butcher et al - 2012 - perceptions of audit service quality and auditor retention

Perceptions of Audit Service Quality1 University of Western Sydney, Australia 2 Macquarie University, Australia Prior research has argued that perceptions of the breadth and quality of a

Perceptions of Audit Service Quality

1 University of Western Sydney, Australia

2 Macquarie University, Australia

Prior research has argued that perceptions of the breadth and quality of auditor-sourced services relative to rival audit firms are of paramount importance to client management when making auditor retention decisions However, the audit quality factors affecting auditor retention are under-researched This study examines the association between perceptions of audit service quality and auditor retention in the compulsory audit tendering context of local government in the Australian state

of New South Wales This context overcomes constraints of unlimited tenure and high costs of changing auditors in prior studies of auditor retention in voluntary tendering contexts, and has high retention rates at tender and evidence of council satisfaction with audit quality We use a questionnaire survey of 48 audit service quality attributes drawn from the audit service quality literature and administered to finance professionals and internal auditors across all New South Wales local councils We hypothesize a positive association between perceived audit service quality and auditor retention due to council satisfaction with audit quality We generate factor scores to test the hypothesis in a logistic regression and find evidence that higher-order audit quality factors of relationship (via the expertise dimension) and service qualities (via the responsiveness to client needs dimension) are associated with auditor retention, and are worthy of attention by audit firms in order to enhance client commitment and their likelihood of being retained as an incumbent auditor.

Key words: Audit service quality, auditor retention, compulsory

audit tendering, local government audit, relationshipcommitment in professional services

SUMMARY

Prior research has argued that perceptions of the

breadth and quality of auditor-sourced services

relative to rival audit firms are of paramount

importance to client management when makingauditor retention decisions However, the auditquality factors affecting auditor retention areunder-researched

This study examines the association betweenperceptions of audit service quality and auditorretention in the compulsory audit tenderingcontext of local government in the Australian state

of New South Wales This context overcomes

Correspondence to: Kym Butcher, School of Accounting,

University of Western Sydney, Sydney, Australia Email:

k.butcher@uws.edu.au

ISSN 1090-6738

© 2012 Blackwell Publishing Ltd

constraints of unlimited tenure and high costs of

changing auditors in prior studies of auditor

retention in voluntary tendering contexts, and has

both high retention rates at tender and evidence of

council satisfaction with audit quality

We use a questionnaire survey of 48 audit service

quality attributes drawn from the audit service

quality literature and administered to finance

professionals and internal auditors across all New

South Wales local councils We draw on the

marketing model used by Ismail et al (2006) to

hypothesize a positive association between

perceived audit service quality and auditor

retention due to council satisfaction with audit

quality

We use the audit service quality taxonomy of

Duff (2009) in confirmatory factor analysis, as well

as an exploratory factor analysis of the 48 attributes,

to generate factor scores to test the hypothesis in

logistic regression where the dependent variable

is auditor retention, measured as a dichotomous

decision, intention to retain or rotate the incumbent

auditor We use the taxonomies of Carcello,

Hermanson and McGrath (1992) and Schroeder,

Solomon and Vickery (1986) in sensitivity analysis

The study provides consistent evidence that

two of Duff’s (2009) four higher-order audit

quality factors, specifically, relationship (via the

expertise dimension) and service qualities (via the

responsiveness to client needs dimension), are of

paramount importance to clients when making

auditor retention decisions and, thus, worthy of

attention by audit firms in order to enhance client

commitment and their likelihood of being retained

as an incumbent auditor

1 INTRODUCTION

Perceptions of the breadth and quality of

auditor-sourced services relative to rival audit firms

are of paramount importance to client management

when making auditor retention decisions

(Hackenbrack & Hogan, 2005: 7) Duff (2009: 401)

argues that current issues of ‘the globalization of

business, commercialization of practice and

stakeholder dissatisfaction with the level of

audit quality are focusing auditors’ minds on

improving the quality of service they offer to their

clients’ Morton and Scott (2007: 18) argue that

clients’ perceptions of service quality are associated

with intentions or decisions to continue to

purchase services from the same provider

Consequently, to improve their likelihood of being

retained, audit firms need to be concerned with

both the quality of services offered and clients’

perceptions of the quality of those services (Guy,Harris & Williams, 1979: 17)

There is a considerable body of research onperceptions of audit quality attributes (e.g.,

Schroeder et al., 1986; Carcello et al., 1992; Aldhizer, Miller & Moraglio, 1995; Davis, 1995; Behn et al.,

1997; Chen, Shome & Su, 2001; Kilgore, 2010) Therehave also been many studies that have examinedclient perceptions of audit quality attributesassociated with auditor rotation and change(e.g., Healy & Lys, 1986; Williams, 1988; Haskins &Williams, 1990; Johnson & Lys, 1990; Woo & Koh,2001), and auditor selection and appointment (e.g.,George & Solomon, 1980; Addams & Davis, 1994;Abbott & Parker, 2000; Velury, Reisch & O’Reilly,2003; Godfrey & Hamilton, 2005) The consensus ofthis literature is that dissatisfaction with the quality

of the incumbent auditor is the main reason forauditor displacement (e.g., Burton & Roberts, 1967;Bedingfield & Loeb, 1974; Williams, 1988; Beattie &

Fearnley, 1995; Behn et al., 1997).

