臺大管理論叢 NTU Management Review VOL.30 NO.2

63 NTU Management Review Vol. 30 No. 2 Aug. 2020 auditors), SHARE is set to 1; 0 otherwise. In addition, this study also uses another proxy, the ratio of assets audited by the component auditors to the total assets ( SHARE% ) to enhance the robustness of the study. For the empirical results, this study finds that a group company with component firms engaging with different audit firms has higher audit adjustment (better audit quality), especially when the proportion of group assets audited by other audit firms is greater. The results indicate that engaging with the same audit firm in the group company may impair auditor independence resulting from an increase in the auditor’s economic dependence on the group company. The empirical results do not support the position of auditing standards which rigorously restrict the shared-opinion audit reports contributed to the improvement of audit quality. It is recommended that follow-up research can use other proxies to measure audit quality, such as abnormal accruals, financial restatements, or going-concern opinion audit reports issued by engagement auditors, to discuss this topic again. It is worth emphasizing how different audit quality measurements may not necessarily be consistent with the empirical results. Different audit decisions entail different costs and benefit considerations for the auditors and therefore it is not appropriate to generalize them. Different audit decisions may be formed in different situations because the auditor costs and legal responsibilities are not the same across all cases. Some examples of differing decisions include: allowing a manager to have more room for earnings management, allowing significant misrepresentation (which may lead to ex-post financial restatements), or whether or not to issue going-concern audit opinions while companies face financial crises. In addition, because the auditing environment in Taiwan has several fundamental differences compared to the U.S. and European countries, the empirical results found in this study should not be used as inferences by other countries. This study offers important contributions to the field. First, the study is the first to investigate the shared-opinion audit reports on audit adjustment, which not only fills the gap in existing literature on shared-opinion audit reports, but also on the scarcity of audit adjustments. Second, the audit quality proxy, audit adjustment, may be a more accurate alternative for the literature. Specifically, audit adjustments, which are misrepresentations of financial statements discovered by the auditors in the final stage of the auditing, require the respondent to correct the figures. This measure is consistent with the definition of audit quality mentioned by DeAngelo (1981a). In comparison with another audit quality measure, abnormal accruals, this measure is not necessarily estimated using a complicated statistical model, and has less measurement error.