MA Erasmus Mundus Master Journalism, Media and Globalisation
(joint degree)
Climate Change, Framing and the Role of Socio-Cognitive Perceptions
in Predicting Pro-Environmental Behavior
by Paris Bethel Student ID: 11896531
Master’s Thesis
Graduate School of Communication Master’s programme Communication Science
Supervisor/Examiner: Lukas Otto Date of completion: 31st May 2017
Abstract
Using an online survey experiment, this study examines how episodic and thematic framing may affect an individual’s socio-cognitive perceptions of climate change as well as their intention to participate politically, change behavior and support policy. Overall, results find limited effects of framing on socio-cognitive perceptions and pro-environmental behavior. However, insights derived from spurious moments of causality serve to reinforce the body of literature linking one’s perception of risk and adaptability to their intention to engage in environmental behavior. Limited impacts suggest that the contentious position of climate change within the United States, as well as its unique relationship to political ideology, make perception of the issue hardlined and difficult to influence. This is discussed from a communicative standpoint, where implications are address and suggestions for future research are made.
Key Words:
Introduction
Within the United States, the perception of climate change has fluctuated over tumultuous grounds during the past two and a half decades. Though identified as a real and measurable phenomenon linked to human activity in the early 1990’s, early-set claims of uncertainty diluted the urgency of climate change within American politics and stunted public perception (Nisbet, 2009). The issue was swept up along political and ideological lines before a base of public understanding could be established, leading conservatives and liberals to position themselves on either side of the scientific consensus (Dunlap, McCright & Yarosh, 2016; Nisbet 2009. For this reason, Liberals within the U.S. overwhelmingly view climate change as a risk and support action, while Conservatives stand in opposition to policy and hold comparatively low levels of concern (Dunlap et al., 2016; Nisbet, 2009). For this reason climate change has become one of the most contentiously debated issues within the U.S., alongside ideologically defining topics such as taxes and gun control (Nisbet, 2009). Today, despite the international discussion on climate change shifting from scientific consensus to exploring the best mitigative and adaptive solutions, the topic remains at a standstill within the U.S.. Falling behind in the international sphere, the U.S. has not passed any significant policy aimed at curtailing emissions and it is unclear if it will successfully do so in the future. This lack of agenda is most evident by the U.S. government’s recent announcement that it will withdrawal from the Paris Climate Accord, a first-of-its-kind international commitment to reducing global CO2 levels with over 195 participating member countries.
Under this backdrop, public perception surround climate change in the U.S. has garnered a lot of research attention, with scholars interested to understand how and why perception of the
issue became so ideologically split. Over time, comprehensive research has identified the media as playing a key role in defining climate change among the public, finding it has amplified polarization and contributed to poor perception among the public (Feldman, Hart & Milosevic, 2017; Hart & Nisbet, 2012; Nisbet 2009). Early coverage of climate change in the 2000’s gave equal consideration to ‘contrarian’ views, which fueled a false impression to the public that the scientific consensus was not overwhelmingly agreed upon among scientists (Nisbet, 2009). While a shift occurred in the following decade, with less media questioning the scientific consensus of climate change, this did not result in an overall improvement of climate change communication. Today the impacts of climate change and the actions to address it are
inconsistently communicated (Feldman, Hart & Milosevic, 2017; Hart & Feldman, 2014).
Additionally, mitigative and adaptive actions on climate change are primarily framed by political conflict, focusing on how Democrats and Republicans are in disagreement over policy (Hart & Feldman, 2014). This type of media coverage is disengaging for the public and undercuts the salience of addressing climate change. More importantly, as Feldman et al. (2017) point out, it risks portraying climate change as an unmanageable problem whose solutions are not possible because of unyielding political disagreement.
For this reason, framing research today looks to examine the ways in which varying messages on climate change affect public perception and support for action. In doing so, it is the intention of scholars to develop communicative strategies which can be used to better convey climate change to the public. To understand public perspective, it is not enough to simply look at how political ideology increases or decreases framing effects, it is also important to examine the interplay of subjective attitudes, motivations and beliefs. Also understood as socio-cognitive
perceptions, these factors serves to explain how an individual views and interacts within their social environment. Research demonstrates that, alongside political ideology, these values contribute to the public’s perception on climate change, helping to explain their willingness to participate in pro-environmental behavior (Hart & Feldman, 2016; Sapiens, Beeton & Walker, 2016); Hart, 2011). The present study seeks to contribute to this growing body of research, examining how episodic and thematic framing affects an individual’s socio-cognitive perceptions of climate change as well as their intention to participate politically, change behavior and support policy. It draws from literature on framing as well as the Model of Private Proactive Adaptation to Climate Change (MPPACC), which conceptualizes risk appraisal and adaptation appraisal as two key socio-cognitive measures which impact decision making. Risk and adaptation appraisal are linked to motivating individuals to engage in a specific behavior and driving the belief that that behavior can be achieved (Grothmann & Patt, 2005). This study is the first, that the author is aware of, to analyze the MPPACC within a framing experiment.
Literature Review
Framing in Political Communication
Politically, framing theory derives from the position that the mass media plays a key role in influencing both the political preferences of the public and their general perception of social problems (Schäfer & O'Neill, 2017; Hurtíková, 2013; Druckman, 2001). Conceptualized, framing is defined as the communicative representation of an issue or topic in which some aspects are selectively emphasized as more significant while others are not (Schäfer & O'Neill, 2017; Nisbet, 2009). From a political perspective, framing caters to the selective promotion of a “particular definition, causal interpretation, moral evaluation, and/or treatment recommendation
(Entman, 1993, p. 52).” Such framing emphasize, contextually alters the importance of considerations regarding an issue, and can thus cause individuals to focus on the emphasized considerations when building their own opinions on the issue presented (Druckman, 2001). From a political communications approach, the implications of framing are key as they carry the ability to alter the level of importance attached to particular beliefs (Druckman, 2001). As discussed by Hart and Feldman (2014) in their climate change research, the influence that framing holds over issue evaluation, makes it an essential tool for both the communication for and promotion of public and civic engagement when it comes to climate change.
