690–755.

28.

Dunlap

Catton

Jr.

1994. “Struggling with Human Exemptionalism: The Rise, Decline, and Revitalization of Environmental Sociology.” The American Sociologist 25:5–30.

29.

Eliasoph

Nina

. 1997. “‘Close to Home’: The Work of Avoiding Politics.” Theory and Society 26(5):605–47.

30.

Epskamp

Sacha

Borsboom

Denny

Fried

Eiko I.

2018. “Estimating Psychological Networks and Their Accuracy: A Tutorial Paper.” Behavior Research Methods 50(1):195–212. doi:10.3758/s13428-017-0862-1.

31.

Fishman

Nic

Davis

Nicholas T.

2022. “Change We Can Believe In: Structural and Content Dynamics within Belief Networks.” American Journal of Political Science 66(3):648–63. doi:10.1111/ajps.12626.

32.

Ford

Allison

Norgaard

Kari Marie

. 2020. “Whose Everyday Climate Cultures? Environmental Subjectivities and Invisibility in Climate Change Discourse.” Climatic Change 163(1):43–62. doi:10.1007/s10584-019-02632-1.

33.

Goldberg

Matthew H.

Gustafson

Abel

Ballew

Matthew T.

Rosenthal

Seth A.

Cutler

Matthew J.

Leiserowitz

Anthony

. 2020. “Predictors of Global Warming Risk Perceptions among Latino and Non-Latino White Americans.” Climatic Change 162(3):1555–74. doi:10.1007/s10584-020-02728-z.

34.

Goldberg

Matthew H.

Gustafson

Abel

Ballew

Matthew T.

Rosenthal

Seth A.

Leiserowitz

Anthony

. 2021. “Identifying the Most Important Predictors of Support for Climate Policy in the United States.” Behavioural Public Policy 5(4):480–502. doi:10.1017/bpp.2020.39.

35.

Grace

Matthew K.

García

Ashley M.

2024. “Racial-ethnic Differences in Anticipatory Stress about COVID-19 Mortality: An Evaluation of Multiple Mechanisms.” Sociology of Race and Ethnicity 10(1):120–38. doi:10.1177/23326492231184837.

36.

Graham

Jesse

Haidt

Jonathan

Nosek

Brian A.

2009. “Liberals and Conservatives Rely on Different Sets of Moral Foundations.” Journal of Personality and Social Psychology 96(5):1029–46. doi:10.1037/a0015141.

37.

Grineski

Sara E.

Collins

Timothy W.

Chakraborty

Jayajit

Montgomery

Marilyn

. 2015. “Hazardous Air Pollutants and Flooding: A Comparative Interurban Study of Environmental Injustice.” GeoJournal 8(1):145–58.

38.

Grineski

Sara E.

Collins

Timothy W.

Ford

Paula

Fitzgerald

Rosa

Aldouri

Raed

Velázquez-Angulo

Gilberto

Aguilar

Maria De Lourdes Romo

Duanjun

. 2012. “Climate Change and Environmental Injustice in a Bi-national Context.” Applied Geography 33:25–35. doi:10.1016/j.apgeog.2011.05.013.

39.

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. doi:10.1177/0093650211416646.

40.

Hegtvedt

Karen A.

Parris

Christie L.

Johnson

Cathryn

. 2019. “Framing and Feeling Fuel Environmentally Responsible Behaviors of Black Residents in the United States.” Sociological Perspectives 62(5): 603–26. doi:10.1177/0731121419852946.

41.

Hershey

Marjorie Randon

Hill

David B.

1978. “Is Pollution ‘A White Thing’? Racial Differences in Preadults’ Attitudes.” Public Opinion Quarterly 41:439–58.

42.

Hohm

Charles F.

1976. “A Human-ecological Approach to the Reality and Perception of Air Pollution: The Los Angeles Case.” The Pacific Sociological Review 19(1):21–44. doi:10.2307/1388740.

43.

Itzigsohn

José

Brown

Karida

. 2020. The Sociology of W.E.B. Du Bois: Racialized Modernity and the Global Color Line. New York: New York University Press.

44.

Iyengar

Shanto

Westwood

Sean J.

2015. “Fear and Loathing across Party Lines: New Evidence on Group Polarization.” American Journal of Political Science 59(3):690–707. doi:10.1111/ajps.12152.

45.

Jefferson

Hakeem

. 2020. “The Curious Case of Black Conservatives: Construct Validity and the 7-point Liberal-conservative Scale.” Retrieved May 27, 2024 ((https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3602209).

46.

Kahan

Dan M.

2007. “The Cognitively Illiberal State.” Stanford Law Review 60(1):115–54.

47.

Kellert

Stephen R.

1984. “Urban American Perceptions of Animals and the Natural Environment.” Urban Ecology 8(3):209–28. doi:10.1016/0304-4009 (84)90036-6.

