1. Introduction
Values, being relatively stable and enduring conceptions of what is desirable and undesirable, are thought to shape everything from attitudes toward social and environmental issues to specific environmental behaviors. Specific values do not influence individuals’ attitudes and behaviors in isolation; rather, specific types of values are highly correlated (Schwartz 2012). For example, individuals who highly value “tradition” and “conformity” also tend to value “benevolence” and “security” (Schwartz 1994). Being socially constructed, ideologies are likely to be formed, modified, and reinforced through the exchanges individuals have in their established information and communication networks. Collectively, when large groups share these highly correlated values, they form abstract, symbolic, and easily identifiable ideologies. Like specific values, ideologies have the potential to shape individuals’ attitudes as well as their psychological and behavioral responses to new information.
The symbolic ideologies individuals identify with, such as “conservative” or “liberal,” can influence their attitudes toward climate change (McCright 2011; McCright and Dunlap 2011) and may affect how they respond to climate-driven biophysical changes in nearby natural resources, such as publically managed forests (Newfont 2012). This research explores the relationships between individuals’ political ideologies, the information sources used to learn about the local impacts of climate change, and individuals’ perceived ability to adapt to those changes.
The investigation is focused on five interrelated hypotheses. The first relates to the use and trust of distinct information sources while the other four relate to individuals’ perceived ability to adapt to increasingly variable climate conditions. These hypotheses are as follows:
Individuals’ political ideologies are not significantly related to the trust they place in specific information sources. Both conservatives and liberals are expected to place more trust in immediate friends and family members relative to other information sources.
Individuals’ political ideologies are significantly related to how severe they believe local climate-related risks are. Liberals are expected to perceive local climate-related risks as more severe relative to conservatives.
Individuals’ political ideologies are significantly related to the ease with which they believe information about local climate-driven environmental change can be found. Liberals are expected to believe information is easier to find relative to conservatives.
Individuals’ political ideologies are significantly related to their willingness to plan for climate-driven changes to local forest ecosystems. Liberals are expected to be more willing to plan for change relative to conservatives.
Individuals’ political ideologies are significantly related to their perceived ability to adapt to altered forest conditions driven by increasingly variable climate conditions. Liberals are expected to believe they are more capable of adapting to change relative to conservatives.
2. Related literature
a. Political ideologies and the use and trust of specific information sources
Political ideologies influence the ways in which individuals seek out information (Iyengar and Hahn 2009) and the trust they place in specific information sources (Lee 2010). Selective exposure to different messages and media can lead to polarized views of many contemporary issues, including climate change. Several social theories address information source selection and can be utilized to guide an analysis of potential differences between conservatives and liberals in their use and trust of distinct climate change–related information sources. Specifically, the use and trust of specific information sources can be explained through an understanding of individuals’ social networks and the social capital contained within them. Social capital specifically refers to the information, trust, and norms of reciprocity in one’s social network (Woolcock 1998). Individuals’ social ties are indicative of the types of information they will have access to and the trust they will place in that information. “Bonding ties” refer to close, generally trusted connections among tightly knit homogeneous groups, such as families, close friends, and neighbors (Marsden and Friedkin 1993; Gittell and Vidal 1998). Comparatively, “bridging ties” are less-trusted connections between individuals or groups in more dissimilar social networks. A large body of conceptual work focused on the sociology of climate change adaptation implies information obtained from bonding ties is more trusted than information obtained from bridging ties to more dissimilar others (Pelling and High 2005; Bodin and Crona 2009). Given this, both conservatives and liberals are expected to seek out and place a large degree of trust in information from their bonding social ties (i.e., immediate friends and family members) relative to the extent to which they seek out and place trust in information from their bridging social ties (hypothesis 1).
b. Political ideologies and perceived adaptive capacity
Individuals’ ability to adapt to increasingly variable climatic conditions is a complex and multidimensional process. Previous empirical work suggests perceived adaptive capacity is composed of at least four interrelated psychological dimensions, including 1) perceived risk, 2) the willingness to learn about potential impacts, 3) the willingness to plan for variable climate futures, and 4) a general perception of self-efficacy (Marshall et al. 2007; Smith et al. 2012a). Previous research on political ideologies can help guide hypotheses 2–5, which focus on the relationship between individuals’ political ideologies and each of these dimensions of perceived adaptive capacity. Within this research, there is a well-documented divide between political ideologies on the issue of global climate change (McCright 2011; McCright and Dunlap 2011). However, the vast majority of previous research is focused on climate change in general and not on the climate-driven changes to local natural resource conditions. The broader literature suggests that liberals are more likely to report beliefs consistent with scientific consensus and express greater concern over climate change than conservatives (Hamilton and Keim 2009; McCright and Dunlap 2011). If the capacity of individuals to adapt to changing environmental conditions is determined in part by their perception of risks associated with climate change, a claim could be made that conservative denials of global climate change also affect perceptions of local climate-related environmental risks. This research explicitly explores this proposition.
