Abstract
This meta-analysis investigates real-world nudging interventions to reduce meat consumption, uniquely focusing on 33 field experiments with 78 effect sizes that use revealed preferences to minimize attitude-behavior bias. Key findings indicate that the effectiveness of nudging interventions is highly heterogeneous (I2 = 99.97 %), with an estimated prediction interval ranging from −50 % to 28 %. Therefore, our focus is on the conditions that moderate potential nudging effects in food service: Interventions targeting decision structure yield a substantial effectiveness with reductions in meat consumption up to 30 %, among them, choice defaults with a 54 % reduction on average. Conversely, strategies based on information provision or decision assistance were found to be ineffective in promoting behavioral change. Our study also identifies a concern: approximately 30 % of interventions resulted in a backfire effect, leading to increased meat consumption. This proportion is higher than reported in some previous meta-analyses and could be attributed to less controlled field study environments or psychological reactance. Future research should continue to explore the complex interplay of moderating variables, investigate the long-term persistence of default effects, examine potential compensation behaviors, and analyze how individual values and attitudes might further moderate nudging success. Understanding these nuances will be crucial for designing and implementing truly effective and context-sensitive strategies to promote more sustainable and healthy dietary patterns.