SUMMARY
This article summarizes a study by Huang, Chen, Kaplan, and Lin (2021) which examines whether increases in co-working experience between the lead and concurring audit partners affect engagement audit quality and audit efficiency. The study uses data from Taiwan, where the identities of lead and concurring audit partners are known, and where the role of the concurring partner is different than that of an engagement quality reviewer. This article describes the findings of Huang et al. (2021) and discusses the study's implications to audits in Taiwan and elsewhere which may be of interest to practitioners, regulators, and academics.
I. INTRODUCTION
An audit engagement is performed by an audit team, which includes one or more partners (e.g., lead, multi-location, subject matter expert, advisory, concurring) and staff members (e.g., managers, seniors). While each audit team is formed to perform a specific audit engagement, partners and staff members commonly serve together on multiple audit engagements. The effect of audit team members repeatedly working together on audit outcomes, like quality and efficiency, is largely unknown. Potentially, repeatedly working together could improve communicating and coordination among team members, resulting in improved audit outcomes. Alternatively, repeatedly working together could enhance familiarity among audit team members, which could degrade audit quality by inhibiting one's ability to objectively evaluate another team member's suggestions (e.g., how to resolve audit-related issues) or the work he or she performs.
Huang, Chen, Kaplan, and Lin (2021) provide evidence of the effects of partner-level auditors repeatedly working together on audit engagements. Specifically, Huang et al. (2021) focus on the prior working relationship (co-working relationship) of lead and concurring audit partners of publicly traded Taiwanese companies.1 In Taiwan, lead and concurring audit partners work together to plan the audit engagement, implement the plan, and determine the type of audit report to issue, and both sign the audit report (Schneider and Messier 2007; Chi and Chin 2011). Auditing standards in Taiwan also require an independent engagement quality reviewer (EQR) who is not involved in the audit engagement to conduct a review of the audit engagement. Beyond having implications for Taiwanese lead and concurring audit partners, Huang et al.'s (2021) findings about the co-working relationships of audit partners in Taiwan have implications for a range of auditor co-working relationships in Taiwan and elsewhere, including the U.S.
II. CO-WORKING EXPERIENCE AND AUDIT OUTCOMES
Co-working experience is one of three different types of experience that has been shown to influence team performance within organizations (Levitt and March 1988; Liang, Moreland, and Argote 1995; Edmondson, Roberto, and Watkins 2003; Argote and Miron-Spektor 2011). The other two types of experience are an individual's experience and the organization's experience. Audit researchers have previously examined both the individual audit partners' experience (e.g., individual audit partner tenure) and firm's experience (e.g., audit firm tenure) on audit outcomes. Huang et al. (2021) extend this line of research by providing initial evidence on the effect of auditor co-working experience on audit outcomes. This research is important, in part, because researchers contend that the experience and interactions among audit team members could affect audit quality (DeFond and Zhang 2014; Dickins, Fay, and Daugherty 2015).2
Huang et al. (2021) focus on the co-working experience between the lead and concurring audit partners. On one hand, increases in co-working experience between the lead and concurring partners may lead to improved communication and better coordination, which in turn, could facilitate knowledge sharing (Kaplan 1979; Hansen 1999; Reagans, Argote, and Brooks 2005), and create process gains that allow concurring partners to focus on providing oversight and monitoring the lead partner's work product, judgments, and conclusions, leading to improved audit quality and efficiency. On the other hand, increases in co-working experience could adversely impact audit quality by interfering with the concurring partners' ability to maintain sufficient professional skepticism and monitor the engagement in an objective fashion.
Huang et al. (2021) measure co-working experience based on the lead and concurring partners' experience working together in these roles on previous audit engagements for clients other than the current engagement from 1998 to 2012. Specifically, co-working experience is calculated as the log of the total number of unique prior audit clients the same two individuals have served as signing partners, tracing back from the current year to 1983. For example, if auditor A and auditor B sign the audit reports of client C and client D from year t–2 to year t and client E from year t–1 to year t, then for client C in year t (excluding the co-working experience related to client C) the co-working experience equals two clients (client D and client E). This approach to calculating co-working experience ensures that individual audit partners' client tenure effects are excluded.