The corollary to this finding is that if clients aresatisfied with, or have a favourable perception of,certain audit quality attributes of their incumbentauditor, they will remain with the incumbent(Pandit, 1999) However, there is little empiricalresearch directly examining the audit qualityattributes associated with auditor retentiondecisions Apart from a small number ofpractitioner studies (e.g., Stanny, Anderson &Nowak, 2000; Rummel, Davidson & Action, 1999;Gabhart & Miller, 1984), only Pandit (1999) andMorton and Scott (2007) have examined directly theassociation between client perceptions of auditservice quality attributes and auditor retention.Pandit (1999) examined the association betweenfive dimensions of audit quality (the audit firm’sresponsiveness to client needs; industry expertise

of the audit firm; the audit firm’s executiveinvolvement in the audit; conduct of the auditfieldwork; and quality of the audit staff) and clientintentions to remain loyal to, and retain, theirincumbent auditor The five dimensions wereselected from the 12 factors identified by Carcello

et al (1992) from their list of 41 attributes affecting

perceptions of audit quality Pandit (1999) foundthat two of the five dimensions (the audit firm’sexecutive involvement and responsiveness to clientneeds) were associated with the intention to retain

an incumbent auditor Although hypothesized,

no association was found between the industry

expertise of the audit firm or the conduct of the

audit fieldwork and intention to retain the

incumbent auditor The result for industry

expertise was acknowledged as contrary to prior

research (e.g., Shockley & Holt, 1983; Beattie &

Fearnley, 1995) It was not possible to test the

hypothesis for the quality of the audit staff because

of low factor loadings for this dimension

Morton and Scott (2007) developed a 28-item

measure of audit service quality based on the

behavioural audit survey literature, exploratory

research and the SERVQUAL scale of Parasuraman,

Zeithaml and Berry (1985) Using a sample of 136

chief financial officers of Australian companies,

Morton and Scott (2007) found only a weak

association between perceived audit service quality

and retention of the auditor They argued that this

result might be because there was no legislative

limit on auditor tenure and because the cost of

changing auditor is high

The purpose of our study is to examine the

association between perceptions of audit service

quality attributes and auditor retention in a context

that comprises a high auditor retention rate and

client satisfaction with audit service quality, and

that also overcomes the limitations of Pandit (1999)

and Morton and Scott (2007) The context is that of

compulsory audit tendering (CAT) in New South

Wales (NSW) local government In this context,

legislation, specifically the NSW Local Government

Act 1993, mandates tenders for audit services be

called by local councils every six years irrespective

of whether the council is satisfied or dissatisfied

with its incumbent auditor, or wants or intends to

change auditors (NSW Government, 1993, Sec 423;

Butcher et al., 2011) The typical outcome of audit

tendering is auditor rotation (Butterworth &

Houghton, 1995), with the voluntary audit

tendering studies of Beattie and Fearnley (1998a,

1998b) and Johnson (1993) reporting rotation rates

of 82 per cent and 73 per cent, respectively

However, a high rate of auditor retention exists

in the CAT context of NSW local government

with empirical research providing evidence that

despite the (compulsory) opportunity to rotate,

an increasing number of NSW local councils have

chosen to retain their audit firm, with the auditor

retention rate increasing from 59 per cent at the

time of the first compulsory tender in 1995 to 86 per

cent following the third (in 2007) (Butcher et al.,

2011) Further, Boon, McKinnon and Ross (2008)

found that NSW local councils reported high levels

of satisfaction with their audit service quality

Additionally, the CAT context does not suffer theconstraints of no legislative limit on auditor tenureand the high cost of changing auditor that Mortonand Scott (2007) argued might have confoundedtheir results and caused the weak association theyfound between audit service quality and auditorretention First, CAT legislates a limit on auditortenure of six years Second, as councils have tocall tenders every six years, the costs of tenderingand auditor bid evaluation and selection have to

be incurred irrespective of whether the councilreappoints its incumbent auditor or appoints a newauditor Our study also overcomes limitations ofPandit (1999) As noted above, Pandit (1999) drew

on the 12 audit service quality factors of Carcello

et al (1992), but selected only five for study One of

the five (quality of the audit staff) could not betested because of low factor loadings and another(industry expertise) produced results contrary toprediction and prior research We use the complete

list of attributes and factors from Carcello et al.

