The accounting literature offers several potential explanations for why firms change auditors (i.e. auditor switching). These explanations include: changes in company management who may then prefer another auditor with whom they have some previous association or to remove an incumbent auditor associated with former management (Burton and Roberts, 1967), companies receiving qualified audit opinions indicating a conflict in the auditor-client relationship (Chow and Rice, 1982; Levinthal and Fichman,1988) or disputes over accounting or reporting methods (DeAngelo, 1982), companies in financial distress whose switching of auditors may be correlated with the explanations just mentioned (Schwartz, and Menon, 1985), and companies going public who change auditors in order to retain a larger accounting firm because of the prestige it may add to the offering (Menon and Williams, 1991). Throughout, the common variables related to auditor switching, and influencing auditor selection, are audit fees and qualified audit opinions; however, in more recent studies non-audit services and auditor-client relationships also appear to be important variables, and more so they have both come to be perceived to impact audit quality by creating potential independence issues. It is the association between auditor switches and auditor-client relationships that is the focus of this review.
In terms of auditor-client relationships as a variable affecting audit quality, audit firms are generally considered to be a major source of professionals for accounting positions outside public accounting supplying a large number of alumni to the market given their high turnover rates. Consequently, former employees of audit firms may become potential clients through their positions as senior management or audit committee members and these relationships have been proposed to have an impact on audit independence (and therefore audit quality). However, regulatory and academic interest has primarily focused on the impact on audit quality of those relationships where employees of audit firms (alumni) were hired directly by their audit clients, a practice that was at onetime not uncommon. The potential quality issues stemming from the practice of hiring audit firm alumni arise as the clients may be too familiar with the audit firm’s methods and procedures and the auditor may perhaps be too comfortable with his/her former colleagues. In the U.S., the Sarbanes Oxley Act of 2002 attempted to restrict the practice of hiring audit firm alumni (termed the ‘revolving door policy’) by imposing a one-year waiting period for audit firm employees who leave their accounting firm to become an executive or board (including its committees) member for a former client. In the U.K., the Ethical Standards set forth by the Auditing Practices Board in December 2004 have taken similar steps by imposing a two-year waiting period for audit partners (engagement, quality control, or key partner) who leave their accounting firm for a director or key management position with an audited entity. However, the effects of these waiting periods on questions of audit quality, including auditor switches, remain an empirical question as do questions largely overlooked by regulators focused on those relationships where employees of audit firms (alumni) were hired in financial oversight positions by firms who were not clients of their alma mater firm at the time of their departure but who may have become clients at some point afterwards.
Also serving to address issues of audit quality, two of the audit committee rules coming out of the U.S. Securities and Exchange Commission as a result of the Sarbanes Oxley Act of 2002 required that (1) all firms have an audit committee comprised of at least three members, including at least one financial expert, and composed entirely of members independent from management and that (2) the audit committee have responsibility for appointing the outside auditor. In the U.K., since November 2003, the Combined Code on Corporate Governance’s rules for firms listed on the London Stock Exchange on the first point are identical to the U.S. regulation. However, on the second point, the U.K. code requires merely that audit committees make recommendations in relation to the appointment of the outside auditor and that, and where the board does not accept the audit committee’s recommendation, the reasons for why the board has taken a different position should be disclosed in the annual report. In both settings, the primary role of the audit committee may be seen as to oversee the financial reporting process with the ultimate objective of ensuring high quality financial reporting, which might be achieved, among other things, through effective monitoring of management and a high quality audit. Therefore the first rule requiring all audit committee members to be independent seems to follow from conventional wisdom that independent directors are better monitors of management behavior than non-independent directors. However, the notion of independence has often been fraught with problems not only due to the difficulty in defining what exactly the concept means but in how it can be observed and measured.
