Rethinking Economic Conditions and Environmental Attitudes: Macroeconomic Effects,Individual Experiences,and Subjectivity

Abstract

The global environmental concern literature has focused on economic development as a contextual predictor of environmental attitudes with conflicting empirical results. Some studies, informed by postmaterialist theory or the affluence hypothesis, find that national affluence increases environmental concern, whereas others find the opposite. In this article, we qualify the economy and environmental attitude relationship by arguing that it has several dimensions: long-term macroeconomic history, short-term economic growth, individuals’ recent economic experiences, individuals’ subjective understandings of the national economy, and their personal economic circumstances. This holistic framework is operationalized with multilevel models and data from the Life in Transition II project, using outcomes for climate change attitudes. We find that recent economic hardships at the individual level (e.g., a job loss) have a positive to null effect. More subjectively, respondents who believe that the recent global recession had a negative effect on their household are more concerned with climate change. At the country level, long-run economic development does little to explain cross-national differences in climate change attitudes whereas short-run, relative economic growth increases both climate change concern and willingness to pay for climate policy. More broadly, future research should de-emphasize long-run, macroeconomic conditions as determinants of environmental attitudes.

Keywords

environment and technology climate change economy multilevel model

Introduction

As the world faces complex and increasingly dire environmental problems like climate change, it is important to understand what factors explain cross-national differences in environmental attitudes, policy support, or behaviors. Much of the scholarly debate has concerned the effect of national affluence with three perspectives dominating the “global environmental concern” literature. Postmaterialism predicts that environmental concern will be higher in wealthier countries because as incomes rise and basic, subsistence needs are met, people become more focused on “postmaterial” issues like human rights, freedom of speech, and environmental quality (Inglehart 1977, 1995). The “affluence hypothesis” relies on orthodox economic theory to similarly conclude that wealthier people care more about the environment (Franzen 2003; Franzen and Vogl 2013a, 2013b). Finally, “global environmentalism” holds that national income is not predictive of environmental attitudes (Brechin 1999; Brechin and Kempton 1994; Dunlap and York 2008).

Despite their theoretical differences, both the postmaterialist and affluence perspectives share the common assumption that individual actors respond to economic change in a linear, deterministic fashion. However, individuals often have limited understanding of the actual state of the economy (e.g., Blaacker, Woods, and Oliver 2012; Blendon et al. 1997). All three perspectives typically emphasize long-run, absolute differences in economic development. Yet short-term, relative economic change, such as the recent global recession, also affects concern for environmental issues like climate change (Brulle, Carmichael, and Jenkins 2012; Kahn and Kotchen 2010; Scruggs and Benegal 2012; Shum 2012). These latter studies use data aggregated at some contextual unit (e.g., U.S. state, country), leading to possible ecological inference problems. This leaves significant gaps in the literature: we know relatively little about how an actor’s subjective understanding of the macroeconomy, one’s recent economic experiences, and short-term macroeconomic growth affect environmental concern.

The purpose of this article is to offer a novel, integrated understanding of how economic conditions affect views on environmental issues—specifically climate change. We argue that this relationship has four dimensions: (1) long-run macroeconomic conditions, (2) abrupt, relative short-term economic change, (3) recent personal or household economic experiences, and (4) subjective impressions of the state of the macroeconomy or personal economic situation. This integrated perspective is operationalized with multinational survey data from the Life in Transition II project (ECB 2013) and multilevel modeling.

By combining individual-level economic experience variables with national-level predictors, we advance global environmental concern research (e.g., Dunlap and York 2008; Franzen 2003) and also contribute to the literature on cross-national climate change attitudes (e.g., Mostafa 2016; Sandvik 2008) while improving on prior operationalizations of postmaterialist theory and addressing the ecological inference problems in the literature. We also contribute to the growing body of work on environmental concern in post-Communist states (e.g., Chaisty and Whitefield 2015; Marquart-Pyatt 2012a). In the section that follows, we describe an integrated, multidimensional conceptualization of the “economy” and how it might affect climate change views.

Theoretical Background

Postmaterialism, Affluence, and Global Environmentalism

Postmaterialist theory originated with Inglehart (1977), who argued that growth in material prosperity above the subsistence level reorients values around “postmaterialist” concerns like human rights, democracy, or the environment. The shift toward postmaterialism does not occur instantaneously; rather, postmaterialist values are the result of long-run economic growth and cohort change coupled with rising individual-level economic security (Inglehart and Abramson 1994). Later, the theory was amended with to include an “objective problems” hypothesis (Inglehart 1995), which argues that environmental problems are likely to increase environmental concern independent of the effect of long-term economic development. Critics of postmaterialist theory charge that the “objective problems” amendment was added in a post hoc manner to address empirical deficiencies with the theory’s original formulation (Dunlap and York 2008).

A similar perspective is called the “affluence hypothesis.” Informed by orthodox economic theory, the affluence hypothesis maintains that growth in national or personal income shifts demand for environmental quality upward (Franzen 2003; Franzen and Vogl 2013a, 2013b). Although the theoretical specifics vary, the implications of both the affluence hypothesis and postmaterialist theory are clear—wealthier populations care more about environmental issues and are more supportive of environmental policy. Studies have shown that economic development (e.g., gross domestic product [GDP] per capita) is positively associated with support for environmental policy (Çarkoğlu and Kentmen-Çin 2015; Gelissen 2007), willingness to sacrifice for the environment (Dorsch 2014; Franzen 2003; Franzen and Vogl 2013a; Harring 2013; Kemmelmeier, Krol, and Kim 2002; Lo 2015), environmental behaviors (Freymeyer and Johnson 2010), and more general environmental concern (Diekmann and Franzen 1999; Dorsch 2014; Franzen 2003).

These perspectives are in opposition to “global environmentalism,” which holds that environmental concern is a global phenomenon poorly explained by economic development. Lower income nations often serve as pollution havens and sinks for ecological harms (Hornborg 1998; Jorgenson 2006; Rice 2007a, 2007b) and poorer populations are often more vulnerable to global environmental problems, like climate change (O’Brien et al. 2004). Several studies have concluded that national affluence has either a null or negative association with environmental concern (Brechin 1999; Brechin and Kempton 1994; Dunlap and York 2008; Fairbrother 2013; Givens and Jorgenson 2011, 2013; Ignatow 2006; Jorgenson and Givens 2014; Knight and Messer 2012; Lo 2015; Marquart-Pyatt 2012b), willingness to make personal sacrifices such as increased taxes (Dunlap and Mertig 1997; Dunlap and York 2008; Fairbrother 2013; Knight and Messer 2012; Marquart-Pyatt 2012b; Mostafa 2012, 2013; Running 2015), or environmental skepticism (Zhou 2015). More specific to climate change attitudes, economic development negatively correlates with public concern and willingness to make behavioral changes to address climate change (Kvaløy, Finseraas, and Listhaug 2012; Mostafa 2016; Sandvik 2008) and climate change risk perceptions (Lee et al. 2015; Tjernström and Tietenberg 2008), though Lo and Chow (2015) reported that national income increases the perceived importance of climate change yet reduces perceived risks.

