The Relative Importance of Audit Quality Attributes Alan Kilgore, Renee Radich & Graeme Harrison T he importance of auditing and perceptions of au- diting to the efficientoperation of capital markets is well recognised. For example, Wallman (1996) and Coffee (2001) argue that without high quality audits the capital market would be inefficient and the cost of capital higher. Similarly, Monroe and Tan (1997: 35) conclude ‘the quality of an audit can affect the reliability of audited financial information, which in turn plays an important role in capital markets’. Audit quality is, therefore, fundamental in providing the confidence that capital market participants require and plays an important role in the effect ive allocation of economic resources. As a result of a series of corporate collapses and audit failures, perceptions of audit quality have been at issue over recent decades. These events and the ensuing investigations have resulted in changes to regulatory arrangements. In Australia the ninth stage of the Corporate Law Economic Reform Program (Audit Reform and Corporate Dis closure) Act 2004 (CLERP 9) proposed a range of measures designed to enhance the general cor- porate disclosure framework. The Australian Securities Exchange’s Corporate Governance Council also issued a report entitled Principles of Good Corporate Governance and Best Practice Recommendations (2003). Similarly, in the United States (US) the Corporate and Auditing Accountability, Responsibility, and Transparency Act of 2002, commonly called the Sarbanes-Oxley Act (2002) (SOX), was enacted in response to financial reporting irregularities (McGowan and Brisendine 2003). Various governance initiatives were also implemented in the United Kingdom (UK), including the issuance of the Combined Code in 2003. More recently the global financial crisis (GFC) has seen policy makers once again focus attention on the importance ofan effective audit function asa key compo- nent in effective capital markets and attempt to identify key drivers of audit quality. For example, in the US, the Advisory Committee on the Auditing Profession (2008) was established to provide advice to the US Treasury Department on the auditing profession. In the UK the Financial Reporting Council released The Audit Quality Framework (2008) and in Australia, the Treasury released Audit Quality in Australia – A Strategic Revie w (2010). Corporate collapses and audit failures have threatened the credibility of the audit function, with audit quality once again being a major issue. Motivated by the significance of auditing and perceptions of audit quality in enhancing the reliability and credibility of financial statements, this article investigates the relative importance of audit team and audit firm attributes in perceptions of audit quality by users of audit services. Data are gathered from 81 users of audit services and analysed using adaptive conjoint analysis in order to measure the relative importance of audit team and audit firm attributes in perceptions of audit quality. The results show that, in general, users of audit services perceive audit team attributes as being relatively more important than audit firm attributes in perceptions of audit quality. The findings of the study have implications for regulators and the accounting profession concerned with improving confidence in corporate governance and the effectiveness and integrity of the audit process, and for audit firms in monitoring and promoting the quality of their audit services. Correspondence Alan Kilgore, Department of Accounting and Corporate Governance, Macquarie University, NSW 2109, Australia. Tel: + 61 2 9850 8564; fax: + 61 2 9850 8497; email: alan.kilgore@mq.edu.au doi: 10.1111/j.1835-2561.2011.00141.x Australian Accounting Review No. 58 Vol. 21 Issue 3 2011 253 The Relative Importance of Audit Quality Attributes A. Kilgore, R. Radich & G. Harrison These investigations and regulatory changes make it clear thatthere has been considerable dissatisfaction with the effectiveness of corporate governance, the quality of the audit process and the roles of auditors and auditing. In response, regulators and the accounting profession have taken a number of policy measures to improve audit quality in both fact and appearance. Recent examples include the SEC’s proposed ban on audit firms undertaking non-audit services (NAS) in 2000 (SEC 2000) and the rapid adoption of SOX following Enron’s collapse (Francis 2004). However, these policy decisions have been m ade despite the fact that the empirical evidence regarding factors that can enhance or impair audit quality is inconclusive and uncertain. This study provides empirical evidence on the factors that are perceived to affect audit quality, specifically the relative importance of audit team and audit firm a ttributes in affecting audit quality as perceived by users of audit services. This study contributes to the literature in three ways. First, there is only limited empirical evidence of the factors that affect perceptions of audit quality by users of audit services. However, research into perceptions of audit quality is important because it is p erceptions that determine the credibility of the audit report (Shockley 1981) and that have the potential to erode public confidence in the integrity of the financial reporting system (Pany and Reckers 1988). Consequently, gaining an understanding of factors that affect perceptions of audit quality is important because it can help regulators and the accounting profession to formulate policy based on empirical evidence rather than on aprioriassumptions (Schelluch and Thorpe 1995). This evidence is also useful in ensuring that policies and practices support confidence and credibility in the audit function by encompassing attributes found to be relatively more important in perceptions ofaudit quality. Second, the findings of this study extend the scope of prior studies by examining the relative importance of audit team and audit firm attributes in perceptions of audit quality by examining whether users place greater importance on one type of attribute compared to the other. Prior studies on perceptions of audit quality have identified the distinction between, and importance of, audit team and audit firm attributes (see, for example, Schroeder et al. 1986; Carcello et al. 1992; Warming- Rasmussen and Jensen 1998). However, these studies have only examined the importance of audit quality attributes in an absolute sense and have not examined the relative importance (how important an attribute is relative to all others) of attributes in perceptions of audit quality. Third, some studies (see, for