While the corporate governance regulations arising in the post-Enron environment lead to a definition of independence that excludes current or former employees of the firm, members of an organization that receives contributions from the firm, material business relations such as those with customers, suppliers, or providers of professional services (legal, consulting, financial) to the firm, relatives of senior management, or members interlocked on another board, these regulations do not consider less traditional, but perhaps equally influential, connections between members which may lead to impairment of independence at either the individual member or board/audit committee group level. As a result, the intent of the second rule requiring that the audit committee assume responsibility, full or otherwise, for appointing the outside auditor, while representing an attempt to put distance between senior management and the external audit process due to potential conflicts of interest, may be compromised by firms replacing members lacking traditional independence with members having unregulated connections to other members of the board and senior management. This create an additional question of independence in terms of the extent such unregulated connections impact the oversight and monitoring responsibilities entrusted to the board/audit committee either by discouraging closely connected members from challenging the group or by encouraging members to have a voice within a closely connected group. Thus, relative to auditor switching, the question becomes whether the connections between the audit committee of the board of directors and senior management will have an impact on the audit committee’s support for or opposition to the selection of a firm with ties to senior financial management
Background and Theory
Theories considering economic exchange provide an initial framework for evaluating the engagement of an external auditor. For instance, Williams (1988) developed a theoretical model to explain auditor switches rooted in the stewardship hypothesis as reflected in the agency theory literature. The agency literature explains the demand for auditing as arising from conflicting interests between managers, shareholders and stakeholders (i.e. other entities contracting with a client). Under the stewardship hypothesis, a manager, assumed to engage an auditor to satisfy his own interests, would do so by seeking an auditor who satisfies the shareholder’s needs for assurance and therefore select an auditor of high quality (Williams). Managers would prefer to select an auditor with whom he or she has an accommodating relationship which will help ensure the manager maintains a favorable image as the good steward of shareholder’s investments (Williams). As agency theory considers the auditor-client relationship to be a nexus of contracts, new contracts between managers and shareholders are formed whenever the client hires a new manager or officer. When there is a change in senior management, they may request an auditor change because the old auditor is associated with the former management or because they prefer to work with an auditor with whom they have had favorable dealings in the past (Williams). However, the audit engagement process is seemingly more complex, and other disciplines provide alternative perspectives which more easily allow that decisions are made by both the prospective client and the proposing audit firm. For example, Simunic and Stein (1990) use portfolio theory to model the auditor’s client selection process which helps explain the auditor-client commitment as a two-way relation rather than a simple buyer-seller relationship in that if firms have private information about client performance, they may be more selective in their choice of client companies and the level of acceptable audit fees. This idea of a two-way relation is approached from a different perspective in the work by Levinthal and Fichman (1988) who view the affiliation between auditor and client as a dyadic inter-organizational relationship following social exchange theory.
Social exchange theory primarily differs from economic exchange theory in the way the actors are viewed. Where economic exchange views actors as dealing with a market, and responding to various market characteristics, social exchange theory views the exchange relationship under imperfect market conditions as the relationships between specific actors whose actions are each contingent on the actions of the other (Emerson, 1987). These relationships can also be considered in terms of social capital theory and techniques developed in social network analysis to explain how an actor’s level of connectedness contributes to access to information, ability to coordinate actions, and efficacy as a monitoring agent. The theory of social capital presumes that the actions of individuals are facilitated and coordinated through social organization, via common trust, norms and networks (Adler and Kwon, 2002). These actions will vary depending on the relationships an actor has with other actors, the structure of relationships among actors within the network, or both. Social network analysis and its techniques for measuring ‘connectedness’ rely on this idea of shared characteristics and experiences to ease communication and facilitate mutual understanding, foster personal connections and impede objective monitoring of management. This study will apply the concepts of social capital and social network analysis to senior management and board relationships to determine their potential association with auditor switching.
Relative to auditor-client relationships, early studies of auditor changes focused on the association between auditor switching and changes in management without looking at the inter-organizational or inter-personal link between them. These studies reported no significant association despite the notion that managers, and particularly new managers, may prefer another auditor with whom they have had some previous association in order to control their monitors. For example, Chow and Rice (1982) document that a change in management negatively affects the likelihood of auditor switching but that this association is not significant and that qualification of the audit report is the only significant variable in explaining switching. Later, Schwartz and Menon (1985) focus on motivations for failing firms in particular to change auditors, identifying management changes as one of the factors that could influence auditor switching. They find that, although failing firms have a greater tendency to switch auditors than healthier firms, neither management changes nor audit qualifications were statistically significant in this association. Finally, Williams (1988) model of auditor switches supported that auditor changes are not triggered by senior management ‘shopping’ for a lenient or accommodating auditor which was reaffirmed by Archambeault and DeZoort (2001) in their study showing audit committee independence and financial expertise minimizes the incidence of suspicious auditor switches that signal opinion shopping. Despite these early results provided evidence refuting that senior management seek auditor changes in order to control the auditors who monitor them, perceptions of this persist and, as noted by Lennox), policies that reduce managerial influence over auditor dismissal and selection decisions remain in the interest of regulators. As such, their potential impact on audit quality has been addressed through studies of accounting firm alumni in senior management and board of directors or audit committee independence.
