You are on page 1of 14

The British Accounting Review 47 (2015) 395e408

Contents lists available at ScienceDirect

The British Accounting Review
journal homepage: www.elsevier.com/locate/bar

Auditors' identification with their clients: Effects on audit
quality

Jan Svanberg 1, Peter Ohman*

Mid Sweden University, Department of Business, Economics and Law, Centre for Research on Economic Relations, SE-851 70 Sundsvall,
Sweden

a r t i c l e i n f o a b s t r a c t

Article history: Although client familiarity is desirable from the auditor's perspective, identifying with
Received 1 November 2013 clients threatens auditor objectivity. This study examines the extent to which non-Big 4
Received in revised form 9 June 2014 auditors identify with clients, the effect of auditoreclient identification on auditors' client
Accepted 1 August 2014
acquiescence to client-preferred treatment, and, finally, whether the harmful effects of
Available online 4 September 2014
auditoreclient identification can be extended to a broader set of reduced audit quality acts.
The responses of 141 practicing auditors at non-Big 4 firms in Sweden support our theo-
Keywords:
retical predictions. We find that auditors tend to identify with their clients, and that an
Auditor objectivity
Client identification
auditor who identifies relatively more with a client is more likely to acquiesce to client-
Professional identification preferred treatment and to commit reduced audit quality acts. While previous research
Auditors' client acquiescence has considered only Big 4 firms, the current findings suggest that the problems with
Reduced audit quality acts auditor identification with clients also hold for non-Big 4 auditors.
Non-Big 4 auditors © 2014 Elsevier Ltd. All rights reserved.
Sweden

1. Introduction

Investors perceive a lower quality of accounting information when the audited company's management exerts influence
on auditors (Dhaliwal, Gleason, Heitzman, & Melendrez, 2008; Lambert, Leuz, & Verrecchia, 2007). At the same time, value-
added audit services e i.e., clienteservice activities resulting from an audit that are not directly related to verifying the
financial statements, such as providing information and advice beyond core audit services e are found to be most abundant
when the auditor has a strong relationship with the client (Herda & Lavelle, 2013). In addition, clients prefer a relational
approach from their auditors (Fontaine & Pilote, 2011, 2012), but client desire to have a committed and relationally oriented

auditor appears difficult to reconcile with the need for auditor objectivity (Ohman, Ha€ckner, & So
€ rbom, 2012). The present
study addresses this problematic issue by examining the impact of auditoreclient identification on auditor objectivity and
audit quality by applying social identity theory to the auditoreclient relationship.
Bamber and Iyer (2002, 2007) suggested that auditors identify with their clients and that this identification reduces
auditor objectivity in such a way that an identifying auditor tends to acquiescence to the client's preferred accounting po-
sition. We extend this research by examining a broader set of behaviours. In particular, we examine whether the effects of
auditoreclient identification extend to reduced audit quality (RAQ) acts. The RAQ acts considered here constitute unethical
time-saving acts having no immediate relationship to client preferences, and previous research has found that RAQ acts are

* Corresponding author. Tel.: þ46 (0)60 148677, þ46 (0)730367948 (mobile); fax: þ46 (0)60 148783.

E-mail addresses: Jan.Svanberg@miun.se (J. Svanberg), Peter.Ohman@miun.se (P. Ohman).
1
Tel.: þ46 (0)729019569 (mobile); fax: þ46 (0)60 148783.

http://dx.doi.org/10.1016/j.bar.2014.08.003
0890-8389/© 2014 Elsevier Ltd. All rights reserved.