48.

Kiley

Kevin

Vaisey

Stephen

. 2020. “Measuring Stability and Change in Personal Culture Using Panel Data.” American Sociological Review 85(3):477–506. doi:10.1177/0003122420921538.

49.

Knight

Kyle W.

2016. “Public Awareness and Perception of Climate Change: A Quantitative Cross-national Study.” Environmental Sociology 2(1):101–13. doi:10.1080/23251042.2015.1128055.

50.

Kreger

Janet

. 1973. “Ecology and Black Student Opinion.” The Journal of Environmental Education 4(3):30–34. doi:10.1080/00958964.1973.10801741.

51.

Kyselá

Eva

Tvinnereim

Endre

Ivarsflaten

Elisabeth

. 2019. “Attitudes to Public Spending on Environmental Risk Reduction: The Role of Temporal and Spatial Distance.” Environmental Sociology 5(4):362–73. doi:10.1080/23251042.2019.1643528.

52.

Lakoff

George

. 2002. Moral Politics: How Liberals and Conservatives Think. Chicago, IL: University of Chicago Press.

53.

Lazri

Adam Mcbride

Konisky

David M.

2019. “Environmental Attitudes across Race and Ethnicity.” Social Science Quarterly 100(4): 1039–55. doi:10.1111/ssqu.12626.

54.

Lee

Sanguk

Goldberg

Matthew H.

Rosenthal

Seth A.

Maibach

Edward W.

Kotcher

John E.

Leiserowitz

Anthony

. 2024. “Climate Change Belief Systems across Political Groups in the United States.” PLOS ONE 19(3):e0300048. doi:10.1371/journal.pone.0300048.

55.

Leiserowitz

Anthony A.

2005. “American Risk Perceptions: Is Climate Change Dangerous?.” Risk Analysis: An International Journal 25(6):1433–42. doi:10.1111/j.1540-6261.2005.00690.x.

56.

Lenz

Gabriel S.

2012. Follow the Leader? How Voters Respond to Politicians’ Policies and Performance. Illustrated ed. Chicago, IL: University of Chicago Press.

57.

Zhengyan

. 2021. “A New Look at the Racial Differences in Environmental Attitudes: The Roles of Risk Perception and Group Consciousness.” Journal of Environmental Management 299:1–10. doi:10.1016/j.jenvman.2021.113616.

58.

Zhengyan

Konisky

David M.

Zirogiannis

Nikolaos

. 2019. “Racial, Ethnic, and Income Disparities in Air Pollution: A Study of Excess Emissions in Texas.” PLOS ONE 14(8):e0220696. doi:10.1371/journal.pone.0220696.

59.

Maldonado

Collins

T. W.

Grineski

S. E.

Chakraborty

2016. “Exposure to Flood Hazards in Miami and Houston: Are Hispanic Immigrants at Greater Risk Than Other Social Groups?” International Journal of Environmental Research and Public Health 13(8):775. doi:10.3390/ijerph13080775.

60.

Martin

John Levi

Desmond

Matthew

. 2010. “Political Position and Social Knowledge.” Sociological Forum 25(1):1–26. doi:10.1111/j.1573-7861.2009.01154.x.

61.

McCright

Aaron M.

2011. “Political Orientation Moderates Americans’ Beliefs and Concern about Climate Change.” Climatic Change 104(2):243–53.

62.

McCright

Aaron M.

Dunlap

Riley E.

2008. “Social Movement Identity and Belief Systems An Examination of Beliefs about Environmental Problems within the American Public.” Public Opinion Quarterly 72(4):651–76. doi:10.1093/poq/nfn051.

63.

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.” The Sociological Quarterly 52(2): 155–94.

64.

McCright

Aaron M.

Marquart-Pyatt

Sandra T.

Shwom

Rachael L.

Brechin

Steven R.

Allen

Summer

. 2016. “Ideology, Capitalism, and Climate: Explaining Public Views about Climate Change in the United States.” Energy Research & Social Science 21:180–89. doi:10.1016/j.erss.2016.08.003.

65.

Mohai

Paul

. 1990. “Black Environmentalism.” The Social Science Quarterly 71(4):744–65.

66.

Mohai

Paul

Bryant

Bunyan

. 1998. “Is There a Race Effect on Concern for Environmental Quality?” The Public Opinion Quarterly 62(4):475–505.

67.

Mohai

Paul

Lantz

Paula M.

Morenoff

Jeffrey

House

James S.

Mero

Richard P.

2009. “Racial and Socioeconomic Disparities in Residential Proximity to Polluting Industrial Facilities: Evidence From the Americans’ Changing Lives Study.” American Journal of Public Health 99(Suppl. 3):S649–S656. doi:10.2105/AJPH.2007.131383.

68.

O’Brien

Daniel T.