Thus, despite stark theoretical differences, a review of the literature suggests support for multiple perspectives. Some of these differences may be related to measurement of the dependent variable. Dunlap and Jones (2002) defined environmental concern as “the degree to which people are aware of problems regarding the environment and support efforts to solve them and/or willingness to contribute personally to their solution” (p. 485). Hence, environmental concern is multidimensional (Dunlap and Jones 2002; Fransson and Gärling 1999; Xiao and Dunlap 2007). Some scholars (e.g., Brechin 1999; Brechin and Kempton 1994) argue that general concern for either local or global environmental problems should be preferred to willingness to pay/sacrifice measures whereas others prefer the latter (e.g., Franzen and Vogl 2013a) and many examine both dimensions (e.g., Dorsch 2014; Marquart-Pyatt 2008, 2012b). Still others consider specific issues such as climate change (e.g., Lo and Chow 2015; Sandvik 2008; the present article). Similar to prior analyses (e.g., Dorsch 2014; Marquart-Pyatt 2012b), we adopt a multidimensional approach including both willingness to pay and concern and test the following hypotheses regarding the effect of long-run economic development on climate change views:

Hypothesis 1: As predicted by postmaterialist theory and the affluence hypothesis, long-term economic development (i.e., GDP or gross national income [GNI]) increases climate change concern (Hypothesis 1a) and willingness to sacrifice for climate change (Hypothesis 1b).

Hypothesis 2: Following the objective problems dimension of postmaterialist theory, exposure to climate hazards will increase climate change concern (Hypothesis 2a) and willingness to sacrifice (Hypothesis 2b).

Typically, the postmaterialist theory is operationalized using GDP per capita in a single year. This approach is not the most valid operationalization as postmaterialist values are theorized to arise from long-run economic evolution and cohort change. In some countries, GDP might increase rapidly for a few years, which is unlikely to give rise to a significant values shift. Other countries may experience deep recessions in the short term, which is unlikely to cause a sudden shift away from postmaterialism. Below we describe how we improve up prior operationalizations via a time-lagged indicator of long-run economic development.

Short-term, relative economic effects

Postmaterialism focuses on long-run shifts in values but the affluence hypothesis does not have a specified temporal dimension. Franzen and Meyer (2010) and Franzen and Vogl (2013a, 2013b) did not explain how quickly economic growth will positively influence environmental attitudes though Franzen (2003) argued that the affluence hypothesis is “longitudinal” without specifying a time frame. Givens and Jorgenson (2011) and Gelissen (2007) both used average percent GDP growth over several years to predict environmental concern and find a positive relationship. Overall, the global environmental concern literature has less to say about the effects of short-term and relative economic change.

Other research, focused on the 2007–2008 global economic recession, suggests that climate change views are sensitive to short-run economic fluctuations, with several studies reporting a negative association between economic slowdowns (e.g., increases in unemployment, loss in GDP) and concern for climate change using samples from countries and regions of the United States (Brulle et al. 2012; Kahn and Kotchen 2010; Scruggs and Benegal 2012; Shum 2012). Shao et al. (2014) provided a lone exception; they report that county-level unemployment does not affect climate change attitudes in the United States.

However, the literature cited in the last paragraphs does not allow for inferences into individual reactions to economic change because these studies use opinion and economic data aggregated at the national, state, or county level. Hence, these types of analyses only allow for ecological inferences and the results may be an artifact of data aggregation. Indeed, Inglehart and Abramson (1994) theorized that individual-level economic prosperity gives rise to postmaterialism—yet this individual-level mechanism has received little attention. Notably, Shao et al. (2014) used individual-level data for economic experiences, and report null findings. Because of these ecological inference problems, it is unclear if individuals actually experiencing the worst effects of recession (e.g., job loss, wage reductions, consumption cutbacks) will have lower levels of concern for climate change or less support for climate policy. The following hypotheses are tested in this article:

Hypothesis 3: Short-term, relative economic growth (e.g., percent growth in GNI) increases climate change concern (Hypothesis 3a) and willingness to sacrifice for climate change (Hypothesis 3b).

Hypothesis 4: Personal and household economic hardship (e.g., job loss) will reduce climate change concern (Hypothesis 4a) and willingness to sacrifice (Hypothesis 4b).

Bringing subjectivity in

Granovetter (1985) critiqued both sociology and economics by arguing that both disciplines adopt an unrealistic notion of the actor. For sociology, actors are often portrayed as oversocialized victims of structural forces and largely devoid of agency, whereas much of economics relies on an undersocialized, highly individuated rational actor. The affluence and postmaterialist perspectives assume an oversocialized yet rational actor whose individual behavior will be affected in a deterministic way by macroeconomic change, that is, actors’ dispositions are assumed to be strongly determined by macroeconomic phenomenon and, simultaneously, actors are assumed to respond in an individuated, rational way to the phenomenon. Borrowing from Granovetter (1985), individuals are better conceptualized as embedded in an economic system but with their own agency, subjectivity, cultural worldviews and ideological biases. As such, economic conditions should be seen as simultaneously experiential and subjective—the subjective meaning ascribed to macroeconomic change might vary from person to person or across social groups. Given this problem of embeddedness, it is imperative to consider how individuals subjectively understand economic conditions.

People often hold inaccurate views about the macroeconomy. For instance, the U.S. public overestimates both the degree of inflation and unemployment (Blendon et al. 1997) cannot accurately describe the amount of income inequality (Chambers, Swan, and Heesacker 2014; Norton and Ariely 2011) and residents of Appalachia exaggerate the regional economic benefits provided by the coal industry (Blaacker et al. 2012). Thus, people are inherently subjective and do not respond to large-scale economic change in a deterministic or linear way. Furthermore, the perception of economic conditions does not necessarily align with actual conditions. The implications for environmental concern are as follows: what people think about the economy or their economic circumstances may be as important as the actual state of the economy. To explore this area, we consider the following hypothesis:

Hypothesis 5: Satisfaction with macroeconomic or personal economic circumstances will increase climate change concern (Hypothesis 5a) and willingness to sacrifice (Hypothesis 5b).

