The Effects of Single versus Joint Evaluations on Causal Attributions

Recent research indicates that norms matter for ordinary causal attributions, although there is a good deal of debate concerning why they matter. One prominent account—that the impact of norms works via the salience of counterfactuals—has received support from a recent paper by Icard et al. (2017) reporting a new effect in cases where two agents perform symmetric actions that are each individually sufficient to bring about an outcome. But in four recent studies (Sytsma under review), I was unable to replicate these findings. In this paper I explore why, investigating a key difference between our studies: Icard et al. asked participants about just one agent (single evaluations), while I asked them about both agents (joint evaluations). I find that this difference helps explain the divergent findings, although the results remain problematic for Icard et al.’s view. Further I identity two evaluation effects: there is a general trend for the causal ratings in these cases to be lower when using single evaluations than when using joint evaluations, and this difference is larger when the agent asked about violates an injunctive norm. I consider four potential explanations of the impact of using single or joint evaluations and argue that determining the correct explanation has important implications for work concerning the effect of norms on causal attributions.