We use information extracted from a major proxy advisory service to test predictions from institutional theory regarding when and why audit committee ( AC ) members experience turnover because of evidence of ineffective governance. First, we broadly categorize AC ineffectiveness concerns as either (i) financial reporting failures or (ii) characteristics of individual AC members. Institutional theory suggests that the visible nature of the first category is more likely to threaten perceptions of AC legitimacy and hence prompt turnover, which is what we find. We then enrich the analysis by interacting the AC ‐member ineffectiveness indicators with the extent of shareholder protest votes, finding that shareholder dissent elevates the turnover effects of both categories of ineffectiveness, as institutional theory would predict. Finally, we find that otherwise effective AC members face an increased likelihood of turnover if they serve on the AC when financial reporting failures are discovered, even if they were not on the AC when the events precipitating the failures occurred. Overall, our findings support the institutional theoretic premise that boards take remedial actions when necessary to restore perceived legitimacy.

We find that the effectiveness of piece‐rate compensation relative to fixed pay in a laboratory letter‐search task hinges on the presence or absence of a nonbinding statement to participants that the experimenter values correct responses. In the absence of the value statement, participants with piece‐rate rewards for correct responses generate more correct and incorrect responses than do their counterparts with fixed pay, correcting errors as they go along to maximize compensation. Essentially, piece‐rate compensation acts as an output control, incentivizing participants to maximize correct responses through a “produce‐and‐improve” strategy. The value statement suppresses this strategy because participants appear to perceive it as an input constraint, prompting greater initial care at the expense of lower overall productivity. As a result, the value statement eliminates the gains in correct responses that piece‐rate incentivized participants otherwise realize. Thus, in settings in which individuals can gain efficiency by working expeditiously and improving quality when necessary, our results suggest the possibility that organizations could be better off just letting incentive schemes operate, rather than emphasizing quality in ways that could overly constrain productivity.