From a theoretical perspective, framing exists within a wide arena— for climate change the most explored frames have been content-oriented (Schäfer & O'Neill, 2017). This means frames which refer to general interpretations of how to view or angle climate change as an issue, like those adapted by Hart and Feldman (2014) in their content analysis of U.S. broadcast and newspaper coverage: environment, public health, the economy and (political) conflict.
Thematic and Episodic Framing
Less explored in the context of climate change is framing through the formal/stylistic lense, which pertains to the general structural presentation of communication on an issue (Schäfer & O'Neill, 2017; Gross, 2008; Hart, 2011). Stylistic framing differentiates itself from the content-oriented approach through its use of contrast— such as ‘diagnostic’ versus
‘prognostic’ or ‘loss’ versus ‘gain’(Schäfer & O'Neill, 2017). A key stylistic frame used within political communication, but less so explored in the context of climate change, is the use of
episodic versus thematic framing (Schäfer & O'Neill, 2017; Hart, 2011). There are varied approaches to defining episodic and thematic framing, though the most prominent definition in
academia has been built off Iyengar’s (1991, 1996) research. Iyengar (1991; 1996) defines episodic frames as those which present an issue by offering a specific example or instance— one which is generally isolated from a broader historical context. While thematic framing places emphasis on the wider contextualization of an issue such as through a historical or geographical lense, with no one specific example used to carry the message (Iyengar, 1991; 1996). What is most unique about communicating an issue either episodically or thematically are there effects on attribution of responsibility (Iyengar 1991, 1996; Kim, 2015; Hart, 2011).
The first main takeaway from research on these frames, is that episodic coverage tends to lead audiences to interpret social problems as individually derived, while thematic coverage leads individuals to view issues as rooted in society (Iyengar, 1991; 1996). The second takeaway is that episodic coverage tends to spur attribution of responsibility for a social issue on the individual, while thematic coverage tends to cause individuals to attribute responsibility for a social issue on society, or the government (Kim, 2015; Hart, 2011; Iyengar, 1991, 1996). These distinct features of responsibility attribution, as they pertain to episodic and thematic framing, is why they tend to be referred to interchangeably as responsibility frames (Kim, 2015). Explored by Kim (2015), responsibility framing offers two contrasting views for social
problems. The first is that the issue is caused mostly by individuals, and thus efforts to address it should be on the behaviors of individuals (Kim, 2015). The second view holds that the issue is due to “flaws in social conditions,” thus requiring societal solutions such as through policy implementation (Kim, 2015, p. 554). Therefore, episodic and thematic framing overlap
conceptually with responsibility framing in how they contrast responsibility and lead individuals to determine the causes and solutions of a problem either individually or societally (Kim, 2015;
Hannah & Cafferty, 2006; Iyengar, 1996).
According to Iyengar (1991), general news reporting tends to be overwhelmingly episodic. For climate change, because research is usually content-oriented and not viewed thematically or episodically, there are not many studies which can speak to this contrast of responsibility orientation in the media today (Schäfer & O'Neill, 2017; Howell et al., 2016). Only two known content analyses on climate change have addressed formal-stylistic framing as it relates to responsibility attribution. A U.S. broadcast news analysis by Hart and Feldman found that climate change impacts and potential solutions were only discussed together 39.8% of the time, while 87.3% of broadcasts that did not discuss potential solutions, talked about impacts (2014, p. 336). A follow up study on U.S. newspapers found similar results, with impacts and actionable solutions only discussed together in articles 36.4% of the time (Feldman et al., 2017, p. 490). Schäfer and O'Neill (2017) argue such inconsistency points towards a stronger episodic representation of climate change in the media, as omitting potential solutions isolates broader context of the issue and fails to attribute societal responsibility. One known experimental study on climate change has examined how episodic and thematic frames influence support for policy and support for individual behavior change when mediated by individual treatment responsibility and government treatment responsibility (Hart, 2011). When mediated by attribution of
government responsibility, the study found that thematic framing over episodic framing significantly increased support for policy, but found no mediated significance for support for individual behavior change (Hart, 2011).
While individual behavior changes are a positive step in the direction of addressing climate change, the most impactful solutionary approaches are very much dependent on
societally spurred and government led action. Past studies on episodic and thematic framing has demonstrating that thematic frames are useful in depicting broad contextualization and in evoking sentiments of societal responsibility. Yet, the content analyses discussed in this review of literature revealed that it is primarily episodic coverage which dominates the discussion of climate change in the media. Given the gap in literature, is important for climate change related research to further explore the responsibility oriented influences of episodic and thematic framing as they pertain to pro-environmental behaviors.
Attitudes and beliefs, A Socio-Cognitive Approach
Framing literature on climate change has been widely explored, both in content analyses and experimental form. For experimental research, aside from framing selection, identifying key variables which will help to explain the relationship between framing effects and
pro-environmental behaviors is imperative. Therefore, this study is not only interested in exploring how framing effects pro-environmental behaviors, but also to see the role that beliefs and values play in influencing that dynamic. The reason for observing such interactions is that objective factors to addressing climate change do not take into account whether or not countries will directly translate their maximum capacity threshold into action (Winkler et al., 2007). Just because an individual, community or government can take certain actions, it does not mean they
will choose to do so— which is why subjective ability, built of values and beliefs, varies greatly
from objective ability (Grothmann et al., 2013; Winkler et al., 2007). This is key when
considering representation of climate change as an issue within the United States, as discussed within the introduction. As the largest economy in the world, the U.S. has the financial and institutional capabilities to mitigate and adapt to climate change— yet the government remains
conflicted and slow moving on taking any steps to do so. Objective considerations do not take into account this social phenomenon, which is incorporating social context is so important to understanding the perception the public has on taking action.
To achieve this, a model developed and conceptualized by Grothmann and Patt (2005) will be used to evaluate these social barriers: Model of Private Proactive Adaptation to Climate Change (MPPACC). The model was adapted from the Protection Motivation Theory (PMT), originally created to assess health threats by Rogers (1983), the PMT holds two key
socio-cognitive variables which Grothmann and Patt (2005) applied to the purposes of climate change: risk appraisal and adaptation appraisal (Rogers, 1983; Grothmann et al., 2013). Risk appraisal determines an individual’s cognitive perception of risk, while adaptation appraisal determines an individual’s perception of their own abilities to adapt as it pertains to an identified risk. Both of these measures are necessary for an individual to engage in an adaptive response (Grothmann et al., 2013; Grothmann & Patt, 2005; Rogers, 1983). The purpose of these
measures is to examine the subjective constraints of human perspective on objective conditions. This is relevant to climate change, as Grothmann et al. (2005, p. 3375) highlight, because “the objective ability or capability of a human actor only partly determines whether that actor will take an adaptive response.”
