Speaker
Description
Keywords: Nutrition behavior, Gender differences, Ethical interventions, choice architecture, Health clinic cafeteria, Sustainability, Decision-making autonomy
Introduction and Objectives: This study delves into the evaluation of two distinct interventions designed to curb meat consumption within cafeteria settings: the forced active choice (requiring explicit meat portion selection) and the default nudging (providing a reduced meat portion by default but allowing an opt-out option). While the former tends to prioritize individual decision-making autonomy, the latter often proves more effective in terms of influencing behavioral change. The primary objectives are to assess the efficacy of these interventions and their acceptability among cafeteria patrons. The study was pre-registered with a focus on three key questions related to reducing meat portions: (a) whether forced active choice is significantly less effective than default nudging, (b) whether both interventions differ significantly in terms of acceptance among cafeteria users, and (c) whether the interventions are significantly more effective for certain types of meat-based dishes.
Methods: Our field study was conducted within the actual confines of a health clinic cafeteria and spanned three distinct observation phases, each lasting six weeks: a baseline phase and two intervention phases. The study encompassed a total of 5,966 food choices made by cafeteria visitors. The forced active choice intervention prompted individuals to select either standard or reduced meat portion sizes. In contrast, the default nudge intervention automatically set the default option to reduced meat portions while informing individuals on the choice to opt-out. Data collection was performed manually by our research team using tablets, with team members disguised as chefs positioned behind the serving stations. Following the intervention phases, we surveyed 125 cafeteria users to assess their acceptance of the interventions. We evaluated the intervention's effects through chi-square tests, a Wilcoxon matched-pairs signed-rank test, and coefficient plots generated from linear regression analysis.
Findings: The results to each research question (a-c) were as follows: (a) Both interventions were effective in increasing the selection of reduced meat portions across all tested meat dishes. However, the default nudge intervention had a significantly greater impact, with 90.6% of individuals opting for reduced portions (Cohen’s d =2.73 vs. baseline), compared to 38.5% in the active choice condition (Cohen’s d=0.7 vs. baseline). The active choice intervention exacerbated gender differences in selecting reduced meat portions (48% for women vs. 13% for men), while the baseline (13% for women vs. 3% for men) and the default nudge mitigated these differences (95% for women vs. 81% for men). (b) While active choice was slightly more accepted (rated at 6.38 out of 7) than the default nudge (rated at 6.08 out of 7), both interventions were perceived as ethical by cafeteria users in this study. (c) Additionally, we observed variations in the effectiveness of the default nudge depending on the type of meat dish. Meat-centered dishes (e.g., steak) responded more favorably to the default nudge compared to mixed-meat dishes (e.g., Bolognese sauce). For meat-centered dishes, kitchen chefs believed that sensory quality could be maintained despite more ambitious reduction targets, i.e. lowering the proposed meat quantitiy by 50% as compared to an average reduction of 35% in mixed dishes. Nevertheless, cafeteria users followed the chef's default choice to a similar extent across all meat dishes.
Discussion and Conclusions: This study underscores that choice architects may sacrifice effectiveness for consumer autonomy when favoring active choice in a setting which is accepting of defaults. The findings illuminate this trade-off, emphasizing that the choice of intervention hinges on the legitimacy of the choice architect's actions in pursuit of a goal and the degree to which that goal is desirable for the group of individuals being influenced. We also emphasize that the survey, which may lack sufficient statistical power, does not reveal lower acceptance for defaults, highlighting the possible superiority of intrusive approaches here. Additionally, our research reinforces the idea that transparent nudging is a morally superior approach and, in some cases, can prove equally or even more effective. It's worth noting that altering meat portions in meat-centric dishes is typically more conspicuous compared to mixed dishes, potentially alerting a larger number of cafeteria users to the changes. Nevertheless, this disparity has not resulted in higher opt-out rates, underscoring the importance of ensuring that portion size adjustments remain transparent, benefiting both ethical considerations and, in some cases, effectiveness.
Limitations: We acknowledge several limitations, including potential sample bias among cafeteria users and the specific setting of our study. Further research is needed to validate our findings across various settings and populations. The relatively large effect sizes as indicated by Cohen‘s d should be interpreted cautiously, especially with regard to acceptance, which may vary substantially in other cafeteria settings.
Implications and Future Research: The insights gained from this study hold valuable implications for cafeteria managers and policymakers striving to reduce meat consumption. In summary, this study offers a nuanced understanding of the trade-offs inherent in different behavioral strategies, offering valuable guidance for the design of ethical and effective food choice architectures.
