Climate Change Risk Perception and Pro-Environmental Behavior. Toward a Comprehensive Model

Abstract

Climate change has become one of the greatest concerns of the 21st century as its ramifications pose a major risk to all life on earth. However, not all individuals are aware of this risk and behavioral engagement to counteract the issue is often still lacking. This research examined the relationship between climate change risk perception and pro-environmental behavior. The aim of this cross-sectional study was twofold. First, studying specific pro-environmental behavior and second, using a methodological approach that addresses the measurement inconsistencies in the literature on how to define and operationalize climate change risk perception. The sample consisted of 141 young educated adults (M = 25.67, SD = 2.93) who responded to an online questionnaire. Results showed that climate change risk perception significantly predicted pro-environmental behavior. Comparing personal worry and general concern as two distinct indicators of climate change risk perception indicated that personal worry was stronger correlated to pro-environmental behavior. This relationship was found to be partially mediated by pro-environmental behavioral intentions. Results of a multiple hierarchical regression showed that including personal worry as an additional predictor variable did not significantly increase the explained variance in pro-environmental behavior, after having controlled for variables from the theory of planned behavior (attitude, subjective norms, perceived behavioral control) and the norm-activation model (personal norms). The findings extend the current literature and contribute to a further understanding of why and how humans behave pro-environmentally. Future researchers are encouraged to study the individual drivers of climate change more closely.

Keywords: climate change, risk perception, pro-environmental behavior, personal worry, general concern, theory of planned behavior, norm-activation model, measurement correspondence

Climate Change Risk Perception and Pro-Environmental Behavior – Toward a Comprehensive Model

“We are in a new geological epoch. We are now the dominating force of change on planet Earth. We exceed the forces of the sun or of volcanic eruptions. We are in the driving seat.”

- Johan Rockström.

On November 13, 2017, over 15,000 scientists gave a public warning about the fate of humanity, yet again. In their letter A Second Notice, they predicted that ongoing unsustainable human actions pressuring Earth’s systems could cause irreversible environmental change. This could lead to conditions on Earth which no longer support human development (Ripple, Wolf, Newsome, Galetti, Alamgir, Crist et al., 2017). Climate change has become one of the greatest concerns of the 21st century because its ramifications pose a major threat to life on Earth. Findings of the Intergovernmental Panel on Climate Change (IPCC, 2014) state that climate change will result with a high probability (67 to 95 percent likelihood) in a rising global temperature, heat waves, more floodings and droughts, more extreme weather events, sea level rise and melting ice-caps as some examples. Many of those impacts are already evident today, and according to predictions, further impacts are inevitable.

Today, there is an overwhelming scientific agreement on climate largely changing due to anthropogenic greenhouse gases, particularly from our reliance on fossil fuels and industrialized forms of agriculture (IPCC, 2014). Another contribution of those emissions comes from specific lifestyle choices and behaviors that remain carbon-intensive and unsustainable (Gifford, 2008; Rockström, Steffen, Noone, Persson, Chapin, Lambin et al., 2009). In order to mitigate climate change, both individuals and corporations need to reduce their carbon footprint. Political steps toward a low-carbon world community have already been taken. As of June 2018, 195 countries have signed the Paris Agreement, a contract that aims to keep the global average temperature well below 2° Celsius. However, recent analyses of current global emission pathways predict that humanity is heading toward a global average temperature of 4° Celsius by 2100 (IPCC, 2014), a condition planet Earth has not been in for the last one million years (Rockström et al., 2009). In order to remain in a state that we know supports human development and at the same time minimize anthropogenic impacts on Earth, it is necessary to substantially reduce greenhouse gas emissions. Accordingly, the need to adopt pro-environmental behavior for a sustainable future is widespread (Pidgeon, 2012).

