We attribute more free will to agents who behave immorally compared to a neutral control; also, when expectations for norm adherence are violated, we infer that an agent expressed their free will to do so

Monroe, Andrew E., and Dominic Ysidron. 2019. “Do Moral Judgements Motivate Free Will Belief?.” PsyArXiv. May 25. doi:10.31234/osf.io/8wu4g

Abstract: Free will is often appraised as a necessary input to for holding others morally or legally responsible for misdeeds. Recently, however, Clark and colleagues (2014), argued for the opposite causal relationship. They assert that moral judgments and the desire to punish motivate people’s belief in free will. In three experiments—two exact replications (Studies 1 & 2b) and one close replication (Study 2a) we seek to replicate these findings. Additionally, in a novel experiment (Study 3) we test a theoretical challenge derived from attribution theory, which suggests that immoral behaviors do not uniquely influence free will judgments. Instead, our nonviolation model argues that norm deviations, of any kind—good, bad, or strange—cause people to attribute more free will to agents, and attributions of free will are explained via desire inferences. Across replication experiments we found no evidence for the original claim that witnessing immoral behavior causes people to increase their belief in free will, though we did replicate the finding that people attribute more free will to agents who behave immorally compared to a neutral control (Studies 2a & 3). Finally, our novel experiment demonstrated broad support for our norm-violation account, suggesting that people’s willingness to attribute free will to others is malleable, but not because people are motivated to blame. Instead, this experiment shows that attributions of free will are best explained by people’s expectations for norm adherence, and when these expectations are violated people infer that an agent expressed their free will to do so.

---
Need for Theoretical Reinterpretation
Study 3 presents a theoretical challenge to the motivated free will belief viewpoint. Clark et al. (2014) predicate their conclusions on the claim that observing immoral behaviors activates a desire to punish the wrongdoers, and thereby causes people to inflate their belief in free will as a means to justify their desire to punish. This critical role of a desire to punish requires that the effect on free will beliefs be unique to people’s response to immoral behaviors—other norm violations, such as strange or morally good behaviors, would not engender such a desire to punish. However, in three experiments (Studies 2a, 2b, 3) we found that the desire to punish failed to mediate the effect of immoral behavior on people’s general belief in free will. Most critically, Study 3 revealed that norm violation more generally, not immorality specifically, explained variations in people’s free will judgments. Agents who committed an immoral act, a praiseworthy act, or simply strange act were judged as having more free will than agent who performed a morally neutral act. Importantly, whereas all three norm-violating behaviors (blameworthy, praiseworthy, and strange behavior) significantly differed from the control behavior, blameworthy behaviors did not differ from the praiseworthy or the strange behavior.

Together these findings argue for a non-moral explanation for free will judgments with norm-violation as the key driver. This account explains people’s tendency to attribute more free will to behaving badly agents because people generally expect others to follow moral norms, and when they don’t, people believe that there must have been a strong desire to perform the behavior. In addition, a norm-violation account is able to explain why people attribute more free will to agents behaving in odd or morally positive ways. Any deviation from what is expected causes people to attribute more desire and choice (i.e., free will) to that agent. Thus our findings suggest that people’s willingness to ascribe free will to others is indeed malleable, but considerations of free will are being driven by basic social cognitive representations of norms, expectations, and desire. Moreover, these data indicate that when people endorse free will for themselves or for others, they are not making claims about broad metaphysical freedom. Instead, if desires and norm-constraints are what affect ascriptions of free will, this suggests that what it means to have (or believe) in free will is to be rational (i.e., making choices informed by desires and preferences) and able to overcome constraints.