Where Do Companies Attempt Earnings Management, and When Do Auditors Prevent It?

Abstract

Research about earnings management (EM) is sometimes criticized as insufficiently useful to standard setters (e.g., Healy and Whalen 1999). Because prior studies necessarily focus on those areas of EM that have resulted in SEC enforcement actions or for which empirical tests are most powerful, little is known about EM with respect to many accounts. Also, relatively little research has examined factors that limit EM. For example, because prior studies focus on post-audit financial statements, they cannot separate managers' attempts to manage earnings and auditors' decisions to waive adjustment of earnings-management attempts. Our study complements prior research by constructing a database of 526 earnings-management attempts (EMAs) that span the issues regulated by SFAS's 1-132 and their precursors. We obtain EMAs by surveying audit partners and managers from one Big-5 firm and eliciting specific experiences they have had with audit clients who they believe were attempting earnings management. This database allows us to examine how managers' decisions to attempt EM and auditors' decisions to waive EMAs are affected by financial-accounting-standard precision, transaction structuring, current-year-income effect, materiality, and/or client size. We also assess the relative frequency of EMAs in various accounts, as well as the relative frequency with which EMAs are waived by auditors to become EM in the audited financial statements. Regarding managers' EMAs, analyses indicate that transaction structuring is infrequent, and tends to occur with respect to precise standards. Of those EMAs that affect current-period income, 60% are income increasing. EMAs are more likely to increase current-period income when standards are precise and/or transactions have been structured. Regarding auditors' waive decisions, analyses indicate that auditors are more likely to waive adjustment of an EMA when it decreases current-year income, is governed by an imprecise standard, is structured (particularly to meet a precise standard), is considered by the auditor to be immaterial, or is attempted by a large client. In our database, EMAs occur most frequently with respect to reserves, followed by, in order of frequency, revenue recognition, business combinations, intangibles, fixed assets, investments, leases, compensation, and many other areas. While revenue-recognition EMAs are second most frequent, auditors infrequently waive them, so they are the third most frequent category of EM. Auditors require adjustment of 43% of EMAs. They waive adjustment of 22% because they conclude the EMA is allowed by GAAP, 18% because the auditor has no convincing evidence that their client's position is incorrect, and 17% for another reason (usually, immateriality). At the conclusion of our paper we discuss limitations and implications of our results for earnings-management research, audit practice, and financial-accounting and auditing regulation.