Since audit quality is not directly observable, Huang et al. (2021) measure audit quality using three measures: the absolute value of discretionary accruals (e.g., Leuz, Nanda, and Wysocki 2003; Chi and Huang 2005; Jaggi, Chin, Lin, and Lee 2006), income-increasing discretionary accruals (e.g., Ashbaugh, LaFond, and Mayhew 2003), and the likelihood of issuing a qualified audit opinion (e.g., Chi and Chin 2011). The first two proxies for audit quality measure the proportion of accruals that cannot be explained by a company's operations. The lower the discretionary accruals, the higher the audit quality. This is because higher-quality auditors should be more likely to prevent client managers from using accruals to manage earnings. Accordingly, a negative (positive) association between the co-working experience and discretionary accruals supports that audit quality is higher (lower) when the two signing partners have more co-working experience.
The third proxy is the propensity of an auditor to issue a qualified audit opinion. A higher propensity to issue a qualified audit opinion is indicative of higher audit quality. The intuition is that controlling for indicators of potential material financial statement misstatement, higher quality auditors are presumed to be able to better-detect material errors and uncertainties, like going concern, and to report these in their audit reports. Accordingly, a positive (negative) association between the co-working experience and the propensity to issue a qualified audit opinion supports that increases in the co-working experience are associated with improved (reduced) audit quality.
To measure audit efficiency, Huang et al. (2021) use the log of audit reporting lags (the number of days between the fiscal-year end date and the audit report date) to proxy for audit efficiency (e.g., Al-Ajmi 2008; Munsif, Raghunandan, and Rama 2012; Pizzini, Lin, and Ziegenfuss 2015). The intuition is that shorter audit reporting lags reflect a more efficient audit engagement. If increases in the co-working experience are associated with improved (reduced) audit efficiency, it is expected to observe a negative (positive) association between the co-working experience and audit reporting lags.
Table 1 provides the descriptive statistics of the main variables from the study. The measure of co-working experience has a high variation. On average, a lead partner and a concurring partner have worked together on 10.75 unique clients, while the first quartile, the median, and the third quartile of co-working experience are three, eight, and 16 unique clients, respectively. The range of co-working relationships is zero to 46 unique clients.
With respect to the audit quality measures, the absolute value and the positive value of discretionary accruals average 9 percent (range = 0.11 to 44.90 percent) and 10 percent (range = 0.07 to 41.02 percent) of total assets, respectively, and 2 percent of firm years receive a qualified opinion (range = 0 or 1). With respect to the audit efficiency measure, the average audit reporting lags are 68.88 days (range = 19 to 119 days).
Table 2 summarizes the main results from the study. Using ordinary least squares regression analysis, Huang et al. (2021) find that higher co-working experience is associated with lower absolute and lower income-increasing discretionary accruals, a higher propensity to issue a qualified audit opinion, and lower audit reporting lags.3 Collectively, the results indicate that audit quality and audit efficiency are higher when the lead and concurring partners have more co-working experience.
In terms of economic significance, moving from the first quartile to the third quartile of co-working experience decreases the absolute value of discretionary accruals by 5.8 percent, and the positive value of discretionary accruals by 11.6 percent. Similarly, moving from the first quartile to the third quartile of co-working experience decreases the audit reporting lag by 3.6 percent, and a one standard deviation increase in co-working experience increases the likelihood of issuing qualified audit opinions by 24 percent. These results suggest that partners' co-working experience has a meaningful, not merely statistical, positive impact on audit outcomes.
Huang et al. (2021) further conduct cross-sectional analyses to explore whether the nature of the co-working experience, client audit risk, and audit firm characteristics moderate the association between co-working experience and audit outcomes. The results indicate that the positive association between co-working experience and audit outcomes is more evident when the co-working relationship is stronger, when the role switches between the two partners are less frequent, when the two audit partners are more accessible to each other, when audit risk is higher, when the audit firm's knowledge of the client's industry is limited, and when the audit firm tenure is shorter.
