Research
Job Market Paper
Several studies show that "soft" environmental policies can alter non-targeted decisions. Still unknown is whether such spillovers are caused by merely doing the targeted pro-environmental action or if they are a pure by-product of policies. Each channel has different policy implications. We model the mechanisms underpinning both channels and provide experimental evidence for their existence. Surveying 5557 English respondents, we test if a social norm message promoting vegetarianism alters a non-targeted action: environmental donations. We find three core results. First, social norm messaging effectively increases vegetarian choices, but not uniformly. A subgroup identified with machine learning drives this effect. Second, using an instrumental variable, we find that choosing vegetarian food increases donations, providing evidence for the behavioural channel. Third, the nudge reduces donations for the subgroup identified with machine learning, providing evidence for the policy channel. Our results suggest social norm messaging works at the risk of crowding out further pro-environmental decisions.
Published works
Banerjee, S. & Picard, J. (2023). Thinking Through Norms Can Make Them More Effective. Experimental Evidence on Reflective Climate Policies in the UK (Journal of Behavioral and Experimental Economics, full text here)
Adopting low-carbon diets is important to meet our climate goals. Prior experimental evidence suggests green nudges help people adopt such diets, more so when they are encouraged to think through them. In this paper, we re-evaluate this role of reflection in a “social norm” nudge to promote intentions for climate-friendly diets in the United Kingdom. Using 5,555 British respondents, we find the social norm nudge increases meal order intentions for low-carbon diets versus the control condition. Asking people to reveal their personal dietary norms, after exposing them to these social norms (“lower-order nudge+”), does not produce any measurable change compared to the nudge. However, when people are subsequently encouraged to think and pledge to climate-friendly diets (“higher-order nudge+”), the effectiveness of the social norm nudge increases by 90% or more.Picard, J., Kobrich, A., Schobin, J. (2024). Experimental evidence on the role of framing, difficulty and domain similarity in shaping behavioral spillovers (Scientific Reports, full text here)
We develop a between-subject online experiment to study when encouraging a climate-friendly behavior spur or hamper further engagement for the environmental cause. First, respondents read a newspaper article dealing with a first pro-environmental behavior (helping scientists as a new form of climate activism). Second, participants choose whether to undertake the behavior emphasized. Third, respondents are offered to do a second, subsequent, pro-environmental behavior (signing a petition to call for political action against climate change). In study 1, we manipulate the narrative of the newspaper articles. It either condemns individual inaction, or encourage being pro-environmental. We compare these narratives against a neutral narrative (control) and an article dealing with an unrelated subject (placebo). In study 2, we investigate how the difficulty of the first pro-environmental behavior moderates behavioral spillovers. In study 3, we test the moderating effects of behavioral distance, i.e. the domain-similarity of the two subsequent behaviors.Working Papers
We act pro-socially whenever we forgo immediate hedonic pleasure to do what is right (e.g., being pro-environmental, helping others). I model pro-social decisions as either intrinsically motivated (done as an end to themselves) or extrinsically motivated (done as a means to an end). Individuals' pro-social nature evolves with the context and their past choices. This model reconciles several contradicting findings in the social psychology literature. It also provides micro-foundations to behavioural spillover effects, backfiring effects, and the heterogeneity of pro-social policies.
Ongoing Projects
Picard, J., Bonan, J., Cerroni, S. Green nudges and information avoidance, an experimental investigation with European farmers
We investigate why farmers actively avoid climate change information and how it affects the effectiveness of information campaigns. We administered a preregistered experiment on a sample of European farmers and foresters from Italy, Belgium, Lithuania, and Finland. We collected two samples: one where respondents could skip climate change information and another composed of a control group and a treatment group where respondents are shown climate change information. Around 34% of farmers and forest owners actively avoid the information in the first sample. Distrust in scientists and low pro-environmental attitudes correlate positively with information avoidance. The information nudge increases respondents' willingness to implement environmentally sustainable practices in the second sample. Using machine learning, we find that those with a similar profile to the information seekers of the first sample drive the effect of the nudge. Our results suggest that the null effects of information nudges found in the literature might be explained by information avoidance, creating hidden heterogeneity.
Picard, J., Bonan, J., Cerroni, S. Unpacking the causes of information avoidance, an experiment on climate change and meat consumption
Why do people avoid climate change information, and how does it affect the effectiveness of information campaigns? We use economic theory to explain why some people voluntarily ignore such information. Then, focusing on the impact of meat-based diets on the environment, we design an experiment to identify the different "types" predicted by the model. Finally, we use predictive machine learning algorithms to identify these types in another experimental sample to understand how these different profiles react to an information nudge.
Picard, J., Balmford, B., Banerjee, S., Groom, B. Pro-social motives and the Effectiveness of matching subsidies
Research shows that subsidising donations increases support for charitable causes but crowds out donations from those who would have given anyway. We investigate whether pro-social motives drive these effects. We consider two types of pro-social motives: deontological (“I support the cause for itself”) and consequential (“I support the cause for the benefits it provides to society”). We design a within-between-subject experiment in which respondents give to biodiversity conservation charities. In the within dimension, we vary the subsidy to estimate the elasticity of respondents’ donations with the subsidy. In the between dimension, we prompt respondents to think about the risk of biodiversity depletion with different framing. Namely, we either emphasise the right to live of species (deontological framing), their usefulness for humans (consequential framing), or stays neutral (control). We investigate whether the elasticity varies with the framing.