example, Schroeder et al. 1986; Carcello et al. 1992) found that audit team attributes were more important in assessing audit quality than audit firm attributes. By contrast, other studies (see, for example, Warming-Rasmussen and Jensen 1998) found a greater emphasis on audit firm attributes compared to audit team attributes. Additionally, there are sufficient variations in the number and type of attributes and subject groups used in prior studies that reduce their comparability and usefulness. Using conjoint analysis, this study makes a unique and innovative contribution to the literature. Conjoint analysis, which has not previously been used to measure the relative importance of audit quality attributes, allows direct evidence to be obtained regarding the relative importance of attributes in perceptions of audit quality. Literature Review and Research Questions There is a vast body of literature relating to audit quality and its measurement. Despite the extent of that literature, no single generally accepted definition or generally accepted measure of audit quality has emerged. Much of the audit quality literature derives from DeAngelo’s (1981) frequently cited definition of audit quality (the probability of both discovering and reporting a breach or misstatement in the accounting system or financial statements). Krishnan and Schauer (2000) observe that the two aspects of the DeAngelo (1981) definition are unobservable. Two approaches have been adopted to measuring audit quality; a direct and an indirect approach. The direct approach is based on the assumption that the probability of discovery and reporting of breaches will be reflected in outcomes of the audit, such as audit errors and financial statement outcomes. Examples of studies using the direct approach include Brown and Raghunandan (1995) and Colbert and Murray ( 1998) (audit errors), Balsam et al. (2003) (abnormal accruals) and Krishnan (2003) (valuation of earnings surprises). Research using the indirect approach is of two types. The first type measures audit quality using surrogates for audit quality, such as audit firm size (see, for example, DeAngelo 1981; Francis 1984; Palmrose 1986; Francis and Simon 1987; Craswell et al. 1995; Colbert and Murray 1998), audit tenure (see, for example, Geiger and Raghunandan 2002), provision of NAS (see, forexample, Wines 1994; Barkess and Simnett 1994; Craswell 1999; Elstein 2001) and industry experience (see, for example, Craswell et al. 1995; Hogan and Jeter 1999). These surrogate studies generally examine attributes of the audit firm, rather than the audit team, and typically examine only one attribute in each study. Thesecondtypeadoptsabehaviouralperspective and assesses audit quality by examining the attributes perceived to be associated with auditquality. Behavioural studies include Shockley (1981), Mock and Samet (1982), Knapp (1985), McKinley et al. (1985), Schroeder et al. (1986), Knapp (1987), Sutton and Lampe (1990), 254 Australian Accounting Review C 2011 CPA Australia A. Kilgore, R. Radich & G. Harrison The Relative Importance of Audit Quality Attributes Gul (1991), Knapp (1991), Carcello et al. (1992), Sutton (1993), Beattie and Fearnley (1995), Moizer (1995), Behn et al. (1997), Warming-Rasmussen and Jensen (1998), Behn et al. (1999), Chang and Monroe (2001) and Duff (2004). These behaviour al studies typically examine combinations of audit quality attributes, and include both audit firm and audit team attributes. Two conclusions can be drawn from the behavioural studies. The first is that users of audit services attach different degrees of importance to particular audit quality attributes in their perceptions of audit quality. The second is that, in general terms, audit team attributes are considered to be more i mportant than audit firm attributes in perceptions of audit quality. Based on this discussion, this study examines the following research questions: RQ1: Do users of a udit services attach different relative importance to audit quality attributes? RQ2: Are audit team or audit firm attributes considered relatively more important in perceptions of audit quality by users of audit services? The research questions are based on two assumptions. The first is that audit quality can be regarded as a set of attributes with differing relative importance for users of audit services. This assumption has support i n both the theoretical and empirical literature. The concept of product quality in the economics-based literature, summarised in Davidson and Monroe (1999), contends that goods and/or services areconsidered to be composed of sets or bundles of many different characteristics or attributes. Goods and/or services are only considered to have value to consumers because of thevarious attributes that combine together to form the good or service. Consumers will seek different attributes in the good or service and will place different values on each attribute depending on the consumers’ utility of the attribute. The ‘quality’ of an economic good or service can therefore be viewed as the sum of the values placed on the desired attributes contained in the good or service. With respect to the empirical literature, Simunic and Stein (1987) find that audit services are not seen as homogeneous but differentiated by users on the basis of differences in the values placed on the attributes of audit services. Similarly, the studies that examine audit quality from a behavioural perspective demonstr ate that there is a long list of attributes perceived to be important in assessing auditquality and there isconsiderable variation in the importance of those attributes to users of audit services. The second assumption is that attributes perceived to affect perceptions of audit quality may be classified as one of two types; audit team and audit firm attributes. Audit team attributes are characteristics of the audit team and audit partner and include the level of partner attention to the audit, communication and quality of working relationships between the audit team and client management, and the skills and experience of the audit team (drawing examples from Schroeder et al. 1986). Audit team attributes also include the knowledge of the audit team and partner both generally and in relation to the client’s industry, and the ethical standards of the audit team (Carcello et al. 1992); and the technical competence of the audit team (Beattie and Fearnley 1995; Zerni 2008). Schroeder et al. (1986) argue that, because the audit service is delivered by a relatively small number of professionals comprising the audit team, audit team attributes are likely to play an important role in perceptions of audit