This study examines the role of the external audit in management's decision about the amount of GAAP financial statement information to disclose in the annual earnings announcement. The earnings announcement is a key disclosure provided by public companies. Yet, there is no requirement that earnings announcements contain audited GAAP numbers; in fact, recent trends indicate that a majority of companies release earnings before the completion of year-end audit fieldwork. I predict and find that companies that wait until the audit is more complete at the earnings announcement date and receive higher quality audits provide more detailed balance sheet, cash flow statement, and overall GAAP disclosures. I also provide evidence that complete audits and higher quality audits impact the information content of the earnings announcement. The combined results indicate that audit completeness and quality help facilitate more detailed earnings announcement disclosures and have implications for the equity market. Data Availability: Data are publicly available from sources identified in the text.

We address whether SOX 404(b) internal control audits under two auditing standards regimes and SOX 404(a) management assessments are associated with improved internal control system quality, an important and largely unstudied potential benefit. In 2013, the PCAOB disclosed that 15 percent of inspected control audits were ineffective, suggesting that the current control auditing standard may not be sufficient to induce implementation of high-quality control systems. We use an indirect measure of internal control system quality—future unaudited accruals quality—to proxy for internal control quality because sustained internal control improvements should be exhibited in future quarterly financial reports unaltered by contemporaneous financial statement audits. We find that internal control audits initially provided internal control quality benefits. After the 2007 auditing standards change, internal control quality deteriorated for ICFR audited versus unaudited firms. Finally, we find limited evidence that management assessments affect internal control quality. Results indicate that recent PCAOB concerns may have merit. Data Availability: Data are publicly available from the sources identified in the text.

When using a tournament in multi-period settings, firms have discretion in selecting the tournament horizon. For example, firms can use a single tournament (a grand tournament) or a sequence of multiple tournaments, each with a shorter horizon than a grand tournament (a repeated tournament). Firms have also begun to use a combination of both in which a repeated tournament is embedded within a grand tournament (a hybrid tournament). Using an experiment, we investigate whether the effect of tournament horizon on performance depends on the dynamic complexity of the task, which reflects the potential for effort in one period to influence the link between effort and performance in future periods. When dynamic task complexity is low, we find that performance is greatest in the hybrid tournament, followed by the repeated and then the grand tournament. In contrast, when dynamic task complexity is high, we find that performance is greatest in the repeated tournament, followed by the grand and hybrid tournaments, with similar performance in the latter two tournaments. More generally, the results of our experiment suggest that the effect of tournament horizon on performance depends on dynamic task complexity. These results can help firms make better decisions when designing their tournaments by reinforcing the need to align the tournament horizon with the task.

Audit practitioners, academics, and attorneys have expressed concern that disclosing critical audit matters (CAMs) will increase jurors' auditor liability judgments when auditors fail to detect misstatements. In contrast, this study provides theory and experimental evidence that CAM disclosures, under certain conditions, reduce auditor liability judgments as jurors perceive that undetected fraudulent misstatements were more foreseeable to the plaintiff (i.e., the financial statement user suing the auditor). However, we find that CAM disclosures only reduce auditor liability for undetected misstatements that, absent CAM disclosure, are relatively difficult to foresee. Finally, CAM disclosures that are unrelated to subsequent misstatements neither increase nor reduce auditor liability judgments relative to the current regime (i.e., where CAMs are not disclosed), but reduce liability judgments relative to reporting that there were no CAMs. As such, we find that, relative to stating there were no CAMs, disclosure of any CAM (i.e., related or unrelated) provides litigation protection in cases of undetected fraud. Consequently, the CAM requirement could incentivize auditors to disclose innocuous boilerplate CAMs, thereby diluting the impact of more warranted CAM disclosures. Data Availabliity: Available from authors upon request.