1981). Second.. 2002. Hackenbrack & Nelson. and Stefaniak et al. these arguments suggest that Big 4 firms conduct higher-quality audits than do non-Big 4 firms. . Oakes. De Cremer & Van Vugt. 2007. 1996a. there is no evidence that auditors' client identification is a problem for non-Big 4 auditors. family. claiming that individuals' social identity is the result of a self-categorization process. Pierce & Sweeney. developing a more comprehensive view of the causes of quality problems in auditing. 1998). 2002. We also examine a model of auditoreclient identification developed by Bamber and Iyer (2007) and essentially based on social identity theory (Tajfel & Turner. auditoreclient identi- fication is a likely cause of RAQ acts. 1998. non-Big 4 auditors) identify with their clients.. 2009).g.. 2007) are examples of social identities. 2000. & Nutt. paying little attention to the bonds caused by social forces such as cognitive-based per- sonal relationships with clients. DeFond. Cianci and Bierstaker (2009) argued that the underlying mechanism of the relationship between TBP and RAQ acts is cognitive bias caused by TBP. they classify themselves according to occupation. The important differences between Big 4 and non-Big 4 firms provide the motivation for the present study. 1996). 1992). Sundgren.. 2003. & Woodliff. 1996). and consider whether auditor rotation is a useful way to reduce the negative impact of client identification. auditors believe that they can take such time- saving shortcuts as superficially reviewing client documents or false signoff without increasing audit risk. 1985. Francis. 2000). Salterio. However. Coram. Jiambalvo. 1994. studies of the effect of client identification on RAQ acts connect the research on auditor objectivity with previous studies of TBP and audit quality. This is followed by the findings. 2006). Ng. First. Since research indicates that non-Big 4 firms conduct lower-quality audits than do Big 4 firms (e. & Zhou. Larger audit firms usually deal with larger clients from a variety of industries.e. Big 4 firms need to protect their reputation (Sundgren.g. Auditoreclient identification and auditors' client acquiescence Previous research has focused mostly on financial incentives (e. increasing the likelihood of a person internalizing a group's norms and values (Lembke & Wilson. van Knippenberg. 2005. Auditors who identify with their clients are likely to suffer from a similar biasing of their judgment. & Hogg. Zhou & Elder. Bamber and Iyer (2007) found that auditors identifying with their clients tend to acquiesce to the client's preferred treatment. audit firm size is used as a proxy for audit quality (e.396 € J. All in all. & Griffeth. Bamber. Otley & Pierce. 1996. providing the potential for stronger relationships with clients than in Big 4 firms with their higher employee turnover rates (Herbohn. develop standardized audit methods and have better audit programs (Blokdijk. Lin.g. 1987). Kennedy. We then outline the research method and sample selection. Consequently. Willett & Page. Otley & Pierce. & Sevcik. The present research on auditoreclient identification is important for at least two reasons. & Harquail. The organizational cultures of non-Big 4 firms are suggested to be relatively less competitive (Anderson-Gough. Hogg & Terry. Haynes. nationality. 1999. When people develop a social identity. there are reasons to believe that the threat to objectivity from RAQ acts may be serious in non-Big 4 firms. Affected by this bias. & Subramanyam. 1998). Drieenhuizen. If the employee turnover is lower in non-Big 4 firms. & Wetherell. 2. The remainder of the paper is structured as follows: In the next section we review the literature and present the hy- potheses. Riordan. Individuals group themselves with others and internalize traits that they perceive as typical of the group (van Knippenberg. Reicher. Jenkins. Schatzberg. Therefore. 2008. Chen. Wan-Huggins. Mayhew. We extend the study of Bamber and Iyer (2007) by testing whether auditors who do not work for Big 4 firms (i. & Stein. Turner. 2001. 2001. and it is possible to have many such identities simultaneously (Markus & Wurf. Literature review and hypothesis development 2. 2012.. Simunic. and empirical evidence supports this contention (Becker. 2004. Dukerich. and larger audit firms have better training programs. & Barefield. Grey. Previous research appears to assume that auditors' identification with their clients is a problem that exists only when audit clients are large. Social identities govern how people think and act. 1987). and the discussion and conclusions end the paper. Dhaliwal et al. 2003. 1992. This suggests that these auditors attain a client-inspired perception of the client's internal control and accounting (Mael & Ashforth. Hollingsworth & Li. 2008). Houston. Social identity theory offers a theoretical framework for examining non-financial dependence. & Peecher. We examine the merits of such an assumption since there are reasons to suspect that non-Big 4 auditors also identify with their clients. Svanberg. an auditor identifying with a client is inclined to act in the interest of that client. De Cremer. Dutton. DeAngelo. Moreover. organization. & Cornell. van Knippenberg. 2009) and are therefore suggested to be less likely to compromise their inde- pendence than are their smaller competitors. Hogg. P. 1989. & Robson. Kadous.1. or age. 1998. auditors have more opportunities to develop close client relationships and this may impair auditor objectivity. 2004). 2004). Ohman / The British Accounting Review 47 (2015) 395e408 caused primarily by time budget pressure (TBP) (e. Accounting firm alumni (Iyer.g. These studies examined auditors in Big 4 firms who typically have large clients with whom auditors may be likely to identify. (2012) found further evidence of this effect.. We also provide further evidence of the extent to which professional identity may counteract any negative effects of auditoreclient identification on audit quality. 1997) and auditors (Bamber & Iyer. Defining the self in collective terms leads to experience of collective interest as self-interest (Ashforth & Mael. Confirm- ing these findings. Stefaniak. making research on non-Big 4 auditors less relevant. 2004. 1996b). which enhances their auditors' skills (O'Keefe & Westort. 2012). while previous studies examine the effects of such identification on auditor objectivity (Bamber & Iyer.