These variables serve to help explain the tendencies which lead some individuals to support adaptive environmental behavior and others to not. The following section outlines both perception of risk and perception of adaptive capacity. Important to note is that Grothmann et. al (2013) refer to them as adaptation motivation and adaptation belief. Risk perception is renamed ‘adaptation motivation’, and perceived adaptive capacity is referred to as ‘adaptation belief’,
because of the a wide range of literature which has confirmed risk perception to motivate
adaptive responses and perceived adaptive capacity to correlate with one’s belief they can adapt. Grothmann et al. (2013) cite reasoning for renaming these concepts for ‘easier
understandability’, as they more directly pertain to how individuals interpret the feasibility and effectiveness of adaptive and mitigative measures. After conceptualizing both terms further, this study will refer to them both as adaptation motivation and adaptation belief throughout the methods, results and discussion sections.
Adaptation Motivation
A key influencer and determinant of adaptive environmental responses, Adger et al. (2007) defines risk perception as a social limitation, as perception is dependent on whether climate change impacts, experienced or future-expected, are (a) perceived as a risk and (b) can be acted on. Studies have found that the public perception of risk associated with climate change is fundamental in motivating and driving policy support (Leiserowitz, 2006; Adger et al., 2007). There are certain elements which make individuals more or less likely to respond to an issue— such as perceiving a risk to be immediately relevant and personally implicating (Singh et al., 2017). These concepts of immediacy and personal relevance are important to climate change because the largest impacts are projected to be in the future, and are in many cases not characterized as being identifiably relevant yet. This can defined by the term, psychological
distance, which refers to an individual’s perception of how far removed a risk is from them being
able to directly experience it. Thus, as theorized by both Grothmann et al. (2017) and Singh et al. (2017, p. 94), perception of risk provides the “motivation and preferences for action” when it comes to adaptive and mitigative responsiveness. However, the nuances which drive these
motivations to support policy are still being unpacked, as one study within the U.S. has challenged the relevance of psychological distance. Leiserowitz (2006, p. 53) found that only 13% of respondents were concerned about the impacts of climate change on ‘themselves, their family or their local community’— instead showing greater concern for non-human nature and ‘geographically distant’ peoples. However, despite low perception scores at the local and community level, and moderate perception for distant places and non-human nature, risk perception was positively associated with a high level of support for national regulation and climate policy (Leiserowitz, 2006). Meaning that individuals do not need to associate the risk of climate change with their own locality in order to demonstrate supportiveness for policy.
Outside of policy support, the role of risk perception in motivating environmental behaviors is less explored, but literature theorizes that adaptive decisions are in general
influenced by perception of risk (Singh et al., 2017; Grothmann and Patt, 2005). This study aims to expand on this through examining the role risk perception plays in influencing political participation and intended behavior change, in addition to policy support. The MPPACC evaluates two key indicators which guide risk appraisal and will be used in this study to
operationalize adaptation motivation. The first, ‘perceived probability’ refers to an individual’s expectancy that they themselves will be exposed to a threat (Grothmann & Patt, 2005, p. 203). While the second, ‘perceived severity’, determines an individual’s expectation of how severe the consequences of a threat will be and whether or not they will affect things she or he values (Grothmann & Patt, 2005, p. 203).
Adaptation Belief
adaptation belief, which pertains to one’s perception of being able to adapt to changes brought about by a particular situation or issue. To conceptualize, adaptation belief requires a number of dimensional considerations, which Grothmann and Patt (2005), break down into three main components. The first is adaptation efficacy, synonymous with both response efficacy and outcome efficacy. These terms relate to an individual’s belief that certain actions or responses will be more or less effective in addressing an issue or threat, such as climate change
(Grothmann et al., 2013; Grothmann & Patt, 2005). The second dimension is self efficacy, and refers to a person’s perceived ability that they themselves can respond to and carry out an
adaptive response, such as changing one’s behavior (Grothmann et al., 2013; Grothmann & Patt, 2005). The third dimension of adaptation belief, is perceived adaptive costs. This interprets assumed costs of taking a certain adaptive response, whether it be monetarily based or rooted in personal time or effort. For the purposes of this study, perceived cost will be omitted from the measurement of adaptive belief because, as Grothmann and Patt (2005, p. 203) point out, it shares a close relationship with self efficacy.
Measures of adaptation efficacy and self efficacy are widely used within environmental and climate change research, as studies continuously find them to be highly relevant towards influencing pro-environmental behavior (Feldman et al., 2017; Hart & Feldman, 2016a; Hart & Feldman, 2016b). In a climate change experiment examining the influence of efficacy messages on efficacy beliefs and political participation, Hart and Feldman (2016b) found self, response and outcome efficacy to all be positively associated on intentions to participate politically. In a follow up study, Hart and Feldman (2016a) tested the effects of climate change-related text and imagery on behavior change and political participation when mediated by perceived efficacy.
Here too, indirect effects of image and text on intended behavior change and political
participation were found significant when mediated by perceived efficacy. This study seeks to expand on the role efficacy plays in motivating environmental behaviors. Rather than measuring self, response and outcome efficacy separately, or relying on one statement to measure perceived efficacy as did Hart and Feldman (2016a)— this study will use the MPPACC model
conceptualization of adaptation belief, which more collectively considers each of these efficacies in its overall score.
Current Study
This study seeks to evaluate the ways in which willingness to support governmental policy, intended political participation, and individual behavior change may be affected by the episodic and thematic presentation of adaptive solutions to climate change. The episodic and thematic conditions will be compared to a control condition in which no media is provided. Aside from examining direct effects, this study will also look at indirect effects, testing for mediation with adaptive motivation and adaptive belief and moderated mediation by also measuring political ideology and climate change concern. These attitudes, beliefs and motivations— otherwise known as socio-cognitive perceptions— are of key interest to the improvement of communicating climate change, as they are changeable barriers to adaptation rather than fixed limitations (Grothmann et al., 2013). Assessing such factors in the context of media framing can therefore be useful in adjusting and improving communications oriented at motivating adaptive responses towards climate change. Thus, the following research questions and hypotheses will be examined:
RQ1: How do the framing conditions differ in their influence on intended political participation,
intention to change behavior, and policy support?