Gridley

Brian

Trlica

Andrew

Wang

Jonathan A.

Shrivastava

Aatmesh

. 2020. “Urban Heat Islets: Street Segments, Land Surface Temperatures, and Medical Emergencies during Heat Advisories.” American Journal of Public Health 110(7):994–1001. doi:10.2105/AJPH.2020.305636.

69.

Parker

Julia Dawn

McDonough

Maureen H.

1999. “Environmentalism of African Americans: An Analysis of the Subculture and Barriers Theories.” Environment and Behavior 31(2):155–77. doi:10.1177/00139169921972047.

70.

Pearson

Adam R.

Bacio

Guadalupe A.

Naiman

Sarah

Romero-Canyas

Rainer

Schuldt

Jonathon P.

2021. “Cultural Determinants of Climate Change Opinion: Familism Predicts Climate Beliefs and Policy Support among US Latinos.” Climatic Change 167(11):1–8. doi:10.1007/s10584-021-03165-2.

71.

Phoenix

Davin L.

2019. The Anger Gap: How Race Shapes Emotion in Politics. Illustrated ed. Cambridge, England: Cambridge University Press.

72.

Quintana

Rafael

. 2023. “From Single Attitudes to Belief Systems: Examining the Centrality of STEM Attitudes Using Belief Network Analysis.” International Journal of Educational Research 119: 102179. doi:10.1016/j.ijer.2023.102179.

73.

Ray

Victor

. 2019. “A Theory of Racialized Organizations.” American Sociological Review 84(1):26–53. doi:10.1177/0003122418822335.

74.

Schuldt

Jonathon P.

Pearson

Adam R.

2016. “The Role of Race and Ethnicity in Climate Change Polarization: Evidence from a U.S. National Survey Experiment.” Climatic Change 136(3–4):495–505. doi:10.1007/s10584-016-1631-3.

75.

Sen

Maya

Wasow

Omar

. 2016. “Race as a Bundle of Sticks: Designs that Estimate Effects of Seemingly Immutable Characteristics.” Annual Review of Political Science 19(1):499–522. doi:10.1146/annurev-polisci-032015-010015.

76.

Suldovsky

Brianne

. 2017. “The Information Deficit Model and Climate Change Communication.” In Hans von Storch, eds. Oxford Research Encyclopedia of Climate Science. New York: Oxford University Press.

77.

Taylor

Dorceta E.

2014. Toxic Communities: Environmental Racism, Industrial Pollution, and Residential Mobility. New York: New York University Press.

78.

Teeger

Chana

. 2023. “(Not) Feeling the Past: Boredom as a Racialized Emotion.” American Journal of Sociology 129(1):1–40. doi:10.1086/725803.

79.

Theodore Washington. 1976. Urban Blacks and Wildlife. Fort Collins, CO: Colorado State University.

80.

Uyeki

Eugene S.

Holland

Lani J.

2000. “Diffusion of Pro-environment Attitudes?” The American Behavioral Scientist 43(4):646–62.

81.

Valencia

Richard R.

2002. “‘Mexican Americans Don’t Value Education!’ On the Basis of the Myth, Mythmaking, and Debunking.” Journal of Latinos and Education 1(2):81–103. doi:10.1207/S1532771XJLE0102_2.

82.

Valentino

Nicholas A.

Brader

Ted

Groenendyk

Eric W.

Gregorowicz

Krysha

Hutchings

Vincent L.

2011. “Election Night’s Alright for Fighting: The Role of Emotions in Political Participation.” The Journal of Politics 73(1):156–70. doi:10.1017/S0022381610000939.

83.

Valentino

Nicholas A.

Gregorowicz

Krysha

Groenendyk

Eric W.

2009. “Efficacy, Emotions and the Habit of Participation.” Political Behavior 31(3):307–30.

84.

Wilde

Melissa

Glassman

Lindsay

. 2016. “How Complex Religion Can Improve Our Understanding of American Politics.” Annual Review of Sociology 42(1):407–25. doi:10.1146/annurev-soc-081715-074420.

85.

Yale Program on Climate Change Communication (YPCCC) and George Mason University Center for Climate Change Communication (Mason 4C). (2020). “Climate Change in the American Mind: National Survey Data on Public Opinion (2008-2018)” [Data file and codebook]. doi:10.17605/OSF.IO/JW79P.

86.

Yukich

Grace

Penny

Edgell

, eds. 2020. Religion Is Raced: Understanding American Religion in the Twenty-first Century. New York: New York University Press.

87.

Zahran

Sammy

Brody

Samuel D.

Peacock

Walter Gillis

Vedlitz

Arnold

Grover

Himanshu

. 2008. “Social Vulnerability and the Natural and Built Environment: A Model of Flood Casualties in Texas.” Disasters 32(4):537–60. doi:10.1111/j.1467-7717.2008.01054.x.