5. Conclusions
Our primary conclusion is that neither auditor-client social bonding nor measurement uncertainty can be fully appreciated in isolation. Rather, evidence from our experiment indicates that these two factors interact, insofar as a friendly social bond reduces the adjustments that auditor-participants require only when audit evidence is characterized by measurement uncertainty, not when such evidence is perfectly diagnostic. Although auditors in our experiment do not adjust fully even when presented with certain evidence, to the extent that our results generalize, we can at least take some comfort that social bonding is unlikely to magnify the extent of underadjustment in settings with known misstatements such as Libby and Brown [2013]. Similarly, one might take some comfort from the finding that measurement uncertainty alone does not influence auditor adjustments in the absence of social bonding. Thus, our results suggest that one way to ensure more conservative auditor evaluations of accounts with measurement uncertainty is to limit auditors’ social-bonding opportunities with their clients.
Our findings are also disconcerting, however, insofar as we find that something as simple as working together on a trivia contest can influence the extent to which auditorparticipants subsequently require an adjustment when an audit signal strongly suggests but does not prove aggressive reporting. One implication of this finding is that audit firms and regulators should consider the influence of social bonding when evaluating the role of in-firm specialists, concurring partners, and other relevant firm personnel involved in assessments of accounts with high measurement uncertainty. The PCAOB [2015] has expressed concern about the sufficiency of oversight when auditors seek guidance from specialists when auditing complex estimates. Notwithstanding this concern, in addition to their technical expertise, specialists could confer the advantage of a more dispassionate perspective that is less susceptible to the social bonds that arise from day-to-day client interactions. This reasoning is consistent with evidence from field interviews conducted by Boritz et al. [2015], who find that members of the audit team tend to be more tolerant of client-preferred accounting positions than are the specialists they consult for advice. .
