Wives influence climate change mitigation behaviours in married-couple households: insights from Taiwan

Mitigating climate change requires collective action of various sectors and on multiple scales, including individual behavioural changes among citizens. Although numerous studies have examined factors that influence individuals’ mitigation behaviours, much less attention has been given to interpersonal influence. Children have been suggested to influence parents’ climate change concerns; however, how the interactions between couples—typically the primary decision-makers in married-couple households—influence each other’s climate change concerns has seldom been discussed. In this study, we surveyed married heterosexual couples to investigate the interdependency of husbands’ and wives’ motivations for behavioural change to mitigate climate change. We found that wives’ psychological constructs, including climate change risk perception, self-efficacy, and gender role attitudes, demonstrated stronger effects on their husbands’ motivation than did husbands’ own constructs on their own motivation, whereas husbands’ psychological constructs did not influence their wives’ motivation. Our results suggest the importance of wives’ role in motivating household climate change mitigation behaviours.


Introduction
A main challenge for the human society in meeting the objectives of the Sustainable Development Goals and Paris Agreement is behavioural change in both public and private sectors (Yeeles 2019). Most studies have employed individualistic approaches to understand factors that promote or inhibit individuals' motivations and the associated mitigation behaviours, including many socio-ideological factors, such as climate change risk perception, self-efficacy, worldviews, and environmental attitudes (Clayton et al 2015). Nevertheless, several recent studies have focussed on the influence of interpersonal relationships on climate change beliefs and mitigation behaviours (Hannibal and Vedlitz 2018, Goldberg et al 2019. Many mitigation behaviours are performed within households; thus, how household members discuss, communicate, make decisions related to, and perform mitigation behaviours warrant investigation (Head et al 2016, Hung 2017. Although a recent study demonstrated how educational interventions of children provide an age-related window that influences climate change concerns among their parents (Lawson et al 2019), studies investigating how interactions between couples influence each other's climate change beliefs and associated behaviours remain scant. Investigating this topic is crucial because married-couple households remain the prevalent household type in many countries worldwide (Child Trends 2015) and married couples are typically the primary decisionmakers for tasks in their households (Treas and Tai 2012).
Here, we performed an actor-partner interdependence model (APIM)-based (Kenny et al 2006, Kenny 2015 survey research on the influence of husbands and wives on each other's motivations for behavioural change to mitigate climate change. APIM is a commonly used model to examine the interrelationships of dyads (or two individuals) (Cook and Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Kenny 2005). APIM estimates 'the extent to which the independent variable of a person influences his or her score on the dependent variable' (Fitzpatrick et al 2016: 75) and 'the extent to which the independent variable of a person influences the dependent variable of his or her partner' (Fitzpatrick et al 2016: 75). We first examined the actor effects, that is, whether the husbands' and wives' psychological constructs predict their own motivation. Next, we examined the partner effects, that is, whether the husbands' and wives' psychological constructs predict their spouses' motivation. Because the two effect types are considered simultaneously in APIM, the actor effects can be estimated while controlling for the partner effects and vice versa (Tambling et al 2011). We used the following constructs in our analysis: climate change risk perception, self-efficacy, and gender role attitudes. Climate change risk perception and self-efficacy are considered the drivers of motivation for behavioural change to mitigate climate change (Brody et al 2012), whereas gender role attitudes is the main factor that influences household behaviours (Davis andGreenstein 2009, Horne et al 2018). Thus, in this study, we individually investigated the dyadic relationships of spouses' motivations to mitigate global climate change with climate change risk perception, self-efficacy, and gender role attitudes, respectively.

Methods
We employed quantitative methods by conducting face-to-face interviews of married heterosexual couples living in Taipei City, Taiwan, during March and April, 2019. In total, 152 married couples answered survey questions on the aforementioned three constructs and their motivations for behavioural change to mitigate climate change. More details regarding the methods, including ethics statement, sampling methods, sample characteristics, questionnaire development, and data analysis are available in the supplementary information online at stacks.iop.org/ ERL/14/124034/mmedia.

