Implications of Divided Responsibility in Audits Involving Component Auditors Free

Company audits frequently involve the participation of “component auditors” that audit components of a client’s operations (such as a consolidated subsidiary), in addition to the “lead” auditor that issues a client’s audit report. Regulators in the United States and abroad have in recent years focused on these audits—termed group audits—because of concerns about audit quality. In the United States, the Public Company Accounting Oversight Board (PCAOB) has been working since 2016 on amending auditing standards relating to group audits and released a final rule (effective in 2024) incorporating these amendments in August 2022 (Public Company Accounting Oversight Board (PCAOB) 2022).¹

Under U.S. auditing standards, the lead auditor must make a choice when signing the audit report: should it accept responsibility for the component auditor’s work or should it divide responsibility identifying the work performed by the component auditor in its audit report? The PCAOB labels the component auditor with divided responsibility as a “referred-to” (RT) auditor (Public Company Accounting Oversight Board (PCAOB) 2016). Correspondingly, we label the component auditor for which the lead auditor accepts responsibility as a nonreferred-to (non-RT) auditor (Adams and Zhou 2022; Mao, Ettredge, and Stone 2020). The PCAOB’s final rule covers both RT and non-RT auditors. Focusing on RT auditors, Krishnan and Li (2023) examines two research questions: What factors determine the use of RT auditors? And do RT auditors impact engagement audit quality and efficiency?

The PCAOB’s motivation for the new rule has been the concern that the involvement of component auditors can compromise audit quality. Multiple accounting scandals involving audits of companies’ subsidiaries (e.g., Satyam Corporation and Sara Lee Corporation), and audit deficiencies identified by PCAOB inspections of accounting firms, revealed situations where audit quality was compromised due to inadequate supervision of component auditors by lead auditors (International Forum of Independent Audit Regulators (IFIAR) 2018; PCAOB 2016). While auditors involved in these scandals were non-RT auditors, these examples illustrate the difficulties that can arise in audits of multinational corporations.

A fundamental difference in the lead auditor’s interaction with RT and non-RT auditors is the degree of supervision. Current auditing standards require lead auditors to “obtain, review and retain” information relating to the non-RT auditor’s audit work but require only minimal audit supervision of RT auditors because they do not include review of audit work. Auditing standards under the new rule will require more rigorous supervision of non-RT auditors than before. In contrast, for RT auditors, the new rule introduces a new auditing standard (Auditing Standard (AS) 1206) that requires more communication and disclosure between the lead auditor and RT auditor but essentially retains the minimal audit supervision requirements (see Section II). Because these policy changes are not just a rearrangement of prior standards, it is not clear a priori how they will affect practice relating to component auditors. Importantly, although the PCAOB’s deliberations over the years have generally focused on audit quality concerns relating to non-RT auditors, it speculates that the new rule can have the unintended consequence of an increase in RT audits and that it is not clear how this would affect audit quality given the “limited research findings” regarding division of responsibility (PCAOB 2022, 45). Thus, the analyses in Krishnan and Li (2023)—although focused on RT auditors only for a period preceding the new rule—can shed light on potential implications of the new rule for the future use of RT auditors and for audit quality and efficiency.

Krishnan and Li (2023) document that, during 2000–2017, the work of component auditors was referred to 1,170 times. Even though the use of RT auditors is infrequent,² firms using RT auditors are often large multinational corporations. RT auditors are used (in about equal proportion) in two settings: audits of clients’ consolidated subsidiaries and equity-method investees. Exhibit 1 provides examples of audit reports for the two settings. Table 1 shows that RT auditors audit on average about 19 percent and 9 percent of the audit engagements in the consolidated subsidiaries and equity-method investees setting, respectively.

AS 1205, which is currently (until the new PCAOB rule becomes effective) the primary relevant standard, (1) discusses conditions under which the lead auditor might decide to divide responsibility with the component auditor, (2) specifies the lead auditor’s responsibility when dealing with RT or non-RT auditor, and (3) requires disclosure of the RT auditor’s work in the audit report. As discussed, the relationship between the lead and RT auditors is one of minimal audit supervision. The lead auditor is required to scrutinize the RT auditor’s credentials (e.g., professional reputation and independence), adopt measures to assure coordination with the RT auditor’s activities, and confirm that the RT auditor’s procedures were performed in accordance with PCAOB standards. The lead auditor does not engage with the RT auditor’s audit processes. In contrast, for non-RT auditors, the lead auditor’s responsibilities in AS 1205 (in addition to those that apply for RT auditors) include detailed supervision (e.g., reviewing the component auditor’s scope of work, significant risks, responses, and the results of related procedures).