Climate change is not only an urgent issue but also a very complex one. Counteracting climate change requires mutual scientific efforts across several disciplines. Using psychological empirical research to examine the individual behavior that fosters or counteracts climate change can be helpful to further understand climate change as a human-caused problem and can eventually lead to recommendations on how to best mitigate the issue. The growing psychological research literature on climate change reflects that psychologists begin to recognize climate change as an important field of study. The literature on pro-environmental attitudes and behavior is rich, however, few researchers have investigated the link between the individual’s perception of climate change as a potential risk and their concrete intend to do something about it (van der Linden, 2017). Although it is widely known that climate change is a serious risk to all life on Earth, not all individuals perceive this risk (Lazo, Kinnel, & Fisher, 2000). This raises the question of whether perceiving climate change as a risk can affect the individual’s pro-environmental behavior. Exploring how climate change risk perception relates to pro-environmental behavior has received little attention in the literature, with most research focusing on behavioral intentions and policy support (Bamberg & Möser, 2007). By studying how specific pro-environmental behavior is linked to the perception of climate change, a further understanding of why individuals engage in pro-environmental behavior could be developed. Therefore, the aim of this research is to examine this relationship in an empirical study, while addressing some of the criticism in previous literature. The overall goal of this research is to provide insights on how climate change risk perception is related to pro-environmental behavior and to contribute knowledge to the existing literature that can help to create a more sustainable future.

Pro-Environmental Behavior

There is a growing body of literature on pro-environmental behavior and with it, a number of definitions. Two meta-analytic reviews find that most of the studies completely miss the definition of what is actually considered pro-environmental within their studied area (Bamberg & Möser, 2007; Hines, Hungerford, & Tomera, 1986/87). For example, De Groot and Steg (2009) provide a rather general definition of pro-environmental behavior and simply describe it as behavior that positively impacts the environment. This definition is vague, as it does not specify which types of behavior are meant. The Department for Environment, Food and Rural Affairs (2008) suggests that pro-environmental behavior can generally be divided into five domains: energy and water conservation, food, transportation, waste disposal and eco-friendly shopping. This distinction has been applied in a number of studies (e.g., Gifford & Nilsson, 2014; Whitmarsh, 2009). A more detailed definition is provided by Stern (2000). He argues that pro-environmental behavior can be defined by its impact – the extent to which it “changes the availability of materials or energy from the environment or alters the structure and dynamics of ecosystems or the biosphere itself” (p. 408). For example, using less water or recycling more has a lower impact on the environment than for example reducing meat consumption (Department for Environment, Food and Rural Affairs, 2008). In contrast to Stern’s (2000) definition, Bamberg and Möser (2007) do not define pro-environmental behavior by its impact but rather by motives. They suggest that pro-environmental behavior can either be motivated by self-interest (i.e., save money by limiting energy use) or by altruism, that is being concerned for other people, other species, or entire ecosystems (i.e., limiting CO2 emissions that affect the global climate). Both motives are also reflected in the two most commonly used theoretical models used to explain pro-environmental behavior. While researches who view self-interest as the more dominant motive, often rely on the theory of planned behavior (Ajzen, 1991), others, who consider pro-environmental behavior as more pro-socially motivated often refer to the norm-activation model (Schwartz, 1977). Both theoretical concepts are described in the following.

Theory of Planned Behavior (TPB)

A widely used theory to explain pro-environmental behavior is the theory of planned behavior (TPB, Ajzen, 1991). It has evolved as an extension of the theory of reasoned action (Ajzen & Fishbein, 1980) and aims to explain social behavior and behavioral change. The TPB is referred to as a rational choice model as it states that attitudes, subjective norms and perceived behavioral control (PBC) are involved in the decision-making processes that determine behavioral outcomes. According to the theory, the most proximal predictor of behavior are behavioral intentions, that is the degree to which individuals are willing to try to perform a certain behavior. When forming behavioral intentions, individuals do not only consider to what extent they hold favorable attitudes toward a certain behavior but also estimate their perceived behavioral control, which is their ability to perform a given behavior. As a third variable determining behavioral intentions, the TPB framework proposes subjective norms. These can be either injunctive or descriptive. Injunctive norms represent the individual’s perception of what important others (i.e., parents, peers, teachers) expect them to do or not to do, whereas descriptive norms are based on their perception of those important others’ own behavior (Rivis & Sheeran, 2003). Taken together, attitudes, PBC, and subjective norms do not affect behavior directly but indirectly via their influence on behavioral intentions. The TPB further emphasizes that when PBC accurately reflects actual control, it can have a direct effect on behavior, too (see Figure 1).

