The relationship between country and individual household wealth and climate change concern: the mediating role of control

Abstract

The relationship between wealth and climate change concern has been a focus of several studies. In this article, we hypothesize that richer households (and countries) are less concerned about climate change because wealth provides a buffer against some of the related risks. This leads people in wealthier countries and households to perceive a greater sense of control over climate change impacts, which in turn results in lower levels of concern. We tested this hypothesis using a unique household survey encompassing 11 OECD countries and over 10,000 households and applying mixed multi-level regression models. Our results confirmed a statistically significant negative relationship between country and household wealth and individuals’ perceptions of the seriousness of climate change. This study contributes to current literature by showing that this relationship is mediated through sense of control, measured at the country level by the country’s readiness index and at the individual household level by the extent of adoption of energy efficiency improvements. These findings raise the question of how best to incentivize action on climate change amongst those with the ability—but not necessarily the motivation—to respond.

Appendices

Appendix A

See Tables 4 and 5.

Appendix B

The final specification of Model 1 is the following:

${\text{CCbelief}}_{i,c} = \beta_{0,c} + \beta X_{1,i} + \beta_{c} X_{2,i} + \gamma X_{3,c} + \varepsilon_{i,c}$, with i, the index for the respondent or household, and c, the country index. The dependent variable, ${\text{CCbelief}}_{i,c}$, measures respondents’ level of climate change concern on a scale from 0 to 10. The best specification (as shown in Table 2) features a random intercept ($\beta_{0,c}$); fixed effects ($\beta$) for variables in the $X_{1,i}$-vector of respondent/household characteristics; random effects ($\beta_{c}$) for variables in the $X_{2,i}$-vector of respondent/household characteristics, and fixed effects ($\gamma$) for the country-specific characteristics gathered in the $X_{3,c}$-vector. We have:

$X_{1,i}$ = (male, age, higher education, personal safety ranking, local environment satisfaction, children, and energy use monitored);

$X_{2,i}$ = (trust experts, charity giving, conservative ideology, income, urban location); $X_{3,c}$ = (gross national income, environmental performance index, extreme climate %, floods, mean temperature − 5 years previous-), mean temperature ratio − 100 years previous-), right party ruling).

Model 2 is the same as Model 1 except that it includes the mediating variable (overall readiness score) in $X_{3,c}$.

Models 3 and 4 feature fixed effects only and include both regional and country-specific variables (see complete list in Table 3).