quality. Audit team attributes were also found to be important in perceptions of audit quality in the behavioural studies identified earlier. Audit firm attributes are characteristics of the audit firm. Both the surrogate and behavioural studies have found audit firm attributes to be important in percep- tions of audit quality. The surrogate studies identify audit firm attributes to include audit firm size, litigation experience, auditor reputation, auditor tenure, provision of NAS, audit structure and industry experience to be associated with audit quality. The behavioural studies identify audit firm attributes including quality control procedures, regulatory experience (Schroeder et al. 1986), audit firm responsiveness to client needs and compliance with general audit standards (Carcello et al. 1992), and firm industry experience and sceptical atti- tude of the firm (Warming-Rasmussen and Jensen 1998). Research Method Adaptive conjoint analysis This study gathers and analyses data using Sawtooth Software’s Adaptive Conjoint Analysis System (ACA). The ACA system, developed by Johnson (1987), is a computer-administered, interactive conjoint method and combines the design of conjoint tasks, data collection and data analysis. ACA uses a computerised questionnaire that consists of four sections, namely, importance, ratings, trade-off and calibration. The importance section asks respondents to indicate which audit quality attributes the y consider most important. In the ratings section, respondents provide a rating preference for different levels of the particular audit quality attribute. In the trade-off section a series of customised paired comparison trade-off questions are presented to respondents in order to obtain the conjoint data so that utility values can be calculated. In the calibration section a series of ‘calibration concepts’ are composed using those attributes determined to be most important. This information is used to calibrate the utility v alues calculated in the earlier sections of C 2011 CPA Australia Australian Accounting Review 255 The Relative Importance of Audit Quality Attributes A. Kilgore, R. Radich & G. Harrison the questionnaire for use in further data analysis. A unique feature of ACA is that the procedure is both ‘adaptive’ and ‘dynamic’ in that the answers already provided by respondents are used at each step to select the next paired comparison question. Audit quality attributes Collection of conjoint data using the ACA system requires identification of the attributes and attribute levels (descriptors) to be included in the sur vey questionnaire. Initially, five audit firm att ributes were identified from the literature as the main audit firm attributes implicated in perceptions of audit quality. These were audit firm size, audit tenure, provision of NAS, audit structure and industry experience. Five audit team attributes were selected: partner/manager attention to audit; communication between audit team and client management; partner knowledgeable about client industry; very knowledgeable audit team; and senior manager/manager knowledgeable aboutthe client industry. These audit team attributes were the five highest rated in the two most significant behavioural studies of perceptions of audit quality (Schroeder et al. 1986; Carcello et al. 1992). ACA survey questionnaire The survey questionnaire was previewed for errors using the ACA system tool that automatically checks the completed survey for errors or warnings. Four faculty members of the Department of Accounting and Finance at Macquarie University also reviewed the draft questionnaire. This process resulted in a number of minor changes being made to the questionnaire, namely the use of three levels for the audit firm s ize attribute (Big N firms, mid-tier and local firms), and audit tenure being specified as audit partner tenure with two levels (five years or fewer and more than five years). The reviewers also reported that the y had difficulty understanding and interpreting the term audit structure. Consulting the relevant literature (see, for example, Cushing and Loebbecke 1986; Ashton and Willingham 1988; Dirsmith and Haskins 1991; Carcello et al. 1995) revealed that audit structure methodologies are not likely to be familiar to users of audit services. Consequently, this attribute was replaced with the attribute ‘audit quality assurance review’ with two levels (internal/external). The inclusion of this attribute was motivated by debate surrounding the efficacy of audit quality reviews and the subsequent formation, in February 2006, of the Audit Quality Review Board (AQRB). 1 Furthermore, a number of empirical studies have attempted to determine whether peer review is effective in improving audit quality. Examples of these studies include Francis et al. (1990), Wallace (1991), Deis and Giroux (1992), Krishnan and Schauer (2000) and Hilary and Lennox (2005). These requirements and the distinction between external and internal quality control reviews are likely to be known by users of audit services and provide further motivation for the inclusion of this attribute with two levels (internal/external). The survey questionnaire was pilot tested by 19 faculty members of the Department of Accounting and Finance at Macquarie University. No difficulties in completing the questionnaire were reported. The final audit firm and audit team attributes, their operational definitions and the attribute levels used in the ACA survey questionnaire are shown in Table 1. A copy of the ACA survey questionnaire used in this study is available on request. Data collection The respondents chosen for this study are audit committee chairs/members and financial analysts/fund managers representing users of audit services. As no readily accessible database of either respondent group is available it was necessary to construct databases from publicly available information. The audit committee chairs/members database was constructed from annual reports of companies listed on the Australian Securities Exchange (ASX) selected at random from Aspect Annual Reports Online Database. Information regarding audit committee membership was obtained from the corporate governance section of the annual reports from which the names of audit committee chairs/members were obtained. Contact details for each audit committee chair/member were obtained from the Business Who’s Who of Australia D atabase, where available. A total of 318 audit committee chairs/members were identified, with 158 representing the Top 500 firms listed on the ASX, and the remaining 160 representing non Top 500 firms. The financial analysts/fund managers database was constructed from information obtained from the Australian Financial Services Directory (AFSD), which provides web addresses for all financial analysts/fund managers listed in the AFSD. Individual websites were accessed to obtain the names, position descriptions and contact details (mailing address, telephone number and email address). Only individuals whose position descriptions clearly identified them as working in financial analysis/funds management areas, or who could clearly be identified as having prior work experience in these areas, were selected for inclusion in the database. In order to obtain a similar sample size for financial analysts/fund managers, the names and contact details of 331 individuals were selected through this process. 