H1: A thematic frame will lead to higher willingness to support climate change policy policies
and intended political participation than an episodic frame, while an episodic frame will lead to higher willingness to change behavior than a thematic frame.
H2: There will be no difference in framing influence on intended political participation, intended
behavior change and policy support across the conditions.
RQ2: How do adaptation motivation and adaptation belief differ in their influence on intended political participation, intended behavior change and policy support.
RQ3: What role does political ideology play in influencing adaptation motivation and adaptation
belief?
Conceptual Map
Methods
ParticipantsThis study sample focused on U.S. millennials. According to the Pew Research Center the cut-off age range for millennials is 22-37, born between the years of 1981-1997 (Dimock, 2018). While focus on a singular generation is uncommon among media effects research on
climate change, such a distinction is of analytical interest. Millennials surpassed Baby Boomers as the largest generation in 2016, and carry unique political characteristics (Dimock, 2018). For example, they are more likely than any other generation to skew politically to the left, and they had their largest voting turnout for the 2016 election, with 34 million casting their votes
compared to 18.4 million in the 2008 (Fry, 2017). In 2020 the millennial vote is expected to surpass Gen X (Fry, 2017). Millennials are picking up momentum in the political sphere, driving the decision for this study to focus on their characteristics and positioning as it pertains to
climate change. The sample for this study was collected via online recruitment, primarily through Facebook and Linkedin.
Stimuli
The design of this study comprised of two experimental conditions and a no-message control condition. Out of 191 completed surveys, 34 responses were deleted for either (a) failing to complete the survey in totality (b) being outside of the 22-37 age group and/or (c) failing the manipulation check— leaving a final total of 157 participants for the study. Each was randomly assigned to one of the three conditions, resulting in (N:57) in the control, (N:44) in episodic, and (N:56) in thematic. Participants assigned to one of the two stimulus conditions read an article on the connection between U.S. consumption habits and climate change. Each article was
constructed specifically for the purposes of this study, built from information by a number of sources including Scientific American, National Geographic and a 2015 report on household consumption by the Journal of Industrial Ecology. Guidelines for differentiating the frames episodically versus thematically were taken by Hart (2011), Iyengar (1991, 1996) and Kim (2015). The episodic condition discussed the importance of individual lifestyle adaptation:
(byline) “Ties between consumption habits, the environment, and climate change lead one
American to change her ways”. While the thematic condition provided statistics about the impact of consumption habits and prioritized the importance of policy in shifting lifestyle adaption: (byline) “Ties between American consumption and the environment are likely to strain the U.S.’s ability to address climate change.” The complete text for these stimuli, and the complete scales for each of the variables below can be found in the appendix.
Independent Variable. The independent variable in this study was the experimental condition
which participants were assigned to randomly, episodic (N = 44), thematic (N = 56) and control (N = 57).
Control Variables. The sample was 61.5% Female, 37.4% Male and 1.1% Non-binary, while the
average age was 26 years old (SD = 3.31). Six levels of education were measured: (1) High school, (2) Trade/Technical/Vocational Training, (3) Associates degree, (4) Bachelors degree, (5) Masters degree and (6) PhD, with the average level completed being a bachelor's degree (M:
4.62 SD = 0.9).
Manipulation Check. Attribution of responsibility was used as a manipulation check following
the stimulus. Participants were asked to answer the following on a 7-point scale (1 = strongly disagree; 7 = strongly agree): ‘this article placed responsibility on the government to take action and address climate change.’ Before conducting tests, data for 16 participants was omitted for failure of the manipulation check. Failure of the manipulation check occurred if someone
exposed to a thematic frame disagreed that the article placed responsibility on the government, or was exposed to an episodic theme and agreed that the article placed responsibility on the
Moderating Variables. Political Ideology was measured following the approach of Hart and
Feldman (2016), asking participants to answer how they view their orientation when it comes to politics on a 7-point scale (1 = strongly conservative; 7 = strongly liberal) (M = 5.61; SD: 1.24).
Climate Change Concern was measured following the 2-statement concern score by Howell, Capstick and Witmarsh (2016): asking participants the extent to which they worry about climate change and find the issue important (1 = not at all; 7 = extremely) (𝛼 = 0.83).
Mediating Variables. Adaptation Motivation. Operationalization of this variable followed the
definition by Grothmann and Patt (2005; Grothmann et al., 2013) who determined two parts necessary: (1) perceived severity, which assesses an individual’s expectancy of how serious the consequences of a threat are— both current and anticipated (2) perceived probability, tied to personal relevance and assesses an individual’s expectancy that they themselves, their community or country will be exposed to and affected by the threat. To operationalize this concept, a scale by Leiserowitz (2006) was shortened and adapted, asking participants the level of seriousness they equated with a number of statements (1 = not at all serious; 5 = extremely serious) (𝛼 = 0.82).
Adaptation Belief. Operationalization of this variable also followed the definition of Grothmann and Patt (2005, Grothmann et al. 2013), which attribute the main determinants of one’s adaptation belief to perceptions of self-efficacy, outcome-efficacy and control beliefs. All statements were asked using a 7-point scale (1 = strongly disagree; 7 = strongly agree). Four self-efficacy measures were adapted from Lauren et al. (2017), three outcome-efficacy measures were adapted from Feldman and Hart (2016), and two control-efficacy measures were created specifically for this variable, adapting the ‘confident’ versus ‘capable’ statement presentation
used by Lauren et al. (2017) in their self-efficacy scale: ‘I feel confident that the Federal
government will take action...’ versus ‘I feel the Federal government is capable of finding ways to take action....’ (𝛼 = 0.71).
Dependent Variables. Intended political participation was measured using a 5-question scale by Feldman & Hart (2016), no adaptations were applied. Participants were asked how likely or unlikely it was that they would personally engage in an action in the next 12 months (1 = extremely unlikely; 7 = extremely likely): ‘contact government officials to urge them to take action to reduce climate change (Feldman & Hart, 2016).’ (𝛼 = 0.85).