Abbildung in dieser Leseprobe nicht enthalten

Figure 1. The theory of planned behavior (Ajzen, 1991, p. 182). PBC=Perceived behavioral control.

There is abundant empirical support for the TPB within the domain of pro-environmental behavior. Numerous studies have demonstrated the theory’s predictive power for several types of behaviors (Abrahamse & Steg, 2011; Abrahamse, Steg, Gifford, & Vlek, 2009; Boldero, 1995; Heath & Gifford, 2002; Schultz, Nolan, Cialdini, Goldstein, & Griskevicius, 2007; Scott, Jones, & Webb, 2014; Sparks & Shepherd, 1992). For example, Boldero (1995) discovered that individuals’ actual pro-environmental behavior – in this case, recycling – was predicted by the attitude toward recycling in the first place. A positive attitude was found to predict the individuals’ intention to recycle which, again, predicted the concrete recycling behavior. In another study about green consumerism, Sparks and Shepherd (1992) showed that all three predictor variables (attitudes toward green consumerism, subjective norms, and PBC) were significantly related to individuals’ intentions to consume organic vegetables. Further empirical support comes from studies about car use (Abrahamse et al., 2009), public transportation (Heath & Gifford, 2002) and energy conservation (Abrahamse & Steg, 2011; Schultz et al., 2007; Scott et al., 2014). A robust finding is that behavioral intentions act as the immediate antecedent of behavior. In their meta-analysis, Bamberg and Möser (2007) compared 57 studies and found that on average pro-environmental behavioral intentions account for 27 % of the variance of pro-environmental behavior. Despite its influencing role in previous literature, the theory of planned behavior has been criticized for missing the explanation of recurrent behaviors (Klöckner, 2013) and for lacking the consideration of moral norms as an important predictor of behavior (Gifford & Nilsson, 2014). In fact, researchers who view pro-environmental behavior as pro-socially motivated, understand moral norms as a key variable and often refer to Schwartz’ (1977) norm-activation model.

Norm-Activation Model (NAM)

As a theoretical model, the NAM was originally developed by Schwartz (1977) to explain altruistic behavior. Like the TPB, it has been extensively used to explain various social behaviors, including pro-environmental behaviors. Contrasting the TPB, the NAM states that not intentions, but personal norms are direct determinants of pro-social behavior. According to Schwartz (1977) “activated personal norms are experienced feelings of moral obligation, not as intentions” (p. 227). He further states that these norms are determined by two factors: being aware that performing (or not performing) particular behaviors has certain consequences and the individual’s feeling of responsibility for performing the specific behavior. Recent work by de Groot and Steg (2009) compared whether the NAM is best interpreted as a moderator or mediator model and argued for the latter. In the mediator model, awareness of consequences affects personal norms via felt responsibility (see. Figure 2). De Groot and Steg (2009) further argue that an individual must be aware of the behavioral consequences before being able to feel responsible for the behavior. Feelings of responsibility then trigger personal norms which directly influence specific behavior.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2. Norm-activation model interpreted as a mediator model, adapted from De Groot and Steg (2009).