(1992) This allows a comprehensive examination ofthe association between audit service quality andauditor retention

In summary, given the CAT context in NSW localgovernment of high auditor retention rates andcouncil satisfaction with audit quality, we use thiscontext to examine the association between auditservice quality and auditor retention We use aquestionnaire survey of 48 audit service qualityattributes drawn from the literature andadministered to finance professionals and internalauditors across all NSW local councils Weformulate a hypothesis based on theory drawn

from the marketing model used by Ismail et al.

(2006) and on Duff’s (2009) AUDITQUAL measure

of audit quality We find that audit firm expertiseand responsiveness to client needs are associatedwith auditor retention and are, therefore,important attributes for audit firms to focus on toenhance the likelihood of being retained as theincumbent auditor

The paper is organized as follows Section 2provides background information on the CATcontext of NSW local government Section 3 draws

on the audit service quality and marketingliteratures to formulate a hypothesis about theassociation between audit quality attributes andauditor retention Sub-hypotheses are formulatedbased on Duff’s (2009) AUDITQUAL model ofaudit quality Section 4 discusses the method,including variable identification, survey designand administration, sample selection, method of

analysis and specification of the model to test the

hypothesis Section 5 presents the results of testing

the hypothesis using AUDITQUAL Section 6

presents the results of sensitivity analysis using

Carcello et al.’s (1992) audit quality attribute

categories as adapted by Behn et al (1997), and

Schroeder et al.’s (1986) audit team and audit

firm attribute categories Section 7 presents the

conclusions, limitations and suggestions for future

research

2 BACKGROUND

This section provides demographic and descriptive

information on NSW local government, and

discusses the context of, and legislative

requirements for, local government auditing in

NSW

New South Wales (NSW) is the most populous

of the eight states and territories in Australia,

having approximately one-third (7.24 million) of

the Australian population of 22.5 million Although

the area of NSW is 800,000 km2, the population

density is just 8.44 per km2, with 63 per cent of

the population living in the state capital, Sydney,

and an urbanization rate of almost 80 per cent

(Wikipedia, 2011) There are 152 local councils in

NSW Local councils are the third tier of

government in Australia, the first and second

being the federal and state/territory governments,

respectively Councils are subject to state

government legislation; in the case of NSW by the

NSW Local Government Act 1993 administered by

the Division of Local Government, Department of

Premier and Cabinet

Because of the demographics of the state, local

council populations and areas differ greatly with

an inverse ratio of population to area between

metropolitan (particularly Sydney) and rural

councils The council with the highest resident

population of 300,000 (Blacktown City Council

within the Sydney metropolitan area) covers an

area of just 240 km2(a population density of 1,250

per km2), while the rural council of Urana has a

population of just 1,300 but an area of 3,357 km2

(a population density of 0.38 per km2) The mean

populations are 47,000 for all councils, 106,700

for Sydney councils and 27,500 for non-Sydney

councils, Revenues correlate positively with

population and population density, with a

minimum of $5.9 million per annum (Urana), a

maximum of $457.8 million (Sydney City) and

a mean (median) of $61.6 million ($38 million)(NSW Government, 2011)

Local councils are funded by their residentialand commercial constituents (the community)through rates, user charges and fees (amounting

to about two-thirds of income from all sources),and by the state government through general andspecific purpose grants (amounting to about

17 per cent of total income) (NSW Government,2011) The community and government delegatedecision-making authority and responsibility tocouncil managers and directors (councillors)for service provision within their areas andconstituencies, typically road and infrastructuremaintenance, community facilities (recreational,parks, libraries), town planning and developmentapplications, and waste collection

The Local Government Act 1993 provides anumber of mechanisms to monitor theaccountability of councils to their government andcommunity stakeholders A major accountabilitymechanism is the annual report and the associatedauditor’s report Section 428 of the LocalGovernment Act requires councils to prepareannual reports which include audited financialstatements The Act requires that the council’sannual report and audit report be submitted to theMinister for Local Government and the Executive

of the Division of Local Government Councils arealso required to present their audited financialstatements at a public council meeting no later thanfive weeks after receipt of the audit report.The legislative requirements for local councilauditing in NSW changed significantly in 1993with the introduction of CAT under the NSWLocal Government Act of 1993, Section 422, andfollowing the recommendations of the 1991 Report

on the Audit of Local Government conducted bythe Public Accounts Committee (Parliament ofNSW, 1991) The Act (Section 424) described CAT asthe process whereby an open tender for auditservices is sought from qualified auditors everysix years, with the incumbent auditor eligible toreapply CAT was introduced to increase localgovernment accountability, and to increase thecompetitiveness of the local government auditmarket in order that councils could obtain the bestpossible audit services in terms of costs and quality(NSW Government, 1992: 8, 36) Prior to 1993, onlyauditors holding a Local Government AuditingCertificate were eligible to audit local councils Inhis submission to the 1991 Public AccountsCommittee, the NSW Auditor-General described

the existing process of auditor appointment as

being ‘far too narrow and restrictive’