This judgment bias. 1991. In total. 1987. 1991). 2004). 1990. and inadequately investigating ac- counting principles (McNair. Willett & Page. Lai. € J. This in turn means that auditors can perform RAQ acts to save time in the belief that these acts do not involve an audit risk increase. these observations suggest that auditors who identify with their clients are more likely to commit RAQ acts than are other auditors.2. Because non-Big 4 firms are smaller than the Big 4 firms. Glover. 2004). They demonstrated that TBP causes auditors to suffer from biased judgment. or total fees to the total audit office revenue (Hollingsworth & Li. we expected to find that auditors in non-Big 4 firms identify with their clients to an extent comparable to that previously found for Big 4 auditors. Furthermore.. 2010). 2002). which in turn causes RAQ acts to become more frequent. The European Commission has expressed similar concerns (European Commission. We state the hypothesis as follows: H1 Non-Big 4 auditors' acquiescence to client-preferred treatment increases with increasing client identification. 2013). . 2010. 1998) and other evidence suggests that auditors can have the goals of being efficient and supporting client-preferred positions (DeZoort & Lord. Bedard. RAQ acts have been so closely associated with TBP that they may be interpreted simply as ways for auditors to save time in order to attain time budgets. Instead. & Ortegren. Cianci and Bierstaker (2009) explained auditors' biased judgment with reference to motivated reasoning theory (Kunda. 2012). 1999). 2003. and Seavey (2013) found that the Big 4 market share has a positive impact on audit quality as long as the market is evenly split between the Big 4 firms. Pierce & Sweeney. a factor that in itself is fertile soil for strong client relationships and may contribute to auditors' development of client identification and impaired objectivity. which means that small clients may be important for small audit firms. Turner et al. clients are more likely to exert influence on the non- Big 4 firms. Their concern was the extent to which adverse effects on audit quality could result from lack of competition. This theory claims that individuals with directional goals conduct biased searches for and overestimate the importance of evidence supporting their preferred conclusion (Ditto. 1997. Similarly. However. Apanovitch. but Cianci and Bierstaker (2009) offered an alternative way of looking at the phenomenon.. Lundgren & Prislin. 2005. Ohman / The British Accounting Review 47 (2015) 395e408 397 The difference in market position between Big 4 and non-Big 4 firms is manifested by the discussion of inappropriate market dominance that has occurred in Europe and elsewhere. Scepansky. 1995). audit fees.e. & McDaniel.. 2010). Consequently. & Biggs. Khurana. i. Willett & Page.. Wilks. Svanberg. Haynes et al. research has found that auditors in some cases reject awkward-looking items from a sample (Coram et al. Moreover. Dalton. Individuals who place more emphasis on prestige and less on workelife balance are more likely to fit in at Big 4 firms (Bagley. Michas. and non-Big 4 auditors are less able to withstand client pressure than are Big 4 auditors (Boone. Bedard. and sign off prematurely (Alderman & Deitrick. Previous research on financial dependence has measured client importance as the proportion of client non-audit service fees. Consistent with motivated reasoning theory. For these reasons. and that the non-Big 4 auditors would be likely to acquiesce to client-preferred accounting positions. Johnstone. 1999. previous evidence indicates that clients’ bargaining power is an important factor affecting auditors’ reporting decisions (Lai. P. the culture is less competitive. efficiency-mediated estimation. We state the following hypothesis: H2 The frequency of RAQ acts increases with increasing audit client identification. Sedikides & Brewer. Pierce & Sweeney.. Auditoreclient identification and reduced audit quality acts Several auditor behaviours identified as RAQ acts have been examined in previous research. 2012). 1987). there is no empirical support for such a concern. the auditors tend to feel more job satisfaction and like their working conditions better than in Big 4 firms (Patten. & Lockhart. means that auditors discount negative information. 1998. Prawitt. 1996a). 1998). 1996). This situation provides good opportunities in non-Big 4 firms for long relationships with clients. Munro. 2013). 2004. Otley & Pierce. accounting research has provided evidence that auditors make judgments in favour of client-preferred alternatives in response to efficiency concerns (Bierstaker. 1997. and make favourable assessments of the client's internal control as demands for efficiency increase. These RAQ acts are closely related to TBP in audit firms and this pressure is the most commonly observed cause of these behaviours. 1996). 2003. The House of Lords in Great Britain issued a call for evidence regarding the consequences of the Big 4 market concentration (House of Lords. Kelley & Margheim. In non-Big 4 firms. including accepting weak client explanations (Coram et al. fail to pursue questionable items (McNair. Of particular importance in this respect is the similarity of these effects to the inclination of auditors to take a col- lective's (group or organization) interest to heart (Hogg. the fact that non-Big 4 firms have smaller average clients does not necessarily mean that the clients are perceived as less important to the auditor. This seems to further support the contention that audit firms with weak market positions have relatively more problems achieving high audit quality. emphasize positive information about the client's accounting as more rele- vant. superficially reviewing client doc- uments (Dalton & Kelley. Deetz (1995) found that professionals who identify with a client are more likely to under-report time. & Raman. which is likely to appear when the auditor identifies with clients. inappropriately rely on the client's inter- nal control (Pierce & Sweeney. Francis. 2001. 1982). 2003. & Biggs. 2.

other accounting specialists are also allowed to be FAR members (Carrington et al. and client image on auditor acquiescence to client-preferred treatment. client importance. For these reasons. H5 Auditors' client identification increases with more positive client image. Shafer & Wang. This study also examines whether client identification mediates the effects of auditor tenure. the number of certified auditors in Sweden has remained relatively stable at around 4000. 1981. we state three hypotheses as follows: H3 Auditors' client identification increases with increasing auditor tenure. who found that the effects of auditor tenure.398 € J. or public watchdogs. Park. only professionals carrying the CA designation were allowed to be members of the national professional institute (FAR). Accordingly. Auditors are sometimes recruited by their clients (Dart & Chandler. Pollock. we hypothesize that: H6 Auditors' acquiescence to client-preferred treatment decreases with increasing professional identity. and accounting firms. Social identity theory and evidence from Bamber and Iyer (2007) suggest that the positive effect on auditors' self-image provided by association with important clients in combination with the amounts of time that auditors allocate to such clients stimulate audit client identification. their behaviour being highly governed by these values and norms. auditors will tend to identify more with clients that they deem important. the more salient the organizational membership becomes for self-categorization. Johed. 2013). In recent decades. Professional identification. Revisorsna €mnden (the Supervisory Board of Public Ac- countants) is the authority that issues authorizations and approvals and supervises authorized auditors. Aranya. Johansson. Dutton et al. 2013). 2. Tajfel & Turner. 2003). on auditor acquiescence were mediated by auditoreclient identification. As indicated above. auditors who identify with their profession tend to internalize the values and norms of the profession. This reasoning is supported by Jenkins and Lowe (1999). Until 2007.3. According to the social identity argument. the formal competence differs between two categories of certified auditors in Sweden. approved auditors. Accordingly. the opposite should hold for professional identification. 1985). who argue that auditors could be either advocates for their clients.1. auditor tenure) is positively related to auditoreclient identification. respectively. increasing indi- vidual self-esteem and image (Haslam.e. Antecedents of auditoreclient identification Three antecedents of auditoreclient identification were examined by Bamber and Iyer (2007). P. 2013. Such clients are often the largest clients of an audit firm (Chung & Kallapur. we expect auditors with a stronger professional identity to be less likely to commit various RAQ acts.. Bamber and Iyer (2007) confirmed that the length of time that an auditor has worked on the audit engagement (i. and auditors invest more time and effort in important clients. though small clients may be important for non-Big 4 firms due to the relative size of the audited company. & Liao. Ohman / The British Accounting Review 47 (2015) 395e408 2. 2007). supporting strong professional identification. For the same reasons that auditors tend to acquiesce less to the client-preferred treatment when they have a stronger professional identity (Bamber & Iyer. or both. 2010). (1994) argued that the longer a person remains with an organization. Context Swedish auditing regulations have a long tradition of being governed by laws. Authorized auditors must meet higher formal education requirements and stricter requirements when qualifying as senior accountants . client importance. & Ohman. & Amernic. and the current laws state that auditors € should act not only on behalf of shareholders but also on behalf of society in general (Ohman & Wallerstedt. 2012). 3. Today. H4 Auditors' client identification increases with increasing client importance. H7 The frequency of RAQ acts decreases with increasing professional identity. Shafer. Carrington. Svanberg. a situation that causes auditors to value clients as potential employers as well. although the hypotheses do not state these relationships.. of whom 55% are authorized and 45% approved auditors (Carrington et al.4. 2002. auditors' client acquiescence.. 1984. Auditors who are less independent should have a higher tendency to neglect the audit rigor compromised by RAQ acts. Finally. Supporting a large body of social psychological research. 2013). Method 3. If it is correct to hypothesize that the propensity to acquiescence to the client's preferred accounting position increases with increasing client identification. which represents Sweden in organizations such as the International Federation of Accountants (IFAC). and client image. and RAQ acts Research on the organizationaleprofessional conflict experienced by auditors has claimed that professional commitment or professional identity is separated from and precedes other commitments for this strongly professional group (Aranya & € Ferris. auditors. we expect to confirm Bamber and Iyer's (2007) results and find that auditors identify more with clients having a positive image in the eyes of the general public. supporting strong client identification. 2001. Furthermore. social identity theory suggests that people identify with groups that have an appealing image.

