When Do People Prefer Carrots to Sticks? A Robust “Matching Effect” in Policy Evaluation

For a policy to succeed, it must not only be effective in changing behavior but must also be accepted by stakeholders. Here, we report seven sets of studies demonstrating strong framing effects on the acceptance of equivalent policies. Policies targeting desirable voluntary behavior are preferred when they are framed as advantaging those who act desirably (rather than disadvantaging those who do not). Conversely, policies targeting obligations are preferred when they are framed as disadvantaging those who fail to act desirably (rather than advantaging those who do). These differences in policy acceptance do not result from common causes of framing effects, such as a misunderstanding of outcomes or insufficient deliberation about the implications. Rather, the framing effects we document follow from beliefs about when punishment is and is not appropriate. We conclude with a field experiment demonstrating framing effects in a setting where policy acceptance directly affects respondents’ outcomes.

Data, as supplemental material, are available at https://doi.org/10.1287/mnsc.2016.2539.

This paper was accepted by Yuval Rottenstreich, judgment and decision making.