Bringing it all together

The primary purpose of this article is to integrate the literature reviewed above to develop a comprehensive understanding of how both long-run and short-term macroeconomic conditions, individual experiences, and subjective impressions of economic conditions affect both concern for climate change and willingness to pay to address climate change. We identified gaps in the literature. First, “global environmental concern” research de-emphasizes short-run economic change and does not grapple with how individuals subjectively understand and experience economic conditions. Rather, this literature, especially that which endorses postmaterialism or the affluence hypothesis, implicitly assumes that macroeconomic conditions exert a predictable and uniform effect on all residents of a particular nation (e.g., Franzen and Meyer 2010). Research on recessions and climate change attitudes suffers from ecological inference problems (e.g., Brulle et al. 2012; Scruggs and Benegal 2012). In the next section, we describe how we use novel data and improved operationalizations to address these gaps and test the hypotheses.

Data, Measures, and Methods

Data

We use the Life in Transition Survey II (LITS II). LITS II was conducted in 2010 as a joint project by the European Bank for Reconstruction and Development and the World Bank. Approximately 35,000 respondents were interviewed in the following 35 countries: Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, France, Georgia, Germany, Great Britain, Hungary, Italy, Kazakhstan, Kosovo, Kyrgyzstan, Latvia, Lithuania, Macedonia, Moldova, Mongolia, Montenegro, Poland, Romania, Russia, Serbia, Slovakia, Slovenia, Sweden, Tajikistan, Turkey, Ukraine, and Uzbekistan.

Compared with more commonly used data sets (e.g., the World Values Survey), the LITS II data have several strengths and weaknesses. As noted earlier, some literature in this area suffers from ecological inference problems because of the lack of individual-level variables to capture economic experiences. A major strength of the data is that they contain variables that can solve this ecological inference problem—in addition to unique variables for subjective understandings of economic conditions. Hence, the LITS II data allow us to address gaps in the literature.

However, LITS II also has several limitations—the sample of countries is the most severe. These countries extend from Central Asia to Western Europe: no countries from several regions of the world—such as North America, South America, or Africa—are included in the data. We caution that the results below do not generalize to all regions and are most applicable to transition states—environmental attitudes in these nations have received relatively little attention in the literature (see Chaisty and Whitefield 2015; Hadler and Wahlkonig 2012; Marquart-Pyatt 2012a, for exceptions). A further limitation is that our dependent variables relate to climate change. LITS II does not include questions about more general environmental concern.

Outcome Variables

We use two outcome variables. The first outcome, Climate Concern, asked about concern for climate change (1 = no concern, 5 = high concern). Most of the variation in Climate Concern is within countries (standard deviation [SD] = 0.77) as opposed to between countries (SD = 0.19). The second outcome variable, Willingness to Pay, is constructed from a survey question that asked respondents if they would be willing to pay higher taxes to combat climate change (0 = no, 1 = yes); 65 percent of respondents answered no and most of the variation is within countries (SD = 0.45) as opposed to between countries (SD = 0.15).

Predictor Variables

Country-level predictors

GNI, gross national product, or GDP per capita in a single year is most often used to operationalize national affluence (Brechin 1999; Dunlap and Mertig 1997; Dunlap and York 2008; Freymeyer and Johnson 2010; Mostafa 2013, 2016). However, this is not an optimal operationalization, as postmaterialist theory argues that long-run economic development shifts values. To improve on this, we use a time-weighted measure of GNI per capita in thousands of dollars.¹ This variable, GNItime($000), was calculated by starting with 2010 and weighting previous years less and less; the first year included for any country is weighted at 0.10 and 2010 is weighted at 1.0.²

To capture short-term economic trends, we calculated the percent change in GNI per capita from 2009 to 2010 for each country. This variable, GNIchange, ranges widely from country to country. GNI per capita grew by 15 percent in Uzbekistan and Moldova, 10 percent in Kazakhstan, but fell 4 percent in Romania and 7 percent in Great Britain. When averaged across all countries in the sample, GNI per capita increased by 3.3 percent between 2009 and 2010.

Prior research has used several different measures of “objective problems” including air quality and water quality (Gelissen 2007), the ecosystem well-being index, ecological footprints, greenhouse gas emissions (Marquart-Pyatt 2012b), and CO₂ per capita (Givens and Jorgenson 2013). We rely on a categorical indicator of climate change’s impact on a country, Climate Impact, from the Climate Change Vulnerability Monitor (http://daraint.org/). This variable is a weighted combination of economic and health impacts from climate change ranging from “low” to “acute” impact—though no countries in our sample reach “acute” status. Most of the nations are “low” impact, while Armenia, Belarus, Bosnia and Herzegovina, Croatia, Georgia, Kyrgyzstan, Moldova, and Ukraine were classified as “moderate” and Mongolia and Tajikistan have experienced high climate change impacts. Montenegro and Serbia were missing data for this variable. To avoid losing these countries to listwise deletion, they were coded as “low” impact.³

Clearly, there are multiple ways to operationalize “objective problems.” We use an indicator of climate change impacts because our outcome variables are specific to climate change, as opposed to general environmental concern. Second, our variable captures human impacts, in terms of the economy or health, which is not as effectively captured by measures like CO₂ emissions. Wealthier nations expel high levels of greenhouse gas emissions, yet have experienced relatively little negative consequences from climate change. As such, CO₂ emissions do not truly operationalize the “objective problems” portion of postmaterialism because emission-intensive nations experience less “objective problems” from climate change.

Individual-level variables

There are three groups of individual-level predictors. These groups of variables capture respondents’ (1) recent economic experiences, (2) subjective assessments of personal finances and the national economy, and (3) relevant sociodemographic controls. Descriptive statistics are displayed in Table 1.

Table 1.

Descriptive Statistics for Explanatory Variables.

Variable	Mean	Standard deviation (between)	Standard deviation (within)	Minimum	Maximum
Country-level economic characteristics
GNItime($000)	82.633	5.210	N/A	4	384
GNIchange	3.113	5.210	N/A	−7	16
Economic experiences and subjective assessments
Impact of crisis	1.637	0.358	1.023	1	4
Staple cutbacks	0.655	0.163	0.448	0	1
Luxury cutbacks	0.433	0.127	0.479	0	1
Wage reduction	0.410	0.167	0.465	0	1
Job Loss (household head)	0.103	0.053	0.300	0	1
Job Loss (other member)	0.116	0.048	0.317	0	1
Personal satisfaction	2.231	0.403	1.066	1	5
Economic satisfaction	2.540	0.593	0.972	1	5
Perceived income decile	3.410	0.539	1.606	1	10
Controls
Education	3.160	1.352	0.573	1	7
Age	2.574	0.450	1.577	1	6
Sex	0.604	0.059	0.486	0	1
Market economy	0.348	0.105	0.465	0	1

Note. GNI = gross national income.