Intended behavior change was adapted from Hart and Feldman’s (2016) four-statement operationalization, and expanded on for a total of 8 statements: ‘Drive less by walking, biking, or taking public transportation more,’ ‘Bring reusable bag to the grocery store,’ ‘Cut down on meat consumption,’ and ‘Buy something second-hand (piece of clothing or household item).’ (𝛼 = 0.71).
Policy Support was conducted following the measurement of a 5-question scale by Singh et al. (2017), asking participants to determine how much they would support climate change related policy on a 5-point scale (1 = not at all; 5 = a great deal). Two of the five questions were adapted to be more closely oriented towards the stimuli articles on consumption habits: ‘a federal policy that provides you with subsidies or other incentives to increase your ability to lower your carbon footprint’. (𝛼 = 0.83).
Analysis
participation (Y1), intended behavior change (Y2) and policy support (Y3). Following this, a mediation analysis was conducted to assess the indirect effects of mediators adaptation motivation (M1) and adaptation belief (M2), as outlined by Hayes (2013) using PROCESS macro. This test was carried out to examine indirect effects of the frame conditions on support for policy, behavior change and intended political participation through the mediators.
Bootstrapped interval calculations were used to identify indirect effects, at a 95% confidence interval, 5,000 sampling iterations were used. The third step is a continuation of the previous, conducting a moderated mediation analysis to further assess direct and indirect effects with added moderators: political ideology (W) and climate change concern (Z).
Preceding the first step, data was assessed for outliers, normal distribution and
homogeneity of variances. No outliers in the data were found, and normal distribution between framing conditions was assessed using the Shapiro-Wilk test (support for policy: thematic, p = .056; episodic, p = .046; control, p = .023) (Intended behavior change: thematic, p = .001; episodic, p = .068; control, p = .025) (intended intended political participation: thematic, p = .001; episodic, p = .001; control, p < .001). Despite abnormal distribution, the ANOVA test was run as the deviations from normality were moderate. Lastly, homogeneity of variances, as assessed using the Levene's test for equality of variances (policy support; p = .782) (intended behavior change: p = .469) (intended political participation: p = .441). To run mediation and moderated mediation within PROCESS macro, indicator (dummy) variables were created in order to better distinguish between the three treatment groups. The episodic indicator compared against thematic and control conditions, while the thematic indicator compared against the episodic and control conditions.
Results
Political ideology for this sample skewed liberally. Participants reported high levels of climate change concern, adaptation motivation and adaptation belief, which led to high resulting scores of each of the dependent variables.
Table 1. Variable Means by framing condition.
Note: All variables are on a 7-point scale (1 = strongly disagree, 7 = strongly agree; 1 = extremely unlikely, 7 = extremely likely), except for adaptation motivation and policy support which are on a 5-point scale (1 = not at all serious, 5 = extremely serious; 1 = not at all, 5 = a great deal); N = 157.
Direct Effects
A one-way ANOVA was conducted to determine if support for policy, intended behavior change and intended intended political participation(CWWS score) varied for participants depending on exposure to different framing condition. Participants were classified into one of three groups: thematic (n = 56), episodic (n = 44), and control (n = 57).There were no
conditions for policy support, F(2, 154) = .37, p = .689. Neither was their statistical differences in CWWS score between the different framing conditions for intended behavior change, F(2, 154) = .075, p = .928, or intended intended political participation, F(2, 154) = .24, p = .786. These findings address RQ1, determining that there is no differentiation of influence on the dependent variables among the framing conditions. Thus, in referring to RQ1 we can reject H1 in favor of H2 which hypothesized that no difference in influence would be found.
Mediation Analysis
First examined were the impacts of framing conditions on adaptation motivation and adaptation belief. No significance was found for any of the three dependent variables, thus rejecting the possibility of either an indirect or direct effect. Second to be examined was the impact of adaptation motivation and adaptation belief on each of the three dependent variables (intended political participation: F(4, 152) = 18.55, p < .001; intended behavior change: F(4, 152) = 12.22, p < .001; policy support: F(4, 152) = 15.49, p < .001). Significance was found for both mediators on each dependent variable, as is seen below in table 2.
Table 2. Adaptation Motivation and Adaptation Belief on dependent variables (b-path).
Analysis of the unstandardized coefficients displayed a trend of adaptation motivation appearing to hold higher impact on each of the dependent variables than adaptation belief. To test this, the unstandardized beta coefficients of both mediators were standardized. This
addressed RQ2, examining how the mediators differ in impact on the three dependent variables.
Table 3. Standardized b-path beta coefficients
The standardized coefficients indeed confirmed that adaptation motivation holds a higher impact than adaptation belief when it comes to influencing intended political participation, intended behavior change and policy support. Following these findings mean levels of adaptation belief and adaptation motivation were conducted against political ideology to answer RQ3. The results found that while political ideology affects adaptation motivation, it does not influence adaptation belief, see figure 1 and 2.
Figure 1. Mean levels of adaptation motivation for different political ideology identifications
Moderated Mediation Analysis
Figure 3. Moderated mediation model showing indirect effects on intended political participation for thematic and episodic indicators (Y).
Note: First beta coefficient refers to thematic indicator, second refers to episodic indicator. Unstandardized beta coefficients are shown, p< .05*, p < .01**.; p < .001***; (N=157).
First examined was the moderated mediation of framing conditions on risk capacity with moderators, political ideology and climate change concern (thematic indicator: F(6, 150) = 15.49; episodic indicator: F(6, 150) = 15.86). There were no moderated causal relationships found between the framing conditions and adaptation motivation. However, political ideology and climate change concern were both found to have correlation with adaptation motivation. Within the episodic indicator, political ideology had a significant positive influence: (B = .78, SE
= .27, p = .004, CI 95% = [0.25; 1.30]). This same positive influence was found for climate change concern: (B = .95, SE = .18, p < .001, CI 95% = [0.59; 1.31]). Results for the thematic indicator demonstrated a similar trend (political ideology: B = .66, SE = .27, p < .017, CI 95% = [0.12; 1.21]) (climate change concern: B = .976, SE = .17, p < .001, CI 95% = [0.64; 1.31]).