256 Australian Accounting Review C 2011 CPA Australia A. Kilgore, R. Radich & G. Harrison The Relative Importance of Audit Quality Attributes Table 1 Audit firm and audit team attributes, operational definitions and attribute levels used in the ACA questionnaire Attribute Operational Definition Attribute Levels Audit firm size Big N/Mid-tier/Local firm 1. Big N auditor 2. Mid-tier firm 3. Local firm Audit partner tenure Duration of auditor–client relationship 1. ≤5 years 2. >5 years Provision of non-audit services (NAS) Percentage of NAS fees to audit fees 1. <30% of audit fees 2. 30–60% of audit fees 3. >60% of audit fees Industry experience Industry specialisation 1. Specialist 2. Non-specialist Audit quality assurance review Audit quality control review 1. External quality control review 2. Internal quality control review Partner/manager attention to audit Activity level of partner/manager 1. High activity level 2. Moderate activity level 3. Low activity level Communication between audit team and Nature and frequency of communication 1. Formalised and regular client management 2. Informal and regular 3. Ad hoc Partner knowledgeable about client Years of experience in client industry 1. ≤3 years industry 2. 4–6 years 3. >6 years Senior manager/manager Years of experience in client industry 1. ≤3 years knowledgeable – client industry 2. 4–6 years 3. >6 years Very knowledgeable audit team Years of experience in accounting and auditing 1. ≤3 years 2. 4–6 years 3. >6 years Data collection was undertaken in two stages. In stage 1, 649 invitation letters and acceptance forms were mailed to all 318 audit committee chairs/members and 331 financial analysts/fund managers. After following the multiple mailing procedure in line with Dillman’s (2000) Tailored Design Method (TDM), 173 responses agreeing to participate in the study were received from audit committee chairs/members (96) and financial analysts/fund managers (77). An overall response rate of 28.8% to the invitation letter was achieved. Stage 2 commenced with each audit committee chair/member and financial analyst/fund manager who agreed to participate in the study being sent an email thanking them and providing details of how to access the survey questionnaire online. Subsequent emails/telephone calls were made to encourage re- spondents to complete the questionnaire. In total, 22 respondents withdrew from the study despite initially agreeing to participate. A further 66 did not respond to repeated emails/telephone calls. A total of 85 respondents attempted the questionnaire, of which four were incomplete, resulting in an overall response rate of 46.8% of respondents that initially agreed to participate. For audit committee chairs/members, 40.6% responded; for financial analysts/fund managers 54.5% responded. Table 2 summarises the response rates for stage 1 and 2 procedures. Results ACA average utility values Prior to analysis, the utility data must be scaled and normalised so that utility values can b e compared across respondents. 2 Table 3 shows the average utility values. The average utility values represent the desirability of the attribute levels within each attribute. 3 Relative importance scores Toenable comparison between attributes, therelative im- portance score for each attribute across all respondents was calculated using the formula: RI i = (MaxU − MinU )i n  i (MaxU − MinU ) (1) Where: RI i = the relative importance of the i th attribute Max U = the maximum utility of the i th attribute Min U = the minimum utility of the i th attribute. Therelativeimportancescoresforeachattribute across all respondents are provided in Table 4 and are C 2011 CPA Australia Australian Accounting Review 257 The Relative Importance of Audit Quality Attributes A. Kilgore, R. Radich & G. Harrison Table 2 Response rates Audit Committee Chairs/ Financial Analysts/Fund Members (% of total) Managers (% of total) Total Percent Stage 1 Procedures Invitations mailed 318 (49.0) 331 (51.0) 649 100.0 Less: No response 192 (60.4) 236 (71.3) 428 64.6 Undeliverable 30 (9.4) 18 (5.4) 48 7.4 Responses received 96(30.2) 77 (23.3) 173 28.8 Stage 2 Procedures E-mails sent 96 (55.5) 77 (44.5) 173 100.0 Less: Withdrawals 15 (15.7) 7 (9.1) 22 12.8 No response 39 (40.6) 27 (35.1) 66 38.1 Complete responses 39 (40.6) 42 (54.5) 81 46.8 Incomplete responses 3 (3.1) 1 (1.3) 4 2.3 Table 3 Average utility values for all levels of all attributes for all complete responses (N=81) Average Attribute Utility Attribute Levels Values a Audit firm size Big N auditor 52.86 Mid-tier firm 8.64 Local firm −61.50 Audit partner tenure 5 years or less 21.54 More than 5 years −21.54 Provision of NAS <30% of audit fees 38.59 30–60% of audit fees −2.22 >60% of audit fees −36.37 Audit firm industry Specialist 50.61 experience Non-specialist −50.61 Audit quality assurance External quality assurance 14.96 review re view Internal quality assurance −14.96 review Partner/manager attention High activity level 58.24 to audit Moderate activity level −1.94 Low activity level −56.3 Communication between Formalised and regular 33.86 audit team and client Informal and regular 15.73 management Ad hoc −49 .59 Partner knowledgeable 3 years’ experience −52 .26 about client industry 4–6 years’ experience 9.72 >6 years’ experience 42 .55 Senior manager/manager 3 years’ experience −56.09 knowledgeable – client 4–6 years’ experience 11.66 industry > 6 years’ experience 44.44 Very knowledgeable 3 years’ experience −57.26 audit team 4–6 years’ experience 14.52 >6 years’ experience 42.74 a Utility values are interval data scaled to an arbitrary additive constant within each attribute using the ‘Zero-Centred Diffs’ method. shown in histogram graphic form in Figure 1. Relative importance scores measure how much influence an attribute has on a person’s choices. An attribute with a high importance score is more influential because the difference between the average utility values at the ‘best level’ and the ‘worst level’ for such an attribute is high. Since the relative importance scores of the 10 attributes total 100, if the attributes were equally preferred by respondents, the importance score of each attribute would be 10. 