Causal thinking and support for climate change policies: International survey findings

Abstract

Few comparative international studies describe the climate change policies people are willing to support and the reasons for their support of different policies. Using survey data from 664 economics and business undergraduates in Austria, Bangladesh, Finland, Germany, Norway, and the United States, we explore how perceived risk characteristics and mental models of climate change influence support for policy alternatives. General green policies such as funding research on renewable technologies and planting trees were the overwhelmingly most popular policy alternatives. Around half the students support carbon reduction policies such as requiring higher car fuel efficiency and increasing taxes on fossil fuels. Least popular were engineering alternatives such as fertilizing the oceans and replacing fossil fuels with nuclear power. Variations among nations are generally small. Support for different policy alternatives corresponds with different causal thinking. Those who hold a pollution model of the causes of climate change, tend to blame environmental harms (e.g., air pollution from toxic chemicals), see general green policy alternatives as effective, and support general green policies. Support of carbon reduction strategies is associated with seeing carbon emissions as the cause and reducing carbon emissions as effective solutions. Support of engineering solutions increases with identifying volcanoes among causes and regarding engineering solutions as effective. Although these international students agree that climate change is a threatening problem, their causal thinking correlates with support for different mitigative policy actions, with the most popular ones not necessarily the most effective.

Introduction

Addressing climate change successfully will require effective public policies to reduce and delay global warming. Governments do not necessarily need broad public support to implement policies designed to reduce greenhouse gases or take other steps to limit climate change (Sabatier and Jenkins-Smith, 1993), but public support facilitates the adoption and implementation of strong policies. The purpose of this research is to explore the nature and origins of support for different types of policies designed to reduce climate change. Although there is much research that reports U.S. public opinion on the realities of global warming (see Nisbet and Myers, 2007), there is a paucity of comparative international studies that describe what policies citizens are willing to support and why they are willing to support those policies. Do people understand climate change, its causes, consequences, and the effectiveness of proposed policies? Once we take into account people's fears and attitudes toward climate change, does their thinking about the causes of climate change and the effectiveness of proposed solutions to the problem influence their willingness to support different types of policies?

Despite widespread speculation and a mounting number of opinion polls and surveys assessing perceptions of climate change (e.g., Leiserowitz, 2006, Leiserowitz et al., 2010), little research has explored the role of causal beliefs in support for climate change policies. Notable exceptions to the above include the work by O’Connor et al., 1999, O’Connor et al., 2002, which shows that knowledge matters, and the recent paper by Shwom et al. (2010), exploring respondents’ own (written) open-ended explanations for their reported preferences over a set of possible climate change policies. Six of the eight survey items they use to elicit policy preferences are derived from O’Connor et al., 1999, O’Connor et al., 2002. All of the survey items in the Shwom et al. study load reliably on a single scale, which is perhaps not surprising as they all directly address carbon dioxide emission reduction, in one form or other. Mental models of climate change, however, are considerably more varied in their assessment of climate change causality (Bostrom et al., 1994, Bostrom and Lashof, 2007, Böhm and Pfister, 2001, Kempton, 1991, Kempton et al., 1995, Read et al., 1994, Reynolds et al., 2010, Whitmarsh, 2009b; see also Löfstedt, 1991). Commonly identified mental models that do not focus on the role of greenhouse gas emissions and their effects per se include a “pollution” model, in which global warming appears to be attributed to all kinds of pollution fairly indiscriminately, and a stratospheric ozone depletion model, according to which the ozone hole and stratospheric ozone depletion are causing global warming by letting more UV radiation into the atmosphere. These studies suggest that mental models play a role in support for potential climate mitigation policies; those who hold stratospheric ozone depletion models and think aerosol spray cans are to blame also support stopping the use of aerosol spray cans as a way to combat global warming.

While numerous studies have found correlations of single questions or indices of climate change beliefs with energy preferences or showed that knowledge about climate change plays a role in individual behaviors (e.g., Greenberg and Truelove, 2011, Lorenzoni et al., 2007, Pidgeon et al., 2008, Visschers et al., 2011, Whitmarsh, 2009a), only a few studies (e.g., O’Connor et al., 1999, O’Connor et al., 2002) have demonstrated how causal thinking systematically influences policy preferences in the environmental realm. We look at three classes of causal thinking: (a) dimensions of understanding the causes of climate change, (b) how likely climate change is to cause various consequences, and (c) effectiveness of proposed solutions to reduce and delay climate change.