The NAM has been applied in the environmental domain and received empirical support in a series of studies. Personal norms were found to be a significant predictor of pro-environmental behaviors like recycling (Guagnano, Stern, & Dietz, 1995), travel mode choice (Hunecke, Blöbaum, Matthies, & Höger, 2001), pro-environmental buying (Thøgersen, 1999), and energy conservation (Black, Stern, & Elworth, 1985). Further support for the role of personal norms in explaining pro-environmental behavior comes from three meta-analyses (Bamberg & Möser, 2007; Hines et al., 1986/87; Klöckner, 2013). All of them report mean correlations between personal norms and pro-environmental behavior ranging from r = .32 to r = .39. However, personal norms have been repeatedly shown to only have an indirect impact on behavior, completely mediated by intentions (Bamberg & Möser, 2007).

Integrated TPB / NAM Framework

Given the empirical support of both models, pro-environmental behavior is probably best viewed as a blend of self-interest and pro-social motives. This has led researchers to combine both theoretical models to explain pro-environmental behavior. For example, Harland, Staats, and Wilke (1999) found that adding personal norms to the TPB framework, increased the explained variance of behavioral intentions by 1 to 10 %. Likewise, by adding behavioral intentions to the NAM framework, explained variance of behavior was found to increase by about 17 % (Bamberg, Hunecke, & Blöbaum, 2007; Bamberg & Möser, 2007).
In sum, these findings suggest that an integrated TPB / NAM model best explains pro-environmental behavior. Klöckner’s (2013) meta-analysis supports this assumption.
A comparison of 56 studies found the key constructs for explaining pro-environmental behavior intentions to be attitudes, subjective norms, PBC and personal norms. Together, these four variables accounted for 55 % of explained variance in intentions.

Although the TPB and the NAM provide useful theoretical models to explain pro-environmental behavior, a relatively large proportion of variance remains unexplained.
A possible explanation for that might be that pro-environmental behavior is a complex construct with other contributing variables outside of the TPB and NAM. This has led researchers to test the predictive power of potentially contributing variables, such as self-identity (Rise, Sheeran, & Hukkelberg, 2010), environmental identity (Clayton, 2003), personality factors (Arbuthnot, 1977) or values (Stern, Dietz, & Kalof, 1993). With the rising psychological research body about climate change, it has been argued that risk perception plays a key role in influencing pro-environmental behavioral intentions (van der Linden, 2015).
At this point, the question arises whether the individual’s pro-environmental behavior is affected by whether or not climate change is perceived as a risk.

Climate Change Risk Perception

Climate change is defined as a long-term change in global and regional weather patterns associated with increasing global average temperatures (Rockström et al., 2009). There is an overwhelming scientific consensus that human activities are influencing climate change particularly by burning fossil fuels, farming livestock and cutting down rainforests (IPCC, 2014). Potential consequences of continuously rising average global temperatures include loss of sea ice, rising sea levels, more extreme weather events such as droughts, floodings, wildfires and tropical storms, and loss of biodiversity (Rockström et al., 2009). Although climate change is undoubtedly a serious risk to all life on Earth, including the human species, not all people equally perceive climate change as a risk (Lazo et al., 2000). Since climate change is a unique phenomenon as it is abstract, complex and not directly perceivable (Bamberg & Möser, 2007), people have to rely on the news media, scientific reports and expert judgments when forming an understanding of climate change (Leiserowitz, 2007; Swim, Clayton, Doherty, Gifford, Howard, Reser et al., 2009). Consequently, the individual’s perception of climate change differs dramatically, ranging from complete denial of its existence on the one end to viewing it as the greatest risk of our time on the other (Leiserowitz, 2005). Perceiving climate change as a risk implies recognizing the probability of its negative consequences (Sundblad, Biel, & Gärling, 2007). Following this traditional view of risk (Slovic, 1999), climate change risk perception can be defined as being aware of climate change and fearing its potential negative effects (Leiserowitz, 2006; Weber, 2006; Whitmarsh, 2008).