In essence, it has created a ‘closed shop’

environment Only auditors holding a Local

Government Auditing Certificate are eligible

for appointment and once appointed it is

very difficult to remove them, even where a

Council is dissatisfied with the service provided

(Parliament of NSW, 1991: 68)

CAT in the context of NSW local council audit

operationalizes the six criteria argued by the US

General Accounting Office (GAO) (1987) to be

desirable in audit procurement: competition,

solicitation, technical evaluation, a written

agreement, multi-year agreements, and specialized

finance professionals CAT requires that audits are

put to competitive and open tender every six years;

tenderers are provided with a tender specification;

tender specifications state that councils are not

required to take the lowest bid and that both price

and service-based quality factors will be taken into

account; the tender specifications and engagement

letter form the basis of the audit contract;

appointment is for a mandated tenure period of six

years; and appointment decisions are made by a

tender committee comprising finance professionals

with extensive local government knowledge

3 THEORETICAL FRAMEWORK AND

HYPOTHESIS DEVELOPMENT

3.1 Audit quality and audit service quality

Duff (2009: 401) notes that the literature does

not provide a consistent definition or

operationalization of audit quality Perhaps the

most agreed upon definition is that of DeAngelo

(1981) who defines audit quality as the probability

that the auditor will both discover irregularities

and breaches (due to the auditor’s technical

competence) and report the irregularities and

breaches (due to the auditor’s independence)

There are various models used in the literature to

operationalize, or measure, audit quality (see

Francis, 2004 and Kilgore, 2007 for reviews of audit

quality research), including those that focus on the

service quality dimensions of audit quality

As for audit quality, there is no consensus on

the conceptualization or operationalization of audit

service quality, as different researchers focus on

different models of service quality (Cronin & Taylor

1992; Ismail et al., 2006) Duff (2009: 401) states that

‘service quality can be thought of as a sustainablemeans of providing clients with what they want

or need, better, and more effectively’ Prominentexamples of models of service quality used

in the auditing literature include SERVQUAL(Parasuraman, Zeithaml & Berry, 1991),AUDITQUAL (Duff, 2004, 2009), and models usingaudit quality attributes, the three most significant

of which, according to Kilgore (2010), are

Schroeder et al (1986), Carcello et al (1992) and

Warming-Rasmussen and Jensen (1998)

Duff (2009) notes that although prior modelsusing audit quality attributes (such as

Schroeder et al., 1986; Carcello et al., 1992 and

Warming-Rasmussen & Jensen, 1998) containsimilar attributes, these models have beendeveloped and used in isolation from eachother By contrast, Duff’s (2004, 2009) model,AUDITQUAL, integrates the attributes fromthe prior audit quality literature into amultidimensional, structured model that has beentested for its psychometric properties over differenttime periods and samples, and has been found

to demonstrate both construct validity andmeasurement equivalence as a measure of auditquality AUDITQUAL consists of nine dimensionswithin four higher-order factors of competence,independence, relationship and service qualities.The competence factor relates to perceptions ofthe auditor to detect errors and comprises threedimensions of reputation, capability and assurance.The independence factor is uni-dimensional, whilethe relationship factor reflects the technicalqualities of the auditor and is two-dimensional,comprising experience and expertise Thefourth factor, service qualities, reflects the

‘often-unobserved nature’ of the auditor–clientrelationship and comprises the three dimensions ofresponsiveness, empathy and non-audit services(Duff, 2009: 404)

3.2 The marketing model

Theoretically, we use the marketing model

employed by Ismail et al (2006) who examined, for

500 publicly listed Malaysian companies, therelationship between audit service quality, clientsatisfaction and loyalty The marketing modelpostulates that these three elements are distinct andthat service quality leads to client satisfaction

which, in turn, leads to client loyalty (Ismail et al., 2006: 742) Ismail et al (2006: 739) cite Oliver’s

(1980) cognitive model of the antecedents of

satisfaction decisions in the marketing literature

to argue that client satisfaction (or dissatisfaction)

‘results from experiencing a service and comparing

that experience with the kind of quality of service

that was expected’ Ismail et al (2006: 741) also cite

Andreassen and Lindestad (1998) in the services

literature who argue that ‘customer satisfaction

is the accumulated experience of a customer’s

purchase and consumption experiences’ Rust and

Zahorik (1993), in the services literature, and

Storbacka, Strandvik and Gronroos (1994), in the

marketing literature, argue that satisfied customers

are more likely than dissatisfied customers to

remain loyal Client satisfaction is seen as

mediating the association between service quality

and loyalty (Ismail et al., 2006: 744).

