Abstract: Meat consumption and the willingness to reduce it is closely linked to one's political ideology and convictions. Contrary to standard beliefs suggesting that left-leaning individuals are predisposed towards meat reduction policies, more than those in the center or right, we show these political divisions depend on the specifics of the policy and shouldn't be taken as the absolute norm. Using a conjoint experiment involving nationally representative samples from Denmark, Germany, and the Netherlands (N=1,505), we find no significant differences in support for a hypothetical meat-tax policy among participants of varying political ideologies. Specifically, a shift towards the ideological right is only associated with diminished support when the policy includes a direct or indirect revenue recycling mechanism, accounts for the negative impacts of meat production in its costs, and emphasises environmental concerns over health. These disparities in political ideology are further explained by differing levels of Schwartz's human values. Based on our findings, we argue that the politicisation of meat should acknowledge these nuanced differences, as electoral support may not uniformly align across all domains.

Abstract: Virtual Reality (VR) enables immersive experiences that can enhance awareness about environmental problems. We measure the effectiveness of VR versus 2D in an environmental campaign using a field experiment in Brazil. 617 passers-by at a mall were randomly assigned to watch a video clip about the Amazon Forest through VR or a traditional 2D device, with some being randomly interviewed before watching the movie (control group). We find that both the 2D and VR interventions increase individuals' propensity to i) contribute to an Amazonian humanitarian campaign, ii) share contact information, iii) interact with a conservation campaign, and iv) state pro-conservation opinions. We find no additional impact of VR compared to 2D, but VR participants were more likely to engage with pro-conservation content online three months later. Our findings provide cautionary evidence about the additional potential of using immersive technologies, like VR, to improve conservation behaviors compared to 2D methods.

Abstract: Vegetarian diets can reduce global ecological costs. Yet meat continues to be culturally dominant, especially in the West. Does culture hamper the adoption of vegetarian diets? In an online experiment, we observed the intentions of 2,775 English participants to choose vegetarian food. We causally test if framing food options with culturally familiar names alters the effectiveness of a nudge promoting vegetarianism. Facing culturally familiar food does not change the effect of the nudge. However, exploratory analyses reveal that participants ask for lower monetary compensation for being forced to choose vegetarian food when it is culturally framed. Thus, welfare losses of hard policies, like bans, may be lower when alternative options look familiar.

Abstract: Behavioural nudges, such as defaults, improve human welfare by reducing choice overload. But how far should policy-makers go in reducing available choices?  Using a preregistered, online survey experiment, we randomly assign 1,518 UK citizens to two versions of the same opt-out default nudge aimed at encouraging charitable donations. In one version, the default was embedded in a list of multiple options, whereas in the other it was presented as a single choice. In both versions, participants could opt-out and choose any preferred donation amount. We find that when the default is embedded in a multiple-choice list, opting out rates and average individual donations are significantly higher compared to the standard single-choice default. Our findings suggest that encouraging active choosing alongside a default can improve agency without necessarily deteriorating behavioural outcomes.

Abstract: Respondents from around the world are financially incentivized to complete surveys in higher paying markets, and some respondents are even willing to falsify their credentials to do so. We propose a new cost-free method for identifying and excluding non-target respondents who see blank JavaScript-programmed experiments due to VPN usage. Using a dataset of 72,200 respondents in the G-7 countries, we show that 94 respondents (0.13\%) in our entire G-7 sample and 49 respondents (0.49\%) who report being from the United States are likely using virtual private networks (VPNs) to disguise their locations to qualify for surveys outside of their country. These non-target respondents provide low-quality responses, affecting external validity. Simple JavaScript programming can reject this rare type of non-target respondent from surveys administered using Qualtrics. 

Nudge+, a nudge for good thinkers.  Joint with Peter John. (Second R&R)