catanach and walker - 1999 - the international debate over mandatory auditor rotation - a conceptual research framework [mar]

Walker This paper contributes to the recent international debate over mandatory auditor rotation by providing a conceptual research framework in which to view the tenure-audit quality re

Rotation: A Conceptual Research Framework

Anthony H Catanach Jr.

Paul L Walker

This paper contributes to the recent international debate over mandatory auditor rotation by providing a conceptual research framework in which to view the tenure-audit quality relation Audit quality is viewed to be a function of auditor performance The auditor’s ability and professional conduct are argued to be major factors affecting performance Economic incentives and market structure have endogenous relationships with both performance and tenure Research implications of the framework suggest that evaluating the efficacy of mandatory auditor rotation is likely to be a complex process, more involved than a simple association test of the tenure-audit quality relation The study also proposes several avenues for future examination: (1) evaluation of assumptions implicit in rotation arguments; (2) testing of magnitudes and effect directions; (3) examination of professional oversight con- trols; and (4) assessment of the costs of compulsory rotation © 1999 Elsevier Science Inc All

Anthony H Catanach Jr.● Villanova University, College of Commerce & Finance, 800 Lancaster Avenue, Villanova, PA 19085–1678; Phone: 610 –519 – 4825; Fax 610 –519 –5204; E-mail:

acatanach@msn.com Paul L Walker ● University of Virginia, McIntire School of Commerce, Monroe Hall, Charlottesville, VA 22903–2493.

Journal of International Accounting, Auditing & Taxation, 8(1):43– 66 ISSN: 1061-9518 Copyright © 1999 by Elsevier Science Inc All rights of reproduction in any form reserved.

attempted to dispel criticism of the profession in the United States (U.S.) byissuing a Statement of Position on Mandatory Audit Firm Rotation in 1992 In

1994, the U.S Securities and Exchange Commission (SEC) also considered theissue in its report draft on auditor independence, and the U.S Senate subsequentlydrafted legislation requiring auditor rotation for all telecommunication compa-nies.1The recent formation of the Independence Standards Board also supports aneed for analysis of regulatory actions that promote auditor independence.2Despite continued professional debate, little research exists to support either theadoption or rejection of mandatory rotation

This paper reviews the compulsory rotation issues raised internationally,summarizes the existing research evidence on this controversial topic, andproposes a conceptual framework to guide future examinations of the tenure-audit quality relation By identifying and organizing the major constructsassociated with the rotation issue, this framework provides a foundation forthe future development of more formalized logical or mathematical models

to yield testable research hypotheses The proposed framework suggests thatendogeneity among many of the constructs adds complexity to the tenure-audit quality relation, particularly when national and cultural differences areconsidered

THE CONTROVERSY SURROUNDING MANDATORYAUDIT ROTATION Arguments for Compulsory Rotation

Over the years major organizational collapses have been attributed topoor audit quality associated with a perceived lack of auditor independence.These alleged “audit failures” were deemed to have occurred because auditorsfailed to either detect or report material errors in the financial statements.Mandatory auditor rotation (compulsory rotation) frequently has been sug-gested as a means of strengthening independence and reducing the incidence

of audit failure.3

Proponents of mandatory rotation argue that it will prevent long-termauditor-client relationships from developing that could impair independence andobjectivity In fact, several studies suggest that longer auditor tenure leads to auditquality decline (DeAngelo, 1981b, 189; Deis & Giroux, 1992, 470; O’Keefe et al.,

1994, 53; Raghunathan et al., 1994, 33) Rotation advocates propose that it wouldreduce audit failures, force clients to adopt conservative accounting practices, andresult in more complete financial statement disclosures (OCA, 1994, 53) Rotationalso would ensure that a company’s representations, particularly those in subjec-tive and judgmental areas (e.g., intangibles), would be reviewed by different audit

