Around the world, laws limit the incentives that can be paid for transactions such as human research participation, egg donation or gestational surrogacy. A key reason is concerns about ‘undue inducement’—the influential but empirically untested hypothesis that incentives can cause harm by distorting individual decision-making. Here I present two experiments (n = 671 and n = 406), including one based on a highly visceral transaction (eating insects). Incentives caused biased information search—participants offered a higher incentive to comply more often sought encouragement to do so. However, I demonstrate theoretically that such behaviour does not prove that incentives have harmful effects; it is consistent with Bayesian rationality. Empirically, although a substantial minority of participants made bad decisions, incentives did not magnify them in a way that would suggest allowing a transaction but capping incentives. Under the conditions of this experiment, there was no evidence that higher incentives could undermine welfare for transactions that are permissible at low incentives.

在世界各地，法律限制了人类研究参与、卵子捐赠或妊娠代孕等交易的激励措施。一个关键原因是对“不当诱因”的担忧——这是一种有影响力但未经实证检验的假设，即诱因可能会通过扭曲个人决策而造成伤害。在这里，我展示了两个实验（n  = 671 和n  = 406），其中一个基于高度本能的交易（吃昆虫）。激励措施导致了信息搜索的偏差——参与者提供了更高的激励来遵守更多的要求，从而鼓励他们这样做。然而，我从理论上证明，这种行为并不能证明激励措施具有有害影响；这符合贝叶斯理性。根据经验，尽管相当少数参与者做出了错误的决定，但激励措施并没有以允许交易但限制激励措施的方式放大这些决定。在本实验的条件下，没有证据表明较高的激励措施可能会损害低激励条件下允许的交易的福利。