3.3. Big 4 auditors exclusively represent the audit profession on Revisorsna €mnden (the Swedish Supervisory Board of Public Accountants). In 2011. the audit firms in Sweden had total revenues of SEK 16 million (equal to EUR 1. 1991). Results for late respondents. All prospective certified auditors therefore need eight years of theoretical and practical training. the modest audit fees paid by these firms limit the time spent on each audit. 3. we selected a random sample of 600 Swedish auditors employed in non-Big 4 firms. three years of university study). 2012).. listed companies in Sweden must establish audit committees. three years of practical training under supervision. were statistically indistinguishable from those for early respondents. Non-respondents' answers were taken to be represented by late respondents' answers. Unlike in many other countries.. the questionnaire was pilot-tested using two authorized auditors. . which represent the vast majority of Swedish companies. Sweden has adopted IFAC's International Standards on Auditing (ISA). 1977). In general. The Appendix presents the measures used in the present study. However. Moreover. x 3e6). rep- resenting a 23. 2013). The Big 4 firms accounted for around 66% of total revenues. This response rate was achieved after four reminders generated by the e-mail survey soft- ware over three weeks. we developed a questionnaire to collect empirical data for the study.2. Based on the pilot test results..000 limited companies in Sweden are listed). and participants responded using five-point Likert- type scales. Late respondents were defined as the last 30 respondents to submit their questionnaires. and to pass an examination adapted to the education and practical training required for this particular category of auditors. 2012). the questionnaires took about fifteen minutes to complete. as the survey software did not enable the researchers to track respondent identity. modelled after that used by Bamber and Iyer (2007) and in previous time pressure and audit quality research (Otley & Pierce. To become an authorized auditor. An ANOVA was computed to ensure that late respondents did not answer the questions differently from early respondents. one must possess a bachelor's or a master's degree (requiring three or four years of university study). the small proportion of listed companies explains why most audit firms in Sweden specialize in owner-managed businesses (only around 300 of the approximately 350. and any difference between late and early respondents was treated as a measure of non-response bias. A potential difference between Big 4 and non-Big 4 firms is that the small-firm auditors may feel somewhat distant from the audit profession. that is. According to the pilot study. there are important similarities between Swedish auditors and auditors elsewhere. and e- mailed them the instrument using the survey software. Although the nature of the audit differs in some respects between Sweden and some other countries. Svanberg.. Moreover. and must pass the same examination before being certified. 141 complete responses were obtained. Sweden's professional audit association (FAR) has traditionally been dominated by auditors € representing the big audit firms (Ohman & Wallerstedt. For example.5% response rate. The possibility of bias in the data was dealt with in the following way. as surrogates for non- respondents (Armstrong & Overton. another two years of practical experience is required to become an authorized auditor.7 million). 1996a). For example. In addition to the three years of practical training required of an approved auditor. while the non-Big 4 firms accounted for the remaining 34% (Carrington et al. To be an approved auditor. an anonymous questionnaire is a method of data collection that allows respondents to reveal sensitive information about their behaviour without fear of consequences. Participation in the study was voluntary and respondents were assured that the information would be used solely for scientific purposes. 1995:665. 2012). We also informed respondents that the collection of responses through the e-mail survey software ensured anonymity. € J. 2013). this is not the case for limited companies not listed on the stock exchange. management may be heavily involved in the selection process that precedes the € appointment of auditors by the general meeting of shareholders (Ohman et al. and to ensure proper translation from English to Swedish. As frequencies of RAQ acts and TBP are to be measured. which demonstrates that there is some assurance against non-response bias in the sample.e. while early respondents were the first 30 respondents. it is the general meeting of shareholders that elects the company's auditors in Sweden. after several reminders. Auditors in this category must pass an examination of professional competence based on the qualifications needed. Measures The measures were adapted from scales validated in previous research. In accordance with the EU's Eighth Company Law Directive. However. the organization and employment structure of the Big 4 firms are comparable. Using a register of Revisorsn€amnden. and these firms dominate in Sweden as they do elsewhere. Sample and data collection To investigate client identification in a natural environment. In these companies. both approved and authorized auditors who have passed a € written examination may audit limited companies (Ohman & Wallerstedt. Before administration. it has been decided that there will be only one category of certified auditors e authorized auditors e in Sweden. The self-administered questionnaire was distributed by e-mail in May 2011. Stakeholders' pressure on auditors may also be fairly limited when auditing small companies. Ohman / The British Accounting Review 47 (2015) 395e408 399 than must approved auditors (Fo €rordning om revisorer. Although the Big 4 firms have higher market shares than do the non-Big 4 firms. This information was provided to attenuate the social desirability response bias that otherwise may occur in behavioural ethics research (Randall & Fernandes. P. Although it is arguably easier to audit small firms. and the same holds for the risk of auditor litigation (Carrington et al. the questionnaire was revised to ensure that the respondents correctly interpreted the items. Of the 600 e-mailings. one needs an education comparable to a bachelor's degree (i.

