Ulrike Thuerheimer

5.1 Summary of main findings
Multiple different engagement-specific, audit firm-specific and contextual factors, as well as incentives, affect auditors’ supply of audit quality and other audit outcomes (IAASB [2014]). In this dissertation I focus on engagement-specific and contextual factors, including inputs to the audit, auditor incentives, and the institutional environment. In the three studies of this dissertation, I examine how audit outcomes are associated with the extent of component auditor involvement in group audits, the commencement of public oversight worldwide, as well as the interaction between regulatory and litigation risk.

In the first study (Chapter 2), I investigate the determinants of component (network and/or unaffiliated) auditor involvement in MNE group audits, and the association between the involvement and audit quality and fees. To explore this question, I make use of the Australian disclosures of audit fees paid to component (network and/or unaffiliated) and principal auditors in annual reports of listed companies. I find that principal auditor characteristics, and MNE complexity and internationalization are associated with the likelihood and extent of involvement of component auditors. I further find that likelihood of network (unaffiliated) component auditors is positively (negatively) associated with Big 4 and large non-Big 4 principal auditors, and that size, and degree of internationalization of MNEs are only associated with network component auditors. I then find that component auditor involvement is negatively associated with audit quality and positively associated with audit fees, regardless of type of component auditor. Lastly, while I find that unaffiliated component auditor involvement decreased after the revision of the group audit standard ISA 600, audit quality or audit fees did not change in response.

In the second study (Chapter 3), I investigate the association between audit quality and the commencement of national inspections of public oversight bodies worldwide. To examine this question, I use a large international sample of publicly listed companies audited by audit firms that are subject to inspections of national public oversight boards of 50 countries. Using a difference-in-differences design exploiting the staggered adoption of inspections internationally, I find that audit quality increases after inspection commencement. I further find that the increase in audit quality after inspection commencement is higher for large (Big 4) auditors, compared to small (non-Big 4) auditors. Since different public oversight boards were designed with different working approaches to inspections, disclosure of inspection results and enforcement ability, I document these differences. Based on these descriptives, I find that the effect of inspection commencement on audit quality is conditional on public oversight boards disclosing inspection results.

In the third study of this dissertation (Chapter 4), I investigate how two auditor incentives, litigation and regulatory risk, are jointly associated with auditor conservative reporting behavior and audit pricing. I measure differences in litigation risk for auditors across different U.S. states, and regulatory risk as the start of PCAOB inspections for the different audit firms. I exploit the U.S. setting which features variation in auditor litigation risk across different U.S. states, both in terms of third-party and damage apportionment standards which capture different dimensions of litigation risk. Further, regulatory risk increased during the sample period 2001-2009 with the introduction of public oversight over the audit profession, and specifically the introduction of inspections of audit firms at different time periods. Using these empirical proxies for litigation and regulatory risk, I find evidence these risks jointly increase audit fees, and conservative auditor reporting. Further, this effect is attributed to non-global audit firm networks (non-GAFN). This suggests that litigation and regulatory risks reinforce each other, rather than acting as substitutes.

5.2 Contribution and implications
Overall, the three chapters of this dissertation contribute to the body of knowledge about engagement-specific and contextual factors that influence audit outcomes. The research questions and settings that I examine in this dissertation are timely and of concern to the audit profession, regulators and standard-setters, and the public at large. The contribution of each study is discussed in the respective chapters in detail. I reiterate the important points here.

The first study (Chapter 2) is a very timely topic since regulators, standard-setters and practitioners worldwide are concerned with the level of audit quality of multinational group audits (e.g., IFIAR [2018], PCAOB [2016], [2017], IAASB [2015a], [2015b], [2019]). These audits are of particular concern when component auditors, especially foreign or unaffiliated, are involved. While this is clearly a topic of concern, there is to date only scant empirical research on group audits, primarily due to data constraints. This study is one of the first studies that documents audit quality and fee consequences of group audits. This is particularly interesting for regulators and standard-setters who are under pressure to address this issue. Our study shows that despite the IAASB’s attempt to increase the work of the principal auditor with the revision of ISA 600, audit quality problems remain even though the involvement of unaffiliated (but not network) auditors decreased. These results can be informative for the PCAOB and IAASB as they aim to revise their group audit standards (see footnotes 11 and of this dissertation for current initiatives of these standard-setting bodies). This study, based on public disclosures of component auditor involvement, also shows that users of group financial statements in settings outside Australia might benefit from more public disclosure on component auditor involvement.

Study two (Chapter 3), is the first comprehensive investigation into the effect of inspection commencement in settings outside the U.S., where PCAOB inspections have been investigated widely. I collect descriptives on the design of public oversight boards that show vast differences. Since public oversight boards have varying working approaches, inspection disclosure decisions and enforcement ability, it is unclear whether the U.S. findings can be extrapolated to other settings. This study generalizes the findings of the various U.S. studies that audit firm inspections by independent public oversight boards improve audit quality, however, it is contingent on inspection disclosure decisions of public oversight boards. These findings and descriptives can be useful for regulators that have ongoing discussions about mutual recognition of public oversight across countries, as well as for the (re)design of public oversight boards to increase their effectiveness in safeguarding the public interest. This study is also of interest to investors who can benefit from understanding the effectiveness of different public oversight designs.

In the third study (Chapter 4), I investigate the interaction between two auditor incentives (litigation and regulatory risk) and the association with audit outcomes. While the