4 Table 4 and Figure 1 indicate that the most important attribute for respondents is ‘Audit firm size’ (relative importance score 13.63). The average utility values (Table 3) indicate that the preferred levels for all respondents are, firstly, ‘Big N auditor’ average utility value 52.86) and then ‘Mid-tier firm’ (average utility value 8.64), since the average utility values for both levels are positive. Other attributes considered important by respondents (relative importance scores greater than 10) in order of relative importance are ‘Partner/manager attention to audit’ (relative importance score 12.72), ‘Senior manager/manager knowledgeable – client industry’ (relative importance score 11.04), ‘Very knowledgeable audit team’ (relative importance score 10.65), ‘Commu- nication between audit team and client management’ (relative importance score 10.62), ‘Audit firm industry experience’ (relative importance score 10.51) and ‘Partner knowledgeable about client industr y’ (relative importance score 10.34). The least important attributes for respondents are ‘Audit quality-assurance review’, ‘Audit partner tenure’ and ‘Provision of non-audit services’ (relative importance scores less than 10). Within each attr ibute the level preferred by respon- dents is provided by the average utility values (Table 3). For the attribute ‘Partner/manager attention to audit’, respondent preference is for a ‘high activity level’ (aver- age utility value 58.24). Respondents prefer a minimum level of knowledge of at least four years for the attributes ‘Partner knowledgeable about client industry’, ‘Senior manager/manager knowledgeable – client industry’, and ‘Very knowledgeable audit team’, since the average utility values for levels ‘4–6 years’ experience’ (average utility values 9.72, 11.66 and 14.52, respectively) and ‘More than 6 years’ experience’ (average utility values 42.55, 44.44 and 42.74, respectively) are positive for all attributes. Respondents prefer regular communication 258 Australian Accounting Review C 2011 CPA Australia A. Kilgore, R. Radich & G. Harrison The Relative Importance of Audit Quality Attributes Table 4 Attribute relative importance scores for all complete responses (N = 81) Relative Importance Attribute Score Audit firm size 13.63 Audit partner tenure 6.39 Provision of NAS 8.94 Audit firm industry experience 10.51 Audit quality assurance review 5.15 Partner/manager attention to audit 12.72 Communication between audit team and client management 10.62 Partner knowledgeable about client industry 10.34 Senior manager/manager knowledgeable – client industry 11.04 Very knowledgeable audit team 10.65 with client management, whether formal or informal (average utility values of 33.86 and 15.73, respectively) for the attribute ‘Communication between audit team and client management’. For the attribute ‘Audit firm industry experience’, the level ‘specialist’ (average utility value 50.61) is preferred by respondents. Discussion The study found that, in general terms, users of audit services consider audit team attributes to be relatively more important than audit firm attributes in their perceptions of audit quality. All five audit team attributes were found to be relatively more important since their relative importance scores were greater than 10. Only two audit firm attributes had relative importance scores greater than 10. These two attributes were audit firm size and audit firm industry experience. The results for each of the 10 attributes are discussed in descending order of relative importance. Audit firm size This attribute was ranked as the most important attribute. In terms of attribute levels, respondents indicated a preference for Big N and mid-tier audit firms compared to local firms. This result is not surprising since audit firm size has been demonstrated to be an important factor in perceptions of audit quality in prior literature (see, for example, DeAngelo 1981; Francis 2004), with audit firm size serving as a proxy for audit quality, based on the assumption that larger audit firms are morecompetent andindependent thansmaller firms, resulting in audits of a higher quality. Partner/manager attention to audit Respondents ranked this attribute second in terms of relative importance. This ranking emphasises the perception by respondents of the importance of the audit partner/manager in providing a high level of audit quality. The attribute level with the highest average utility value was ‘high activity level’. This result supports findings of studies such as Knechel (2000), who observes that, since auditing is inherently a judgement and 15.0010.005.000.00 Audit firm size. Partner/manager attention to audit Senior manager/manager knowledgeable -client industry Very knowledgeable audit tea m Communication between audit team and client ma nagement Audit firm industry experience Partner knowledgeable about client industry Provision of non-audit services Audit partner tenure Audit quality ass urance review Figure 1 The relative importance of attributes for all respondents C 2011 CPA Australia Australian Accounting Review 259 The Relative Importance of Audit Quality Attributes A. Kilgore, R. Radich & G. Harrison decision-making process, audit quality is particularly contingent upon both the judgement and decision- making qualities of the members of the audit team and the audit partner. Senior manager/manager knowledgeable – client industry Respondents ranked this attribute third most important in their perception of audit quality. Greater average utility values were assigned to the levels ‘4–6 years’ experience’ and ‘More than 6 years’ experience’ compared to the level ‘3 years’ experience’, thereby rating more experienceas more importantin perceptions of audit quality. The importance accorded to this attribute is, again, likely explained by recognition that the audit manager is closely connected with the individual audit procedures and pra ctices that affect audit quality. Audit managers are responsible for the conduct of the audit and the supervision and review of the work of the audit team. Consequently, audit managers are pivotal in ensuring that the audit is conducted to a high standard. Very knowledgeable audit team Respondents ranked this attribute fourth most im- portant. They assigned hig her average utility values to the levels ‘4–6 years’ experience’ and ‘More than 6 years’ experience’ than to the level ‘3 years’ experience’, thereby indicating more experience as more important in perceptions of audit quality. This finding supports the results of Zerni (2008), w ho argues that audit quality is related to the characteristics of the audit engagement staff; that is, to audit team characteristics, which include factors such as the technical competence, knowledge and experience of the audit team. Communication between audit team and client management This attribute was ranked fifth in terms of relative importance. Respondents assigned the highest average utility value to the level ‘Formalised and regular’ communication and higher average utility values to the level ‘Informal and regular’ compared to ‘ad hoc’, further emphasising the importance of regular communication, whether formal or informal, as opposed to ad hoc communication. Audit firm industry experience Respondents ranked this attribute sixth in terms of relative importance, indicating that audit firm industry experience is considered less important than other attributes in perceptions of audit quality. Within the attribute respondents assigned thegreatestaverageutility value to the audit firm being an industry specialist. Prior studies (see, for example, Hogan and Jeter 1999; Deis and Giroux 1992; Solomon et al. 1999) support a link between audit firm industry experience and audit quality. These studies argue that because industry specialists have more experience, have financial savings (economies of scale) and a higher concentration of clients, they will make better judgements and therefore provide audits of higher quality. Further, studies by Craswell et al. (1995), Knechel et al. (2007) and Lowensohn et al. (2007) suggest that participants in the audit market perceive differences in audit quality due to factors such as industry specialisation. While the result in this study supports prior research in finding that respondents assigned the greatest utility to the ‘specialist’ attribute level, it also extends earlier research by finding that industry experience is relatively less important than other attributes in assessing audit quality. Partner knowledgeable about client industry Respondents ranked this attribute seventh in terms of relative impor tance, indicating that it was somewhat less important in perceptions of audit quality. The attri bute levels with the highest/lowest average utility values were ‘More than 6 years’ experience’ and ‘3 years’ experience’, respectively. These findings indicate that users of audit services, while not considering partner knowledge about the client industry to be particularly important in their assessments of audit quality, nevertheless prefer a more experienced partner compared to a less experienced one. An explanation for this finding is that the role of the audit partner in the conduct of the audit is somewhat different from that of the audit manager and audit team. While the audit partner has overall responsibility for the audit engagement, it is the responsibility of the audit manager to coordinate andsuperv ise the execution of the audit program, and the responsibility of the audit team to undertake the audit procedures. Users of auditservices appear to be aware that while the audit partner bears the responsibility for the audit, it is the other members of the audit team and their connection to individual audit procedures and practices that play a more significant role in achieving an audit of high quality. Provision of non-audit services Respondents ranked this attribute eighth in relative importance, indicating that the y did not consider this attribute to be part icularly important in their perceptions of audit quality. However they assigned 260 Australian Accounting Review C 2011 CPA Australia A. Kilgore, R. Radich & G. Harrison The Relative Importance of Audit Quality Attributes the greatest average utility value to the level ‘Less than 30% of audit fees’. The result regarding the average utility value is not surprising given the significant and ongoing attention that has been given to the issue of NAS provision by audit firms in both the academic and professional literature and by regulators and legislators in recent years. In part icular, the provision of NAS has been the subject of considerable scrutiny and recent legislative and regulatory change in Australia (CLERP 9) and elsewhere (e.g., SOX in the US). By indicating a preference for limiting the provision of NAS by audit firms, respondents appear to support recent legislative and regulatory efforts on this issue. Audit partner tenure Respondents ranked this attribute ninth in relative importance in their perceptions of audit quality, indicating thatthey did not consideraudit partner tenure to be very important. However, respondents assigned the highest average utility vales to the level ‘5 years or less’. The result regarding the average utility value is most likely due to audit partner tenure having received significant attention by regulators and the auditing profession. Specifically, CLERP 9, Australian Securities and Investments Commission’s Policy Statement PS 187 Auditor Rotation, and standards set by the Accounting Professional and Ethical Standards Board mandate lead audit and audit review partner rotation every five years. The intent of these regulatory and professional mandates is to restore public confidence in the auditing process and in the quality of that process. Nevertheless, it is interesting to note that respondents consider this attribute to be relatively less important in their perceptions of audit quality. Audit quality assurance review In terms of relative importance respondents ranked this attribute the lowest for all attributes (tenth). Within the attribute respondents assigned the greatest average utility value to the level ‘External quality assurance review’. Recent attention has been given to the issue of audit quality review. For example, the Audit Quality Review Board (AQRB) was established in 2006 with the stated objective ofundertaking independent, transparent monitoring of the quality of assurance services in order to enhance audit quality. The intent of this initiative is aimed at restoring public confidence in the auditing process and in the quality of that process. Given this intent, it is interesting to find that respondents consider this attribute to be relatively less important in their perceptions of audit quality. Implications To date, there is limited empirical evidence of the attributes that affect perceptions of audit quality and limited evidence regarding the relative importance of those attributes for users of audit services. The findings from this study are important because it is these