Research (poster at the bottom!)
Research (poster at the bottom!)
We explore the interplay between two key drivers of individual green consumption: environmental preferences and social image concerns, when both conspicuous goods and green goods contribute to social image. We consider four lifestyles crossing two characteristics (Green/Brown) x (Conspicuous/Ordinary), all differing in environmental impacts and status benefits. Our working hypothesis is that the more one cares about the environment, the more one weighs green consumption relative to conspicuous consumption in social image benefits, allowing environmentally sensitive agents to achieve moral consistency. Our model generates two general insights under such type of preferences. First, under moral consistency in image benefits, the environmental impact of consumption does not necessarily decrease monotonically when environmental preferences increase. This impact depends on relative impact-intensities, i.e., on pollution per dollar rather than per unit of goods, giving rise to possible volume effects and behavioral rebound. When Green Conspicuous goods have relatively low impact intensity, increasing environmental concern above some threshold may paradoxically raise pollution when image is morally consistent. Second, Green Conspicuous goods smooth the behavioral change from Brown Conspicuous to Green Ordinary lifestyles when environmental preferences change, offering individuals with intermediate environmental concern and strong image motives greener options while maintaining social image from conspicuous consumption. These results show that image-driven green consumption does not automatically reduce environmental damage, and highlight both the potential and the limits of conspicuous green goods in reconciling Eco and Ego in consumer behavior.
Click here to read the latest version.
Presentations of this work: FAERE Doctoral Workshop (March 2024), Internal PhD Seminar (June 2024), FAERE Annual Conference (September 2024), 30th EAERE Conference in Bergen (June 2025), 24th Journées LAGV in Marseille (July 2025), 14th Mannheim Conference on Energy and the Environment (May 2026), World Congress of Environmental and Resource Economists (WCERE) in Lisbon (July 2026)
We investigate how social image concerns and pro-environmental motivation shape vegetarian food choices using a lab-in-the-field experiment in a university canteen (N=238, with N=118 treated). Participants chose between meat and vegetarian dishes, including an organic vegetarian option that was served on a distinctive green tray in treated groups, increasing the visibility of this choice. We find weak evidence of a positive average treatment effect on vegetarian choices, with no effect on the organic vegetarian option. This average effect masks substantial heterogeneity: the effect is concentrated among individuals with stronger image-oriented behaviors, again only on the vegetarian margin. This suggests the intervention shifted the decision to avoid meat rather than the willingness to pay for a more visible organic choice. This average effect is sensitive to how baseline meat consumption is measured: it holds only under the pre-registered revealed measure from canteen records, not the stated survey measure, pointing to social desirability bias in self-reports. This self-report bias shows the necessity to measure treatment effects by controlling at least for some revealed preferences, not purely stated ones. The main heterogeneity pattern is more consistent with heightened environmental salience or partial image crowding-out than with pure social shaming. A positive association between the perceived meat descriptive norm and vegetarian choice, that is also supported by survey evidence, further suggests norm-sensitive students are distinction-seeking rather than conformists. Overall, our findings suggest that visibility-based nudges for prosocial behavior may operate through heterogeneous psychological channels, highlighting the importance of accounting for individual differences in image-related motivations when designing behavioral interventions.
Presentations of this work: Paris Junior Experimentalists Meetings (November 2024), Environmental Economics Seminar, University of Gothenburg (May 2025), CREST-PSAE Workshop (April 2026), PSAE Internal Seminar, Journées LAGV in Aix-en-Provence (June 2026), 16th Annual Conference of the French Association of Experimental Economics (ASFEE) in Lille (June 2026)
Upcoming presentations: 13th Annual Conference of the French Association of Environmental and Resource Economists (September 2026)
This paper develops a sequential model of pro-environmental (prosocial) behavior with imperfect self-knowledge and self-signaling. Individuals differ in their intrinsic prosocial motivation and choose sequentially between two pro-environmental actions. Individuals differ in their intrinsic prosocial motivation and choose sequentially between two pro-environmental actions. In addition to deriving material benefits from contributing to a public good, agents value seeing themselves as intrinsically motivated. Building on the self-signaling framework of Bénabou and Tirole (2011), we characterize the Perfect Bayesian Equilibria and identify conditions under which motivational types separate or pool as a function of material costs and the relative efficacy of the two actions.
The key result is that consistency-driven positive behavioral spillovers between prosocial actions can imply policy spillovers of both signs, as increasing incentives for an initial prosocial action may not monotonically increase the propensity to take subsequent prosocial behavior, even when people seek consistency in their self-image. The effect depends on the interaction between relative efficacy, material costs, and intrinsic motivation, through two channels: updating of self-beliefs about one's own type and changes in the material returns to subsequent prosocial behavior. As a result, even a policy systematically increasing adoption of a target action may entail negative net spillovers on the non-target action, with the sign potentially varying across motivational types and economic contexts. More broadly, the paper provides a self-signaling framework for understanding how policies targeting one prosocial behavior can lead to heterogeneous and sometimes counterintuitive behavioral effects on subsequent prosocial actions.
Recently presented at the PSAE PhD Seminar and at the Parisian PhD Seminar in Environmental Economics.
Next presentation: Workshop on Cognitive Economics in Herrsching am Ammersee (August 24-25)