In addition to the well-documented differences between conservatives and liberals in the general level of concern given to global climate change, previous research suggests there are general psychological and social differences between the two groups that alter their perceptions of climate change–related risks and, potentially, their adaptive capacity. Relative to individuals who hold conservative ideologies, liberals tend to be more sensitive to threats to generalized others (Choma et al. 2013). Liberals also tend to rely more on moral rules in order to evaluate consequences of their behavior (Graham et al. 2009). Conversely, conservatism is associated with greater threat sensitivity and higher levels of perceived risks to one’s self and similar others (Graham et al. 2009; Choma et al. 2013). This general trend may partially explain why liberals tend to perceive the risks associated with global climate change (commonly thought of as being marginal and temporally/spatially distant) as more severe when compared to conservatives. Additional evidence suggests liberals will believe they are more capable of adapting to variable environmental conditions relative to conservatives. Findings from experimental psychological studies show liberals have a greater preference for flexibility and progress, while conservatives tend to prefer stability and tradition (Jost et al. 2003, 2008). Given this, significant differences were expected between conservatives and liberals on each of the four interrelated psychological dimensions of perceived adaptive capacity (hypotheses 2–5).
3. Methodology
The five hypotheses above were addressed with data collected via a mail-back survey administered to full-time resident homeowners living in three western North Carolina cities. Data were collected on 1) the sources individuals use when seeking out information about local environmental issues, 2) the extent to which individuals trust each of those information sources, 3) individuals’ self-reported political ideology, 4) the extent to which individuals believe local climate-driven environmental change poses a risk to themselves, 5) the ease with which individuals believe information about local climate-related environmental impacts is to find, 6) individuals’ willingness to plan for climate-driven changes to local resource conditions, and 7) individuals’ perceived ability to adapt to local environmental change. The latter four items are intended to capture individuals’ perceived capacity to adapt to climate-related impacts to local forest ecosystems (Smith et al. 2012a).
a. Study populations and data collection
The three study populations were residents living in Waynesville, Franklin, and Spruce Pine, North Carolina, United States. These three cities were chosen because of their geographic proximity to large forest ecosystems projected to change notably due to a variety of factors driven and exacerbated by changes in climate. Specifically, southern pine forests will face increasingly severe and frequent outbreaks of pine beetle infestations because of warmer, drier, and longer winters (Logan et al. 2003). Additionally, drier and warmer summer periods are projected to substantially alter fire risks (Dale et al. 2001). The three cities were also selected because their populations have historically been dependent on forest resources, whether through timber harvesting or the processing of forest products (Smith 2013). Over the past several decades, the entire region has transitioned away from extractive industries and become more heavily dependent on the area’s recreational and aesthetic amenities, which are also dependent on the health of local forest ecosystems (Jones et al. 2003). The survey was administered to 300 full-time resident homeowners in each city; 900 surveys were sent in total, following the Tailored Design Method (Dillman et al. 2008).
b. Measures
1) Political ideology
Political ideology was measured by asking respondents to indicate whether they considered themselves to be a conservative, a moderate conservative, a moderate, a moderate liberal, or a liberal. Individuals were also given an open-ended “other” response category. To avoid the possibility of priming bias, the political ideology question was asked near the end of the survey instrument, along with other typical sociodemographic characteristics questions.
It is important to note here that this study is measuring what public opinion researchers refer to as symbolic political ideology (Page and Shapiro 1992). “Symbolic” refers to more general and abstract ideological constructs that can be manifested through self-identification and the use of labels such as “left” or “right” (Jost et al. 2009). Symbolic political ideology differs from “operational” political ideology, which refers to more specific, concrete, and issue-based opinions and beliefs. Individuals can, and often do, hold specific issue-based opinions and beliefs that differ from the more general ideological group they identify with. As an example, an individual may hold conservative opinions and beliefs in regard to fiscal issues, but still identify as a liberal because they hold stronger liberal opinions and beliefs on a wider set of issues (social policy, international trade, migration, the environment, etc.; Stimson 2004).
2) Information sources
The survey solicited information about where individuals obtain their information about the local environmental issues and the extent to which they trust the information obtained from each source, if it was used. The survey was populated with 13 different nonexclusive information sources:
local newspapers or periodicals,
close friends,
friends you interact with infrequently,
coworkers,
immediate family members,
extended family members,
local television news,
national television news,
civic groups (Kiwanis, Rotary Club, etc.),
educational groups [parent–teacher associations (PTAs), school boards, etc.],
local officials (city council members, chamber of commerce representatives, etc.),
local churches,
online news sources, and
online non-news sources (blogs, Facebook, etc.).
Given that the hypotheses are concerned solely with trusted information sources, this research only focused on whether or not an individual always trusted information from a particular source. Responses were collapsed into two categories, the first containing information sources that were always trusted and the second containing information sources that were partially trusted, distrusted, or not used. Responses were binned following the guidelines provided by Linacre (2002) to ensure a strict interpretation of the trust concept, with the goal of yielding the most objective (i.e., context independent) measure possible.