Empirical research in the marketing literature

has found support for this model In reviewing

that literature, Ismail et al (2006) cite studies that

provide consistent evidence that customer or client

satisfaction is affected by perceptions of service

quality (e.g., Yi, 1990), and that satisfaction is

positively associated with client loyalty (e.g., Stauss

& Neuhaus, 1997) Despite these findings, the

evidence concerning the association between

service quality and client loyalty is mixed in

the marketing literature, perhaps due to the

complexity of the loyalty construct and the

different ways of measuring the construct,

including cognitive, attitudinal and behavioural

measures (Ismail et al., 2006: 741) Studies in the

marketing area have found positive (Boulding

et al., 1993; Bloemer, de Ruyter & Wetzels, 1998),

negative (de Ruyter, Martin & Bloemer, 1998), and

no relationship (Cronin & Taylor, 1992) between

service quality and loyalty

In the auditing literature, there is considerable

support for the model in terms of the relationship

between service quality and satisfaction Attributes

of audit service quality have consistently being

found to be associated with customer satisfaction

(e.g., Behn et al., 1997, 1999; Samelson, Lowensohn

& Johnson, 2006; Boon et al., 2008), and service

quality (operationalized using Parasuraman et al.’s

(1991) SERVQUAL model) has also been found to

lead to client satisfaction (e.g., Ismail et al., 2006:

743)

With the exception of Ismail et al (2006), however,

there is little research in the auditing literature on

the relationship between client satisfaction and

client loyalty Additionally, there is little research

on the relationship between audit service quality

attributes and client loyalty expressed as auditor

retention The research that exists shows mixedresults Three practitioner studies (Gabhart &

Miller, 1984; Rummel et al., 1999; Stanny et al., 2000)

provide no common attributes Rather, these studiesfind a range of attributes associated with experiencewith the incumbent auditor to be important,including data processing capabilities and theability of the audit partner to communicateeffectively (Gabhart & Miller, 1984), staff

personalities and meeting deadlines (Rummel et al., 1999) and overall quality of services (Stanny et al.,

2000) And, as noted in the introduction to the paper,the two academic studies of Pandit (1999) andMorton and Scott (2007) that have examined therelationship between audit service quality andretention also provide mixed results

Ismail et al (2006) examined the relationship

between audit service quality, client satisfactionand client loyalty They hypothesized positiverelationships between service quality andsatisfaction and between satisfaction and loyalty,arguing that clients who were satisfied with theirauditor would retain that auditor’s services for alonger period and would buy additional non-audit

services from the auditor Ismail et al (2006) used

the five-dimensional marketing model SERVQUAL

(Parasuraman et al., 1991) to measure audit service

quality, with the five dimensions being tangibility,reliability, responsiveness, assurance and empathy.They used hierarchical regression to examinethe mediating effect of client satisfaction on theassociation between audit service quality andloyalty They found that the service quality factors

of tangibility, reliability and empathy wereassociated with satisfaction, satisfaction wasassociated with client loyalty, and that satisfactionpartially mediated the relationship between auditservice quality and client loyalty via the reliabilitydimension of the five-dimensional model (Ismail

et al., 2006: 750) Their finding supported the

conjecture that client satisfaction is a mediator ‘tothe extent that it carries the influences of audit

service quality to client loyalty’ (Ismail et al., 2006:

749)

3.3 Hypothesis

Based on the foregoing theory drawn from themarketing model that service quality is associatedwith client satisfaction and that satisfaction withaudit quality is associated with client loyalty (Ismail

et al., 2006), we expect a positive and significant

relationship between perceptions of audit service

quality and auditor retention in our context of

compulsory audit tendering (CAT) in NSW local

councils This context has been demonstrated to be

one in which clients, local councils, are satisfied

with audit service quality (Boon et al., 2008), and

one in which the CAT regime is associated with

audit quality CAT has been proposed as a means of

enhancing audit quality and auditor independence

by improving the information set available during

the auditor appointment process, and altering the

supply of audit firms available to an organization

from which to appoint an auditor (Jensen & Payne,

2005a, 2005b)