As a control variable. and tenure was measured using one item. Client importance was measured using one item previously used by Bamber and Iyer (2007). Items Factor loadings Client identification When someone praises this client. The two-factor solution accounted for 54. Respondents were asked about the likelihood that they would accept the client-preferred treatment and not require that the liabilities be recorded in the financial statements. 1994). We caution that the amount of field work may cause variation in RAQ acts as well as variation in client identification. Respondents were asked to assume that the case involved their largest client. The variable “position” indicates whether the auditors are juniors/seniors or managers/partners. 1996a) and its internal reliability has previously been confirmed in several studies.835 When someone praises my profession. The internal reliability of the client image scale was low. the two-factor structure described in Table 1 as the solution after the item “I am very interested in what others think about this client” has been excluded. The natural logarithm of this value was used in the correlations and regressions as the control for client size. so this control variable is a way of ensuring that a relationship between client identification and RAQ acts is not simply a matter of different levels of exposure to audit field work. 0. and when we found that the highest Cronbach alpha score we could obtain was 0...3 with the two factors. For control reasons. In the latter case. & Salterio. which we considered too low. it feels like a personal compliment. Svanberg. The factor analysis presented in Table 1 revealed that the items loaded on two factors representing the client identification scale and the professional identification scale. 0. we constructed the client identification measure in accordance with Bamber and Iyer (2007). Respondents were to indicate their response on a probability scale ranging from 0 (“very low likelihood”) to 100 (“very high likelihood”). it feels like a personal insult. we performed a principal-component factor analysis of the items in the constructs.9% of the variance. 2007.66. the respondents were asked to think about their largest client and indicate the length of time the auditor had been employed by that client.849 . 0. 2007.451. Knapp. The likelihood of accepting the client-preferred treatment is the variable “auditor acquiescence”. it feels like a personal compliment.670 I am very interested in what others think about my profession. 0. Accordingly. Responding to this case provides a measure of auditor objectivity. 0. I usually say “we” rather than “they”. The professional identification scale consisted of five items but the client identification scale initially had one item that had cross loadings above 0. The internal reliability (Cronbach alpha) of the client identification scale was 0.846 When someone criticizes this client.79. respondents were asked whether they were juniors/seniors or whether they occupied a manager or partner position in their audit firm. 1985). 1992. These results suggest that the measures assess two different constructs. 0. 1992) as a measure of client image. 0. Table 1 Results of principal components factor analysis. The case describes a situation in which the auditor's conclusion is that unrecorded liabilities are material but the client's management strongly disagrees. Ohman / The British Accounting Review 47 (2015) 395e408 We adapted the measure of client identification and professional identification from the organizational identification scale (Mael & Ashforth. we measured the size of the office in which the auditor is employed using the item “How many auditors work at your office?” Measuring office size is a way of controlling for the possibility that managers and partners who work in small offices tend to devote more time to audit field work because they have less audit staff to manage. The questionnaire also contained a short case adapted from earlier research on auditors' behaviour in an audit conflict situation (Bamber & Iyer. we excluded this construct from further analysis. The instrument has a demonstrated level of construct validity (Otley & Pierce. TBP was included in the questionnaire.697 My profession's successes are my successes.644 When I talk about my profession. As demonstrated in Table 1. The specific question was “How many years have you audited this client?” Auditor experience was measured as the number of years the respondent has been an auditor. P. it feels like a personal insult. To determine whether client identification and professional identification are distinct constructs. Gibbins & Newton. Wan-Huggins et al. respectively. Gibbins. I usually say “we” rather than “they”. 0. 0. Several studies have successfully used this recall method to elicit responses from auditors (Bamber & Iyer. A three-item version of the organizational image scale was used (Iyer et al.687 This client's successes are my successes. McCracken. a higher score representing a higher likelihood of the auditor acquiescing to the client's demand and not maintaining an objective judgment. TBP was measured in accordance with Otley and Pierce (1996a) using a three-item scale.400 € J.755 Professional identification When someone criticizes my profession. 2005.777 When I talk about this client. Mael & Ashforth. 1997. 1998) and rephrased the five-item scale to a professional orientation so that it could measure client identification as well as professional identification. Conse- quently.77 and the corresponding value for the professional identification scale was 0. Another control variable was the size of the auditors' largest client measured as the annual turnover (million SEK). while managers and partners in larger offices tend to supervise an audit team and devote relatively less time to audit field work. we were unable to test H5. Pierce and Sweeney (2004) reported a Cronbach alpha of 0. The specific question was “How many years have you worked as an auditor?” Moreover.