Accounting firm alumni in senior management
One line of research has investigated how accounting firm ‘alumni’ in senior management positions of audit clients can potentially compromise audit quality through auditor switching. This line of research brings the inter-organizational link into the picture, as seen in the study by Iyer et al. (1997) who find in a survey U.S. accounting firm alumni that alumni are inclined to provide economic benefits to their former audit firms. One way that alumni in senior management positions of audit clients have been seen to benefit their former firm is to influence the appointment of that firm as the company’s auditor thereby directing business to their former audit firms. Additionally, the concern is that the familiarity between accounting firms and their former employees may impair auditor skepticism and objectivity if the auditor is responsible for auditing a company whose senior management was previously employed by the audit firm. Finally, because alumni have a good understanding of their former accounting firm’s audit methodology, there is the potential for them to more easily deceive auditors.
These concerns are supposed to have a role in certain restrictions that came out of the Sarbanes Oxley legislation on the ‘revolving door’ of accounting firm personnel to clients. Empirical studies of the U.S. market such as the one by Lennox (2005) find that auditors are less likely to issue modified audit reports for clients with accounting firm alumni in top-level management positions and by Menon and Williams (2004) document that abnormal accruals are larger for clients with former accounting firm partners as employees implying greater earnings management. These two studies provide evidence that audit behavior is more lenient towards clients with highly placed alumni. While this provides evidence of lower audit quality, Lennox (2005) reports that only 10% of audits in his sample involve alumni in senior management positions and only 7% of the sample in Menon and Williams (2004) have former audit partners. Further, Geiger et al. (2005) presents results for U.S. data which indicate that earnings management, in the form of increased accounting accruals, is no greater immediately before or after hiring in the companies engaging in the practice of hiring their financial reporting executives (CFO, VP-Finance or Controller) directly from their external audit form compared to three separate control groups hiring individuals from other sources or retaining their incumbent financial reporting executives.
Despite the seemingly low prevalence of alumni in senior management positions and mixed results, Geiger et al. (2008) continue on this subject determining the prevalence of the revolving door hiring practice and investigating the market’s reaction to companies hiring their senior accounting and finance officers directly from their external audit firms. They find, consistent with the two studies mentioned above, that the proportion of revolving door hires is relatively low at 6% of their sample, but that when they did occur, the market valued the revolving door appointments more positively than other appointments. So despite the impact of the ‘revolving door’ phenomenon may be less significant than commonly thought, there is indication that auditor-client relationships may have consequences for financial reporting and audit quality.
Board and audit committee independence
In the second line of research, concerning studies of board of directors or audit committee independence, the concern that audit quality can be impaired if the company is audited by a firm that formerly employed one of the company’s senior officers is seen to be mitigated to the extent a company’s audit committee is independent, as an independent audit committee might perceive these connections as a potential threat to audit quality, thereby preventing the appointment of officers’ former firms. Prior U.S. research linking audit quality with the boards of directors, and the audit committees of boards of directors, shows that audit quality is higher when boards and audit committees are more independent (i.e. having more outside directors). For example, Carcello and Neal (2000) show that in the presence of more independent boards, auditors are more likely to issue going concern reports for financially distressed firms and less likely to be replaced by the company following their issuance. Abbott and Parker (2000) report evidence that boards who are more independent boards and who are more active (i.e. meet more frequently) are more likely to demand higher audit quality, a demand which is met through hiring industry specialists as their auditors. And, finally, Klein (2002) present a negative relation between board and audit committee independence and earnings management as measure by abnormal accruals. Yet these studies do not specifically consider auditor alumni factors.
More directly, empirical studies such as the one by Lennox and Park (2007) find that an audit firm is more likely to be appointed if the company has an officer who is an alumni of the firm, especially if the officer has recently left the audit firm. However, they also find that this likelihood to be mitigated by the extent the company’s audit committee is independent (i.e. the more independent the company’s audit committee, the less likely the company is to appoint an officer’s former audit firm). Finally, in a study of alumni audit partners appointed to company audit committees in the U.S., Naiker and Sharma (2009) examine the association between internal control deficiencies reported under Section 404 of the Sarbanes Oxley legislation and the presence of former audit partners on the audit committee who are affiliated and unaffiliated with the company’s external auditor finding a negative association for both affiliated and unaffiliated former partners. Naiker and Sharma (2009) interpret their findings to suggest that both affiliated and unaffiliated former partners on the audit committee are associated with more effective monitoring of internal controls and financial reporting and that concerns about ‘revolving door’ appointments are less applicable to the audit committee setting because audit committee members are sensitive to the possibility of affiliations threatening audit quality.