Climate Change Risk Perception Model

In an attempt to model climate change risk perception, van der Linden (2015) proposed four key theoretical dimensions as immediate determinants. These are cognitive factors (i.e., knowledge about climate change), experiential processing (i.e., personal experience and affective evaluations), socio-cultural influences (i.e., social norms) and socio-demographic characteristics. Empirically, the combination of those four dimensions explains about 70 % of the variation in climate change risk perception (van der Linden, 2017). The Climate Change Risk Perception Model (CCRPM) provides an overview of the single factors involved in forming individuals’ risk perception about climate change (see Figure 3).

Abbildung in dieser Leseprobe nicht enthalten

Figure 3. A simplified version of the Climate Change Risk Perception Model (CCRPM) by van der Linden, 2015, p. 5.

Cognitive Factors. Van der Linden (2015) questions “if one has no basic awareness of the climate change problem, then how can a judgment about the issue be formed?” (p. 6). According to this line of thought, possessing knowledge about climate change seems to be a prerequisite of climate change risk perception. Consequently, if the individual understands that climate change is currently happening and that it has negative ramifications for life on Earth, he or she should be concerned about it. Additionally, one might expect that more knowledge would lead to higher risk perception and thus, more pro-social engagement. However, the relation between knowledge and risk perception is still unclear. The literature provides mixed results. For example, Heath and Gifford (2006) found a positive association between knowledge and concern, whereas Brody, Zahran, Vedlitz, and Grover (2008) found no relation at all. Kellstedt, Zahran, and Vedlitz (2008) however, report that knowledge and climate change risk perception are in fact negatively associated, indicating that the more informed the individual is, the less he or she feels concerned about and personally responsible for climate change. According to van der Linden (2017), these inconsistent findings are partly due to the different ways knowledge is defined and operationalized. For example, the use of self-reported items to assess knowledge rather reflects subjective knowledge than accurate knowledge. In that case, knowledge represents what people think is true, instead of what is accurately true. Studies using objective measures to assess how much people accurately know about climate change generally report a significant positive association between knowledge about climate change and public concern (Guy, Kashima, Walker, & O’Neill, 2014; Milfont, 2012; Shi, Visschers, Siegrist, & Arvai, 2016). In fact, a large-scale study covering 119 countries found that educational achievement and viewing climate change as a human-caused problem are important determinants of risk perception (Lee, Markowitz, Howe, Ko, & Leiserowitz, 2015). Although there is no scientific consensus in the literature yet, knowledge is most likely a necessary but not a sufficient condition for feeling concerned about climate change (van der Linden, 2015).

Experiential processing. An individual is not only able to have cognitive knowledge about climate change but can also experience it in emotional terms (van der Linden, 2015). One of today’s widely acknowledged theories, the dual-process theory by Kahneman (2011), suggests that information processing occurs in two fundamentally different ways. The theory suggests that thought can arise as a result of two different processes. These can be divided into cognition, indicated by slow, conscious, controlled, analytical and rule-based processes, and affect, indicated by fast, unconscious, automatic and associative processes. According to Loewenstein, Weber, Hsee, and Welch (2001), affective climate change risk judgments are often more dominant than cognitive judgments. For example, studies examining individual’s first thought response when thinking about climate change often find negative affects related to the impacts of climate change to be the first thing that comes to mind (Leiserowitz, 2006; Smith & Leiserowitz, 2012). Since climate change is an intangible phenomenon that cannot be experienced directly (Whitmarsh, 2008), the individual’s emotional judgment of it is mostly influenced either by media coverage or their social environment (Swim et al., 2009). Although it is hard to actually grasp climate change on a sensory level (Pawlik, 1991), research shows that people do detect changes in local weather patterns. For example, Howe, Markowitz, Lee, Ko, & Leiserowitz (2013) report that individuals living in places with rising temperatures are more likely than others to perceive local warming. Other studies show that people are more concerned about climate change on hot days than on cold days (Schuldt & Roh, 2014; Zaval, Keenan, Johnson, & Weber, 2014). Greater concern on hot days and less concern on cold days is also known as the local warming heuristic. However, the individual cannot infer any reasonable judgments about a complex phenomenon like climate change by simply relying on available information, such as the daily temperature (Zaval et al., 2014). The problem is that changes in local weather patterns are not necessarily reflective of long-term trends in Earth’s climate, which makes the local warming heuristic an unreliable tool for the individual’s risk judgment about climate change (van der Linden, 2015). Hence, climate change risk perception research suggests avoiding the local warming effect by instructing individuals to think about trends rather than current events (Druckman, 2015).

