Group differences in agency modulate error monitoring

Mistakes can lead to aversive outcomes. Error monitoring may help prevent mistakes, but it might be maladaptive for individuals who lack control over aversive outcomes, as it consumes cognitive processing resources that could be allocated elsewhere. Here, we examined the effect of agency (i.e., control over punishment) on error monitoring using the error‐related negativity (ERN), an event‐related potential measure of error monitoring and error rate. Ninety unselected participants performed a flanker task in which they were shocked according to their own errors (controllable punishment, n = 47) or were shocked in accordance with another participant's errors (uncontrollable punishment, n = 43). Participants without agency over punishment showed smaller ERNs and higher error rates compared with participants with agency. Furthermore, punishment only reduced error rates for participants with agency. Together, these results provide the first experimental evidence that agency modulates error monitoring and suggest an adaptive process in which error monitoring is increased/decreased depending on its utility. If aversive outcomes occur regardless of behavioral adjustments, error monitoring may be maladaptive, because it expends resources needlessly. In a yoked experiment, participants without agency (i.e., who were punished according to another person's errors) showed reduced error‐related negativities and higher error rates versus participants with agency (i.e., who were punished only for their own errors). Results provide the first experimental evidence that agency modulates error monitoring and suggest that error monitoring is increased/decreased depending on its utility.