3) Perceived adaptive capacity to climate-driven changes to local forest conditions
Individuals’ perceived adaptive capacity can be measured through the use of psychometric scales asking questions about key adaptation dimensions. As noted in the literature review, previous research has suggested perceived adaptive capacity is composed of at least four interrelated dimensions, including 1) perceived risk, 2) the willingness to learn about potential impacts, 3) the willingness to plan for variable climate futures, and 4) a general perception of self-efficacy (Smith et al. 2012a; Marshall et al. 2007). A perceived adaptive capacity measurement instrument was developed based on this previous research, with each of the four dimensions being measured through three statement items (12 total individual statement items). The instrument was tailored to focus on changes to local forest ecosystems as a result of climatic change (e.g., increased frequency of severe storm events and longer, drier, and hotter summers).
c. Data analysis
1) Political ideologies and trust in individual information sources
To examine the first hypothesis—that individuals’ political ideologies are not related to the trust they place in specific information sources—χ2 tests were used to assess the relationship between the political ideology variable and the 13 dichotomous trust variables corresponding to each information source. This relationship was examined with both the five original conservative to liberal categories and a collapsed variable where self-identified moderate conservatives were binned with the self-identified conservatives and self-identified moderate liberals were binned with the self-identified liberals; moderates were omitted. To preview the findings, the results are robust to either classification scheme.
2) Adaptive capacity scale evaluation
Hypotheses 2–5 were explored through a two-step process, first constructing and evaluating the measurement model for the multidimensional adaptive capacity scale and then using multivariate multiple regression to examine the relationships addressed in each hypothesis.
The scale’s construct validity was determined by testing the reliability of individual latent dimensions as well as satisfying the requirements of convergent and discriminant validity. Cronbach’s α values and reliability coefficients were checked for each of the four adaptive capacity dimensions; values above 0.70 were considered acceptable (Nunnally and Bernstein 1994). There is no established cutoff for reliability coefficients, which measure the proportion of an item’s variance explained through its latent dimension; however, the higher the coefficient, the more reliably it represents the underlying latent dimension (Kline 2011).
Convergent validity was tested by examining factor loadings and acceptable fit of the measurement model. Factor loadings greater than 0.60 were deemed acceptable (Hair et al. 2010). Model fit was determined through a confirmatory factor analysis using the relative χ2, the root-mean-square error of approximation (RMSEA), the comparative fit index (CFI), and the Tucker–Lewis index (TLI). Values less than 3.0 indicate an acceptable relative χ2 (Carmines and McIver 1981); values up to 5.0 may be acceptable in exploratory work (Schumacker and Lomax 2004). RMSEA values below 0.06 and values at or above 0.90 for both the CFI and TLI suggest the data fit the hypothesized model structure (Hu and Bentler 1999). Once convergent validity was demonstrated for the pooled dataset, the model was reestimated using data from each sampling effort (i.e., individual cities) individually. Multigroup estimation yields a configural model with summed χ2 statistics and reestimated fit indices that provide a baseline against which models with equality constraints can be compared. The measurement models’ measurement weights, intercepts, and covariances were constrained across datasets to test for measurement invariance (whether or not the scale is gauging the same latent constructs across groups). Invariance is supported if the CFI values between models changes less than 0.01 (Cheung and Rensvold 2002).
In the final test of construct validity, the perceived adaptive capacity scale’s discriminant validity was verified by analyzing factor correlations. Factor correlations below 0.80 reveal acceptable discriminant validity (John and Benet-Martinez 2000). All analyses of the measurement model were completed with IBM software SPSS Amos, version 22 (Arbuckle 2013).
3) Multivariate multiple regression
With reliability, convergent, and discriminant validity established, confirmatory factor analysis predictions for the four perceived adaptive capacity dimensions were saved and those factor scores were used as dependent variables in subsequent analysis. The political ideology variable and the 13 trust variables were included as independent variables. Controls for the basic sociodemographic characteristics of age, gender, education, and income were included. The multivariate multiple regression analysis was performed with the mvreg command in Stata 11 (StataCorp 2009).
4. Results
a. Descriptive statistics
Of the 900 surveys sent in the initial mailing, a total of 40 were returned undeliverable. Of the successfully delivered surveys, 420 were returned completed (Waynesville = 159, Franklin = 158, Spruce Pine = 103), which tabulates out to a 48.8% response rate. The sociodemographic characteristics of respondents did not differ significantly when compared to 2010 U.S. Census data on owner-occupied housing units within each city (Table 1).
Sociodemographic characteristics of respondents by study city.
Regarding political ideologies, just over one-third (33.9%) of all respondents identified as being conservative, with another fifth (20.3%) identifying themselves as moderate conservatives. Self-identified moderates accounted for 15.9% of all respondents. Nearly one-fifth (19.7%) identified as being moderate liberals, with a smaller proportion (10.2%) identifying as being liberal.
Descriptive statistics regarding where individuals obtain their information about local environmental issues, and whether or not they always trust the information from those sources, is reported in Table 2. The most used information sources in descending order were local television news (91.7%), local newspapers or periodicals (91.1%), national television news (89.3%), close friends (88.7%), and immediate family members (85.4%). Regarding the most trusted information sources, individuals placed the most trust in immediate family members, with 38.8% of respondents indicating they always trust information from their immediate family members (i.e., their bonding ties). This proportion is notably higher than any other source and was expected based on conceptual, nonempirical work focused on the role of social capital in climate change adaptation (Pelling and High 2005; Bodin and Crona 2009). Among the remaining information sources, individuals indicated they placed the most trust in close friends (25.3%), local newspapers or periodicals (25.0%), local churches (23.5%), and local television news (21.7%).
Descriptive statistics for use of, and complete trust in, different sources of information.