Board and audit committee ties
Recently, the independence of the board and the audit committee has been considered in terms of its internal connectedness, an application of social network analysis, which focuses on the professional and social connections within the board or between the CEO and other board members. There are two dimensions to consider in measuring internal connections: 1) the degree and 2) the type (professional or social). The degree can be defined at two levels of connection: a) a first degree connection, i.e., two members know each other because they sit or have sat on another board together be it of a public or private company or of an organization (charitable, university, art, sporting, political, etc.), have been employed by the same company at the same time or they have direct social ties such as membership in the same organizations or attendance at the same university, and b) a second degree connection where two board members share a common professional or social tie with a third person or again they share a common backgrounds (same alma mater, same nationality, same religion, same professional background (e.g., both are engineers, CA, served in the military). The literature in this area is concerned with issues of corporate governance such as the board’s impact on corporate decision making or internal corporate governance practices. For example, the Hwang and Kim (2009a) study expands on the work of Klein (2002) to consider the impact of social ties between the CEO and audit committee members showing that these ties are associated with higher levels of earnings management and higher CEO bonuses. They also suggest that internally connected audit committee members have been replacing financially or family connected members since the regulations governing audit committee independence were enacted. Using the same dataset, Hwang and Kim (2009b) suggest that CEO compensation is lower and more sensitive to pay performance when boards are conveniently and socially independent from the CEO. Their measure of social ties is quite broad with the traditional definition of independence serving as a direct measure (current employee of the firm, former employee of the firm, member of an organization that receives contributions from the firm, business relations (customer/supplier) with the firm, offers professional services (legal, consulting, financial) to the firm, relative of the CEO, interlocked on another board) as well as what is categorized above as second degree characteristics serving as an indirect measure (CEO and member have serve in the military, share the same alma mater, same regional origin, studied the same discipline, worked in the same industry, common third party connection).
Measuring these connections, while possible using biographic information found in annual reports or proxy statements, is quite tedious (refer to Appendix B for more information on measurement). More recently, different database mangers have begun collecting the data and allowing broader research, one of these being BoardEx. Fracassi and Tate (2009) examine the external network connections between the CEO and his board members using BoardEx data for the S&P 1500 companies between 2000 and 2007. They look at the current and past employment network, the education network as the other activities network which they aggregate into a Social Network Index. They find that more powerful CEOs tend to appoint directors with whom they share social ties. Further, firms with higher internal connectedness have lower market valuations and engage in more value destroying acquisitions in the absence of compensating governance mechanisms. Schmidt (2009) examines the cost and benefits of ‘friendly boards’ during mergers and acquisitions. Using BoardEx data, he defines friendly boards as those where greater numbers of directors are connected to the CEO via clubs, fraternities, not-for-profit organizations, background (religious organizations, military), network organizations, or having done their MBA together. He finds that friendly acquiring boards lead to higher abnormal announcement returns when advisory needs are higher but the opposite when monitoring needs are greater.
Despite an existing literature on audit quality in terms of auditor switching and a growing literature on board and audit committee connections in the finance and economics discipline (refer to Appendix A summarizing the literature reviewed above), previous studies of audit quality and board connections in the U.S. have been limited to professional (alumni) affiliations of the board of directors or senior management of the company to the incumbent auditor and to board/audit committee independence issues but have not considered the possibility of affiliation, and therefore possible impairment of independence of a different nature, through non-professional (i.e. social) means. In addition, there are few studies looking at the relationship between auditor switching and auditor-client connections through the closeness of the client’s management and audit committee members, in particular in the environment pre- and post-changes to the corporate governance regulations on revolving door hires and audit committee independence in the U.K. In this study, we propose to examine whether connections, in terms of auditor-client affiliations as well as professional and social ties between senior management and audit committee members, have an impact on auditor switching not in terms of causing the auditor switch so much as influencing the selection of the new auditor. In particular, this study will answer the questions of (1) whether senior financial management (CEO and CFO being the most influential) of the company appoint their former audit firms given a change in auditors and (2) whether the extent of senior management and audit committee connections facilitates or inhibits such appointments. These two questions will be considered in light of relevant changes in the U.K. corporate governance code designed to encourage audit quality by promoting auditor independence.
In a preliminary stage, we start with a sample of 563 auditor changes in the U.K. between 1997 and 2009 and investigate the role of auditor-client relationships in the auditor selection decision through the study of former audit alumni in financial oversight roles who select their alma mater firm as auditors. Later, we may investigate whether the closeness of the senior management and audit committee members, as measured by their professional and social ties, has an impact on this selection. We examine the U.K. setting since 1) it represents a slightly different corporate governance and institutional environment than that of the U.S. and 2) we expect greater connections in the UK, at least prior to the enactment of changes to the UK code, due to its smaller size and therefore smaller network.