Socio-cultural factors. Besides the importance of cognition and affect when forming risk perception, it is important to focus on the social and cultural context in which risks are formed (Weber & Hsee, 1999). For example, interpersonal relations, mass media, and norms clearly influence the process of forming risk perception. Particularly the sources individuals get their information about climate change from and the way this information is communicated can influence risk perception (Arlt, Hoppe, & Wolling, 2011).

Socio-demographic characteristics. The literature about the influence of socio-demographics on climate change risk perception provides mixed results. On the one hand, studies have found that higher education can predict a stronger risk perception of climate change (Hornsey, Harris, Bain, & Fielding, 2016; Lee et al., 2015). On the other hand, studies did not find any education effect (Brody et al., 2008; Kellstedt et al., 2008) or even an inverse relationship between higher education and concern about climate change (Malka, Krosnick, & Langer, 2009; Slimak & Dietz, 2006). Given the inconsistent effect of socio-demographics, van der Linden (2015) mainly includes them as control variables “to assess the net influence of cognitive, experiential and socio-cultural factors on risk perception” (p. 117) in his model.

In summary, it can be said that the CCRPM provides a solid framework of what constitutes climate change risk perception. In his original version of the CCRPM, van der Linden (2015) included heuristics and biases as a fifth dimension to explain climate change risk perception. This dimension was omitted later because the heuristics and biases could be included in the existing categories. Although they are not included in the presented version (Figure 3), they help to understand how individuals form a risk perception and are therefore briefly discussed below.

Previous research found that individuals utilize a number of these heuristics and biases when forming their (risk) perception of climate change (van der Linden, 2015). Many of them result from the already mentioned abstract and intangible nature of climate change. This leaves individuals relying on other information than personal experience to build an understanding of climate change. Besides the local warming effect, individuals also seem to be unrealistically optimistic about the future (Sharot, 2011). This so-called optimism bias is indicated by overestimating the probability of positive life events while underestimating the probability of negative life events. Regarding climate change, research found that people generally believe that it is a more serious issue for other people in other places than for themselves (Gifford, Scannel, Kormos, Smolova, Biel, Boncu et al., 2009; Leiserowitz, 2005). Additionally, individuals tend to believe that climate change is more likely to happen in the distant future than in the near present (Spence, Poortinga, & Pidgeon, 2012). Therefore, the optimism bias might influence people in the sense that they might view climate change less as a personal risk and more as a global risk for others.

To further understand the concept of climate change risk perception it is important to consider that risk perception can operate on different levels. According to van der Linden (2015), it is relevant to distinguish between personal-level and societal-level risk judgments. Ignoring this consideration can result in incongruent definitions and operationalizations of climate change risk perception. Numerous studies have used different indicators of climate change risk perception. While some measured perceived seriousness about climate change, others asked how generally concerned individuals are about it (van der Linden, 2017). Yet others assessed how much individuals personally worry about the issue or even combined some of these indicators (Brody et al., 2008; Kellstedt et al., 2008; Spence et al., 2012; van der Linden, 2015). Clearly, the different indicators do not reflect the same level of concern. For example, there is a notable difference between general concern and personal worry about an issue. While the latter is an active emotional state, often linked to conforming behavioral responses aiming to reduce a particular threat, general concern can be expressed without any specific motivational or emotional content (Smith & Leiserowitz, 2014). Conclusively, an individual can be generally concerned about climate change without personally worrying about it. In other words, general concern may be a necessary but not sufficient condition for personal worry.

[...]