Next examined was the moderated mediation of framing conditions on adaptation belief (thematic indicator: F(6, 150) = 6.67; episodic indicator: F(6, 150) = 6.04).While no effect was found for either indicator group on adaptation motivation, a moderated causal effect was found within the thematic indicator on adaptation belief, seen above in figure 4. Conditional effects were found for the thematic indicator on adaptation belief, appeared to only be significant for people who skew conservatively for political ideology but hold high levels of climate change concern (B = 4.13, SE = 1.89, p = .03, CI 95% =[ 0.40; 7.86]) (B = 6.55, SE = 2.65, p = .014, CI
95% = [1.32; 11.78]). No moderated causal relationship was found for adaptation belief within the episodic indicator, as is seen in figure 4. However, climate change concern trended towards a significant causal moderation on adaptive capacity (B = -.93, SE = .53, p = .082, CI 95% = [-01.98; 0.12]). Climate change concern did correlate with adaptation belief (B = -1.78, SE = .35, p < .001, CI 95% = [1.09; 2.46]), while political ideology did not (B = -53, SE = .51, p = .30, CI 95% = [-1.53; 0.48]).
Summary
No direct or indirect effects were found, thus the frames were not predictive of any of the three dependent variables. However, when looking at the tables and the model in figure 3 there are instances of spurious causality between variables, whose partial-connections provide further insight. The first to note is that adaptation belief and adaptation motivation in particular, both
had significant effects on all three dependent variables: intended intended political participation, intended behavior change, and support for policy. Although adaptation motivation and
adaptation belief were positioned in this study as mediators, results show they instead are playing the role of predictors. The second takeaway is that the standardized beta coefficients showed that across all three dependent variables, adaptation motivation has about twice as much impact as adaptation belief on participant’s intention to participate politically, change behavior and support policy. Upon this finding the mean levels of adaptation motivation and adaptation belief were compared to political ideology (figure 1 & 2), finding only adaptation motivation to be
significantly correlated F(5, 151) = 9.70, p < .001 (adaptation belief : F(5, 151) = 1.17, p = .326).
The last main takeaway are the findings on total moderated mediation, which found that the conditional effect of the thematic indicator on the three dependent variables via adaptation belief is lowered by those who skew more politically conservative and is increased by those with a high level of climate change concern.
Discussion
This study sought to contribute to the growing body of research examining how framing affects an individual’s socio-cognitive perceptions of climate change as well as their intention to participate politically, change behavior and support policy. No direct or indirect effects were found between the frames and the pro-environmental behaviors tested, thus in regards to RQ1, H1 was rejected in favor of H2. The lack of direct or indirect effects may in part be characteristic of the perception of climate change within the United States. Public perception is highly
polarized, making it susceptible to ‘hardened opinions that are resistant to change (Hart &
be significantly correlated to political ideology in this study. Keeping in mind the polarization of the issue, one explanation for a lack of indirect effect through adaptation motivation could be that because motivation is linked to ideology, it is more difficult to influence. The lack of indirect effects through adaptation belief, which pertains to one’s perceived efficacy, could be explained with reference to a study by Hart and Feldman (2016). They found limited effects of efficacy-framed messages, with only two of six experimental conditions shifting efficacy perceptions. The fact that efficacy-framed media messages were not more successful in increasing perceived efficacy highlights the difficulty of effectively increasing this
socio-cognitive factor through media messaging. Aside from these observations, the lack of effects may also be related to the sample of participants. Use of a control condition in this study highlighted that in general the sample had high mean scores for climate change concern,
adaptation motivation, adaptation belief and each of the dependent variables. This could be attributed to the fact that the millennial sample skewed both liberally and highly educated. Had there been a more diverse pool of participants, the findings may have resulted differently.
Expanding on RQ1, the conditional effects of thematic framing on adaptation belief when mediated by political ideology and climate change concern support past literature which
highlight the complexity of communicating climate change within the United States. While the thematic effect on pro-environmental behavior via adaptation belief was raised for those with a high level of climate change concern, the same messaging effect was lowered by those who skew more politically conservative. The thematic stimuli used in this study positioned the effects of consumer consumption on climate change in a broad context, and angled responsibility towards the government. The explanation for why this frame would have a negative effect for those who
skew conservatively can be attributed to conservative political values in the U.S., which do not generally support governmental policies which strongly regulate businesses or the public. Similar communicative concerns were found by Feldman et al. (2015, p. 18), where they express concern over how frames can behave as a ‘double-edged sword’— mobilizing some individuals while ‘deactivating’ others. This was also found in a study by Howell et al. (2016), where government action framing disengaged participants with a low level of climate change concern. Thus, there is a general sentiment among literature that certain frames will only reinforce the partisan divide, causing conservatives to detach from the issue. This unique ideological chasm in the U.S. as it relates to climate change is why the common communicative strategies for framing usually center around the need for tailored messaging according to audience (Howell et al., 2016; Feldman et al., 2015). Specifically, findings suggest that framing should be oriented towards the beliefs and values of audiences, and how they correspond to the audience’s understanding of climate change (Howell et al., 2016). The largest challenge in putting this communicative strategy into action is that the media communicate to the general public— they do not have the logistical capabilities to fully carry out such a recommendation. The second challenge is finding the right way to reach audiences that skew conservatively. Mitigating climate change and adapting to its effects requires governmental policy on both the public and private sphere. However, as seen in this study and past literature, climate change messaging that thematically frames responsibility on the government directly confronts conservative ideology. As the U.S. strives to move forward to act on climate change, these communication barriers are going to continue to play a role in shaping the perception of the public.
The direct effects of modeled mediators on each of the pro-environmental dependent behaviors, contribute to the growing body of literature which find risk perception (adaptation motivation) and efficacy (adaptation belief) to be correlated with action (Grothmann & Patt, 2005). What this study offers as supporting insight is the ways in which these socio-cognitive measures were found differ in their influence on pro-environmental behavior, as explored through RQ2. Adaptation motivation played twice as significant a role as adaptation belief in influencing intended political participation and support for policy, and nearly as much for intended behavior change. Climate change research places a lot of focus on the importance of efficacy and risk individually or as they relate to each other, but this study was able to provide a side by side comparison to see which played a more influential role. A further insight looking into adaptation motivation and adaptation belief, was the role of political ideology explored by RQ3. Adaptation motivation was directly affected by ideology, correlating positively as ideology moved from conservative to liberal. However, political ideology had no significant impact on adaptation belief. Hart and Feldman (2016) found similar results, where political ideology was significantly associated with perceived issue importance, but did not interact with perceived efficacy. This may be the reason why no moderated mediation effects were found for frames on adaptation motivation, as it is linked to ideology and therefore harder to influence. It is
speculated that because participants generally scored high for adaptation motivation and adaptation belief, that episodic framing was not influential. The episodic frame positioned responsibility on the individual to address climate change, but the sample already had a high level of awareness and efficacy they could engage in actions to address climate change.
