摘要：

Challenges faced by nudges in practice have prompted scholars to reflect on effectiveness and applicable conditions of nudge. A common approach is to conduct meta-analyses on the results of behavioral intervention experiments involving nudges. However, existing meta-analyses have significant limitations. On one hand, due to differences in the selection of effect size indicators, the results of different meta-analyses lack comparability, preventing a comprehensive view of the overall effects of various nudging tools across common behavioral domains. On the other hand, most existing meta-analyses categorize nudges based on the “MINDSPACE” framework, which does not encompass all nudge interventions.

Drawing inspiration from Hansen and Jespersen (2013), this study establishes an analytical framework from a dual perspective of cognitive pathway and transparency, encompassing commonly used nudging interventions in practice. Based on this framework, a meta-analysis is conducted to compare the differences in effectiveness across nudge categories, providing a comprehensive and objective evaluation of the real effects of nudges under different cognitive pathways and levels of transparency. This helps address academic disputes regarding the effectiveness of nudges under different cognitive pathways and transparency levels.

This study focuses on the effectiveness of nudge in the actual policy domain, conducting a meta-analysis on nudge intervention experimental research published in the two professional journals: Behavioural Public Policy (2017-2022) and Behavioral Science & Policy (2015-2022). Using the standardized mean difference (Cohen's d) as an effect size index, Comprehensive Meta Analysis 2.0 is employed to calculate the effect size. All articles published in these two journals were included (a total of 330) in our sample frame, and a thorough reading of the full text was used to screen the literature one by one. The experimental data included in the meta-analysis had to meet the calculation requirements of CMA 2.0, containing basic statistical data such as means, standard deviations, t-values, and F-values. After screening, 40 sample articles were ultimately selected, yielding 108 research results. Considering the differences in research sample sizes, a random effects model was applied to calculate the overall effect of nudging, and then compare the effectiveness of nudges that work through different cognitive pathways and are presented with different levels of transparency. Subsequently, a regression analysis model is established using SPSS25 to test the effects of three classes of heterogeneity factors (research design, behavioral characteristics, and behavioral domains) with eight variables (sample size, field experiment, data type, behavioral motivation, monetary change, health, consumption, finance, public interest) on nudging intervention effects; Finally, an interaction regression analysis model is constructed to analyze the interactions between nudge cognitive pathways, transparency, and heterogeneity factors.

The results showed that: (1) Nudge research in the field of behavioral public policy has a relatively small overall effect and still faces the risk of failure in practice. (2) Transparent nudges can also be effective in guiding people to change their behaviour, and there is an interactive effect between different nudges that work through different cognitive pathways and are presented with different levels of transparency. (3) The effects of nudges are influenced by study design and vary across different behavioral domains. (4) Complex interactions are observed between the cognitive pathways through which nudges operate, the transparency of nudge design, and the factors affecting effect size heterogeneity.

The academic contributions of this paper are: (1) By conducting a meta-analysis on nudge behavioral effects from the dual perspectives of “cognitive pathway” and “transparency”, this research helps separate the actual effects of various psychological cognitive processes and intervention techniques, aiding in forming an objective understanding of the real effectiveness of different types of nudges. (2) Evaluating the effects and differences of various nudges, which not only addresses the academic debate on whether nudge designs should be transparent but also compares the differences in the effects of two cognitive pathways on behavior change, providing empirical evidence for the selection of nudging tools in subsequent policy practices. (3) This study's analysis of the root causes of heterogeneity in nudge effects enriches the literature on the factors influencing nudge effectiveness and applicable conditions, promoting the accumulation of knowledge in the field of behavioral public policy.

关键词:

behavioural public policy, nudge, meta-analysis, policy effectiveness, cognitive pathway, transparency

原文载于：https://journal.psych.ac.cn/xlkxjz/CN/10.3724/SP.J.1042.2023.02275