The PCAOB’s new rule replaces the existing AS 1205 with a new auditing standard, AS 1206, for RT auditors and a revised AS 1201 for non-RT auditors. The overall tenor of the new rule is a substantive increase in supervision stringency of non-RT auditors and relatively small change in the audit supervision of RT auditors.³ For non-RT auditors, new supervision requirements include communications regarding the risks of material misstatement, level of tolerable misstatement, and the threshold at which misstatements are considered immaterial and do not require aggregation. Further, the lead auditor must review the non-RT auditor’s planned audit procedures, obtain a written affirmation that the non-RT auditor performed its procedures as planned, review audit documentation to determine whether further procedures are necessary to address previously unidentified risks of misstatement, and evaluate the first non-RT auditor’s supervision of the second non-RT auditor’s work in multitiered engagement teams.

In contrast, AS 1206 does not increase, or even discuss, supervision requirements for RT auditors’ work. Unlike the previous requirements, the lead auditor should now disclose the RT auditor’s name in its audit report. The lead auditor must document, in more detail than under AS 1205, its relationship with the RT auditor—communicate, in writing, the plan to divide responsibility, obtain representation regarding whether the RT auditor is registered with the PCAOB for certain situations, and disclose whether conversion adjustments, if any, between different financial reporting framework (e.g., IFRS to U.S. GAAP) have been reviewed and, if so, by which auditor. Finally, the lead auditor should, if applicable, disclose the division of responsibility for internal control audits. Notably, these added requirements are not audit supervision requirements (e.g., instructions or review of the RT auditor’s audit work).

Lead auditors of firms with components (consolidated subsidiaries or equity-method investees) must incorporate audits of the components into their client’s audit. The components may be audited by component auditors or by the lead auditor’s audit team. The decision to use a component auditor for a subsidiary is made by the lead auditor or the subsidiary (Burke, R. Hoitash, and U. Hoitash 2020; Carson, Simnett, Thurheimer, and Vanstraelen 2022). However, an equity-method investee is likely to appoint its own auditor because the investor (the lead auditor’s client) has less than 50 percent ownership in it.

Krishnan and Li (2023) posit (following AS 1205) that the decision to divide responsibility with the component auditor (recall that this requires minimal audit supervision) likely depends on two factors: (1) the difficulty of the lead auditor to supervise the component auditor and (2) the materiality of the component. Krishnan and Li (2023) use three variables as proxies for supervision difficulty: the presence of foreign operations, the presence of merger and acquisition activities, and the lead auditor not having access to a strong global network. Component materiality is measured by the ratio of minority interest to total assets in the consolidated subsidiaries setting and by the ratio of equity investments to total assets in the equity-method investees setting. The results in Krishnan and Li (2023) indicate (see Table 2) that supervision difficulty is positively associated with the use of RT auditors for consolidated subsidiaries but not for equity-method investees, while component materiality is positively associated with the use of RT auditors in both settings.

Audit Quality: Are clients with consolidated subsidiaries (equity-method investees) that are audited by RT auditors associated with lower audit quality than clients with consolidated subsidiaries (equity-method investees) that are not audited by RT auditors?

An RT auditor is essentially an agent of the lead auditor (Carson et al. 2022; Mao et al. 2020). This arrangement may engender agency problems and, ultimately, impair audit quality. First, the absence of direct supervision by the lead auditor could reduce the RT auditor’s incentive to conform to the lead auditor’s audit requirements. Second, working with RT auditors can be more challenging than working with non-RT auditors; there is often less history of collaboration between lead and RT auditors. Finally, fee arrangements for RT auditors may differ from those for non-RT auditors. Downey and Westermann (2021) document that when lead auditors use a decentralized audit approach, the fees are negotiated between the component audit team and the component client. Thus, the RT auditor’s audit quality choices could be driven by its client’s (i.e., the component’s) requirements rather than the requirements of the lead auditor.

However, the RT auditor’s identity is disclosed in its audit report that is attached with the lead auditor’s report. This creates an “accountability effect” for its audit work (for example, in the event of audit failure) potentially engendering good audit quality, countering any negative effects of heightened agency problems.

Krishnan and Li (2023) employ three proxies for audit quality (which is unobservable) that are commonly used in academic research: discretionary accruals, Big R restatements, and little R restatements (DeFond and Zhang 2014). Discretionary accruals capture accruals that are not generated by normal operations, the idea being that they reflect choices made by a firm’s managers (e.g., earnings management). Big R restatements reflect corrections of previously reported financial statements in a Form 8-K filing with the SEC. Little R restatements incorporate prior-period errors in current year’s financials without an 8-K filing (Tan and Young 2015). For all three audit quality measures, a higher value indicates lower audit quality.