Variables Frequency Mean SD Median Min Max Male 100 Female 41 Authorized auditor (CPA) 67 Approved auditor (AA) 74 Junior/senior 39 Manager/partner 102 Age (years) 46.58 3. The Cronbach alpha for this measure was previously reported to be 0.46 1.83 0.5 21 Office size (number of auditors) 8. P. 4.69 2. approved auditor. Ohman / The British Accounting Review 47 (2015) 395e408 401 Table 2 Descriptive statistics.90 0.69 0.27 0. Most respondents are men. and in the present study the Cronbach alpha for the measure was found to be 0. professional identification.40 2.00 1.00 4.68 8 3 80 Client importance 2.00 Sig.68 2..572 Separate RAQ acts (1e5)b Accepted weak client explanations 2. “Auditor acquiescence” measures the extent to which an auditor acquiesces to the client-preferred treatment.82 (Pierce & Sweeney.00 4. Just under half of the respondents possess the highest certification available in Sweden (i.33 TBP (junior/senior) 1.60 Professional identification (manager/partner) 3. “TBP” is the perceived time budget pressure. Several variable scores are reported separately for the groups “junior/senior” and “manager/partner”. “Client identification” measures the extent to which an auditor identifies with a client.82 0. Results 4.26 3.00 Client identification (manager/partner) 2.17 2.33 TBP (manager/partner) 2.17 2.73 0.474 Client identification (1e5) 2.e.e.1.71 0.69.80 1.2 Table 2 also shows that most respondents are either managers or partners.00 5. and in the present study we find a Cronbach alpha of 0.71 60 0 100 Sig.e.40 Sig. “Client importance” measures the perceived importance of the client for the auditor's firm. the average scores on the central variables (i.94 9.17 3.27 70 0 100 Auditor acquiescence (manager/partner) 56.54 2.67 8 0. On average.72 0.73 for TBP.83 1. differencea 0. the respondents are 47 years old.71 2. The five-item measure of RAQ acts was adapted from Otley and Pierce (1996a) and has been used in several previous TBP studies.67 17 2 35 Auditor tenure (years) 10.42 1.00 Sig.40 1.80 1. A variable name followed by junior/ senior or manager/partner means that the score is reported separately for the group's junior/senior and manager/partner. and have worked for their largest client for an average of 10 years.00 TBP (1e5) 1.69 3.00 1.31 0. differencea 0.01 2 1 5 Client size (turnover MSEK) 128.40 1.629 RAQ acts (1e5) 1. “Client size” is the logarithm of the annual turnover (MSEK) of the client. AA).00 31. differencea 0. client identification. Descriptive statistics Demographic information about the respondents is summarized in Table 2.00 4. auditor 2 No significant differences between authorized and approved auditors were found.66 3. As can be seen.99 9.83 2 1 4 Failed to research an accounting principle 1.24 0. and we therefore excluded this variable from further analysis.50 1. “RAQ acts” is the average value of five reduced audit quality acts.39 0. 2004).34 2.00 1.76 32.81 85. “Professional identification” is the extent to which the auditor identifies with the audit profession.02 0. authorized auditor. .80 1.91 33.86 45 27 65 Experience (years) 16.40 Client identification (junior/senior) 2.50 RAQ acts (junior/senior) 1.65 2 1 3 Superficial reviews of client documents 2.00 4..00 4.64 2 1 3 Signed off prematurely 1. “Auditor tenure” is the length of time the auditor has audited his or her largest client.33 3.00 0. differencea 0.02 142. differencea 0.00 900.00 4.60 2 1 3 Reduced work below a reasonable level 1.83 1. “Experience” is the number of years the auditor has worked as an auditor.00 4. have been auditors for 17 years.362 Auditor acquiescence (0e100) 57.00 Professional identification (junior/senior) 3.33 Sig.24 5. “Office size” is the number of colleague auditors working in the same office as the auditor. Svanberg. Variables: “Age” is the age of the auditor in years.62 0.846 Professional identification (1e5) 3.86 0. b The frequency scores for the five RAQ acts that together make up the RAQ acts score are reported separately.00 5.70 2.61 2 1 3 a Two-tailed t-test.76 2.97 0.50 RAQ acts (manager/partner) 1.96 60 0 100 Auditor acquiescence (junior/senior) 60..73 0.00 1. € J. CPA) and the rest of them have the lower certification (i.