For recent economic experiences, we use a series of variables to assess negative outcomes related to the recession of circa 2008. These five variables asked if the respondent, or anyone in their household, has experienced a reduction in wages, a cutback in the consumption of staple goods, a cutback in luxury consumption, and if the head of the household or another member of the household lost one’s job as a result of the global recession (0 = no, 1 = yes).

The second set of variables assesses understanding of economic conditions. Impact of Crisis is developed from a survey question that asked about the effect of the global recession on a respondent’s household (1 = not at all, 4 = a great deal). LITS II does not include a personal income variable comparable across nations. Hence, we use a measure (Perceived income decile) for perceived income decile (1–10). The final two subjectivity variables asked respondents to rate their national economy and personal finances using a five-item Likert-type agreement scale. The first, which we call Economic Satisfaction, asked people how much they agree (1 = disagree, 5 = agree) with the following statement: “On the whole, I am satisfied with the present state of the economy.” The second variable, Personal Satisfaction, asked respondents how much they agree with this statement: “All things considered, I am satisfied with my financial situation as a whole” (1 = disagree, 5 = agree).

Last, we include a set of standard sociodemographic controls for age (1 = 18–24, 6 = 65+), sex (0 = male, 1 = female), and education (1 = no education, 7 = master’s degree or PhD). Our data do not include a measure of political affiliation comparable cross-nationally. In lieu of political ideology, we control for the preference for a free market over other economic systems as this has been connected to climate change disbelief (Heath and Gifford 2006; Lewandowsky, Oberauer, and Gignac 2013; McCright 2011).

Analytical Strategy

Standard regression approaches (e.g., ordinary least squares) assume that observations are independent. This assumption is clearly violated in the LITS II data as individuals are nested within countries. Because of the data structure, we estimate multilevel binary and ordinal logistic regression models with random intercepts (Luke 2004; Raudenbush and Bryk 2002).⁴ Each model includes the sociodemographic controls but we do not report these variables in the regression tables for the sake of brevity—full tables are available in the online appendix. Remaining variables were entered in three nested models. For each outcome, Model 1 includes country-level characteristics, Model 2 includes individual recent economic experiences, and Model 3 adds subjective economic assessments. Akaike information criterion (AIC) and Bayesian information criterion (BIC) statistics inform model selection. Coefficients and standard errors are reported in Tables 2 and 5; following the advice of Klein (2004), we report p values instead of flagging statistically significant coefficients. Because logistic regression coefficients are difficult to interpret directly, we display predicted probabilities derived from our best-fitting models to better understand the practical significance of our results (Mood 2010; Pampel 2000). We do not use random coefficients (i.e., effects that vary across countries) or cross-level interactions because these estimates will likely be biased to an unknown degree due to the relatively small number of Level 2 units (Bryan and Jenkins 2013; Stegmueller 2013).

Table 2.

Multilevel Ordinal Logistic Regression Models for Climate Concern.

Predictors	Model 1			Model 2			Model 3
Predictors	b	SE	p	b	SE	p	b	SE	P
Country-level characteristics
GNItime($000)	0.00	0.00	.53	0.00	0.00	.28	0.00	0.00	.31
GNIchange	0.04	0.01	.01	0.04	0.01	.01	0.04	0.01	.00
Climate impact (ref. low)
Moderate	0.19	0.16	.25	0.19	0.15	.22	0.21	0.15	.18
High	0.53	0.41	.20	0.55	0.42	.20	0.55	0.45	.22
Recent economic experiences
Job loss (household head)				0.12	0.06	.06	0.06	0.07	.38
Job loss(other member)				0.09	0.06	.12	0.03	0.06	.55
Staple cutbacks				0.12	0.03	.00	0.11	0.04	.00
Luxury cutbacks				0.15	0.05	.00	0.13	0.05	.01
Wage reduction				0.06	0.03	.06	−0.01	0.04	.81
Subjective economic assessments
Impact of crisis (ref. not at all)
Just a little							0.14	0.05	.00
A fair amount							0.19	0.06	.00
A great deal							0.28	0.08	.00
Economic satisfaction (ref. strongly disagree)
Disagree							−0.02	0.05	.63
Neither agree nor disagree							−0.07	0.06	.25
Agree							0.04	0.07	.55
Strongly agree							−0.08	0.12	.53
Personal satisfaction (ref. strongly disagree)
Disagree							−0.05	0.05	.28
Neither agree nor disagree							−0.12	0.05	.03
Agree							−0.10	0.07	.19
Strongly agree							−0.09	0.09	.27
Perceived income decile							0.07	0.02	.00
AIC	98,031.3			97,899.7			97,739.7
BIC	98,216.3			98,126.8			98,034.1
Variance component	.165			.155			.157
N	33,159			33,159			33,159

Note. All models include controls for education, age, sex, and free market preference. AIC = Akaike information criterion; BIC = Bayesian information criterion; GNI = gross national income.

Results

Climate Concern

Results for the Climate Concern multilevel ordinal logistic regression are reported in Table 2. The AIC and BIC are lowest in Model 3, indicating the best fit, so this model will be the focus of our discussion.⁵

The influence of GNItime($000) on climate concern is statistically significant, but the practical size of this effect is negligible. To better understand the role of country-level phenomenon, we calculated the probability of achieving the highest category (5) of Climate Concern and plotted these probabilities against GNItime($000) and GNIchange using violin plots in Figures 1 and 2. These plots show the distribution of predicted probabilities by country identified on the horizontal axis—see the online appendix for the list of countries and their codes. The central dot in each distribution or “violin” represents the mean predicted probability within that particular country with vertical bars for the interquartile range. These probabilities were calculated by holding all variables at their observed values and using only the fixed portion of the model (i.e., without the random intercepts).

Figure 1.

Predicted climate change concern = 5 by GNItime($000).

Figure 2.

Predicted climate change concern = 5 by GNIChange.

Figure 1 shows that GNItime$(000) has a minimal effect across on the probability of the highest level of Climate Concern as there is a significant overlap between countries and no obvious upward sweep in the probabilities. For instance, Germany’s GNItime($000) is 325 whereas it is only 17 for Armenia, indicating enormous differences in economic history between the two nations, yet the average of the predicted probabilities for the highest level of Climate Concern (i.e., 5) is .17 in Germany and .20 in Armenia. Thus—counter to Hypothesis 1a—long-run economic development has little relationship with concern.

Alternatively, Figure 2 indicates that the recent growth in GNI (GNIchange) has a positive effect. For example, the United Kingdom lost 7 percent of GNI between 2009 and 2010 and the average of the predictions of climate change concern = 5 is .13 among UK respondents. GNI per capita grew by 15 percent in Uzebekistan from 2009 to 2010 and Uzbekistanis average .23. Thus, these results lend support to Hypothesis 3a, in which we suggested that short-term economic growth increases concern.