Table 3 displays the descriptive statistics for the statements included in the adaptive capacity scale. On average, respondents agreed with statements regarding risks associated with climate-related changes to local resources; mean levels of agreement were above 3.5 on the five-point scale, which ranged from strongly disagree (1) to strongly agree (5). Respondents were ambivalent regarding the ease at which information about potential threats to local forests could be found (means for the two corresponding statements were 2.82 and 2.73). However, respondents did indicate that learning about changes to local resources was important to them (M = 3.93). Respondents also indicated that they were willing to plan for increasingly severe weather conditions (M = 3.64) and changes to local forest conditions (M = 3.56). Finally, respondents believed they could adapt to increasingly severe weather conditions (M = 3.47) and cope with changes in forest-related industries (M = 3.42).
Psychometric properties of perceived adaptive capacity scale.
b. Political ideologies and trust in individual information sources
The first hypothesis posited that both conservatives and liberals would place more trust in immediate friends and family members relative to other information sources. This hypothesis was rejected. As can be seen in Fig. 1, conservatives and moderate conservatives placed more trust in friends contacted infrequently [χ2 (4) = 9.9, p < 0.05; χ2 (1) = 8.5, p < 0.001], educational groups [χ2 (4) = 10.6, p < 0.05; χ2 (1) = 10.9, p < 0.001], local churches [χ2 (4) = 26.4, p < 0.001; χ2 (1) = 23.3, p < 0.001], and online non-news sources such as blogs [χ2 (4) = 3.6 p = 0.464; χ2 (1) = 3.4, p < 0.05]. These results suggest that political ideologies are related to the trust placed in some, but not all, information sources.
Proportion of respondents by political ideology indicating they always trust specific information sources.
Citation: Weather, Climate, and Society 8, 1; 10.1175/WCAS-D-15-0046.1
c. Psychometric evaluation
The two-step process of addressing hypotheses 2–5 first called for an evaluation of the construct validity of the adaptive capacity scale. Tests for internal reliability of the scale’s intended measurement constructs suggested that two statement items should be removed as their inclusion resulted in a substantial decrement in α; recalculated α values were acceptable (Table 3). Evaluation of squared multiple correlation values indicate that a priori latent dimensions explain a substantial amount (>0.56) of the variance in their measurement items. Collectively, the α values and the high squared multiple correlations support the internal reliability of the scale’s four conceptual dimensions and their respective measurement items.
Convergent validity was tested through an examination of factor loadings and fit statistics for the four-dimensional measurement model. As shown in Table 3, all factor loadings were acceptable (>0.72). Fitting the correlation–covariance matrix of the data to the hypothesized measurement model revealed acceptable fit statistics (relative χ2 = 3.14; RMSEA 90% CI = 0.06–0.10; CFI = 0.96; TLI = 0.93). These results support the convergent validity of the adaptive capacity scale using the pooled dataset. The scale’s ability to consistently measure the four dimensions of perceived adaptive capacity across distinct samples was tested through multigroup invariance tests. The baseline configural fit statistics indicate an acceptable fit (relative χ2 = 2.37; RMSEA 90% CI = 0.05–0.08; CFI = 0.93; TLI = 0.87). Constraining the measurement weights, intercepts, and covariances to be equal across the three sample populations revealed no substantial decrement to model fit (i.e., changes in the CFI were less than 0.01), indicating that the scale can successfully measure the same latent constructs across groups and offers further support for the scale’s convergent validity.
Finally, discriminant validity was tested through an analysis of factor correlations (Table 4). All correlations were at or below 0.44 and nonsignificant, suggesting that the scale’s measurement items can adequately discriminate between the four intended latent constructs.
Factor correlations across perceived adaptive capacity dimensions.
d. Multivariate multiple regression
The results of the regression model used to test hypotheses 2–5 are presented in Table 5 and Fig. 2. The second hypothesis posited that individuals identifying as more liberal would perceive climate-driven environmental changes as more severe relative to conservatives. However, the data suggest just the opposite. Individuals identifying as more conservative tended to perceive climate-driven changes to local forest conditions as more severe after controlling for all other covariates [coefficient (Coef.) = −0.111, p = 0.024]. Consequently, hypothesis 2 was rejected.
Results from multivariate multiple regression estimates. Single asterisk denotes p < 0.05; double asterisk denotes p < 0.01.
Fitted estimates of individual adaptive capacity dimensions regressed on political ideologies (all other model covariates are held constant).
Citation: Weather, Climate, and Society 8, 1; 10.1175/WCAS-D-15-0046.1
The third hypothesis postulated that liberals would believe that information about climate-driven environmental changes is easier to find relative to conservatives. The results failed to yield any significant relationships between individuals’ political ideologies and the willingness to learn about potential climate-related impacts (Coef. = 0.015, p = 0.766). Hypothesis 3 was not rejected. Given previous empirical work by Kahlor and colleagues documenting a significant relationship between climate change knowledge, perceived information insufficiency, and the willingness to search for information about climate change (Kahlor 2007; Kahlor and Rosenthal 2009), the results are unexpected. Liberals, assuming they are more knowledgeable about climate change, would be expected to be more willing to learn about potential climate-related impacts to local forest resources. The findings suggest either that liberals and conservatives in the sample did not differ in their perceived level of information insufficiency or that the results of Kahlor, which were focused on global climate change impacts, do not extend to localized impacts to forest resources. Future research is needed across multiple scales of climate-related impacts to test this directly.
