Alternative Approaches to Climate Change Policy Design using Understandings from Environmental Psychology and Behavioural Economics

The IPCC states in its latest, 6th synthesis report that human activities have ‘unequivocally’ resulted in a rise in the surface temperature of the earth, inducing changes in the atmosphere, oceans, cryosphere and biosphere which have had adverse effects on both the natural world and humanity (IPCC, 2023). In order to limit warming to 1.5oC, the IPCC posits that annual greenhouse gas (GHG) emissions must decrease by 45% of their 2010 levels by 2030, and drop to net zero by 2050 (Masson-Delmotte et al., 2019). At present, 65% of global GHG emissions can be directly and indirectly linked to consumption within households (Ivanova et al., 2016). Consequently, individuals can have a profound effect on the levels of GHG’s emitted into the atmosphere. Thus, changing individual behaviour within households represents a significant leverage point (Ivanova et al., 2016; Pörtner et al., 2022). It is known that further actions may be taken by individuals to reduce the carbon footprint of households (Hall et al., 2018; Steg, 2018). However, it remains unclear how to induce individuals to undertake such additional pro-environmental behaviours.

Currently, climate change policies have recruited economic understandings of behaviour to design interventions that incorporate fiscal motivations and deterrents as part of their structure (Cooper, 2011). For example, emission trading schemes (ETS), initially envisaged within the Kyoto Protocol, have been separately engineered in different regions of the world (Leining et al., 2020; Skjærseth & Wettestad, 2008; X.-Q. Wang et al., 2022). These policies have had negative consequences in the form of furthering existing inequities between countries (Liverman, 2009), as well as issues in quantifying their additionality as compared to a business-as-usual scenario (Teixidó et al., 2019; Verde, 2020). Furthermore, ETS are necessarily targeted at the social actors who participate in the engineered markets, rendering such policies ineffective for directly targeting the 65% of global GHG emissions embodied in individual/household behaviours.

The research in this PhD steps away from economic understandings of (environmental) behaviour, to explore the utility of other conceptualizations of human action. The work done here examines theories from environmental psychology and behavioural economics, to explore how psychological phenomena such as attitudes, and binds on rationality such as the recruitment of heuristics, direct decision making in relation to environmental issues. By examining alternative drivers of behaviour, it is hoped that different environmental policies directly targeting behaviour within households could be designed in the future.

Prior to the work undertaken here, the multitude of hypothesised determinates of environmental behaviour precluded the effective integration of alternative perspectives from environmental psychology and behavioural economics into policy discourses. As Shove (2010, pp. 1275) pointed put there was no ‘obvious limit to the number of possible determinants and no method of establishing their…precise role in promoting or preventing different behaviours’. Through the course of this research, multiple models of human behaviour from within and between the disciplines of environmental psychology and behavioural economics have been interrogated. These models have been examined in relation to both self-reported and observed measures of individuals’ environmental action, in an effort to delimit the critical determinates of pro-environmental behaviour - and curtail the number of possible pre-cursors lamented by Shove.

The first set of studies undertaken as part of this doctoral research focused on models of behaviour from environmental psychology. The numeracy of the models recruited within environmental psychology to explain behaviour ensured there existed considerable redundancy between these concerning the variables they incorporated into their structures. For example, biospheric values are cited in the norm activation model (Schwartz, 1977), value belief norm theory (Stern, 2000), and recent models concerning an individual’s connection to nature (Martin & Czellar, 2017). Work was initially undertaken to create a database documenting solely the variables from within these models. This was done for clarity, and to eliminate the aforementioned redundancy. Individual meta-analyses were run for each of the psychological variables to generate an effect size estimate for each. Effects size estimates were then compared between the psychological variables, to determine which of these correlated most strongly with pro-environmental actions and intentions. The five key variables with the largest (lower-bound) effect size estimates for both pro-environmental intentions and behaviour were revealed to be: attitudes, perceived behavioural control, injunctive norms, descriptive norms, and personal moral norms. Following the meta-analyses, an online study was run to investigate whether participants’ reported: attitudes, perceived behavioural control, injunctive norms, descriptive norms, and personal moral norms, were related to an observed (as opposed to a self-reported) measure of pro-environmental action. This was done to determine whether the results obtained in the meta-analyses were robust when participants could not simply report their own actions. In the online study, participants were asked how much (real) money they would like to donate to an environmental charity (the ‘Clean Air Task Force’), to observed directly whether they behaved in alignment with their proclaimed environmental beliefs, attitudes and norms. Regression analyses and structural equation modelling revealed that personal moral norms were the most critical antecedent to the observed measure of pro-environmental behaviour, with attitudes and injunctive norms also demonstrating strong relationships with monetary donations. Perceived behavioural control and descriptive norms did not demonstrate a significant direct association with participants’ charitable donation behaviours.

In a final online experiment, insights regarding human behaviour from behavioural economics were introduced. The experiment took place over two time points. At the first point in time, the three psychological variables that previously exhibited significant direct relationships with participants’ observed monetary donations (attitudes, injunctive norms and personal norms), were elicited from participants’, and their influence over behaviour was assessed in comparison with the effects of a social norms nudge. This investigation was undertaken to evaluate whether critical ingrained psychological variables (attitudes, injunctive norms and personal norms) were associated with environmental behaviour to a lesser or greater extent than an applied nudge, which manipulated the salient social norm and thus operated through the activation of a simple imitation heuristic within participants. At the second point in time the longevity of the effect of the social norms manipulation was investigated. This was done by comparing the subsequent behaviour of participants who were, and were not, exposed to the social norm nudge in part one. At time two, behaviour was evaluated in relation to the same action as before, and also in relation to a second action from a different domain of environmental behaviour. Discrepancies between participants’ who had, or had no prior exposure to the social norms manipulation at time one, were used to comment on the temporal and behavioural spillover effects of the social norms manipulation. At the first point in time, only the three psychological variables (attitudes, injunctive norms and personal norms) exhibited a significant association with the observed environmental behaviour. No effect over behaviour of the social norms nudge was observed at time one. The lack of significant influence demonstrated by the social norms nudge at time one, portended the lack of spillover effects detected for the nudge at time two. This research demonstrated the central importance of personal moral norms, attitudes, and injunctive norms for determining individuals’ substantive pro-environmental behaviours. This thesis advanced previous work by (i) using meta-analyses to summarise and compare the relationships between numerous heterogeneous psychological determinates of pro-environmental action, and self-reported environmental intentions and behaviours; (ii) advancing theory by using structural equation modelling to demonstrated that personal moral norms were more closely affiliated with an observed measure of environmental behaviour than other variables such as attitudes; and (iii) revealing the critical determinates of pro-environmental action in order to design appropriately targeted behaviour change intervention policies: the variables that most closely correlated with observed pro-environmental acts were not nudge interventions, but instead individuals internalised moral norms and attitudes. The results of this work are necessarily limited to some extent by the methodological decisions made in service of achieving the ambitions of this doctoral research. The chosen observed behaviour (charitable giving), the selected nudge (provision of information concerning a social norm), and the techniques used to elicit participant responses (online questionnaire), all represent valid techniques and constraints on the broader applicability of my findings. These should be taken into consideration when extrapolating any results of this work to alternative scenarios.