This study, as with any, has limitations that should be acknowledged. The results of the experiment would likely have benefited from a larger and more diverse sample, with a higher percentage of conservatively skewing participants. The high scores across the three dependent variables also brings into question whether future research should include measures to detect social desirability bias. It is possible that participants over-reported their likelihood to engage in behaviors because they view them as favorable. Future research with the scope to do so, might consider conducting studies with conservative-only samples. Framing research often concludes that it is audiences who skew conservatively that are negatively affected, not liberal audiences. As climate change becomes an increasingly important international issue, it is important to look past what is the most effective messaging for all audiences— and focus on what is the most effective messaging for those audiences with a low level of climate change concern. A second consideration for future research, not feasible within the scope of this study, would be to take a longitudinal approach with multiple message exposures. The polarized nature of public
perception on climate change has made the issue vulnerable to hardlined positioning, and both adaptation motivation and adaptation belief are made up of deeply rooted beliefs and values. For an issue like climate change which resonate within the public’s ideology in the same manner as gun rights or taxes, short-term media exposure may be less generous in yelding results.
Works Cited
Aarøe, L. (2011). Investigating frame strength: The case of episodic and thematic frames. Political Communication, 28(2), 207–226.
https://doi.org/10.1080/10584609.2011.568041
Adger, W. N., Dessai, S., Goulden, M., Hulme, M., Lorenzoni, I., Nelson, D. R., … Wreford, A. (2009). Are there social limits to adaptation to climate change? Climatic Change, 93(3–4), 335–354. https://doi.org/10.1007/s10584-008-9520-z
Capstick, S. B. (2013). Public understanding of climate change as a social dilemma. Sustainability (Switzerland), 5(8), 3484–3501. https://doi.org/10.3390/su5083484
Dimock, M. (2018). Where Millennials end and post-Millennials Begin. Retrieved May 30, 2018, from http://www.pewresearch.org/fact-tank/2018/03/01/defining-generations-where-millennials-end-and-post-millennials-begin/
Druckman, J. N. (2001). On the Limits of Framing Effects : Who Can Frame ? Author ( s ): James N . Druckman Source : The Journal of Politics , Vol . 63 , No . 4 ( Nov ., 2001 ), pp . 1041-1066 Published by : The University of Chicago Press, 63(4), 1041–1066. https://doi.org/10.1016/j.proenv.2015.01.005
Dunlap, R. E., McCright, A. M., & Yarosh, J. H. (2016). The Political Divide on Climate Change: Partisan Polarization Widens in the U.S. Environment, 58(5), 4–23. https:// doi.org/10.1080/00139157.2016.1208995
Feldman, L., Hart, P. S., Leiserowitz, A., Maibach, E., & Roser-Renouf, C. (2017). Do Hostile Media Perceptions Lead to Action? The Role of Hostile Media Perceptions, Political Efficacy, and Ideology in Predicting Climate Change Activism. Communication Research, 44(8), 1099–1124. https://doi.org/10.1177/0093650214565914
Feldman, L., Hart, P. S., & Milosevic, T. (2017). Polarizing news? Representations of threat and efficacy in leading US newspapers’ coverage of climate change. Public Understanding of Science, 26(4), 481–497. https://doi.org/10.1177/0963662515595348 Fry, R. (2017). Millennials expected to outnumber Boomers in 2019. Retrieved May 30,
2018, from http://www.pewresearch.org/fact-tank/2018/03/01/millennials-overtake-baby-boomers/
Gross, K. (2008). Framing persuasive appeals: Episodic and thematic framing, emotional response, and policy opinion. Political Psychology, 29(2), 169–192. https://doi.org/ 10.1111/j.1467-9221.2008.00622.x
Grothmann, T., Grecksch, K., Winges, M., & Siebenhüner, B. (2013). Assessing institutional capacities to adapt to climate change: Integrating psychological dimensions in the
adaptive capacity wheel. Natural Hazards and Earth System Sciences, 13(12), 3369– 3384. https://doi.org/10.5194/nhess-13-3369-2013
Grothmann, T., & Patt, A. (2005). Adaptive capacity and human cognition: The process of individual adaptation to climate change. Global Environmental Change, 15(3), 199–213. https://doi.org/10.1016/j.gloenvcha.2005.01.002
Hart, P. S., & Feldman, L. (2016). The Impact of Climate Change–Related Imagery and Text on Public Opinion and Behavior Change. Science Communication, 38(4), 415–441. https://doi.org/10.1177/1075547016655357
Hart, P. S., & Feldman, L. (2014). Threat Without Efficacy? Climate Change on U.S. Network News. Science Communication, 36(3), 325–351. https://doi.org/ 10.1177/1075547013520239
Hart, P. S., & Feldman, L. (2016). The influence of climate change efficacy messages and efficacy beliefs on intended political participation. PLoS ONE, 11(8), 1–16. https:// doi.org/10.1371/journal.pone.0157658
Hart, P. S., & Nisbet, E. C. (2012). Boomerang Effects in Science Communication: How Motivated Reasoning and Identity Cues Amplify Opinion Polarization About Climate Mitigation Policies. Communication Research, 39(6), 701–723. https://doi.org/ 10.1177/0093650211416646
Hart, P. S. (2011). One or many? The influence of episodic and thematic climate change frames on policy preferences and individual behavior change. Science Communication, 33(1), 28–51. https://doi.org/10.1177/1075547010366400
Howell, R. A., Capstick, S., & Whitmarsh, L. (2016). Impacts of adaptation and responsibility framings on attitudes towards climate change mitigation. Climatic Change, 136(3–4), 445–461. https://doi.org/10.1007/s10584-016-1627-z
Hurtíková, H. (2013). Framing theory in the field of Political Communication : Theoretical and Methodological Conceptualization. Contemporary European Studies, 27–44. Retrieved from https://www-ceeol-com.proxy.uba.uva.nl:2443/search/viewpdf? id=250229
Kahan, D. M. (2012). Ideology, Motivated Reasoning, and Cognitive Reflection: An Experimental Study. SSRN Electronic Journal, 8(4), 407–424. https://doi.org/10.2139/ ssrn.2182588