“experts.” If the tenure period were limited, auditors also would have greaterincentives to resist management pressures (AICPA, 1992, 1–2) Finally, support-

ers of rotation suggest that it would foster a more competitive market However,

to date the evidence is mixed on this issue.4

Arguments Against Mandatory Rotation

Opponents of compulsory rotation generally acknowledge the potential efits of its adoption, but argue that implementation costs (both to the auditor andclient) greatly exceed the limited gains perceived (AICPA, 1992, 2) The CohenCommission (1978) indicated that duplication of start-up costs, including the timeauditors spend gaining familiarity with the client company, would increase auditorcosts considerably.5Lost audit efficiencies also may increase costs Arrunada andPaz-Ares assert that predecessor auditors will not be able to transfer their knowl-edge of the client, its accounting system, and market to successors, and this “value

ben-is destroyed by rotation” (European Accountant, 1995b, 7) Additionally,

Euro-pean professionals note that rotation may reduce auditor incentives to invest in

specific industries (European Accountant, 1995b, 7; Petty & Cuganesan, 1996,

41) If auditors must periodically abandon an industry that requires specialknowledge, they may be less willing to specialize in it

The accounting profession argues that mandatory rotation denies auditors theability to assess the true financial situation of a company because their under-standing of the client’s business, operations, and systems would be limited to only

a few years (European Accountant, 1995a, 4) This view is supported by a recent

AICPA study that found audit failures occurring more frequently when auditorswere performing their first or second audit of a company (AICPA, 1992).6Clients also may face increased costs if compulsory auditor change isrequired Management faces the potentially disruptive, time-consuming, andexpensive process of selecting new auditors, and familiarizing them with theorganization’s operations, procedures, systems, and industry (AICPA, 1992, 3).Increased auditor costs also might lead to reduced competition and higher client

audit fees (European Accountant, 1995b, 7) However, according to the Cohen

Commission, mandatory rotation also could result in “excessive competition”because clients would have difficulty evaluating audit quality differences acrossfirms This heightened competition might lower audit prices and create pressuresthat increase the likelihood of faulty audit work In addition, clients might beforced to accept a lower quality of service from an auditor who is a generalist, iffewer auditors invest in specialized industries (Petty & Cuganesan, 1996, 41).Finally, accounting professionals believe that several key factors in theauditing industry make mandatory rotation unnecessary The AICPA (1992, 5)suggests that auditors are motivated to maintain objectivity and independence bytheir desire to protect their reputation and client revenues Professional behavior

is further assured by quality control standards (e.g., peer review programs andperiodic partner rotation) implemented by accounting and auditing organizationsinternationally Accounting organizations also note that external market forces(e.g., litigation costs and negative publicity) provide controls on auditor behavior

THE INTERNATIONAL DEBATE AND EXPERIENCE

Countries with Active Compulsory Rotation Discussions

While most professional accounting organizations have been reluctant toaccept compulsory rotation, active debates on the issue continue in at leastfour countries: Australia, Germany, the United Kingdom (U.K.), and theUnited States.7 Research into recent audit failures in Australia finds a lack ofindependence as a major contributing factor (Walker, 1991, 82) Responding

to problem audits witnessed during the 1980’s, the Ministerial Council forCorporations appointed a working party in 1993 to review auditor regulation.While fixed appointment terms were recognized to have some merit, theaccounting profession did not support auditor rotation.8 Not surprisingly, theworking party recommended that “the law not impose fixed terms of appoint-ment for auditors or otherwise mandate the rotation of auditors” (Common-wealth Government, 1996, 87)

The auditor rotation question is not new to Germany, but the current debatemarks the first time that the issue has been discussed so widely The near collapse

of the Metallgesellschaft group in 1994 due to dubious oil derivatives contractshas renewed interest in the topic In the past few years, the president of theGerman Shareholders’ Association has called for the mandatory rotation of

auditors every five years (Corporate Accounting International, 1994; European

Accountant, 1995a, 4) As in many countries, the German accounting profession

is universally opposed to rotation However, Germany’s central bank, the bank, recently introduced auditor rotation to encourage German companies to

