Understanding how audit market competition affects audit quality is of significant interest to academics, practitioners, and regulators, with important capital market and policy implications. Regulators around the world have expressed concerns that companies have too few choices in the audit market and believe that the lack of choice is resulting in low competition. As a result, regulators conjecture that audit firms supply low-quality audits and/or charge high fees. Proposals aimed at increasing audit market competition include breaking up the largest four audit firms to increase the number of audit firms available for companies to choose from.
From a theoretical perspective, it is unclear what effect audit market competition has on equilibrium audit quality. On the one hand, greater competition can foster audit process innovation and reduce auditor complacency, such that audit procedures are more rigorous and financial statements strictly adhere to Generally Accepted Accounting Principles (GAAP) (e.g., Government Accountability Office [GAO] 2008). Further, to the extent audit clients bear the costs of a low-quality audit (e.g., misstatement) and demand high-quality audits as a result, competition can strengthen auditors’ reputational incentives to supply high audit quality. On the other hand, competition can lower audit quality if it leads auditors to focus on appeasing clients by reducing professional skepticism and allowing clients excessive financial reporting discretion.
Several prior studies examine the relation between audit market concentration and audit quality/fees, but they find mixed results depending on how audit market concentration is measured (e.g., at the national- or city-level, within the Big-4 vs. between the Big-4 and the non-Big-4, etc.). Further, most prior studies do not directly address endogeneity concerns related to how markets become more or less competitive/concentrated and are thus unable to make any causal statements.
In our recent paper, we examine the relation between audit market competition and audit quality using a quasi-natural experiment in China. Specifically, we rely on the observation that audit production requires audit staff to spend significant amounts of time at the client location, and the time spent commuting between the auditor’s office and the client’s office affects the cost of conducting an audit.
Our analyses focus on companies headquartered in prefectural cities because these are smaller cities that are home to few audit firm offices. The median prefectural city that is home to at least one public company headquarters had no audit firm office located in the same city and only two audit firms with offices within a 100-mile radius. The introduction of bullet trains put companies headquartered in the median prefectural city within a two-hour train ride of most of the audit firms authorized to audit listed companies in China. We use the staggered inception of bullet train connectivity to prefectural cities in China as shocks to the number of audit firm choices available to companies located in these cities. By reducing travel time to the city, bullet train connectivity makes it economically viable for a significantly larger pool of audit firms to compete for the audit business of companies located in connected cities. Thus, bullet train connectivity significantly increases the threat of competition for incumbent auditors of companies located in cities with few audit firm offices.
Above, we assume that an audit requires audit firm personnel to be physically present in their clients’ offices for a significant portion of the audit. Performing an audit requires auditors to access their clients’ accounting records, such as invoices, contracts, general and subsidiary ledgers, journal entries, adjustments to the financial statements that are not reflected in formal journal entries, etc. In addition, audit procedures include physical inspections of records and assets, and observation of activities by auditors (e.g., inventory counting by the company’s personnel). In fact, audit firm personnel are typically provided working space in their client’s office during the course of an audit. Thus, we believe the assumption that audits are typically done in-person is justified and descriptive of practice, especially pre-COVID.
One noteworthy advantage of the Chinese setting is that it allows us to measure audit quality using proxies relatively less susceptible to measurement error. Specifically, one of our proxies for audit quality is the adjustments made to pre-audit earnings during year-end audits, which are based on confidential data obtained from the Chinese Ministry of Finance (MOF). The ability to access audit adjustment data helps us isolate the effect of an audit on earnings, controlling for all time-varying and time-invariant factors (e.g., performance, growth, etc.) that affect audited earnings via their effect on pre-audit earnings.
Using a generalized difference-in-differences design, we find that bullet train connectivity increases the incidences of GAAP violations by 4.5 percentage points (pp), decreases the likelihood that a client receives a modified audit opinion by 1.7 pp, and decreases the magnitude of income-decreasing adjustments to earnings during year-end audits by 1.6 pp (but no change in income-increasing adjustments). These results are consistent with the hypothesis that competition increases auditors’ focus on client retention such that auditors allow clients greater reporting discretion and reduce professional skepticism, lowering audit quality.
We conduct two sets of cross-sectional analyses. First, we exploit cross-sectional variation in the impact of bullet train connectivity on changes in audit market competition. We find that the negative relation between bullet train connectivity and audit quality is stronger when bullet trains impose greater competitive pressure on the incumbent audit firm. Second, we examine how cross-sectional variation in clients’ ex-ante preference for audit quality affects the relation between bullet train connectivity and audit quality. The ex-ante pairing of clients and auditors is not random but rather the outcome of a two-sided matching game where clients select auditors based on audit quality supplied and fees charged, and auditors select clients based on the audit risks they impose and the fees they are willing to pay. By increasing audit market competition, bullet train connectivity increases the client’s bargaining power relative to the auditor. We find that the negative effect of competition on audit quality is weaker when clients demand high-quality audits ex ante.
We conjecture that increases in competition lower audit quality because audits possess characteristics of a credence good, where buyers of such services (e.g., companies, investors) find it difficult to assess the quality of the service that was performed. As a result, when incumbent auditors are faced with a greater threat of competition, they can respond by lowering quality (i.e., reducing professional skepticism, pushing back less on clients’ accounting choices, etc.) to retain their clients. Due to the credence good aspects of auditing, such decreases in audit quality might not catch the attention of the client company and/or its investors, barring instances of ex-post audit failure (e.g., restatements, fraud, etc.). If regulators took policy initiatives that helped outsiders better evaluate audit quality, then perhaps greater competition would not negatively affect audit quality.
Yue Pan is a professor of Finance at the School of Economics, Xiamen University. She is also the Director of the Corporate Finance Department.
Nemit Shroff is the School of Management Distinguished Professor at the MIT Sloan School of Management.
Pengdong Zhang is an Assistant Professor at the School of Business in Sun Yat-sen University.
This post is adapted from their paper, “The Dark Side of Audit Market Competition,” available on SSRN.
The views expressed in this post are those of the author and do not represent the views of the Global Financial Markets Center or Duke Law.
 We recognize that this assumption might be weakened post-COVID due to improvements work-from-home technologies. However, our analyses pre-date COVID. In addition, our interest lies in examining the effect of competition on audit quality, and changes in travel time is simply a proxy for changes in audit market competition. It is not clear that COVID affects the economic incentives that shape the relation between competition and audit quality.