Krishnan and Li (2023) compare each sample of RT auditors with a matched-control sample of non-RT auditor arrangements. The results (see Table 3) show that the presence of RT auditors for consolidated subsidiaries is negatively associated with audit quality measured by two of the three proxies: discretionary accruals and Big R restatements. The presence of RT auditors for equity-method investees is negatively associated with audit quality only for little R restatements. Thus, RT auditors for both consolidated subsidiaries and equity-method investees are associated with lower audit quality, but the audit quality concerns seem to be less severe in the equity-method investee setting.

Audit Efficiency: Are clients with consolidated subsidiaries (equity-method investees) that are audited by RT auditors associated with lower audit efficiency than clients with consolidated subsidiaries (equity-method investees) that are not audited by RT auditors?

Accountability and agency effects can also impact audit efficiency. On the one hand, accountability might benefit efficiency due to a sense of responsibility on the part of the RT auditor. On the other hand, this same sense of responsibility may lead to over-auditing. Agency problems discussed earlier can lead to efficiency losses, but this effect is likely to be smaller for equity-method investees when compared to consolidated subsidiaries. While audit efficiency of RT auditors for equity-method investees can be impacted negatively by the lack of coordination with the lead auditor, other factors generating agency costs for RT auditors of consolidated subsidiaries (for example, difference in supervision and fee allocation complexity) do not apply to the equity-method investees setting.

Two variables are considered as proxies for audit efficiency: audit lag and nontimely 10-K reporting. Audit lag measures the number of days between a company’s financial year-end and the signing of the audit report. Nontimely 10-K reporting indicates whether the filing went beyond the mandated deadline.⁴ For both measures, a higher value indicates lower audit efficiency. As before, each RT sample is compared with a matched non-RT control sample. The results (Table 3) provide evidence of decreased audit efficiency when an RT auditor is present for both the consolidated subsidiaries and the equity-method investees settings.

The findings in Krishnan and Li (2023) have practical implications. First, from a regulatory perspective, Krishnan and Li (2023) provide evidence that is directly relevant to the issues concerning the PCAOB. The board concluded that division of responsibility is infrequent but is necessary in some situations where it is not possible for the lead auditor to supervise a component auditor (e.g., acquisitions occurring late in a fiscal year). The finding in Krishnan and Li (2023) that difficulty of supervision drives the lead auditor’s decision to divide responsibility in the consolidated subsidiaries setting corroborates the PCAOB’s statement on the necessity of keeping the practice of division of responsibility.

Second, despite the institution of a new standard for RT auditors, the PCAOB’s discussion throughout the process of developing the new rule does not indicate concern about audit quality for audits involving them. However, as discussed, Krishnan and Li (2023) find some evidence of lower audit quality and efficiency in both RT settings. These findings are concerning because one potential unintended consequence of the updated policy could be an increase in the use of RT auditors. The PCAOB notes that lead auditors may view the new rigorous requirements for supervision of non-RT component auditors as too costly and decide instead to divide responsibility with component auditors more often (PCAOB 2016, 42; PCAOB 2022, 45). In fact, the Center for Audit Quality (CAQ) notes an important omission in the new rule: it no longer provides guidance on whether to make reference and “only addresses the requirements when making reference to another auditor.” This can lead to “unintended changes from current practice.”⁵

Further, in feedback to the PCAOB, the CAQ and three Big 4 accounting firms (Deloitte, EY, and PwC) note that the use of RT auditors could increase in the future due to the adoption of mandatory auditor rotation in some countries, forcing lead auditors to switch from their affiliates, which are generally non-RT auditors, to RT auditors.⁶ PwC’s assertion asserts that it does not “believe this result would benefit audit quality,” which is consistent with the finding in Krishnan and Li (2023) about lower audit quality and efficiency in RT settings. Thus, although much regulatory attention has been focused on the audit quality effects of non-RT auditors (and potential improvements resulting from the new rule), regulators should monitor changes in the relative use of RT/non-RT auditors and impact on audit quality. Relatedly, the PCAOB’s new standard requires more active verification by lead auditors but does not assign more responsibilities for the RT auditor’s audit work. In short, AS 1206 may not have the mechanisms necessary to fix the audit quality and efficiency issues documented by Krishnan and Li (2023). Future research should examine changes in audit quality and efficiency in RT settings, post AS 1206 implementation.