“Client ID” is a variable describing the extent to which an auditor identifies with a client. “Client size” is the logarithm of the annual turnover (MSEK) of the client.457** Client imp. 0.077 0.” is the number of years the auditor has worked as an auditor. Examining the relationship between client identification and RAQ acts. we sample auditors who work in relatively small offices in which there is less division of labour. Age Gender Exp. **p < 0.05. sample that Bamber and Iyer (2007) used.055 0.061 0.079 TBP 0.220* 0. Variables: “Auditor acq. Margheim.193* 0.143 0.050 0. significantly above the scale midpoint (p < 0.087 0.039 0. & Pattison. The “managers” and “partners” in the present sample are largely employed in relatively small firms e the average office size is eight persons e in which the tasks of managers have more in common with those of senior auditors in Big 4 firms than with those of the managers or partners of such firms.006 0. who found average scores for RAQ acts of “close to two” for four of the RAQ acts considered in the present study.402 € J.052 0. 0.062 0.037* 0.014 0. This interpretation is further supported by Otley and Pierce (1996b).006 Age 0. by choosing non-Big 4 auditors.001).160 Exp. reported RAQ act frequencies suggesting a total average value of only marginally “above two” for junior auditors. In cases in which one variable is dichotomous and one is ordinal. Ohman / The British Accounting Review 47 (2015) 395e408 Table 3 Pearson productemoment correlations. The present sample appears suitable in that the sampled auditors are actively engaged in audit practices in which RAQ acts are apparently committed. Variables RAQ acts Client ID Prof ID TBP Auditor tenure Client imp. RAQ acts are also below the scale midpoint.146 0.05) below the scale midpoint.015 0. We therefore ensured that the data would be suitable for this analysis in three ways.037 0.024 0.027 0. which is between the midpoint (“sometimes”) and the endpoint (“never”). . The average RAQ acts value (mean ¼ 1. despite the fact that most describe themselves as managers or partners. “Pos. the phi coefficients are presented.4 units higher than in the present study. and RAQ acts) do not differ significantly between the two groups of auditors.049 Client ID 0.01. despite the fact that the average level of RAQ acts in the sample indicates that the auditors are sufficiently involved in audit field work.042 RAQ acts 0. “Gender” is 0 for women and 1 for men.168 0. “Professional ID” is the extent to which the auditor identifies with the audit profession.031 0. “Age” is the age of the auditor in years.69) indicates considerable variation in client identification between auditors.044 0.121 0. we find that the variation in the client identification level is higher in our sample of non-Big 4 auditors than in the Big 5 firm sample examined by Bamber and Iyer (2007). The average professional identi- fication is 3. Svanberg & Ohman. “Client imp.377** 0.410** 0.675** 0.128 0.83) refers to the frequency labelled “rarely”. “Auditor tenure” is the length of time the auditor has audited his or her largest client.186* 0.133 0.97) is significantly (p < 0. significantly below the scale midpoint (p < 0. At the same time. The control variable “office size” was employed to ensure that a relationship between client identification and RAQ acts would not be observed because of the varying involvement in audit field work associated with varying office size.037 0.156 0.501** 0. The frequency scores for the five RAQ acts that together constitute the RAQ acts score are reported separately.047 0.001 0. Auditor acq. and the same holds for professional identification.003 0.001). “RAQ acts” is the average value of five reduced audit quality acts. 0.73.222** 0. 1999. “Exp. and the variation between auditors (standard deviation 0. Third. “TBP” is the perceived time budget pressure.075 0.3 This indicates that the sampled auditors report committing RAQ acts.037 0. Thus.061 0.044 0.66) is comparable to the variation in client identification. First. Office size Client size Pos.008 0.064 Auditor tenure 0.002 0.495** 0.090 0.” is the perceived importance of the client to the auditor's firm. we dealt with the possibility that the level of RAQ acts may vary between different managers and partners because of varying involvement in field work. based on the same RAQ acts scales as are used here.241** 0. who found both values to be above the scale midpoint and reported identification levels approximately 0.031 0. client identification seems lower among these non-Big 4 Swedish auditors than in the Big 5 U.275** 0.606** Gender 0.125 0.” measures the extent to which an auditor acquiesces to the client-preferred treatment.202* 0. In comparison.142 *p < 0. Table 2 shows that TBP (mean ¼ 1.005 Prof. we initially cautioned that a sample consisting mostly of managers and partners would be less suitable because the highest levels of TBP as well as RAQ acts tend to be observed among juniors and seniors (Kelley.31.014 0. Table 2 shows that the average client identification is 2.317* 0. That study also used a five-point Likert scale with a score of two indicating “rarely”. in cases in which two variables are dichotomous.263** 0.728** Office size 0. P.038 0.102 0.063 0. These values can be compared with those obtained by Bamber and Iyer (2007). acquiescence.045 0.032 0.” is a dichotomous variable that is 0 if the auditor is junior or senior and 1 if the auditor is manager or partner (junior/senior or manager/partner is the auditor’s own classification).S.191* 0.269** 0.294** 0. The standard deviation (0.111 0. Second.090 0. we compare the levels of TBP and RAQ acts in the present study with those found in previous studies using comparable measurement instruments. the correlations are point-biserial correlation coefficients. ID 0. indicating that auditors in this sample find time budgets relatively attainable. two-tailed tests. Svanberg.316** 0.022 0.197* Client size 0.060 0. 3 The RAQ acts measure was reverse coded. Otley and Pierce (1996a). “Office size” is the number of colleague auditors working in the same office as the auditor. € 2013).010 0.