Regarding Hypothesis 2a, Climate Impact has relatively large though statistically insignificant coefficients, indicating that as the human impacts of climate change increase, so does climate change concern. To understand the practical implications of Climate Impact, we calculated average predicted probabilities for both outcomes and present these in Table 3. The probabilities suggest the effect of Climate Impact is substantively important. For instance, the probability of the highest category of Climate Concern (5) for individuals in low-impact nations is .167, whereas individuals in high-impact nations average .284—a substantial difference.

Table 3.

Predicted Climate Concern and Willingness to Pay by Climate Impact.

Variable	Climate impact
Variable	Low	Medium	High
High climate concern	.1671	.2056	.2849
Willingness to pay	.3149	.3958	.4836

Recent negative economic experiences increase Climate Concern and, among these variables, a loss of luxury consumption has the strongest effect. To better understand the effect of these variables, Table 4 presents predicted probabilities, again calculated at observed values using the fixed portion of the model. These predicted probabilities suggest that the effect of recent economic experiences is practically small. For instance, the average probability of high climate change concern (i.e., Climate Concern = 5) for an individual whose head of household lost his or her job is .189, whereas that same figure for individuals whose head of household did not lose his or her job is .180. Thus, there is no support for Hypothesis 4a—individual economic hardships appear to increase or have a null effect on concern, rather than reduce concern.

Table 4.

Predicted Probabilities of Recent Economic Experiences for Climate Concern and Willingness to Pay.

Variable	Job loss (HH)		Job loss (other)		Staple cut		Luxury cut		Wage reduction
Variable	Yes	No	Yes	No	Yes	No	Yes	No	Yes	No
High climate concern	.189	.180	.185	.181	.191	.176	.192	.172	.180	.182
Willingness to pay	.386	.337	.185	.180	.189	.180	.361	.327	.355	.337

Regarding Hypothesis 5a, subjective assessments of economic conditions have inconsistent effects. Individuals who report high impacts of the global recession report higher levels of Climate Concern. Macroeconomic conditions and personal economic satisfaction both have a negative effect, but the coefficients are small and not statistically significant. Income, captured subjectively with Perceived Income Decile, increases concern.

Model 3 suggests that recent negative economic experiences, like a reduction in consumption, or perceived negative effects from the recent global recession are positively associated with concern about climate change, that is, individuals who have experienced negative personal or household impacts of the recession are more likely to be highly concerned with climate change. Furthermore, short-term economic growth positively influences climate change concern but long-term economic evolution has very little effect. Last, subjective assessments of economic conditions have a more inconsistent role.

Willingness to Pay

Results for the Willingness to Pay models are shown in Table 5. As with Climate Concern, we look for an improved model fit as more variables are included, with AIC and BIC again indicating the fully elaborated Model 3 as the best fit. Our discussion will center on this model.

Table 5.

Multilevel Binary Logistic Regression Results for Willingness to Pay.

Predictors	Model 1			Model 2			Model 3
Predictors	b	SE	p	b	SE	p	b	SE	p
Country-level characteristics
GNItime($000)	0.00	0.00	.04	0.00	0.00	.04	0.00	0.00	.09
GNIchange	0.06	0.02	.00	0.06	0.02	.00	0.06	0.02	.01
Climate impact (ref. low)
Moderate	0.43	0.32	.18	0.45	0.32	.17	0.51	0.34	.14
High	0.58	0.35	.10	0.61	0.36	.09	0.56	0.42	.19
Recent economic experiences
Job loss (household head)				0.12	0.06	.07	0.23	0.07	.00
Job Loss (other member)				0.05	0.06	.41	0.12	0.07	.08
Staple cutbacks				−0.12	0.08	.11	0.03	0.08	.73
Luxury cutbacks				0.12	0.05	.01	0.16	0.05	.00
Wage reduction				0.03	0.06	.53	0.08	0.05	.12
Subjective economic assessments
Impact of crisis (ref. not at all)
Just a little							0.02	0.06	.71
A fair amount							−0.08	0.07	.24
A great deal							−0.14	0.09	.13
Economic satisfaction (ref. strongly disagree)
Disagree							0.10	0.05	.06
Neither agree nor disagree							−0.02	0.08	.81
Agree							0.13	0.08	.13
Strongly agree							−0.15	0.16	.34
Personal satisfaction (ref. strongly disagree)
Disagree							0.15	0.07	.03
Neither agree nor disagree							0.19	0.08	.02
Agree							0.50	0.09	.00
Strongly agree							0.62	0.12	.00
Perceived income decile							0.07	0.01	.00
AIC	35,888.07			35,855.37			35,476.21
BIC	36,046.03			36,054.89			35,758.87
Variance component	0.43			0.43			0.46
N	30,134			30,134			30,134

Note. All models include controls for education, age, sex, and free market preference. AIC = Akaike information criterion; BIC = Bayesian information criterion; GNI = gross national income.

Country-level characteristics have effects similar to the climate change concern models. Again, GNItime($000) has a very limited influence while short-term economic growth (GNIchange) increases willingness to pay.

To better understand these relationships, we display violin plots that show the distribution of predicted probabilities for each country. As with Climate Concern, the effect of economic history is minimal. For instance, Figure 3 shows that GNItime($000) for Kosovo is about 10 whereas it is 317 for France. The average of the predictions for French respondents is .37 whereas for Kosovoans the same value is .29—even enormous differences in economic history have a modest effect. Figure 4 suggests that GNIchange increases a respondent’s willingness to pay higher taxes to combat climate change. For example, GNI per capita increased by 15 percent in Moldova and fell by 4 percent in Romania. Moldovans have an average predicted willingness to pay of .54 whereas for Romanians that same amount is only .18. These results lend no support to Hypothesis 1b, in which we suggested the long-run economic development would improve Willingness to Pay but largely corroborate Hypothesis 3b as short-term, relative economic growth increases Willingness to Pay.

Figure 3.

Predicted willingness to pay by GNItime($000).

Figure 4.

Predicted willingness to pay by GNIchange.

Moving to Hypothesis 2b, climate change impacts have a substantial impact on Willingness to Pay. Table 3 shows that individuals living in low-impact nations have an averaged predicted probability of .316 whereas that same figure for individuals in high-impact nations is .534—a substantive difference.

Negative economic experiences (e.g., job loss, staple cutback, luxury cutback, wage reduction) all have a slight positive association with Willingness to Pay. As shown in Table 4, the largest observed effect is for cutbacks in luxury consumption. Here, predicted Willingness to Pay is .36 for respondents whose household has experienced a loss of luxury consumption and .33 for those who have not. Overall, negative recent economic experiences increase Willingness to Pay, but the effect is substantively small. Thus, these findings do not provide evidence for Hypothesis 5b, in which we suggested that economic hardship would reduce Willingness to Pay.