Bundes-consider the idea (European Accountant, 1996b, 4).9

During the 1980’s several high profile episodes occurred in the U.K thatsuggested deficient auditing practices: DeLorean, Johnson Matthey, Milbury,Westminster Property, Lloyd’s, and Alexander Howden In response, thegovernment proposed reforms (the EC Eighth Directive) that would curtailnon-audit services by auditors and require rotation However, lobbying by theaccounting profession and its corporate clients succeeded in postponing gov-ernmental intervention Auditing practices again came under scrutiny in theearly 1990’s in such cases as Polly Peck, Maxwell, BCCI, and Levitt.10 Thistime, the profession commissioned two investigations (the Cadbury Commit-tee and the McFarlane Report) to investigate the role of auditing in thesescandals and the need for self-regulation While the Cadbury Committeeopposed the rotation of auditors, the McFarlane report suggested that auditors

be appointed for a fixed five year term

During the last twenty years, the U.S Congress has held extensive hearings

on the public responsibilities of the accounting profession and its ability toregulate itself (Leibman & Kelly, 1992, 362) During the 1970’s, outside auditorsfailed to detect criminal activity among several large U.S corporations As a

result, the Moss-Metcalf inquiries examined the profession’s ability to detectmanagement fraud and its effectiveness in disclosing bribes to foreign officials(Previts & Merino, 1979, 318) The Metcalf Committee staff report concluded thatinsufficient independence between auditors and their clients may have contributed

to these conditions, and suggested rotation as a solution to the auditor dence problem.11 However, in 1978 the Cohen Commission, an independentcommission established by the AICPA, concluded that rotation should not berequired because its potential benefits would be offset by associated costs (CohenCommission, 1978, 108)

indepen-Audit failures stemming from the bank and thrift crises of the 1980’sreignited discussion of the effectiveness of the U.S accounting profession’sself-regulation program The Dingell Committee specifically investigated thepotential conflict of interest inherent in the auditor-client relationship The ac-counting profession responded by implementing several organizational changes topromote self-regulation including the adoption of a peer review program Nev-ertheless, critics of the profession continued to call for mandatory auditor rotation,standardization of fees, and other constraints on the auditor-client relationship(Chatov, 1985, 173) In 1991, the U.S Comptroller General discussed with BigSix audit firms the possibility of requiring banks to change their auditors on a

periodic basis (World Accounting Report, 1991, 257) In 1993, Walter Schuetze,

then Chief Accountant of the SEC, proposed mandatory auditor rotation as asolution to the accounting profession’s inability to self regulate (Craig, 1993, 18).Despite continued SEC interest in the independence issue, neither the SEC nor theAICPA, has endorsed the idea of audit firm rotation.12

International Experience with Mandatory Rotation

Italy currently requires the mandatory rotation of external auditors on listedcompanies, insurers, investment houses, newspaper publishers, state-owned busi-nesses, and companies benefiting from state aid These companies can keep thesame auditor for a maximum of three, three-year terms (or nine total years) TheItalian securities regulatory agency, Consob, also has the authority to approve theselection of auditors and to set fees Consob’s oversight may minimize any excesscompetition (e.g., fee cutting) caused by rotation, that could result in fewer auditprocedures and lower audit quality (OCA, 1994, 54) Although continuing cor-porate scandals in Italy suggest the ineffectiveness of compulsory rotation, its

“true” effects remain unknown.13

Israel adopted audit rotation on “government companies” in the early 1970’s

in response to calls by smaller accounting firms for more work While these firmsare required to change auditors every three years, the requirement appears to beloosely enforced Recently, management of the Electricity Corporation, one of thelargest and most profitable companies in Israel, successfully postponed a change

in auditors indefinitely (World Accounting Report, 1996, 20).

Prior to 1996, the Spanish government required companies to change tors every three, six or nine years with the total tenure period limited to a

audi-maximum of nine years (European Accountant, 1995b, 7) Unlike other countries,

Spain apparently did not adopt rotation in response to concerns over auditorindependence Instead, rotation was implemented to allow local firms to gain

market share (European Accountant, 1996a, 9) Spain recently has abandoned

mandatory rotation and some have considered its decision to be evidence thatrotation is not needed elsewhere.14

In summary, auditor rotation generally appears motivated by public concernsover large, high-profile corporate collapses As companies fail, auditors arecriticized for their performance and perceived lack of independence, and callsoriginate for regulation of the accounting profession As evidenced in Spain,mandatory rotation also may increase competition in the audit market