The fourth hypothesis suggested that liberals would be more willing to plan for climate-driven changes to local forest ecosystems relative to conservatives. This hypothesis was not rejected as the results failed to discern any significant relationships between individuals’ political ideologies and their willingness to plan for potential climate-related impacts to local forest resources (Coef. = −0.032, p = 0.535). Similarly, the fifth hypothesis was not rejected; it suggested that liberals would believe they are more capable of adapting to altered forest conditions driven by increasingly variable climate conditions relative to conservatives. The relationship between political ideology and the self-efficacy dimension of the perceived adaptive capacity scale was not significant (Coef. = −0.020, p = 0.689).
Further inspection of the results revealed several significant and notable findings. Irrespective of political ideology, individuals who completely trust the information they obtain from national television news were significantly more likely to perceive climate-driven changes to local forest conditions as more severe relative to individuals who either do not seek out, or do not always trust, information from national television news outlets (Coef. = 0.358, p = 0.050). The same relationship was found for individuals who completely trust the information they obtain from online news sources; they are more likely to perceive climate-driven changes to local forest conditions as more severe than individuals who do not seek out or do not always trust information from online news sources (Coef. = 0.405, p = 0.050).
The analysis revealed that individuals who place complete trust in the information they receive from local churches were less likely to believe that information about local climate-driven environmental changes is easy to find relative to those individuals who either do not obtain, or do not always trust, information from local churches (Coef. = −0.326, p = 0.050).
The analysis also revealed that individuals who place complete trust in the information they receive from local television news outlets are significantly less willing to plan for climate-driven changes to local forest conditions relative to individuals who do not obtain, or do not always trust, information from the local news media (Coef. = −0.446, p = 0.050). Conversely, individuals who place complete trust in national television news are significantly more willing to plan for climate-driven changes to local forest ecosystems when compared to those individuals who either do not use, or do not always trust, information from the national television media (Coef. = 0.511, p = 0.024).
Finally, the results suggest that women hold significantly stronger beliefs about their ability to adapt to altered forest conditions when compared to men (Coef. = 0.323, p = 0.014).
5. Discussion
This investigation set out to test five hypotheses focused on individuals’ political ideologies, their trust in distinct information sources, and the effects of both these factors on their perceived ability to adapt to climate-driven changes to local forest conditions. The analyses revealed some support for a potential relationship between political ideology and trust in specific information sources. Specifically, political ideologies were related to the trust placed in some, but not all, information sources. The results suggest that conservatives tended to place more trust in their infrequently contacted friends, educational groups, local churches, and online non-news sources such as blogs. This finding holds several implications for how information about climate-related impacts to local forest conditions are disseminated to the public. The analyses also revealed that conservatives tended to perceive climate-driven changes to local forest conditions as more severe than liberals. This finding suggests the relationship between political ideologies and attitudes toward climate-driven environmental changes should not be conceptualized as context independent. Rather, the spatial scale (e.g., local versus global) and context of the environmental change itself is likely to play a substantial role in risk perceptions.
a. Scientists and forest management agencies need to consider alternative outlets for disseminating climate-related information
The results have implications for scientists and resource management agencies seeking to disseminate climate change information more effectively. In creating new messages and fine-tuning how audiences are addressed, scientists and managers need to consider the extent to which individual information sources are trusted. By selecting specific information conduits for different target audiences, scientists and managers can speak to those individuals more directly, through media they understand and trust. More practically, scientists and forest management agencies need to establish an informal dialogue with multiple information outlets to ensure forest management and climate change–related information can be delivered through those outlets in the first place. Establishing a diverse array of connections through which information can be delivered will not only help forest management agencies to disseminate information about changes to resource conditions or important management issues, but it can also aid in building trust between the community and the agency. High levels of trust between forest management agencies and local communities is essential to successful public engagement and collaborative management processes (Lachapelle and McCool 2012; McFarlane et al. 2012; Olsen and Sharp 2013; Smith et al. 2013; Winter et al. 2004; Parkins and McFarlane 2015). With trusted lines of communication at their disposal, scientists and forest management agencies can focus more fully on designing messages tailored to specific groups’ needs and interests (Andreasen 1995; McKenzie-Mohr 2011).
Speaking exclusively to the data and study communities, scientists and forest management agencies responsible for the dissemination of information regarding how changes to long-term climate variability affect local resource conditions should target conservative populations through educational groups, local churches, and online non-news sources such as blogs; these sources were most frequently used and more trusted among those individuals with conservative ideologies. Scientists and forest managers should also be acutely aware of why information disseminated directly from state and federal agencies is not likely to be trusted and used. This is particularly true in this study area, where a large portion of the public distrusts state and federal agencies because of previous conflicts over social and environmental policies (Smith et al. 2012b). Alternative communication strategies do not have to be elaborate to be effective. The U.S. Department of Agriculture’s Forest Service, for example, has realized benefits in other forest-associated communities across the southeastern United States from simple efforts such as participating in local community events (festivals, parades, etc.) and allowing local youth and religious groups to use agency facilities (Wondolleck and Yaffee 2000; Smith 2012).
b. The effect of ideologies on attitudes is not universal
Unlike earlier studies, this research focused on individual perceptions of local climate change risks caused by changes to local forest conditions; this is distinct from the well-studied perceptions of general climate change risks. The results are unique in that they offer some insight into how to start communicating local risks to individuals with different political ideologies. For forest management agencies, targeting individuals with different ideologies through specific sources can result in individuals receiving information they are more likely to trust. Despite this implication, resource management agencies may struggle when attempting to disseminate climate-related information through nontraditional conduits such as civic and religious groups. Past research suggests small, amenity-rich communities are often very insular and unwilling to openly accept information from infrequently contacted and anonymous nonlocal groups such as federal or state forest management agencies (Lachapelle and McCool 2012; McFarlane et al. 2012; Olsen and Sharp 2013; Parkins 2010; Smith et al. 2012b).