Subjective economic assessments do not have the same effect on Willingness to Pay as they do on Climate Concern. Individuals who believe that the economic crisis has affected them or their household greatly have lower willingness to pay, but this effect is not statistically significant. Satisfaction with national macroeconomic conditions has an inconsistent effect across its categories but none reach statistical significance. Personal satisfaction with one’s financial situation, however, has a consistently positive effect and reaches statistical significance for each of its thresholds. Thus, some, but not all, types of subjective economic assessments are consequential for willingness to pay for climate change policy.

Discussion and Conclusion

The purpose of this article is to integrate and amend previous scholarship on the economy-environment relationship by developing a holistic, multidimensional understanding of how economic conditions affect environmental attitudes using the case of climate change views. As such, this article contributes to the long-standing debates on global environmental concern (e.g., Brechin 1999; Dunlap and York 2008) and the emergent literature on cross-national climate change attitudes (Scruggs and Benegal 2012; Tjernström and Tietenberg 2008). In this section, we will discuss our results in a way that reflects our integrated understanding of the relationship between economic conditions and environmental concern.

Absolute differences in economic development have almost no influence—dramatic differences in economic history do little to explain divergence in climate change concern or willingness to pay across countries. This is in line with the argument of “global environmentalism” scholars (e.g., Brechin 1999; Brechin and Kempton 1994; Dunlap and York 2008) who maintain that environmental attitudes are not dependent on economic development. Climate change impacts increase concern and willingness to pay, but have a stronger effect on willingness to pay. These results support a qualified version of the “objective problems” argument in that exposure to impacts from climate change influences policy or behavioral responses more than general concern.

However, short-term relative economic growth increases climate change concern and willingness to pay. These findings are consistent with previous research (e.g., Brulle et al. 2012; Kahn and Kotchen 2010; Scruggs and Benegal 2012; Shum 2012), which found that relative, short-run economic change affects attitudes toward climate change. This could be interpreted as qualified support for a version of the affluence hypothesis in which relative economic growth in the short term, as opposed to long-run economic development, increases both concern and willingness to pay.

We included several individual-level variables that capture personal or household experiences with the recent global recession. Individual negative economic experiences have either a null effect or slightly increase concern and willingness to pay. This is an important finding because studies using data aggregated at a contextual unit (e.g., countries) find that growth in unemployment or loss in GDP reduces concern for climate change (Brulle et al. 2012; Kahn and Kotchen 2010; Scruggs and Benegal 2012; Shum 2012). Our results suggest that this seemingly close relationship between economic recession and climate change attitudes is likely an ecological artifact.

Given that some negative individual economic experiences (such as a job loss) increase concern and willingness to pay and subjective economic impressions have almost no effect, further discussion is warranted. We suspect that recessions galvanize the opinions of people who were less concerned about climate change or not supportive of climate change policy prior to the recession, regardless of their personal economic experiences. Individuals who fail to recognize the importance of climate change may simply interpret recessions as yet another reason to avoid action on climate change, even if the recession has not had much impact on them. Schor (2014) emphasized the importance of discursive framing in her review of recession and public concern for climate change. Regardless of personal economic experiences, some individuals might frame policy action on climate change as a threat to economic growth and subsequently deny its reality or the necessity of a policy response.

The microfoundation of postmaterialist theory is that individual economic security engenders postmaterialist values, which in turn create environmental concern—leading to significant changes in aggregate opinion. Similarly, the affluence hypothesis suggests that as individuals achieve greater economy prosperity, they increasingly demand environmental quality. The null or positive effects of individual economic insecurity call into question these microfoundations.

More work is needed to theoretically articulate how people interpret short-term, relative economic change. Future research should develop a better understanding of how embedded actors conceptualize the broader economy and their personal economic situation and, in turn, how these variables affect environmental views. More broadly, environmental views are affected by many noneconomic variables, including perceptions of global citizenship (e.g., Running 2013) and social trust (e.g., Harring 2013). Nations differ by far more than just economic development—such as the quality of their institutions and degree of democracy. Moving beyond economic explanations could be especially fruitful for global environmental concern scholarship. Furthermore, environmental viewpoints are themselves highly subjective and at least somewhat influenced by contextual conditions (e.g., Auyero and Swistun 2009; Hamilton, Colocousis, and Duncan 2010; Shriver and Kennedy 2005). Climate change beliefs may be partially reflective of dependence on extractive industries, proximity to a coast or other contextual variables less often considered in the cross-national literature.

This analysis also has implications beyond theoretical debates internal to the social scientific literature. As Schor (2014) noted, political leaders use economic downturns to justify inaction on climate change and political discourse often portrays climate action as a “luxury good.” This research suggests that economic hardship does not drive individuals to de-prioritize climate change. Rather, at the individual level, concern about climate change and support for climate change action endures in the face of hardship.

Footnotes

Acknowledgements

The authors thank Dr. Lynn Hempel and Dr. Stephanie Malin for comments on earlier drafts.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental material

Supplementary material for this article is available online.

Notes

References

Auyero

Javier

Swistun

Débora Alejandra

. 2009. Flammable: Environmental Suffering in an Argentine Shantytown. Oxford, England: Oxford University Press.

Blaacker

Debra

Woods

Joshua

Oliver

Christopher

. 2012. “How Big Is Big Coal? Public Perceptions of the Coal Industry’s Economic Impact in West Virginia.” Organization & Environment 25(4):385–401.

Blendon

Robert J.

Benson

John M.

Brodie

Mollyann

Morin

Richard

Altman

Drew E.

Gitterman

Daniel

Brossard

Mario

James

Matt

. 1997. “Bridging the Gap between the Public’s and Economists’ Views of the Economy.” The Journal of Economic Perspectives 11(3):105–18.

Brechin

Steven R.

1999. “Objective Problems, Subjective Values, and Global Environmentalism: Evaluating the Postmaterialist Argument and Challenging a New Explanation.” Social Science Quarterly 80(4):793–809.

Brechin

Steven R.

Kempton

Willett

. 1994. “Global Environmentalism: A Challenge to the Postmaterialism Thesis?” Social Science Quarterly 75(2):245–69.

Brechin

Steven R.

Kempton

Willett

. 1997. “Beyond Postmaterialist Values: National versus Individual Explanations of Global Environmentalism.” Social Science Quarterly 78(1):16–20.

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 US, 2002–2010.” Climatic Change 114 (2): 169–88.

Bryan

Mark L.

Jenkins

Stephen P.

2013. “Regression Analysis of Country Effects Using Multilevel Data: A Cautionary Tale.” Retrieved December 27, 2015 (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2322088).

Çarkoğlu

Ali

Kentmen-Çin

Çiğdem

. 2015. “Economic Development, Environmental Justice, and Pro-environmental Behavior.” Environmental Politics 24(4):575–97.