Results
First, for the climate change risk perception-motivation relationships ( figure 1(a)), the actor effects were positive and significant for both husbands (standardised effect=.195, p=.01, r=.208) and wives (standardised effect=.553, p<.001, r=.431); by contrast, the partner effects from wives to husbands were positive and significant (standardised effect=.218, p=.006, r=.22), but not those from husbands to wives (standardised effect=−.148, p=.101, r=.134). Second, for the self-efficacymotivation relationships ( figure 1(b)), the actor effects were positive and significant for wives (standardised effect=.552, p<.001, r=.451), but not for husbands (standardised effect=.09, p=.241, r=.096). By contrast, the partner effects from wives to husbands were positive and significant (standardised effect=.313, p<.001, r=.314), but not those from husbands to wives (standardised effect=.012, p=.888, r=.012). Third, for the gender role attitudes-motivation relationships ( figure 1(c)), the actor effects were significant for neither husbands (standardised effect=−.079, p=.358, r=−.075) nor wives (standardised effect=.09, p=.393, r=.07); similarly, the partner effects from husbands to wives were nonsignificant (standardised effect=−.157, p=.141, r=−.12), but those from wives to husbands were positive and significant (standardised effect=.189, p=.028, r=.178). As a higher score for gender role attitudes reflects a more traditional attitude, the positive partner effects from wives to husbands indicated that with their stronger traditional gender role attitudes, the wives increase the motivation of their husbands.

Discussion
The results of our analysis highlight the important role of wives in facilitating household climate change mitigation behaviours by influencing their own and their husbands' motivations for behavioural change to mitigate climate change. Husbands' motivation is influenced by their own climate change risk perceptions and self-efficacy; notably, it is also influenced by their wives' climate change risk perceptions and selfefficacy, and these partner effects are even stronger than husbands' own actor effects. Furthermore, husbands' motivation is not influenced by their own gender role attitudes, but by their wives' gender role attitudes. This may be due to both husbands and wives associating 'household' climate change mitigation behaviours with traditional women's tasks within households and thus making the wives the main initiators for household climate change mitigation behaviours (Organo et al 2013). Many studies on household decision-making have suggested that the person who brings resources to the households is likely to be the primary decision-maker to household tasks (Belch andWillis 2002, Xu andLai 2002). By contrast, here, only 1% (n=2) of the study population indicated wives as the primary financial contributor to their households; moreover, 71% (n=119) and 27% (n=41) indicated husbands as the primary financial contributor and husband and wife as equal contributors, respectively. In other words, wives' role in affecting their husbands' motivation is not associated with the wives' financial contribution to their families.
The crucial role of wives' constructs in influencing their own and their husbands' motivation has implications for policymakers and practitioners. The promotion of mitigation behaviours by overcoming individuals' barriers remains crucial (Gifford 2011), as suggested by the current significant actor effects for both husbands and wives. On the other hand, because of the significant partner effects for both husbands and wives, considering interpersonal perspectives and understanding how household members influence each other-particularly the influence of wives on husbands-in terms of climate change beliefs or climate change mitigation behaviours could be more important. Climate change-related education, -information, and-mitigation campaigns that target wives are likely to be more efficient in increasing the motivation for behavioural change to mitigate climate change than are those that target husbands.

Conclusion
Our study indicated the importance of considering interpersonal perspectives and examining intrahousehold dynamics (Hung 2017) in understanding the motivation for behavioural change to mitigate climate change, a key antecedent to actual climate change mitigation behaviours (Kollmuss and Agyeman 2002). Our APIM-based analysis and its results for identifying the roles of wives in motivating behavioural changes is straightforward, we suggest that future studies include a large sample size and divide their samples into groups based on different factors, such as climate change beliefs and political ideology (Lawson et al 2019). In addition, our study population was exclusively from Taiwan, which is heavily influenced by Confucianism (Zhang et al 2005). Thus, our results are mostly generalisable in the contexts of other East Asian cultures. It should however be noted that other factors such as income disparity between husbands and wives, and differences in national education systems should be taken into account when replicating this study in other East Asian contexts. Additional studies replicating this study in other cultural contexts to understand the effects of cultural differences are warranted. Furthermore, because household structures worldwide are becoming diverse (Van de Kaa 1987), investigating how intrahousehold dynamics affect mitigation behaviours in other household types is essential. Finally, as the classic APIM framework only included 4 variables in the analysis (2 variables from each of the two actors), it is hard to unveil the complex intrahousehold dynamics that affect household mitigation behaviours. Future studies could use structural equation modelling to include more variables from both husbands and wives, or apply qualitative methods to uncover the social and cultural processes underpinning household decision-making and mitigation behaviours.