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

29 NTU Management Review Vol. 30 No. 2 Aug. 2020 observations and 0 for single-signature observations. “ CV ” indicates control variables. These control variables include the company’s firm size, leverage, market to book value, and audit firm size. In equation (3), the coefficient of “ EARN*DUAL ” indicates the incremental effect of a dual-signature system on earning information. β 3 is expected to be positive. In the next equation (4), the sum of the coefficients on “ EARN*DUAL ” and “ UEARN*DUA L” ( γ 4 + γ 5 ) indicates the incremental ERC (Earnings Response Coefficient) for a dual-signature report relative to the single-signature report. If investors perceive a dual-signature system as bringing in higher audit quality, then the incremental ERC will be positive. 4. Findings Empirical results indicate that the dual-signature system is associated with higher audit quality as represented by audit opinions. Shareholders focus largely on public trading companies. Therefore, such an association exists only for companies listed on Taiwan’s Stock Exchange but not for non-publicly traded companies. In addition, the informativeness of earnings and earnings response coefficients are higher under the dual- signature requirement, suggesting that investors perceive higher earnings quality under the dual-signature requirement. 5. Research Limitations This study has the following limitations. First, the sample observations are collected between 1981 and 1985. The legislation system, and the capital market have not been developed well at that time. Therefore, the results of this study may not be applicable in the current environment. Second, because the data is from 1981, only the effects of a dual- signature system two years before and after the change of regulation can be compared. This also limits the use of other audit quality measurements in the study. 6. Contribution This study presents a number of contributions. First, this study contributes to the audit literature by providing archival evidence that audit quality is higher under a dual- signature requirement. In addition, the results indicate that auditor decisions are different when they audit public trading companies. The empirical results bring understanding in regards to auditor behaviors and support the dual-signature system, which can have policy implications.