III. IMPLICATIONS FOR PRACTICE, REGULATORS, AND ACADEMICS
Implications for Practitioners
Audits, whether in Taiwan or elsewhere, include an engagement review process. The engagement review process plays a critical role in ensuring that the quality of each audit engagement conforms to professional and firm standards (Epps and Messier 2007; Emby and Favere-Marchesi 2010). The main results in Huang et al. (2021) indicate that the co-working experience between the lead and concurring audit partners positively affects audit quality and audit efficiency. While the evidence is only about lead and concurring audit partners, similar benefits are likely to accrue for other co-working relationships whether in Taiwan or elsewhere. For example, lead audit partners have strong working relationships with multi-location partners, technical subject matter partners, and advisory partners. It is expected that increases in co-working experience between the lead and these other partners will also increase audit quality and audit efficiency. Accordingly, audit firms should consider the co-working experience between partners when making engagement staffing decisions. The benefits arising from assigning partners with substantial co-working experience must also be balanced against the need to create opportunities for partners with less co-working experience to work together. The findings of Huang et al. (2021) also have implications for audit firms' human resource policies by highlighting the potential benefits of retaining experienced audit partners.
Implications for Regulators
Audit partner tenure has received considerable attention from regulators because of concerns that auditor independence may be impaired by an audit partner with an overly long tenure with a client. The International Federation of Accountants' (IFAC) Code of Ethics indicates that a lead audit partner with a prolonged tenure on an audit engagement may become overly familiar with the client and its management, threatening auditor independence. To reduce this concern, the IFAC proposed that the lead partner should rotate off the engagement after a predefined period (IFAC 2007). This type of auditor rotation policy has been implemented in several jurisdictions, including Argentina, Australia, China, Denmark, Hong Kong, Mexico, the Netherlands, Norway, Russia, Taiwan, the United Kingdom, and the United States (Moroney 2016). Regulators should consider the benefits of partners' co-working experience when establishing auditor tenure and rotation regulations.
Implications for Academics
The results in Huang et al. (2021) highlight several fruitful venues for future research. To the extent possible, the effect of co-working experience among audit team members other than audit partners should be further explored. For example, increases in co-working experience may affect the quality and/or nature of the interactions between staff members, which, in turn, could impact audit quality. Understanding the influence of co-working experience on the interactions within the entire audit engagement team is particularly important because of the challenges audit firms face in recruiting and retaining qualified employees (AICPA 2011, 2014; Hermanson, Houston, Stefaniak, and Wilkins 2016; Khavis and Krishnan 2021).
In addition to the interactions within the engagement team, future research may explore the interactions within the audit firm. For example, Duh, Knechel, and Lin (2020) show that knowledge sharing in an audit firm increases audit quality and audit efficiency. Potentially, the benefits of knowledge sharing to an audit engagement may depend on the co-working experience between audit partners and/or other staff members. In addition, future research is encouraged to investigate the effect of co-working experience from the perspective of the entire engagement team. Survey methods are particularly well suited for generating theory-based evidence where archival evidence is largely unavailable. Finally, experimental methods could be used to manipulate the lead and concurring partners' co-working experience to examine whether it influences the judgments and decisions of other engagement team members.
REFERENCES
Auditing standards and regulations in Taiwan have required the “lead” and “concurring” audit partners to sign audit reports (dual signature) since 1983. The term “lead partner” is used here to refer to the partner responsible for the day-to-day management of the audit engagement. The term “concurring partner” is used here to refer to the second signing partner responsible for reviewing the audit engagement, as it is the term commonly used in Taiwan, the setting for the study (Chi, Huang, Liao, and Xie 2009; Chin and Chi 2009; Chi and Chin 2011). A difference between concurring partners and engagement quality reviewers is that concurring partners are involved in the audit engagement, but engagement quality reviewers are not.
Relatedly, Dickins et al. (2015) surveyed audit partners familiar with engagement quality reviews and found that the PCAOB's Auditing Standard No. 7 changed the nature and extent but not the timing of engagement quality reviews.
Huang et al. (2021) construct a model of audit outcomes as follows: Audit Quality/Efficiency = β0+ β1Co-Working Experience + Controls + Year + Industry + Exchange + ε. Huang et al. (2021) use the absolute value of discretionary accruals, the positive value of discretionary accruals, and an indicator of issuing a qualified audit opinion to proxy for audit quality, and use the log of audit reporting lags to proxy for audit efficiency. They use the log of the total number of unique prior audit clients the same two individuals have served as signing partners to proxy for the co-working experience between the two signing partners, and include several client and auditor characteristics as control variables in the model. Year, industry, and stock exchange fixed effects are also included in the model. An ordinary least squares regression is estimated when the dependent variable is the absolute value of discretionary accruals, the positive value of discretionary accruals, or the log of audit reporting lags, and a logistic regression is estimated when the dependent variable is an indicator of issuing a qualified audit opinion.