DEVELOPMENT OF A FRAMEWORK

Despite the historical controversy over compulsory rotation, little analytical orempirical research exists to either support or reject its adoption.15In fact, the AICPA(1992, 2) states “there is no empirical evidence to support the perceived benefit ofmandatory audit firm rotation.” Moreover, the potential impact of national andcultural differences on mandatory rotation also has been ignored Proponents ofrotation argue that long-term client relationships lead to lower audit quality and,ultimately, audit failure To support their position, they point to empirical studies ofthe tenure-audit quality connection To date, the preponderance of empirical evidencereveals a negative association between tenure and audit quality Using audit engage-ment hours as a proxy for audit quality, Palmrose (1989, 496) finds that qualitydeclines with firm tenure Copley and Doucet (1993, 32) report that the likelihood of

a substandard audit increases with the length of auditor tenure O’Keefe et al (1994,52) show initial engagements to be associated with fewer violations of generallyaccepted accounting standards than repeat engagements Raghunathan et al (1994,33) reveal that audit failures are more likely to occur in the first year or after the fifthyear, further suggesting a tenure-audit quality relation In their examination of law-suits filed against auditors, St Pierre and Anderson (1984, 249) reveal that only 23percent are associated with audit-client relationships of three years or less Deis andGiroux (1992, 474) and Giroux et al (1995, 76) also indicate that audit qualitydeclines with the length of auditor tenure

These investigations generally have assumed simple associations betweentenure and professional conduct (usually independence), as well as betweenprofessional conduct and audit quality Consequently, a simple causal relationbetween auditor tenure and audit failure is implied However, rotation supportersacknowledge that audit failures may be associated with factors other than tenure,suggesting that the link between longer terms and audit quality is not likely sosimple or direct A framework that accumulates and integrates the various

elements associated with the tenure-audit quality issue likely will assist bothresearchers and accounting professionals in addressing the compulsory rotationquestion Moreover, such a structure represents a necessary first step towardsfuture model development and empirical investigation The framework outlined inthis paper synthesizes the relevant audit tenure and quality literature and considersanecdotal arguments both in favor of and against rotation In doing so, itspecifically incorporates both micro and macro economic influences relevant tothe debate Framework development begins with a general discussion of auditquality and its determinants Next, the potential associations among audit quality,performance, and tenure are discussed The paper then reviews the possibleendogenous relations of economic incentives and market structure with perfor-mance and tenure A summary of the proposed framework’s major constructs andrelated professional factors is presented in Table 1.

Determinants of Audit Quality

Proponents of compulsory rotation view it as a means to reduce the

likeli-hood of audit failure However, audit failure is simply the observed outcome of

TABLE 1

Summary of Constructs and Related Professional Factors

Audit Quality

Knowledge (training, education) Experience (professional, industry, client-specific) Adaptability

Technological proficiency Professional Conduct (Reporting) Independence

Objectivity Integrity Due professional care Conflicts of interest Judgement

Economic Incentives General (fees, costs, profits)

Efficiency Innovation Management advisory services Litigation

Market Structure Competition (market share, concentration)

Supply and demand Entry barriers (economies of scale, product differentiation and diversification)

Professional regulatory mechanisms

poor audit quality DeAngelo (1981a, 115) defines audit quality as the probability

that an auditor will both discover and truthfully report material errors,

misrepre-sentations, or omissions in a client’s accounting system The probability ofdiscovering (detection) such breaches depends on an auditor’s technical compe-

tence or ability (Deis & Giroux, 1992, 464) Prior studies have suggested that the

probability of reporting is a function of independence (DeAngelo 1981a, 116;Raghunathan et al., 1994, 35), honesty (Watts & Zimmerman, 1981), and integrity(Johnson & Lys, 1990, 281) Therefore, this paper views the probability of

reporting errors as dependent on a much broader concept: the auditor’s

profes-sional conduct This approach allows the framework to encompass such additional

factors as independence, objectivity, due professional care, conflicts of interest,and judgment, all of which are cited by the U.S accounting profession as critical

to ensuring the public trust (AICPA, 1991, 4 –12)

