Which theoretical approach most effectively reduces consumption of meat and animal products (MAP)? We answer this by characterizing and meta-analyzing the most rigorous, policy-relevant MAP reduction research: randomized controlled trials with at least 25 subjects in treatment and control that measure MAP consumption directly at least a single day after treatment begins. Although the theoretical space of MAP reduction literature is vast, the 26 papers and 59 studies meeting our inclusion criteria (N = 38,071) embody five approaches: nudges, social norms appeals, and persuasion efforts centered around the environment, health, and animal welfare. We find an overall null effect of the assembled interventions on MAP consumption (59 studies, ∆ = 0.02). While studies aimed at reducing consumption of red and processed meat are substantially more effective (12 studies, ∆ = 0.22), we cannot say anything definitive about the net effect of these interventions on animal welfare or the environment due to a lack of data on substitution to other MAP products. While nudges appear to be the most effective approach on average (6 studies, ∆ = 0.269), this effect is highly uncertain because of a universal lack of follow-up data on potential compensatory behavior, i.e. eating more meat at a subsequent meal. Overall, no approach leads to consistent, robust behavioral changes. In light of these findings, we advise scholars to design future studies to remedy existing measurement gaps and to evaluate promising approaches that have yet to be rigorously tested, such as price changes or time spent with farm animals.

Seth Green is a non-resident fellow at the Kahneman-Treisman Center at Princeton University and an affiliate of the Humane and Sustainable Food Lab at Stanford. He has written two previous meta-analyses and two papers on computational reproducibility. He is broadly interested in the intersection between emerging technology and scientific credibility. He has an M.A. in political science from Columbia University.