The focus on location-specific climate-related risks also offers theoretical insights related to variations in climate-related risk perceptions between individuals who hold differing political ideologies. Previous research suggests general psychological and social differences between conservatives and liberals might alter their perceptions of climate change risks and influence their perceived ability to adapt (Jost et al. 2003, 2008; Graham et al. 2009; Choma et al. 2013). Using different information sources can help to ameliorate the consequences of those differences, but framing the messages in different ways can also have a substantial impact on whether or not the substantive information is perceived as valid (Nisbet 2009). For conservative individuals, who tend to be more concerned about perceived risks to themselves and similar others, framing a message about local climate change impacts could focus more on geographically proximate impacts and concerns. Asking about climate change–related risks in a geographically specific context might elicit conservatives’ general tendencies to be more concerned about threats to other individuals in their in-group. Similarly, framing a message about local climate change impacts through the lens of equity and the uneven distribution of impacts could influence liberals to consider climate-related messaging more carefully.
c. Limitations
There are several limitations to the inferences made from the analyses. First, it is uncertain whether self-identified conservatives will be concerned about the local impacts of climate change in other communities or regions. The three communities analyzed in this study were selected primarily because they are small (populations <10 000) and have undergone a relatively rapid transition from extraction-based economies (primarily mining and forestry) to recreation- and tourism-based economies. It is possible that residents of these amenity-transition communities are more sensitive to local environmental and economic changes than communities with more stable and diverse economies. Additionally, the southern Appalachia region is widely recognized to have a strong local cultural identity (Salstrom 1997). Local opposition toward several federal policies has contributed to strong and deeply rooted personal identities tied to the place (Smith et al. 2012b). It is possible that the communities’ strong local identities contribute to their high level of sensitivity to localized climate-related environmental changes. Residents of more transient communities adjacent to the study region, such as those in exurban areas around Atlanta and Charlotte, may not perceive or attach as much meaning to climate-related environmental risks. More focused investigations need to be conducted that either systematically select study communities along “length of residency” gradients (as opposed to the urban-to-rural gradients that are used more often by geographers and sociologists) or apply novel statistical scaling techniques to accurately estimate local attitudes and beliefs through nationwide surveys modeled in conjunction with U.S. Census data (Howe et al. 2015).
Another limitation of this study is an inability to identify the explicit information sources that individuals were using to obtain information about local environmental issues. Rather, this study was only able to quantify more general “types of information sources.” There is obviously a considerable amount of variability in the trust individuals place in explicit information sources within any one type of information source. Future research utilizing alternative data collection strategies, such as Internet-based surveys, can overcome this limitation and yield a substantially more detailed level of inference. Additional insights could also result from data collection methods that distinguish between specific information sources and information channels. Technologically integrated communication strategies allow for the same information to be disseminated via national television broadcasts, Facebook posts, Twitter feeds, blogs, and a whole host of other outlets. The amount of trust individuals place in information from a particular source might be mediated by the communication channel through which it is disseminated. Both theoretical and practical insights can be gained from research focused on this question.
6. Conclusions
This research set out to examine differences between conservatives and liberals in their use and trust of distinct information sources. The research also examined the relationship between individuals’ political ideologies and several key dimensions of perceived adaptive capacity. The analyses suggested that political ideologies were significantly related to some, but not all, of the information sources asked about. Specifically, the analysis revealed significant differences for several information sources (educational groups, religious groups, infrequent friends, and online non-news sources); differences among these sources offer potential avenues through which scientists and forest management agencies can deliver targeted climate-related information that meets the needs and interests of their constituents. The analysis also suggested that political ideologies are related to the perceived severity of local climate change–related risks. Conservatives in the study perceived climate-driven changes to resource conditions as more severe relative to liberals, which contradicts previous studies focused on more general or abstract climate-related risks. This finding highlights a need for future research to examine explicitly how the spatial scale, and the framing of impacts or affected groups, shape climate change risk perceptions.
Acknowledgments
This research was supported by a grant from the Sociology Program area within the U.S. National Science Foundation’s Social, Behavioral and Economic Sciences Directorate (Award #1030395).
REFERENCES
Andreasen, A. R., 1995: Marketing Social Change: Changing Behavior to Promote Health, Social Development and the Environment. Jossey-Bass, 368 pp.
Arbuckle, J. L., 2013: IBM SPSS Amos 22 user’s guide. IBM, 673 pp. [Available online at ftp://public.dhe.ibm.com/software/analytics/spss/documentation/amos/22.0/en/Manuals/IBM_SPSS_Amos_User_Guide.pdf.]
