Voluntary audit committees, auditor selection and audit quality: evidence from Taiwan

This study aims to examine the Taiwan setting, where audit partners’ names are presented in the audit report and where audit committee formation is voluntary in the initial stage of audit committee reform. This paper investigates the effects of the formation of voluntary audit committees on the selection of individual audit partners, and, in turn, the audit quality. This contrasts with previous studies investigating the relationship between audit committees and auditor selection at the audit firm level.

This paper samples all of Taiwan’s publicly listed firms for the period 2007–2012 and uses Heckman’s (1979) two-stage estimation model to achieve our objectives.

Using different characteristics of individual engagement partners as proxies for a higher quality auditor, the main empirical results show that voluntary audit committee formation is positively related to an industry specialist lead partner and a lead partner that has a larger number of clients. In addition, this paper also finds that voluntary audit committee formation has a positive impact on audit quality (proxied by discretionary accruals). The results suggest that the voluntary formation of an audit committee contributes positively to both auditor selection and audit quality. Furthermore, an additional test shows that the main empirical results are robust to a validity threat that firms that have good corporate governance prior to the formation of voluntary audit committees tend to select high-quality audit partners.

The paper contributes to the audit committee literature in the following ways: this paper takes advantage of Taiwan’s unique setting, where forming an audit committee is not compulsory in the initial stage of audit committee reform, to investigate the voluntary audit committee, auditor selection and audit quality; this paper expands on Abbott and Parker’s (2000) study of audit committee characteristics and auditor selection at the audit firm level by examining this relationship at the individual audit partner level; this paper responds to the call by Church et al. (2008) and DeFond and Francis (2005) who propose more studies on audit quality at the individual engagement partner level.