The six-year mandatory tenure period in the

NSW local council context is also associated

with audit quality because, while it means the

incumbent auditor should face no dismissal risk for

the period, at the end of the period the council

might not reappoint (retain) the incumbent, but

might change auditors, meaning that the quality of

the predecessor’s audit will be subject to scrutiny

(Vanstraelen, 2000) This scrutiny is likely to be

intense because new auditors have an incentive to

provide superior audit service in the initial years of

engagement to demonstrate their ability to deliver

a quality audit (Craswell, Francis & Taylor, 1995)

Our study employs the Duff (2009) model of

AUDITQUAL to hypothesize and measure

audit quality We use AUDITQUAL because

of its theoretical properties as an integrated

multidimensional model of audit quality and

its psychometric properties of construct validity

and measurement equivalence As noted earlier,

AUDITQUAL consists of nine dimensions

within four higher-order factors of competence,

independence, relationship and service qualities

Our study also employs 48 individual audit quality

attributes drawn from the audit service quality

literature The attributes allow us to hypothesize

and measure seven of the nine dimensions in

AUDITQUAL, including the three dimensions of

the competence factor (reputation, capability and

assurance), independence, the two dimensions of

the relationship factor (experience and expertise)

and one dimension of the service qualities factor

(responsiveness) The other two dimensions of this

factor (empathy and non-audit services) are not

measured within the 48 attributes and are not

hypothesized We hypothesize as follows:

H 1 There is a significant and positive

relationship between perceptions of audit service

quality and auditor retention

This overall hypothesis is stated as sevensub-hypotheses using seven dimensions of Duff’s(2009) four-factor model of audit service quality.The sub-hypotheses are as follows:

H 1a There is a significant and positiverelationship between reputation and auditorretention

H 1b There is a significant and positiverelationship between capability and auditorretention

H 1c There is a significant and positiverelationship between assurance and auditorretention

H 1d There is a significant and positiverelationship between independence and auditorretention

H 1e There is a significant and positiverelationship between experience and auditorretention

H 1f There is a significant and positiverelationship between expertise and auditorretention

H 1g There is a significant and positiverelationship between responsiveness and auditorretention

4 METHOD

4.1 Variable identification

Figure 1 provides a comprehensive list of 48attributes drawn from the audit service qualityliterature The attributes are shown in Column 2and the literature sources in Column 3 Figure 2uses AUDITQUAL (Duff, 2009) to classify the

48 attributes into four higher-order categoriesand seven sub-categories: (i) competence(sub-categories of reputation, capability andassurance), (ii) independence (sub-category ofindependence), (iii) relationship (sub-categories ofexpertise and experience) and (iv) service qualities(sub-category of responsiveness)

Columns 1, 2 and 3 of Figure 2 show the categoryname, sub-category number and sub-categoryname, respectively Column 5 shows the individualattributes that are used in the survey questionnaire,classified into the seven sub-categories and fourcategories Column 4 shows the numbering of theattributes in the questionnaire The numbers in

Attribute

(survey item)

no.

1 The audit firm has been performing the audit for at least 2–3 years Carcello et al (1992), Chen et al (2001)

2 The audit firm is considered to be a specialist in local government

audit

Carcello et al (1992), Pandit (1999), Chen et al (2001)

3 The audit engagement partner has been on the audit for at least 2–

3 years

Carcello et al (1992)

4 The partner assigned to the audit engagement is very

knowledgeable about the industry

Carcello et al (1992), Behn et al (1999), Pandit (1999), Chen et al (2001)

5 The audit firm participates in the peer review process, and its most

recent peer review report was a clean one

Schroeder et al (1986), Carcello et al (1992), Aldhizer et al.

(1995)

6 The audit firm actively encourages staff members to take courses

and attend seminars in fields where the firm has major clients

Schroeder et al (1986), Carcello et al (1992)

7 The audit firm that is conducting the audit provides no consulting

services for the client

Carcello et al (1992), Chen et al (2001)

8 The audit firm is skillful in devising acceptable accounting

treatments for transactions that generate results that council

management wants

Carcello et al (1992), Pandit (1999), Chen et al (2001)

9 The audit firm has a policy on the maximum number of hours per

day and per week that its staff can work

Carcello et al (1992), Chen et al (2001)

10 The audit firm develops stringent time budgets for each audit area

and expects its people to meet them

Carcello et al (1992), Pandit (1999), Chen et al (2001)

11 The audit firm has a high audit staff turnover rate Chen et al (2001)

12 The audit firm conducts a thorough study of the client’s system of

Carcello et al (1992), Pandit (1999), Chen et al (2001)

14 The audit firm’s attitude is one of a sceptic, not one of a client

advocate

Carcello et al (1992), Behn et al (1999), Pandit (1999), Chen et al (2001)

15 The audit firm is agreeable to completing the audit by a date the

client has set

Schroeder et al (1986), Carcello et al (1992), Davis (1995),

Pandit (1999)

16 The auditors assigned to the engagement are very knowledgeable

about accounting and auditing standards

Carcello et al (1992), Davis (1995), Behn et al (1999),

Pandit (1999)

17 The audit team members as a group always exercised due care

throughout the engagement

19 There is frequent communication between the audit team and the

council’s audit committee

Schroeder et al (1986), Carcello et al (1992), Behn et al.