DeAngelo (1981a, 116) indicates that the probabilities associated with tection and reporting are unlikely to be separable Deis and Giroux (1992, 464)note that prior audit quality studies generally assume auditor competence, andfocus on only one professional conduct factor: independence.16 However, theyacknowledge that without information about the auditor’s technical capabilities,the complex set of interrelationships between auditor performance and auditquality cannot be fully understood Previts (1985, 155) confirms that indepen-dence, skill and legality, professionalism, and adaptability play major roles in theattest function Therefore, this paper depicts audit quality as a function of twobroad performance constructs: ability and professional conduct

de-Yardley et al (1992, 153) indicate that the audit firm’s performance is afunction of economic considerations (fees, costs, profits), efficiency, and innova-tion Raghunathan et al (1994, 37–38) argue that economic incentives may

influence both detection and reporting They propose that large, fee-paying clients

are likely to be more difficult to audit, resulting in an increased likelihood of afailure to detect financial statement errors, despite adherence to professionalconduct standards They also suggest that the possibility of being terminated by

a client could affect an auditor’s reporting decision Consequently, the proposed

framework includes economic incentives as a factor influencing auditor

perfor-mance, specifically both ability and professional conduct

Both sides of the rotation debate argue that such a policy will affectcompetition Potential cost increases associated with rotation could increase fees,lead to price cartels, and reduce competition Alternatively, restricting auditortenure might promote increased fee competition if clients are less able to dis-criminate between the audit quality of different firms Proponents point to theItalian experience as evidence that competition increases with rotation However,the Spanish market provides evidence to the contrary Nevertheless, these expe-riences suggest that audit market structure must be considered in any attempt toexplain the audit function Elitzur and Falk (1996a, 260) concur, acknowledgingthat supply and demand may play a role in planned audit quality Additionally,

DeAngelo (1981b, 189) relies on competitive pressures to provide auditors witheconomic incentives to reduce audit quality Therefore, the proposed framework

includes a construct to capture market structure influences on both auditor

performance and audit quality

To summarize, this paper argues that audit quality is a function of twoperformance constructs (ability and professional conduct) each of which may beaffected by an auditor’s economic incentives Market structure also interacts withboth auditor performance and audit quality The following sections describe theinterrelationships among these factors in the context of the compulsory rotationissue Figure 1 depicts the conceptual framework of the tenure-audit qualityrelation proposed by this paper

Performance Determinants

Ability. Since the probability of discovering irregularities in a client’ssystem of accounting depends on an auditor’s capabilities, ability must be con-sidered when examining the compulsory rotation issue In fact, the first generalstandard of U S auditing states that “the audit is to be performed by a person orpersons having adequate technical training and proficiency as an auditor” (AICPA,

1972, AU Section 150.02) Deis and Giroux (1992, 464) argue that informationabout the technical capabilities of an auditor is critical to distinguishing betweenprofessional conduct (i.e., independence) and ability This construct is intended toinclude such aspects of auditor ability as knowledge (e.g., training, education),experience (e.g., professional, industry, client-specific), adaptability, and techno-logical proficiency

Empirical research results suggest that auditor ability is positively associatedwith audit quality (path B3F, Figure 1) St Pierre and Anderson (1984, 256)show that 72 percent of audit litigation cases allege a problem with the auditor’sinterpretation of accounting principles or auditing standards, or allegations offraud The auditor’s ability to address complex accounting issues seems particu-larly critical to ensuring audit quality For example, both Stice (1991, 530) andFeroz et al (1991, 112) document relationships between companies with complexfinancial accounting issues (receivables and inventory) and some type of problemaudit (suits against auditors and SEC enforcement actions, respectively).17Opponents of compulsory rotation contend that it will hinder specialization,

thus reducing auditor expertise (European Accountant, 1995b, 7).18 O’Leary(1996, 21) indicates that an auditor’s client-specific knowledge grows as thelength of the relationship increases These arguments suggest a positive relationbetween tenure and ability or expertise (path A3B).19

Elitzur and Falk (1996b)also suggest that longer tenure periods motivate the auditor to introduce techno-logical improvements Petty and Cuganesan (1996, 41) note that clients maychange auditors if they become dissatisfied with the quality of technical adviceprovided during an engagement This suggests that ability also may be a positiveinfluence on tenure (path B3A, Figure 1)

Professional Conduct. Professionalism is a key element in any model ofthe attest function (Previts, 1985, 155) However, its boundaries are unclear anddefining the construct is troublesome Previts (1985, 156) suggests the followingcharacteristics as descriptive of professionalism:

Most important might be a definable body of knowledge, individuality (decisions are personal, not collective), ethical constraints (self-discipline), altruism (placing the well-being of others above self-interest), and judgment (decision-making in the face of uncertainty).