Karlsson-Vinkhuyzen, S. I., Friberg, L., & Saccenti, E. (2017). Read all about it!? Public accountability, fragmented global climate governance and the media. Climate Policy, 17(8), 982–997. https://doi.org/10.1080/14693062.2016.1213695
Kim, S. H. (2015). Who Is Responsible for a Social Problem? News Framing and Attribution of Responsibility. Journalism and Mass Communication Quarterly, 92(3), 554–558. https://doi.org/10.1177/1077699015591956
Leiserowitz, A. (2006). Climate change risk perception and policy preferences: The role of affect, imagery, and values. Climatic Change, 77(1–2), 45–72. https://doi.org/10.1007/ s10584-006-9059-9
Nisbet, M. C. (2009). Communicating Climate Change: Why Frames Matter. Environment: Science and Policy for Sustainable Development, 51(2), 12–23. https://doi.org/10.3200/ ENVT.51.2.12-23
O’Brien, K. L. (2009). Do values subjectively define the limits to climate change adaptation? Adapting to Climate Change - Thresholds, Values, Governance, 164–170. https:// doi.org/10.1017/CBO9780511596667.011
Preston, B. L., Mustelin, J., & Maloney, M. C. (2013). Climate adaptation heuristics and the science/policy divide. Mitigation and Adaptation Strategies for Global Change, 20(3), 467–497. https://doi.org/10.1007/s11027-013-9503-x
Retzbach, J., Otto, L., & Maier, M. (2016). Measuring the perceived uncertainty of scientific evidence and its relationship to engagement with science. Public Understanding of Science, 25(6), 638–655. https://doi.org/10.1177/0963662515575253
Sapiains, R., Beeton, R. J. S., & Walker, I. A. (2016). Individual responses to climate change: Framing effects on pro-environmental behaviors. Journal of Applied Social Psychology, 46(8), 483–493. https://doi.org/10.1111/jasp.12378
Schäfer, M. S., & O’Neill, S. (2017). Frame Analysis in Climate Change Communication (Vol. 1). https://doi.org/10.1093/acrefore/9780190228620.013.487
Singh, A. S., Zwickle, A., Bruskotter, J. T., & Wilson, R. (2017). The perceived psychological distance of climate change impacts and its influence on support for adaptation policy. Environmental Science and Policy, 73(September 2016), 93–99. https://doi.org/10.1016/ j.envsci.2017.04.011
Tabernero, C., & Hernández, B. (2011). Self-efficacy and intrinsic motivation guiding environmental behavior. Environment and Behavior, 43(5), 658–675. https://doi.org/ 10.1177/0013916510379759
Valkenburg, P. M., & Peter, J. (2013). The differential susceptibility to media effects model. Journal of Communication, 63(2), 221–243. https://doi.org/10.1111/jcom.12024
Weber, E. U., & Stern, P. C. (2011). Public Understanding of Climate Change in the United States. American Psychologist, 66(4), 315–328. https://doi.org/10.1037/a0023253
Appendix A: Variable Scales
Adaptation Motivation
Please indicate the level of seriousness you associate with the following statements: How serious do you feel the CURRENT impacts of climate change around the world are? How serious do you feel the FUTURE impacts of climate change around the world will be? (1 = not at all serious, 5 = extremely serious)
How likely do you think it is that each of the following will occur during the next 50 years? Worldwide, many people will be negatively affected
The United States will NOT be negatively affected People I know will be negatively affected
I personally will be negatively affected (1 = not at all likely, 5 = extremely likely) Adaptation Belief
Please indicate the level of which you agree or disagree:
I feel confident that I can find ways to reduce my environmental impact I feel capable of finding ways to reduce my environmental impact
I do NOT feel capable of changing my lifestyle to help protect the environment I feel confident that I can conserve energy in everyday life
If I were to change my behaviors to be more environmentally friendly, this would NOT be effective in reducing the negative impacts of climate change
If American citizens were to collectively change their behaviors to be more environmentally friendly, this would be effective in reducing the negative impacts of climate change
If government officials were to pass laws to reduce America’s dependence on fossil fuels, this would be effective in reducing the negative impacts of climate change
I feel confident that the Federal government will take action to pass laws to reduce America’s dependence on fossil fuels
I feel the Federal government is capable of finding ways to take action and address climate change
Policy Support
How much would you support the following?
Increasing federal aid for developing countries who are vulnerable to natural disasters that have increased due to climate change?
Implementing zoning laws in 10 years that prohibit building of homes and commercial buildings in areas that are threatened by natural disasters such as wildfires, coastal flooding, or hurricanes?
A federal policy that provides you with subsidies or other incentives to increase your ability to lower your carbon footprint?
Increasing funding for cities to implement projects aimed at lowering their carbon footprint? Increasing funding to programs in your state that help protect or restore natural habitats that are endangered because of climate change?
(1 = not at all, 5 = a great deal)
Behavior Change
How likely or unlikely it is that you will personally engage in each the following actions over the next 3 months:
Use less air conditioning in the summer and less heat in the winter Drive less by walking, biking, or taking public transportation more Buy locally, like going to the farmers market
Buy something second-hand (piece of clothing or household item) Regularly recycle plastic and paper items
Bring reusable bag to the grocery store Prioritize use of a reusable water bottle Cut down on meat consumption
(1 = extremely unlikely, 7 = extremely likely)
Political Participation
How likely or unlikely is it that you will personally engage in each of the following actions in the next 12 months:
Contact government officials to urge them to take action and address climate change Sign a petition in support of taking action on reducing the impacts of climate change
Participate in a rally or protest in support of action to reduce effects of climate change Join or volunteer with an organization working to reduce climate change
(1 = extremely unlikely, 7 = extremely likely)