percep- tions that determine the credibility of the audit report and are associated with confidence in the integrity of the financial report ing system. This confidence is important to capital market participants and plays a prominent role in the effective allocation of economic resources (Coffee 2001; Monroe and Tan 1997; Wallman 1996). The findings of this study have implications at an international level. As previously indicated, corporate collapses and audit failures and the ensuing investiga- tions have resulted in significant changes to regulatory arrangements in an effort to enhance the reliability and credibility of financial statements (e.g., CLERP 9 in Australia, SOX in the US and the Combined Code in the UK). More recently the GFC has seen policy makers again focus attention on the importance of an effective audit function as a key component in effective capital markets. Examples include the Audit Quality in Australia – A Strategic Re view (2010) by the Treasury in Australia, the final report of the Advisory Committee on the Auditing Profession in the US and the release in the UK of The Audit Quality Framework (2008). These reports have identified key drivers (indicators) central to achieving high audit quality and, in particular, have recognised the impor tance of the culture within au- dit firms andaudit partner and staffskills (i.e., audit team attributes) as important drivers of audit quality. The International Forum of Independent Audit Regulators (IFIAR) has also endorsed these drivers of audit quality. These developments have further encouraged regulators and policy makers at an international level to focus on these drivers of audit quality. The findings of this study that users of audit services consider audit team attributes to berelatively more important thanaudit firm attributes in their perceptions of audit quality, will assist regulators and policy makers internationally to address any real or perceived threats to these drivers of audit quality. The findings also have implications for regulators, the accounting profession and audit firms. The findings have implications forregulatory and professional bodies, which, in seeking to improve audit quality, have focused on audit firm factors such as audit tenure (both firm and partner), the provision of NAS and audit quality assurance reviews. This focus, however, has been determined more by aprioriarguments and normative assertions than by empirical evidence (Schelluch and Thorpe 1995). The findings of this study provide direct C 2011 CPA Australia Australian Accounting Review 261 The Relative Importance of Audit Quality Attributes A. Kilgore, R. Radich & G. Harrison evidence that, relative to audit team attributes, the only audit firm attributes relatively more important in perceptions of audit quality are audit firm size and audit firm industry experience. The findings of the study are therefore useful in assisting regulatory and professional bodies in formulating policy based on direct empirical evidence of the relative importance of attributes perceived to affect audit quality by users of audit services. The finding that, in general terms, audit team attributes are relatively more important than audit firm attributes is also relevant to audit firms in communicating and promoting themselves with clients and potential clients. Despite the greater importance accorded to audit team attributes, generally speaking, users of audit services must rely on attributes they can observe, particularly audit firm attributes (specifically audit firm size). Nevertheless, the opportunity exists for audit firms to make available through their com- munications and promotions with clients, particularly prospective clients, information about their audit team and itscharacteristics. Making audit team characteristics, particularly those characteristics that have been found in this study to be relatively more important in perceptions of audit quality, more publicly visible and available to clients may be a more effective means of demonstrating and signalling audit quality than promoting audit firm characteristics. This opportunity may be of particular importance to mid-tier and local audit firms, who may differentiate themselves and demonstrate audit quality through emphasising audit team attributes such as the level of partner involvement in the audit and their audit teams’ knowledge both generally and in specific client industries. This may enable non-Big N audit firms to signal to potential purchasers the quality of their services and obtain fee premiums not previously available to them. The findings also haveimplications forthe literature in terms of understanding the nature of audit quality. First, consistent w ith prior research, the study finds that audit quality is perceived as a multi-dimensional construct. This was demonstrated by the relative importance assigned by respondents to both sets of audit quality attributes and the trade-offs they made within and between these sets. Second, the finding that audit team attributes are perceived as relatively more important in perceptions of audit quality than audit firm attributes is consistent with the findings of Schroeder et al. (1986), Carcello et al. (1992), Herrbach (2001) and Knechel et al. (2007) but inconsistent with the findings of Warming-Rasmussen and Jensen (1998), who found audit firm factors to be more important. The current study lends weight to the greater per- ceived importance of audit team factors and sup- ports the suggestion that the results of Warming- Rasmussen and Jensen (1998) may have been culturally driven. Alan Kilgore, Renee Radich and Graeme Harrison are in the Department of Accounting and Corporate Governance at Macquarie University. Notes 1 The AQRB’s function is to review the systems, processes and controls applied by audit firms to ensure that they meet the independence and quality standards required in audits of listed entities. 2 The ACA system rescales utility data using a normalisation method termed ‘Zero-Centred Diffs’ so that each respondent has equal impact when computing average utility values for all respondents. 