Prior to the 2002 passage of accounting standards designed to increase financial accounting quality, approximately 75% of annual earnings announcements were released on or after the audit report date, but today, most US public companies release annual earnings prior to the completion of audit fieldwork. Although firms releasing annual earnings prior to the audit report exhibit lower financial reporting quality, the authors explore factors potentially contributing to why auditors accept the risk of adverse capital market reaction, lower future audit fees, and reputational damage if announced unaudited accounting information is subsequently revised in audited 10-K filings with the SEC as a result of an auditor adjustment. They provide evidence of the unintended consequences of such legislation as the Sarbanes–Oxley Act of 2002, which added audit requirements related to internal control audits and audit workpaper documentation that had the opposite impact of what was intended.

Using a controlled experiment, the authors investigate whether the audit-judgment quality of highly experienced audit partners and senior managers is affected by the differential timing of firms’ annual earnings releases and financial statement audit completion. Controlled experiments enjoy the comparative advantage of being able to hold exogenous factors of real-world audit settings constant, allowing them to complement and extend evidence from the archival studies that examine the historical records and source documents of actual corporations where the percentage of the total audit hours remaining to be completed varies. They ask 114 highly experienced auditors with a mean audit experience of 17 years to evaluate a complex scenario involving a typical late adjustment related to deferred taxes. The auditors are asked to address the reasonableness of the income tax provision and determine whether there is a need for an adjustment to the valuation allowance. Using a hypothetical company enables the authors to ensure comparability by precisely manipulating such auditor influences as the timing of the earnings announcement in relation to the completion of the audit and characteristics of the audit committee itself. The expectation is that their results are generalizable to other subjective and complex accounting issues.

Auditors are significantly less likely to push for adjustments for aggressive financial reporting when earnings have already been released. They frequently exhibit biased decision processing and lower judgment quality after adopting the client’s financial reporting goal of avoiding any adjustments. Auditors are more likely to support aggressive client-preferred accounting treatment when engagement risk is relatively low, when the auditor perceives a threat of losing the client, or when they are close to the client.

It is unreasonable to force the companies to wait until the audit is completed to report earnings, but the authors found that a strong audit committee with independence and technical expertise reduced the likelihood of the auditors going along with management’s aggressive assumptions on deferred taxes. Auditors with higher professional identification are less likely to cave to client preferences and tend to make more objective judgments that identify with the norms and values of the accounting profession. Auditors with low professional identification focus more on client service than on their public duty and obligations to the capital markets.