Purpose The purpose of this paper is to examine whether firms use different earnings management approaches when facing financial difficulties and the effects of industry-specialist auditors in constraining those choices. The empirical results suggest that (1) firms with lower risk of business failure but with stronger incentives to adjust earnings upward tend to use real earnings management (REM) income-increasing approaches while (2) at the same time, using discretionary accruals for income-decreasing earnings management, due to constraints imposed by specialist auditors on the use of accrual-based earnings management (AEM). This is consistent with the findings of Chi et al., and the authors do not find similar evidence for the firms with higher risk of failure. Also, (3) regardless of the level of failure risk, firms turn to REM while interestingly, such REM behavior is effectively curbed by industry-leading specialist auditors (specialist auditors with the highest client market share) on financially distressed firms. These results extend the findings of Chi et al. (2011), suggesting that industry-specialist auditors have different tolerance levels for earnings management approaches by firms with different levels of business failure risk. That is, when auditing clients with higher risk of failure, specialist auditors are more likely to maintain higher audit quality through more stringent audit testing and use of more audit staff time to prevent an occurrence of audit failure. Design/methodology/approach The authors examine earnings management behavior across firms in Taiwan with different levels of business failure risk and the effects of audit partner industry specialization in constraining that behavior. Chi et al. (2011) studied low-risk firms with incentives to adjust earnings upward and found firms use REM when the auditors constrain AEM. The authors extend the work of Chi et al. and observe firms with different levels of failure risk. Findings The authors find (1) lower risk firms may use discretionary accruals to adjust earnings downward while the authors find no similar evidence for financially distressed firms, (2) lower risk firms may use REM when their industry-specialist auditors curb AEM and (3) the industry leaders among specialist auditors do the same for the financially distressed firms. The results demonstrate the extent to which industry-specialist auditors apply different tolerance levels for earnings management behaviors across firms with different levels of failure risk. Originality/value The study contributes to the literature in the following three ways: first, the authors fill a gap in the existing literature by comparing firms with higher risk of business failure to firms with lower risk of business failure to explore the possible difference in the two different kinds of earnings management behavior; second, the authors extend the findings of Chi et al. (2011) and examine whether specialist auditors, when auditing firms with higher risk of business failure, will input more audit effort to constrain their clients' use of REM and third, since business failures have a significant impact on the capital markets and any associated audit failures can have an even greater negative impact on investor confidence, the study provides information useful to auditors and regulators in the formation of salient policy regarding the use of REM by firms experiencing high risk of business failure.