3 Horng (2005) indicates that due to the arbitrary origin within each attribute, the utility values of levels between attributes (e.g., ‘Audit-partner tenure of 5 years or less’ versus ‘External quality assurance review’) cannot be directly compared. Horng (2005) also points out that since utility values are interval data, they do not support ratio operations. Therefore, when comparing utility values within the same attribute, a level with a utility value of 30 is not twice as desirable as a level with a utility value of 15. However, the directionality of attribute-level utility values does reveal the overall preferences of respondents for levels within an attribute. For example, for the attribute ‘Audit firm size’ it can be concluded that, on average, respondents prefer a ‘Big N auditor’ (average utility value 52.86 in Table 3) to either a ‘Mid-tier firm’ (average utility value 8.64) or ‘Local firm’ (average utility value −61.50) 4 Horng (2005) points out that,unlike average utility values, relative importance scores are ratio data, have a meaningful zero point and are relative to the other attributes. Therefore, an attribute with a relative importance score of 20 is twice as important as one with a relative impor tance score of 10. The relative importance scores reported in Table 4 can be interpreted in the following way. Using the attributes ‘Audit firm size’ and ‘Audit partner tenure’ as an example, the ‘Audit firm size’ attribute (relative importance score 13.63) is seen as being more than twice as important as the ‘Audit partner tenure’ attribute (relative importance score 6.39). References Advisory Committee on the Auditing Profession 2008, Final Report, Department of the Treasury, Washington. Available at http://ustreas.gov. Ashton, R.H. and Willingham, J.J. 1988, ‘Using and Evaluating Audit Decision Aids’, Auditing Symposium IX (University of Kansas): Proceedings of the 1988 Touche Ross/University of Kansas Symposium on Auditing Problems. School of Business, University of Kansas, Kansas. Aspect Annual Reports Online. Accessible at: http:/www.aspectfinancial.com.au/af/home?xtm-licensee = annualreportsonline. Australian Financial Services Directory. Accessible at: http://www.afsd.com.au/directory.htm. Australian Securities and Investments Commission 2007, ‘Policy Statement 187 Auditor Rotation’, Australian Securities and Investments Commission, Sydney. 262 Australian Accounting Review C 2011 CPA Australia [...]... Accounting, Auditing and Finance, 5, 3: 369–78 Francis, J.R and Simon, D.T 1987, ‘A Test of Audit Pricing in the Small-Client Segment of the US Audit Market’, The Accounting Review, 62, 1: 145–57 Geiger, M and Raghunandan, K 2002, ‘Auditor Tenure and Audit Quality , Auditing: A Journal of Practice and Theory, 21, 1: 187–96 Australian Accounting Review 263 The Relative Importance of Audit Quality Attributes. .. Horizons, 10, 4: 76–97 C 2011 CPA Australia A Kilgore, R Radich & G Harrison Warming-Rasmussen, B and Jensen, L 1998, Quality Dimensions in External Audit Services – An External User Perspective’, European Accounting Review, 7, 1: 65–82 Wines, G 1994, ‘Auditor Independence, Audit Qualifications C 2011 CPA Australia The Relative Importance of Audit Quality Attributes and the Provision of Non -audit Services:... The Differentiation of Quality Among Auditors: Evidence from the Not-For-Profit Sector’, Auditing: A Journal of Practice and Theory, 19, 2: 9–25 Sutton, S.G 1993, ‘Toward an Understanding of the Factors Affecting Audit Quality and the Audit Process’, Decision Sciences, 24, 1: 88–105 Lowensohn, S., Johnson, L.E., Elder, R.J and Davies, S.P 2007, ‘Auditor Specialization, Perceived Audit Quality and Audit. .. Intermediaries, Auditor Independence and the Governance of Accounting’, Colombia Law and Economics Working Paper No 191, Colombia Law School Colbert, G and Murray, D 1998, The Association Between Auditor Quality and Auditor Size: An Analysis of Small CPA Firms’, Journal of Accounting, Auditing and Finance, 13, 2: 135–50 Commonwealth of Australia 2010, Audit Quality in Australia – A Strategic Review’, The Treasury,... and Samet, M 1982, ‘A Multi-attribute Model for Audit Evaluation’, Proceedings of the VI University of Kansas Audit Symposium, May 20–21, School of Business, University of Kansas, Kansas Moizer, P 1995, ‘An Ethical Approach to the Choices faced by Auditors’, Perspectives on Accounting, 6, 5: 415–31 Monroe, G.S and Tan, E 1997, The Relationship Between Audit Firm Size and Audit Quality: An Empirical Investigation... http://www.treasury.gov.au/documents/1745/PDF /Audit_ Quality_ in_Australia.pdf C 2011 CPA Australia Financial Reporting Council 2008, The Audit Quality Framework’, London Available at: http//www.frc.org.uk Francis, J.R 2004, ‘What Do We Know About Audit Quality , The British Accounting Review, 34, 4: 345–68 Francis, J.R., Andrews, W and Simon, D 1990, ‘Voluntary Peer Review, Audit Quality and Proposals for Mandatory Peer Review’, Journal of. .. Measurement System for Audit Quality , Proceedings of the 1990 University of Southern California Audit Judgement Symposium, 19–20 February, Newport Beach, CA Wallace, W 1991, ‘Peer Review Filings and Their Implication in Evaluating Self-regulation’, Auditing: A Journal of Practice and Theory, 10, 1: 53–68 Wallman, S 1996, The Future of Accounting, Part III: Reliability and Auditor Independence’, Accounting... Comparison of Factors that Affect Auditors’ Directors’ and Creditors’ Perceptions of Audit Quality , Paper presented at 13th Asian-Pacific Conference on International Accounting Issues, Rio de Janiero Francis, J.R 1984, The Effect of Audit Firm Size on Audit Prices: A Study of the Australian Market’, Journal of Accounting and Economics, 6, 2: 133–51 Coffee, J 2001, The Acquiescent Gatekeeper: Reputational... Independence Perspectives on Contemporary Auditing, Audit Centre of Excellence, Australian Society of CPAs Schroeder, M., Solomon, I and Vickery, D 1986, Audit Quality: The Perceptions of Audit Committee Chairpersons and Audit Partners’, Auditing: A Journal of Practice and Theory, 5, 2: 86–94 Securities and Exchange Commission 2000, ‘Proposed Rule: Revision of the Commissions Auditor Independence Requirement’,... Raghunandan, K 1995, Audit Quality in Audits of Federal Programs by Non-Federal Auditors’, Accounting Horizons, 93: 1–10 Deis, D.R and Giroux, G.A 1992, ‘Determinants of Audit Quality in the Public Sector’, The Accounting Review, 67, 3: 462–79 Business Who’s Who of Australia Accessible at: http://bww dnb.com.au/fact.asp Dillman, D.A 2000, Mail and Internet Surveys: The Tailored Design Method, 2nd ed., John . ‘Auditor Tenure and Audit Quality , Auditing: A Journal of Practice and Theory, 21, 1: 187–96. C 2011 CPA Australia Australian Accounting Review 263 The Relative Importance of Audit Quality Attributes. review Figure 1 The relative importance of attributes for all respondents C 2011 CPA Australia Australian Accounting Review 259 The Relative Importance of Audit Quality Attributes A. Kilgore, R their perceptions of audit quality. However they assigned 260 Australian Accounting Review C 2011 CPA Australia A. Kilgore, R. Radich & G. Harrison The Relative Importance of Audit Quality Attributes the