10.

Chaisty

Paul

Whitefield

Stephen

. 2015. “Attitudes towards the Environment: Are Post-communist Societies (Still) Different?” Environmental Politics 24(4):598–616.

11.

Chambers

John R.

Swan

Lawton K.

Heesacker

Martin

. 2014. “Better Off than We Know Distorted Perceptions of Incomes and Income Inequality in America.” Psychological Science 25(2):613–18.

12.

Diekmann

Andreas

Franzen

Axel

. 1999. “The Wealth of Nations and Environmental Concern.” Environment & Behavior 31(4):540–49.

13.

Dorsch

Michael T.

2014. “Economic Development and Determinants of Environmental Concern.” Social Science Quarterly 95(4):960–77.

14.

Dunlap

Riley E.

Jones

2002. “Environmental Concern: Conceptual and Measurement Issues.” Pp. 482–524 in Handbook of Environmental Sociology, edited by Dunlap

Michelson

London, England: Greenwood.

15.

Dunlap

Riley E.

Mertig

Angela G.

1997. “Global Environmental Concern: An Anomaly for Postmaterialism.” Social Science Quarterly 78(1):24–29.

16.

Dunlap

Riley E.

York

Richard

. 2008. “The Globalization of Environmental Concern and the Limits of the Postmaterialist Values Explanation: Evidence from Four Multinational Surveys.” The Sociological Quarterly 49(3):529–63.

17.

ECB. 2013. “Life in Transition II Study.” Retrieved December 20, 2013 (http://www.ebrd.com/news/publications/special-reports/life-in-transition-survey-ii.html)

18.

Fairbrother

Malcolm

. 2013. “Rich People, Poor People, and Environmental Concern: Evidence across Nations and Time.” European Sociological Review 29(5):910–22.

19.

Fransson

Niklas

Gärling

Tommy

. 1999. “Environmental Concern: Conceptual Definitions, Measurement Methods, and Research Findings.” Journal of Environmental Psychology 19(4):369–82.

20.

Franzen

Axel

. 2003. “Environmental Attitudes in International Comparison: An Analysis of the ISSP Surveys 1993 and 2000.” Social Science Quarterly 84(2):297–308.

21.

Franzen

Axel

Meyer

Reto

. 2010. “Environmental Attitudes in Cross-national Perspective: A Multilevel Analysis of the ISSP 1993 and 2000.” European Sociological Review 26(2):219–34.

22.

Franzen

Axel

Vogl

Dominikus

. 2013a. “Acquiescence and the Willingness to Pay for Environmental Protection: A Comparison of the ISSP, WVS, and EVS.” Social Science Quarterly 94(3):637–59.

23.

Franzen

Axel

Vogl

Dominikus

. 2013b. “Two Decades of Measuring Environmental Attitudes: A Comparative Analysis of 33 Countries.” Global Environmental Change 23(5):1001–1008.

24.

Freymeyer

Robert H.

Johnson

Barbara E.

2010. “A Cross-cultural Investigation of Factors Influencing Environmental Actions.” Sociological Spectrum 30(2):184–95.

25.

Gelissen

John

. 2007. “Explaining Popular Support for Environmental Protection a Multilevel Analysis of 50 Nations.” Environment & Behavior 39(3):392–415.

26.

Givens

Jennifer E.

Jorgenson

Andrew K.

2011. “The Effects of Affluence, Economic Development, and Environmental Degradation on Environmental Concern: A Multilevel Analysis.” Organization & Environment 24(1):74–91.

27.

Givens

Jennifer E.

Jorgenson

Andrew K.

2013. “Individual Environmental Concern in the World Polity: A Multilevel Analysis.” Social Science Research 42(2):418–31.

28.

Granovetter

Mark

. 1985. “Economic Action and Social Structure: The Problem of Embeddedness.” American Journal of Sociology 91(3):481–510.

29.

Hadler

Markus

Wohlkönig

Patrick

. 2012. “Environmental behaviours in the Czech Republic, Austria and Germany between 1993 and 2010: Macro-level trends and individual-level determinants compared.” Sociologický časopis/Czech Sociological Review 48(3):467–92.

30.

Hamilton

Lawrence C.

Colocousis

Chris R.

Duncan

Cynthia M.

2010. “Place Effects on Environmental Views.” Rural Sociology 75(2):326–47.

31.

Harring

Niklas

. 2013. “Understanding the Effects of Corruption and Political Trust on Willingness to Make Economic Sacrifices for Environmental Protection in a Cross-national Perspective.” Social Science Quarterly 94(3):660–71.

32.

Heath

Yuko

Gifford

Robert

. 2006. “Free-market Ideology and Environmental Degradation the Case of Belief in Global Climate Change.” Environment & Behavior 38(1):48–71.

33.

Hornborg

Alf

. 1998. “Towards an Ecological Theory of Unequal Exchange: Articulating World System Theory and Ecological Economics.” Ecological Economics 25(1):127–36.

34.

Ignatow

Gabriel

. 2006. “Cultural Models of Nature and Society Reconsidering Environmental Attitudes and Concern.” Environment & Behavior 38(4):441–61.

35.

Inglehart

Ronald

. 1977. “Values, Objective Needs, and Subjective Satisfaction among Western Publics.” Comparative Political Studies 9(4):429–58.

36.

Inglehart

Ronald

. 1995. “Public Support for Environmental Protection: Objective Problems and Subjective Values in 43 Societies.” Political Science & Politics 28(1):57–72.

37.

Inglehart

Ronald

Abramson

Paul R.

1994. “Economic Security and Value Change.” American Political Science Review 88(2):336–54.

38.

Jorgenson

Andrew K.

2006. “Unequal Ecological Exchange and Environmental Degradation: A Theoretical Proposition and Cross-national Study of Deforestation, 1990–2000.” Rural Sociology 71(4):685–712.

39.

Jorgenson

Andrew K.

Givens

Jennifer E.

2014. “Economic Globalization and Environmental Concern: A Multilevel Analysis of Individuals within 37 Nations.” Environment & Behavior 46(7):848–71.

40.

Kahn

Matthew E.

Kotchen

Matthew J.

2010. Environmental Concern and the Business Cycle: The Chilling Effect of Recession. National Bureau of Economic Research. Retrieved September 3, 2016 (http://www.nber.org/papers/w16241).

41.

Kemmelmeier

Markus

Krol

Grzegorz

Kim

Young Hun

. 2002. “Values, Economics, and Proenvironmental Attitudes in 22 Societies.” Cross-cultural Research 36(3):256–85.

42.

Klein

Rex

. 2004. Beyond Significance Testing: Reforming Data Analysis Methods in Behavioral Research. Washington, DC: American Psychological Association.