23 The audit firm keeps council management informed during the

year about accounting and financial reporting developments that

affect the council

Carcello et al (1992), Davis (1995), Chen et al (2001)

24 The audit engagement partner and manager make frequent visits

to the council during the conduct of the audit

Carcello et al (1992), Davis (1995), Pandit (1999)

25 The percentage that the council audit fee represents to the total

audit fee revenue of the audit firm is not material

Schroeder et al (1986), Carcello et al (1992), Chen et al.

(2001)

26 The personnel on the engagement below manager level have

passed the professional bodies’ exams

Carcello et al (1992), Chen et al (2001)

27 The audit partner on the engagement is a CA/CPA Carcello et al (1992), Aldhizer et al (1995)

28 Before accepting a new client, the CPA firm conducts a

pre-engagement investigation and goes through risk control

procedures including the conduct of a background search on

senior management of the prospective client

Carcello et al (1992), Chen et al (2001)

29 The audit firm reports internal control deficiencies and the

auditors’ recommendations on internal control are useful

32 The work performed by inexperienced members of the audit team

is supervised by the audit team manager

Davis (1995)

Column 4 are not sequential as the attributes were

randomly distributed in the questionnaire

4.2 Survey design

The survey questionnaire (designed following

Dillman’s (2000) Tailored Design Method)

consisted of two sections The first section

provided respondents with the list of 48

attributes Respondents were requested to:

‘Assume that you have been asked to evaluate the

quality of audit services provided by your

incumbent auditor Please indicate the extent

to which you agree or disagree that each attribute

will impact on your evaluation of the quality of

audit services provided by your incumbent

auditor’ Consistent with Butcher et al (2011) and

Beattie and Fearnley (1995), the request was

framed as a generic task in order that respondents

were not primed about the dependent variable of

the study (intention to retain the incumbent

auditor), and thus to avoid measurement

invariance (Brown, 2006) that might confound the

results of the study A seven-point Likert-type

scale was used ranging from -3 (strongly disagree

that the attribute would impact their perceptions

of audit service quality) to +3 (strongly agree) Theorder of the 48 attributes was randomized toavoid any potential bias resulting from demandcharacteristics or halo effect

The second section of the questionnairecontained demographic questions about therespondent (e.g., extent of experience) and therespondent’s council (e.g., size and location).The final question asked respondents: ‘If yourcouncil had to make an auditor appointmentdecision now, would it prefer to (i) retain theincumbent audit firm or (ii) appoint a new auditfirm?’ Responses to this question formed thedependent variable of auditor retention

We used preferred intention to retain or rotatethe incumbent auditor as the dependent variable

to reflect the dichotomous nature of the decision.The demand-side studies of auditor selection alsouse this dichotomy (e.g., Woo & Koh, 2001) Ameasure of preferred outcome follows Pandit(1999), and is based on research that showspreference follows attitude and precedesbehaviour, and is a commonly used approach inthe broader behavioural literatures to measure the

33 In all your dealings with the audit firm and individual audit team

members, the audit firm and audit team members never engaged

in any actions that would compromise its/their independence,

either in fact or in appearance

Behn et al (1999)

34 The audit firm has strict guidelines on the procedures that must be

completed before signing the audit report

Schroeder et al (1986), Chen et al (2001)

35 The cost to the audit firm of different audit procedures in terms of Carcello et al (1992), Pandit (1999), Chen et al (2001)

time expended is the major criterion as to whether a procedure is

used

36 The audit firm has rarely been found negligent in lawsuits brought

against it (alleging inadequate audit performance)

Schroeder et al (1986), Carcello et al (1992), Chen et al.

(2001)

37 The audit team members as a group have an adequate

understanding of the operations of the council

Aldhizer et al (1995), Sucher et al (1998), Chen et al.