Bodin, Ö., and Crona B. I. , 2009: The role of social networks in natural resource governance: What relational patterns make a difference? Global Environ. Change, 19, 366–374, doi:10.1016/j.gloenvcha.2009.05.002.
Carmines, E. G., and McIver J. P. , 1981: Analyzing models with unobserved variables: Analysis of covariance structures. Social Measurement: Current Issues, G.W. Bohrnstedt and E.F. Borgatta, Eds., Sage Publications, 65–116.
Cheung, G. W., and Rensvold R. B. , 2002: Evaluating goodness-of-fit indexes for testing measurement invariance. Struct. Equation Model., 9, 233–255, doi:10.1207/S15328007SEM0902_5.
Choma, B. L., Hanoch Y. , Gummerum M. , and Hodson G. , 2013: Relations between risk perceptions and socio-political ideology are domain- and ideology-dependent. Pers. Individ. Differ., 54, 29–34, doi:10.1016/j.paid.2012.07.028.
Dale, V. H., and Coauthors, 2001: Climate change and forest disturbances. BioScience, 51, 723–734, doi:10.1641/0006-3568(2001)051[0723:CCAFD]2.0.CO;2.
Dillman, D. A., Smyth J. D. , and Christian L. M. , 2008: Internet, Mail, and Mixed-Mode Surveys: The Tailored Design Method. 3rd ed. Wiley, 512 pp.
Gittell, R., and Vidal A. , 1998: Community Organizing: Building Social Capital as a Development Strategy. Sage Publications, 206 pp.
Graham, J., Haidt J. , and Nosek B. A. , 2009: Liberals and conservatives rely on different sets of moral foundations. J. Pers. Soc. Psychol., 96, 1029–1046, doi:10.1037/a0015141.
Hair, J. F., Black W. C. , Babin B. J. , and Anderson R. E. , 2010: Multivariate Data Analysis. 7th ed. Prentice Hall, 816 pp.
Hamilton, L. C., and Keim B. D. , 2009: Regional variation in perceptions about climate change. Int. J. Climatol., 29, 2348–2352, doi:10.1002/joc.1930.
Howe, P. D., Mildenberger M. , Marlon J. R. , and Leiserowitz A. , 2015: Geographic variation in opinions on climate change at state and local scales in the USA. Nat. Climate Change, 5, 596–603, doi:10.1038/nclimate2583.
Hu, L., and Bentler P. M. , 1999: Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct. Equation Model., 6, 1–55, doi:10.1080/10705519909540118.
Iyengar, S., and Hahn K. S. , 2009: Red media, blue media: Evidence of ideological selectivity in media use. J. Commun., 59, 19–39, doi:10.1111/j.1460-2466.2008.01402.x.
John, O. P., and Benet-Martinez V. , 2000: Measurement: Reliability, construct validation and scale construction. Handbook of Research Methods in Social and Personality Psychology, H. T. Reis and C. M. Judd, Eds., Cambridge University Press, 339–369.
Jones, R. E., Fly J. M. , Talley J. , and Cordell H. K. , 2003: Green migration into rural America: The new frontier of environmentalism? Soc. Nat. Resour., 16, 221–238, doi:10.1080/08941920309159.
Jost, J. T., Glaser J. , Kruglanski A. W. , and Sulloway F. J. , 2003: Political conservatism as motivated social cognition. Psychol. Bull., 129, 339–375, doi:10.1037/0033-2909.129.3.339.
Jost, J. T., Nosek B. A. , and Gosling S. D. , 2008: Ideology: Its resurgence in social, personality, and political psychology. Perspect. Psychol. Sci., 3, 126–136, doi:10.1111/j.1745-6916.2008.00070.x.
Jost, J. T., Federico C. M. , and Napier J. L. , 2009: Political ideology: Its structure, functions, and elective affinities. Annu. Rev. Psychol., 60, 307–337, doi:10.1146/annurev.psych.60.110707.163600.
Kahlor, L. A., 2007: An augmented risk information seeking model: The case of global warming. Media Psychol., 10, 414–435, doi:10.1080/15213260701532971.
Kahlor, L. A., and Rosenthal S. , 2009: If we seek, do we learn? Predicting knowledge of global warming. Sci. Commun., 30, 380–414, doi:10.1177/1075547008328798.
Kline, R. B., 2011: Principles and Practice of Structural Equation Modeling. 3rd ed. Guilford Press, 427 pp.
Lachapelle, P. R., and McCool S. F. , 2012: The role of trust in community wildland fire protection planning. Soc. Nat. Resour., 25, 321–335, doi:10.1080/08941920.2011.569855.
Lee, T.-T., 2010: Why they don’t trust the media: An examination of factors predicting trust. Amer. Behav. Sci., 54, 8–21, doi:10.1177/0002764210376308.
Linacre, J. M., 2002: Optimizing rating scale category effectiveness. J. Appl. Meas., 3, 85–106.
Logan, J. A., Régnière J. , and Powell J. A. , 2003: Assessing the impacts of global warming on forest pest dynamics. Front. Ecol. Environ, 1, 130–137, doi:10.1890/1540-9295(2003)001[0130:ATIOGW]2.0.CO;2.
