Noticing negligence or noticing numbers? Preference reversals in punishment judgments of moral luck cases

Negligence cases are a unique and interesting class of moral events, for which punishment judgments are affected by both the degree of negligence and the severity of the resultant outcome, even if the latter is merely a matter of luck. Adapting methods from consumer behavior research, we present participants with scenario-pairs in which individuals commit a negligent act that causes a negative outcome, and show preference reversals in judgments of deserved punishment across three normatively equivalent tasks: choice, rating, and matching. Additionally, across studies, we alternate which attribute is presented qualitatively versus quantitatively, negligence (e.g., ‘was extremely drunk’ versus ‘drank eight beers’) or outcome (e.g., damaged a parked car versus killed two people). We test two competing hypotheses derived from prior research: (a) cognitively simple tasks (choice and rating) lead participants to focus on outcomes while cognitively difficult tasks (matching) lead participants to focus on mental states, or, (b) choice and rating draw attention to qualitative information while matching draws attention to quantitative information. We find mixed support for both possibilities, but the most consistent findings were that preferences in matching were driven by the quantitative attribute, and preferences in choice were driven by outcomes, while preferences in rating were driven more by degree of negligence. These results show that punishment judgments in negligence cases can reverse across elicitation tasks, adding to a growing literature on the constructed nature of moral judgments and raising questions about whether and how moral preferences can and should be measured.