The Adaptive Logic of Moral Luck: Punishment’s Function Shapes the Processes Supporting It

Abstract

Moral luck is a peculiar aspect of our moral psychology – why do we condemn unintended outcomes (i.e. accidents)? In this dissertation, I present a novel analysis of moral luck, focusing on the argument that moral luck serves an adaptive purpose. First, I demonstrate a key aspect of moral luck – that accidental outcomes have a particularly strong influence on punishment judgments, compared to other moral judgments. Here, I contrast punishment with a partner choice decision and find that whereas punishment is sensitive to moral luck, partner choice is almost totally insensitive to accidental outcomes. Why then might punishment be uniquely sensitive to moral luck? I argue that moral luck is a consequence of punishment’s ultimate adaptive purpose: To change others’ behavior. I next test a key prediction of this hypothesis: We should observe moral luck only when the agent causing accidental harm could change their behavior. When behavior cannot be changed, punishment cannot serve its adaptive function, and so moral luck should be lessened. Focusing on manipulations of behavioral control as a test case, I find supporting evidence – when an agent had no control over their behavior, punishment in cases of moral luck is reduced. Finally, I focus on the implications the pedagogical hypothesis has for the motivations underlying punishment. This perspective predicts that punishment in contexts where pedagogy is most effective need not be motivated through other means – the benefits of changing a partners’ behavior are sufficient motivation. In contrast, we might be expect to find punishment motivated externally in cases in which pedagogy is less effective. Borrowing from the distinction between punishing as a harmed victim (2nd party punishment) and as an observer (3rd party punishment), I explore this question using both behavioral economic and vignette-based methods. I predict that punishment as an observer will be externally motivated – others will punish observers who fail to punish, thus motivating such 3rd party punishment. In contrast, 2nd party punishment will not be externally motivated, given that it is intestinally motivated by the benefits of pedagogy. Across both methods, I find the predicted pattern of results: Punishment as a 3rd party is externally motivated, whereas punishment as a victim is not. In total, I shed light on the adaptive logic of moral luck – we punish others to change their behavior and so punishment is sensitive to (unintended) outcomes. In other words, punishment’s adaptive purpose shapes the psychological processes supporting it.