Comparing transparent and covert nudges: A meta-analysis calling for more diversity in nudge transparency research

• Nudge transparency is conducive to preserving democratic rights in behavioural public policy. • Our meta-analysis suggests that transparency for some nudge types may have positive effects on choice outcomes in certain domains. Negative effects are not supported by our data. • However, these findings cannot be generalised, as 88 % of effects are from default nudges and 89 % from online studies. • Systematically testing nudges in real-life settings and improving the diversity of studies testing the effects of nudges can help understand better the relationship between transparency and nudge effectiveness. Do transparent and non-transparent nudges have similar effects? The question is central in recent research on behavioural public policy, as it leads to ethical and practical implications regarding policy-maker responsibility, citizen agency, and nudge design. We meta-analysed results from 23 publications designed to compare transparent to covert nudges including 117 effect sizes and found a positive effect of transparency on behavioural outcomes, but no effect on non-behavioural outcomes. The moderator analyses revealed that studies conducted online, manipulating the decision structure, and conducted in the domain ‘other’ tended to exhibit significantly positive transparency effects for behavioural outcomes. We note that all but two studies were conducted online or in the lab, and that there is an over-representation of research on default nudges (88 % of total effects), severely limiting the generalizability of the findings. Thus, we call for an improvement of research conducted on transparent nudges and the inclusion of more nudge types, preferably in a field setting. We also stress the importance of defining the form of transparency that societies require for respecting their citizen's autonomy.