In this study, we examine a setting where principals use past performance to annually revise performance targets, but do not fully incorporate the past performance information in their target revisions. We argue that this situation is driven by some principals and agents having an implicit agreement where the principal “allows” the agent to receive economic rents from positive performance-target deviations that are the result of superior effort or transitory gains by not revising targets upward, while the agent “accepts” target revisions by not restricting output when these revisions are the result of structural changes in the operation's true economic capacity. Although both the principal and the agent can benefit from an implicit agreement, we argue that for the implicit agreement to be maintainable, the principal either needs information on the cause of the performance-target deviation or there needs to be trust between the principal and the agent. Using archival data across multiple years and independent bank units, we find a pattern of ratchet attenuation and output restriction that is consistent with the existence of implicit agreements for those principal-agent dyads where information asymmetry is sufficiently reduced or mutual trust exists. Data Availability: Data used in this study cannot be made public due to a confidentiality agreement with the participating firm.

Prior research suggests that audit seniors' judgments are sometimes biased by their affect toward (i.e., feeling of personally liking or disliking) client personnel. We examine how experienced audit reviewers respond when reviewing an audit preparer's judgment that appears to be biased by the preparer's affect toward a client's controller. In our experiment, reviewers are provided with a preparer's judgment that appears inconsistent with the audit workpapers. We then examine the effect of providing versus not providing reviewers with a cue about the preparer's positive or negative affect toward the controller. We find that despite reviewers' belief that affect biases a preparer's judgment, reviewers who are informed of the preparer's affect do not rely less on the preparer's judgment. Instead, they actually rely more on the preparer's judgment than do those who are not informed about the preparer's affect. This result is consistent with Wegner's (1994) ironic rebound effect, which predicts that sometimes when individuals are trying not to rely on information, they ironically rely on it more. Our findings suggest a potential limitation of the audit review process.

This study examines the impact of the California Nonprofit Integrity Act of 2004 on CEO compensation costs in affected organizations. Contrary to the stated objective of the Act that executive compensation is “just and reasonable,” we find that CEO compensation costs for affected nonprofits during the post-regulation periods have increased by about 6.3 percent when compared with a control group of comparable unaffected nonprofits. In addition, the relative increase in CEO compensation appears to come from nonprofits that have experienced greater regulatory cost increases. We do not find evidence that the Act resulted in a change in CEO pay performance sensitivity. The observed CEO pay increase is not systematically different across nonprofits that underpaid versus those that overpaid their CEOs during pre-Act periods. Overall, this paper highlights the unintended consequences of regulatory attempts to enhance governance in the not-for-profit sector.