Marsden, P. V., and Friedkin N. E. , 1993: Network studies of social influence. Sociol. Methods Res., 22, 127–151, doi:10.1177/0049124193022001006.
Marshall, N. A., Fenton D. M. , Marshall P. A. , and Sutton S. G. , 2007: How resource dependency can influence social resilience within a primary resource industry. Rural Sociol., 72, 359–390, doi:10.1526/003601107781799254.
McCright, A. M., 2011: Political orientation moderates Americans’ beliefs and concern about climate change. Climatic Change, 104, 243–253, doi:10.1007/s10584-010-9946-y.
McCright, A. M., and Dunlap R. E. , 2011: The politicization of climate change and polarization in the American public’s views of global warming, 2001–2010. Sociol. Quart., 52, 155–194, doi:10.1111/j.1533-8525.2011.01198.x.
McFarlane, B. L., Parkins J. R. , and Watson D. O. T. , 2012: Risk, knowledge, and trust in managing forest insect disturbance. Can. J. For. Res., 42, 710–719, doi:10.1139/x2012-030.
McKenzie-Mohr, D., 2011: Fostering Sustainable Behavior: An Introduction to Community-Based Social Marketing. 3rd ed. New Society Publishers, 192 pp.
Newfont, K., 2012: Blue Ridge Commons: Environmental Activism and Forest History in Western North Carolina. University of Georgia Press, 400 pp.
Nisbet, M. C., 2009: Communicating climate change: Why frames matter for public engagement. Environment, 51, 12–23, doi:10.3200/ENVT.51.2.12-23.
Nunnally, J., and Bernstein I. , 1994: Psychometric Theory. 3rd ed. McGraw-Hill, 736 pp.
Olsen, C. S., and Sharp E. , 2013: Building community–agency trust in fire-affected communities in Australia and the United States. Int. J. Wildland Fire, 22, 822–831, doi:10.1071/WF12086.
Page, B. I., and Shapiro R. Y. , 1992: The Rational Public. University of Chicago Press, 506 pp.
Parkins, J. R., 2010: The problem with trust: Insights from advisory committees in the forest sector of Alberta. Soc. Nat. Resour., 23, 822–836, doi:10.1080/08941920802545792.
Parkins, J. R., and McFarlane B. L. , 2015: Trust and skepticism in dynamic tension: Concepts and empirical refinements from research on the mountain pine beetle outbreak in Alberta, Canada. Hum. Ecol. Rev., 21, 133–153.
Pelling, M., and High C. , 2005: Understanding adaptation: What can social capital offer assessments of adaptive capacity? Global Environ. Change, 15, 308–319, doi:10.1016/j.gloenvcha.2005.02.001.
Salstrom, P., 1997: Appalachia’s Path to Dependency: Rethinking a Region’s Economic History, 1730–1940. University Press of Kentucky, 240 pp.
Schumacker, R. E., and Lomax R. G. , 2004: A Beginner’s Guide to Structural Equation Modeling. 2nd ed. Psychology Press, 536 pp.
Schwartz, S. H., 1994: Are there universal aspects in the structure and contents of human values? J. Soc. Issues, 50, 19–45, doi:10.1111/j.1540-4560.1994.tb01196.x.
Schwartz, S. H., 2012: Basic personal values and political orientations. Improving Public Opinion Surveys: Interdisciplinary Innovations and the American National Election Studies, J. H. Aldrich and K. M. McGraw, Eds., Princeton University Press, 63–82.
Smith, J. W., 2012: Barriers and bridges to U.S. Forest Service—Community relationships: Results from two pilot tests of a rapid social capital assessment protocol. Forests, 3, 1157–1179, doi:10.3390/f3041157.
Smith, J. W., 2013: Adaptive participation in forest planning contingent upon a hypothetical large-scale forest disturbance. For. Sci., 59, 637–648, doi:10.5849/forsci.12-043.
Smith, J. W., Anderson D. H. , and Moore R. L. , 2012a: Social capital, place meanings, and perceived resilience to climate change. Rural Sociol., 77, 380–407, doi:10.1111/j.1549-0831.2012.00082.x.
Smith, J. W., Moore R. L. , Anderson D. H. , and Siderelis C. , 2012b: Community resilience in southern Appalachia: A theoretical framework and three case studies. Hum. Ecol., 40, 341–353, doi:10.1007/s10745-012-9470-y.
Smith, J. W., Leahy J. E. , Anderson D. H. , and Davenport M. A. , 2013: Community/agency trust and public involvement in resource planning. Soc. Nat. Resour., 26, 472–477, doi:10.1080/08941920.2012.742606.
StataCorp, 2009: [r] mvreg. Stata: Release 11, StataCorp LP, College Station, TX, 1121–1127.
Stimson, J. A., 2004: Tides of Consent. Cambridge University Press, 208 pp.
Winter, G., Vogt C. A. , and McCaffrey S. , 2004: Examining social trust in fuels management strategies. J. For., 102, 8–15.
Wondolleck, J. M., and Yaffee S. L. , 2000: Making Collaboration Work: Lessons from Innovation in Natural Resource Management. Island Press, 304 pp.
Woolcock, M., 1998: Social capital and economic development: Toward a theoretical synthesis and policy framework. Theory Soc., 27, 151–208, doi:10.1023/A:1006884930135.