Consequently, this construct embraces such attributes as independence, integrity,objectivity, judgment, and due care

Proponents of mandatory rotation argue that professional conduct affectsaudit quality Walker (1991, 83) cites auditor gullibility and a willingness toaccept management explanations of questionable transactions without question, ascontributing factors to recent audit failures in the U.K In a survey of U.K financedirectors, Hussey (1994, 21) finds that the “personal chemistry” between theauditor and client are important in auditor selection However, O’Leary (1994, 10)reveals that close personal relationships between auditors and their clients alsocan adversely impact audit quality as in the case of Deloitte, Haskins & Sells andAWA Ltd., an international trading company During AWA’s 1986 audit, theengagement partner became aware of problems in the company’s foreign ex-change department which he suppressed at the request of his close, personalfriends, the company’s general manager and chief internal auditor.20 Furthercomplicating the professional relationship issue is the common practice of ac-counting firms placing former employees with clients as part of their placementservices (Liebman & Kelly, 1992, 363) These examples appear to support apositive association between professional conduct and audit quality (path C3F,Figure 1)

Feroz et al (1991, 114) support a connection between professional conductand audit quality in an unrestricted tenure environment They report that almosttwenty percent of the SEC enforcement releases in their study address auditorindependence issues Stice (1991, 529) also reveals that the likelihood of auditorlitigation (presumably evidence of poor audit quality) is negatively associatedwith auditor conduct (i.e., independence)

The behavioral auditing literature suggests that the two performance minants also are related Specifically, ability (e.g., technical expertise, experience)affects judgment quality Frederick and Libby (1986, 270) demonstrate that theknowledge base of auditors affects their judgment Libby and Frederick (1990,350) find that more experienced auditors exhibit more complete knowledge offinancial statement errors Anderson et al (1991, 52) suggest that auditor expe-riences affect their cognitive representations Therefore, a positive associationbetween these two performance factors is shown (path B3C, Figure 1).Mautz and Sharaf (1961, 5) and Raghunathan et al (1994, 40) indicate thatover a long association with a client, the auditor may become less challenged andless likely to use innovative audit procedures, or may fail to maintain an attitude

deter-of prdeter-ofessional skepticism This suggests a negative association between tenureand the professional conduct of an auditor (path A3C, Figure 1) However,Magee and Tseng (1990, 315) find that the auditor’s value of incumbency presents

a threat to independence only under limited circumstances These conditions are:(1) auditors in the market must disagree among themselves about the appropri-ateness of the client’s desired reporting policy; and (2) the reporting issue mustaffect the client for more than one reporting period These results suggest thatprofessional oversight (a component of market structure, discussed below) mayplay a mediating role in the relation between tenure and professional conduct

Economic Incentives

This paper portrays economic incentives as a mediating factor betweentenure and auditor performance Positive (negative) economic incentives are thosethat auditors perceive will increase (decrease) the net cash flows of their firms

Ability Effects. Raghunathan et al (1994, 38) note that clients who payauditors substantial fees are “likely to be larger and/or more difficult to audit.”They argue that despite adherence to professional conduct standards, the com-plexity of these large fee engagements may adversely impact the auditor’s ability

to detect misstatements Therefore, the positive economic incentives provided by

auditors frequently expend economic resources (cash outflows) to specialize orinvest in technology to increase their ability (path D3B, Figure 1) This asso-ciation is likely negative since auditor economic resources are depleted to increaseability Alternatively, Sikka et al (1993a, 17) propose that technological devel-opments, as well as the standardization of audit methods, reduce start-up andlearning costs for auditors Therefore, ability may positively influence economicincentives as well (path B3D, Figure 1) Although Yardley et al (1992, 166)note no significant association between audit fee and expertise (when marketshare proxies for expertise), Craswell et al (1995, 297) report that specializedindustry auditors in Australia earn fee premiums over non-specialists This review

of the literature suggests that all audit firms in the market are equally efficient orinefficient with respect to all clients However, such an assumption is unlikely tohold across different countries Therefore, future researchers may wish to examineunder what circumstances and cultures the framework’s efficiency assumptionsremain valid

Professional Conduct Effects. The accounting profession has long agreedthat an auditor’s future economic interest in a client may affect that auditor’sprofessional conduct Raghunathan et al (1994, 36) view the auditor as aneconomic agent who makes self-interested decisions They argue that an auditor