43.

Knight

Kyle W.

Messer

Benjamin L.

2012. “Environmental Concern in Cross-national Perspective: The Effects of Affluence, Environmental Degradation, and World Society.” Social Science Quarterly 93(2):521–37.

44.

Kvaløy

Berit

Finseraas

Henning

Listhaug

Ola

. 2012. “The Publics’ Concern for Global Warming: A Cross-national Study of 47 Countries.” Journal of Peace Research 49(1):11–22.

45.

Lee

Tien Ming

Markowitz

Ezra M.

Howe

Peter D.

Chia-Ying

Leiserowitz

Anthony A.

2015. “Predictors of Public Climate Change Awareness and Risk Perception around the World.” Nature Climate Change 5(11):1014–1020.

46.

Lewandowsky

Stephan

Oberauer

Klaus

Gignac

Gilles E.

2013. “NASA Faked the Moon Landing—Therefore, (Climate) Science Is a Hoax an Anatomy of the Motivated Rejection of Science.” Psychological Science 24(5):622–33.

47.

Lo, Alex Y . 2015 “National income and environmental concern: Observations from 35 countries.” Public Understanding of Science. doi: 0963662515581302.

48.

Alex Y.

Chow

Alex T.

2015. “The Relationship between Climate Change Concern and National Wealth.” Climatic Change 131(2):335–48.

49.

Luke

Douglas A.

2004. Multilevel Modeling. Thousand Oaks, CA: Sage Publications.

50.

Marquart-Pyatt

Sandra T.

2008. “Are There Similar Sources of Environmental Concern? Comparing Industrialized Countries.” Social Science Quarterly 89(5):1312–35.

51.

Marquart-Pyatt

Sandra T.

2012a. “Environmental Concerns in Cross-national Context: How Do Publics in Central and Eastern Europe Compare with Other Regions of the World?” Czech Sociological Review 48(3):441–66.

52.

Marquart-Pyatt

Sandra T.

2012b. “Contextual Influences on Environmental Concerns Cross-nationally: A Multilevel Investigation.” Social Science Research 41(5):1085–99.

53.

McCright

Aaron M.

2011. “Political orientation moderates Americans’ beliefs and concern about climate change.” Climatic Change 104(2):243–53.

54.

Mood

Carina

. 2010. “Logistic Regression: Why We Cannot Do What We Think We Can Do, and What We Can Do about It.” European Sociological Review 26(1):67–82.

55.

Mostafa

Mohamed M.

2012. “Does Globalisation Affect Consumers’ Pro-environmental Intentions? A Multilevel Analysis across 25 Countries.” International Journal of Sustainable Development & World Ecology 19(3):229–37.

56.

Mostafa

Mohamed M.

2013. “Wealth, Post-materialism and Consumers’ Pro-environmental Intentions: A Multilevel Analysis across 25 Nations.” Sustainable Development 21(6):385–99.

57.

Mostafa

Mohamed M.

2016. “Post-materialism, Religiosity, Political Orientation, Locus of Control and Concern for Global Warming: A Multilevel Analysis across 40 Nations.” Social Indicators Research 128:1273–98. doi:10.1007/s11205-015-1079-2.

58.

Norton

Michael I.

Ariely

Dan

. 2011. “Building a Better America—One Wealth Quintile at a Time.” Perspectives on Psychological Science 6(1):9–12.

59.

O’Brien

Karen

Leichenko

Robin

Kelkar

Ulka

Venema

Henry

Aandahl

Guro

Tompkins

Heather

Javed

Akram

Bhadwal

Suruchi

Barg

Stephan

Nygaard

Lynn

West

Jennifer

. 2004. “Mapping Vulnerability to Multiple Stressors: Climate Change and Globalization in India.” Global Environmental Change 14(4):303–13.

60.

Pampel

Fred C.

2000. Logistic Regression: A Primer. Thousand Oaks, CA: Sage Publications.

61.

Raudenbush

Stephen W.

Bryk

Anthony S.

2002. Hierarchical Linear Models: Applications and Data Analysis Methods. Thousand Oaks, CA: Sage Publications.

62.

Rice

James

. 2007a. “Ecological Unequal Exchange: Consumption, Equity, and Unsustainable Structural Relationships within the Global Economy.” International Journal of Comparative Sociology 48(1):43–72.

63.

Rice

James

. 2007b. “Ecological Unequal Exchange: International Trade and Uneven Utilization of Environmental Space in the World System.” Social Forces 85(3):1369–92.

64.

Running

Katrina

. 2013. “World Citizenship and Concern for Global Warming: Building the Case for a Strong International Civil Society.” Social Forces 92(1):377–99.

65.

Running

Katrina

. 2015. “Towards Climate Justice: How Do the Most Vulnerable Weigh Environment–Economy Trade-offs?” Social Science Research 50:217–28.

66.

Sandvik

Hanno

. 2008. “Public Concern over Global Warming Correlates Negatively with National Wealth.” Climatic Change 90(3):333–41.

67.

Schor

Juliet B.

2014. “Climate Discourse and Economic Downturns: The Case of the United States, 2008–2013.” Environmental Innovation and Societal Transitions 13:6–20.

68.

Scruggs

Lyle

Benegal

Salil

. 2012. “Declining Public Concern about Climate Change: Can We Blame the Great Recession?” Global Environmental Change 22(2):505–15.

69.

Shao

Wanyun

Keim

Barry D.

Garand

James C.

Hamilton

Lawrence C.

2014. “Weather, Climate, and the Economy: Explaining Risk Perceptions of Global Warming, 2001–10*.” Weather, Climate, and Society 6(1):119–34.

70.

Shriver

Thomas E.

Kennedy

Dennis K.

2005. “Contested Environmental Hazards and Community Conflict over Relocation.” Rural Sociology 70(4):491–513.

71.

Shum

Robert Y.

2012. “Effects of Economic Recession and Local Weather on Climate Change Attitudes.” Climate Policy 12(1):38–49.

72.

Stegmueller

Daniel

. 2013. “How Many Countries for Multilevel Modeling? A Comparison of Frequentist and Bayesian Approaches.” American Journal of Political Science 57(3):748–61.

73.

Tjernström

Emilia

Tietenberg

Thomas

. 2008. “Do Differences in Attitudes Explain Differences in National Climate Change Policies?” Ecological Economics 65(2):315–24.

74.

Xiao

Chenyang

Dunlap

Riley E.

2007. “Validating a Comprehensive Model of Environmental Concern Cross-nationally: A US-Canadian Comparison.” Social Science Quarterly 88(2):471–93.

75.

Zhou

Min

. 2015. “Public Environmental Skepticism: A Cross-national and Multilevel Analysis.” International Sociology 30(1):61–85.