Characteristics of risks, such as how threatening or dread the risk is, are predictive of risk acceptance (e.g., Fischhoff et al., 1978, Slovic et al., 1980, Slovic, 1987, Slovic, 2000). Perceived consequences, benefits, controllability (e.g., McDaniels et al., 1995, McDaniels et al., 1996) and morality (Lazo et al., 1999, Böhm and Pfister, 2001) are among the risk characteristics that appear important to explain variance in risk attitudes and support for risk mitigation policies. The long-researched psychometric paradigm for risk perceptions (e.g., Fischhoff et al., 1978, Slovic, 2000), and extensions of this line of research (e.g., Lazo et al., 1999, Böhm, 2003, Böhm and Pfister, 2001) include dimensions that characterize the perceived nature of the risk, which have also been identified as relevant risk characteristics in the domain of climate change (Lorenzoni and Pidgeon, 2006): (a) whether the risk engenders dread, (b) the degree to which the risk is understood, (c) the degree to which the consequences of the risk are controllable, (d) whether people feel moral concerns related to the risk, and (e) whether people feel that issues of equity are related to the risk.

We hypothesize that there are people who dread the consequences from uncontrollable, poorly understood climate change and have moral and equity concerns related to climate change. We think they will support policies to address climate change, although we are uncertain which sort of policies they will prefer. Previous research (e.g., Leiserowitz et al., 2009) has shown that people who anticipate a dystopian world of dire consequences to society from climate change are likely to support efforts to avoid such catastrophic outcomes. Once we take into account how people think about the characteristics of climate change does their causal thinking explain what policies they are willing to support?

In sum, our basic assumption is that perceived risk characteristics together with causal beliefs shape policy preferences. We generally expect that people tend to support policies which they expect to be effective in reducing the highest perceived risks. We are, however, uncertain about the specific relationships between these variables and the relative weights of risk judgments and causal beliefs in predicting policy support.

Previous research has identified international differences with respect to the extent to which people have heard about climate change, attribute the problem to human actions, and feel threatened (Leiserowitz, 2007/2008, EC, 2009). Older research (Inglehart, 1995) found that mass support for strong environmental policies is greatest in countries that face the highest objective risks (e.g., Bangladesh for climate change) and in wealthy countries with high levels of postmaterialist values (e.g., the Nordic countries). An international climate change survey in 28 countries in 2001 found that 60% of respondents overall and majorities in developing countries (e.g., China and India) as well as in developed countries (e.g., Australia and Italy) would be willing to pay a ten percent tax on petrol to reduce air pollution, with least support in France and South Africa (Globescan 2001 as cited by Leiserowitz, 2007/2008). Research in the United States has found that people in more vulnerable places are more likely to rate climate change as risky (Brody et al., 2008) and more likely to support policies to address climate change (Zahran et al., 2006). Thus it is plausible to assume that countries differ in the absolute level at which their residents endorse certain risk characteristics or causal beliefs, or support certain policies. Nevertheless, with respect to the relationships between these various constructs, we do not expect much variation across countries. Further, our data are not representative samples for Austria, Bangladesh, Finland, Germany, Norway, and the United States, but economic and business majors at universities. As described in greater detail in the next section, by limiting the sample to economic and business students, we have increased the homogeneity of educational experiences in order to see if international differences found in other research will remain among subjects with similar educational training. The alternative hypothesis is that some beliefs are shared globally among students trained in the same popular major.

We explore these issues with data from Austria, Bangladesh, Finland, Germany, Norway, and the United States. The selection of these six countries is not the result of a conscious decision, but a serendipitous consequence.¹² The dataset is a convenience sample that cannot serve for inferences to students in those countries, much less to adult public opinion. Still, on variables duplicated in scientific surveys the results seem broadly consistent. The findings are suggestive and we hope have heuristic value.