(2001)

38 The audit team members conducted the audit fieldwork in an

appropriate manner

Behn et al (1999)

39 The audit firm makes extensive use of statistical techniques in

conducting the audit

Carcello et al (1992), Pandit (1999), Chen et al (2001)

40 The audit report and work papers receive a second partner review Aldhizer et al (1995)

41 The auditor adds value to the entity in terms of generating useful

ideas for improvement

Davis (1995), Sucher et al (1998)

42 The size of the audit firm in terms of its total revenue and number

of auditors is much larger than the average size in the region

Chen et al (2001)

43 The audit manager has been on the audit for at least 2–3 years Carcello et al (1992), Chen et al (2001)

44 The audit supervisor has been on the audit for at least 2–3 years Carcello et al (1992), Chen et al (2001)

45 The audit manager and supervisor assigned to the engagement are

very knowledgeable about the industry

Carcello et al (1992), Behn et al (1999), Pandit (1999),

46 The auditors are mindful of how busy the council’s key finance

staff are and contact these individuals only to the extent necessary

Carcello et al (1992), Pandit (1999), Chen et al (2001)

47 The external auditors co-operate with the internal auditors Davis (1995)

48 The number of hours spent by the audit team to complete the audit

(from the beginning of fieldwork to the audit report date) is

commensurate with a quality audit

Aldhizer et al (1995)

Category name Sub-category

number

Sub-category name

Survey attribute no.

30 31 36

42 10 12 13 17 26 27 29 35

37 38 39

48

5 6 9 11 21 28

32 34

Competence

40 7 8 14 18 25

33 2 4

16 20 45 1 3 43

44 15 19 22 23

24 41 46 Service qualities 7 Responsiveness

47

Survey item

The audit firm tends to have decentralized offices rather than centralized offices

The overall reputation of the audit firm is positive

The audit firm has rarely been found negligent in lawsuits brought against it (alleging inadequate audit performance)

The size of the audit firm in terms of its total revenue and number of auditors is much larger than the average size in the region The audit firm develops stringent time budgets for each audit area and expects its people to meet them

The audit firm conducts a thorough study of the client’s system of internal control

The audit firm makes extensive use of computers in conducting the audit

The audit team members as a group always exercised due care throughout the engagement

The personnel on the engagement below manager level have passed the professional bodies’ exams

The audit partner on the engagement is a CA/CPA

The audit firm reports internal control deficiencies and the auditors’ recommendations on internal control are useful

The cost to the audit firm of different audit procedures, in terms of time expended, is the major criterion as to whether a procedure is used

The audit team members as a group have an adequate understanding of the operations of the council

The audit team members conducted the audit fieldwork in an appropriate manner

The audit firm makes extensive use of statistical techniques in conducting the audit

The number of hours spent by the audit team to complete the audit (from the beginning of fieldwork to the audit report date) is commensurate with a quality audit

The audit firm participates in the peer review process, and its most recent peer review report was a clean one

The audit firm actively encourages staff members to take courses and attend seminars in fields where the firm has major clients The audit firm has a policy on the maximum number of hours per day and per week that its staff can work

The audit firm has a high audit staff turnover rate

Audit team members are rotated off the audit periodically

Before accepting a new client, the CPA firm conducts a pre-engagement investigation and goes through risk control procedures including the conduct of a background search on senior management of the prospective client

The work performed by inexperienced members of the audit team is supervised by the audit team manager

The audit firm has strict guidelines on the procedures that must be completed before signing the audit report

The audit report and work papers receive a second partner review

The audit firm that is conducting the audit provides no consulting services for the client

The audit firm is skillful in devising acceptable accounting treatments for transactions that generate results that council management wants

The audit firm’s attitude is one of a sceptic, not one of a client advocate

The audit staff assigned to the engagement have very high ethical standards

The percentage that the council audit fee represents to the total audit fee revenue of the audit firm is not material

In all your dealings with the audit firm and individual audit team members, the audit firm and audit team members never engaged in any actions that would compromise its/their independence, either in fact or in appearance

The audit firm is considered to be a specialist in local government audit

The partner assigned to the audit engagement is very knowledgeable about the industry

The auditors assigned to the engagement are very knowledgeable about accounting and auditing standards

The audit firm conducting the audit has other local council audit clients

The audit manager and supervisor assigned to the engagement are very knowledgeable about the industry

The audit firm has been performing the audit for at least 2–3 years

The audit engagement partner has been on the audit for at least 2–3 years

The audit manager has been on the audit for at least 2–3 years

The audit supervisor has been on the audit for at least 2–3 years

The audit firm is agreeable to completing the audit by a date the client has set

There is frequent communication between the audit team and the council’s audit committee

There is frequent communication between the audit team and council management

The audit firm keeps council management informed during the year about accounting and financial reporting developments that affect the council

The audit engagement partner and manager make frequent visits to the council during the conduct of the audit

The auditor adds value to the entity in terms of generating useful ideas for improvement

The auditors are mindful of how busy the council’s key finance staff are and contact these individuals only to the extent necessary The external auditors co-operate with the internal auditors