Sig.843 0. and we use TBP as a control variable because previous studies have found that a substantial part of the variation in RAQ acts can be explained by TBP (e. 1996a. Sig.008 0.675 0.525 1.. Because these two variables provide similar information.e. auditor acquiescence in models 2 and 3.009 0. 2003. are significant at p ¼ 0. client importance.072 0.227 0.111 0. Svanberg.062 0.601 0. Ohman / The British Accounting Review 47 (2015) 395e408 403 Table 4 Multiple regression of the effect of client identification on auditors' client acquiescence and RAQ acts. Dependent variable Predicted sign Model 1 Model 2 Model 3 Model 4 Model 5 Client identification Auditor acquiescence Auditor acquiescence RAQ acts RAQ acts Coeff.090 F 2. “Auditor tenure” is the length of time that the auditor has audited his or her largest client.215 6. “RAQ acts” is the average value of five reduced audit quality acts. suggesting no multicollinearity.g. 4.004 0.380 0.046 e0.000 0. 0.210 0. .304 0. “TBP” is the perceived time budget pressure. “Office size” is the number of colleague auditors working in the same office as the auditor. regarding collinearity statistics.009 1. We note that the present study is not concerned with the relationship between TBP and RAQ acts.062 0. the effect of the precedent variables (i.013 0. As expected.131 0.0651 0. we excluded age from the multiple regression models to avoid multicollinearity.063 0. As expected.e. Variables: “Client identification” is a variable describing the extent to which an auditor identifies with a client. auditor tenure. Sig.013 0.639 0. Sig.167 Adj.2.183 0.835 0.05 or less.037 0.002 0.180 0. auditor tenure and client importance) is mediated by client identification.0103 0. Otley & Pierce. and client identification.8. then client identification may act as a mediator between the precedent variables (i. 4 If it can be assumed that client identification is caused by auditor tenure or client importance and if it can be assumed that auditor acquiescence is caused by auditor tenure.110 0.141 0. F for change between 0.161 Sig.205 0. Coeff.150 0.176 0. R2 0. P.767 Position  0.121 0.097 0.692 2. our focus is the relationship between client identification and RAQ acts. However.4 We tested this possibility using hierarchical multiple regression and by assessing the incremental contribution of client identification. Variable Client identification þ 1.023 Client size þ 0.003 models 3 and 2 Sig. Coeff. However.022 0. Pierce & Sweeney.910 TBP þ 0. suggesting that time pressure does not affect audit quality in this sample of auditors.027 0.359 0. the correlation between age and experience is high.146 0. Client size is the logarithm of the annual turnover (MSEK) of the client..0 and 1. “Client importance” is the perceived importance of the client to the auditor's firm.556 R2 0.057 0.220 0. and the extent to which client identification mediates the effect of auditor tenure and client importance on auditor acquiescence.006 0. there is no significant relationship between RAQ acts and TBP. This finding is supported by previous research similar to the present study (Bamber & Iyer.022 0. Table 4 presents data from the multiple regressions.023 0.152 0.220 Client importance þ 0. and RAQ acts in models 4 and 5.080 0.050 0. The correlations between RAQ acts and client identification and auditor tenure. and client importance.031 Sig.007 0.208 Office size þ/ 0.019 0. Using the same variables as did Bamber and Iyer (2007) e except for a different definition of auditor experience e we examined the possible relationships between client identification and auditor acquiescence.371 0. the correlations reveal potentially significant relationships between several variables. F for change between 0. “Experience” is the number of years the auditor has worked as an auditor.000 0.188 0.073 0.000 1. tolerance values were all below 1 and VIF values were all between 1.280 0. The dependent variable is client identification in model 1.048 0.952 0. Collinearity statistics were generated to verify whether the relatively high correlations between auditor tenure and experience indicate multicollinearity in the regression model..003 0.137 0.086 0. 1996b).0673 0. respectively.. instead.023 Professional identification  0.745 7. are significant at p ¼ 0. We decided not to include the variable gender in the multiple regressions because the correlations did not indicate that gender would be related to the central variables of the study.696 0.610 1.009 0.0296 0.063 0. auditor acquiescence). 2007.e.182 0. We used a test of mediation described by Baron and Kenny (1986).023 models 5 and 4 Significance values refer to two-tailed tests.071 0.638 0.234 0.013 0. “Auditor acquiescence” measures the extent to which an auditor acquiesces to the client-preferred treatment.064 0. “Professional identification” is the extent to which the auditor identifies with the audit profession.357 0. 2004).275 e0. The test of mediation is the statistical confirmation of causal modelling. auditor tenure and client importance) and the outcome variable (i.564 0.235 Auditor tenure þ 0.032 0. client identification.212 1.009 0. If this is the case..01 or less. Coeff.414 0. “Position” is a dichotomous variable that is 0 if the auditor is junior or senior and 1 if the auditor is manager or partner. Hypothesis test results Table 3 presents Pearson productemoment correlation coefficients for the variables.011 0. The structural equation model developed by Bamber and Iyer (2007) demonstrated that the effects of auditor tenure and client importance on client acquiescence were mediated by client identification. Coeff.937 Experience  0.066 0. The correlations between auditor acquiescence scores and RAQ acts. Coram et al.000 0.525 1.031 0. respectively. € J.523 1.020 0. Sig.

previously examined in time budget pressure (TBP) research (e. This effect was demonstrated using a measure of the extent that the auditor's judgment is biased. Our results on this point are further indications that audit firm management of auditor tenure may be useful. van Knippenberg..05).057 between models 4 and 5. 2013. The significant positive regression coefficient for client identification (p < 0. and client image as suggested by Bamber and Iyer (2007). We found that identification with clients varies substantially between auditors. client importance. 2008. 2007). this effect counteracting the impact of auditor tenure on auditor acquiescence. We found that auditors' client identification serves to impair auditor objectivity because auditors who identify relatively more with their clients are more likely to acquiesce to the client-preferred treatment of a material accounting issue. 2012) and Herda and Lavelle (2013). Ohman / The British Accounting Review 47 (2015) 395e408 The multiple regression analysis presented in model 1 in Table 4 confirms a positive relationship between auditor tenure and client identification (p < 0. Partly contrary to this result. Therefore. Baron & Kenny. 1996a. The hypothesized relationships between RAQ acts and the central variables of this study are the same as we hypothesized for auditors' acquiescence to client- preferred treatment.. supporting H2.086 to 0. i. we examine whether their findings can be extended to a sample of auditors in non-Big 4 firms. However. H6. with the magnitude of the coefficient decreasing from 0.404 € J. P.01).g. We hypothesized that client identification was caused by auditor tenure. The control for client size suggests that auditors acquiesce more to the preferences of larger than smaller clients. 5.01) and that the relationship is confirmed in model 3. 1996). auditors should establish a relational approach to their clients. there is some indication of mediation because the coefficient for auditor tenure drops from 0. The significant increase in the R2 value. We contribute to the research on the threat to auditor objectivity posed by client identification in two ways. no significant relationship was found between professional identification and RAQ acts. and the relationship between auditor tenure and auditor acquiescence suggests that auditor rotation could be beneficial because client identification likely takes time to develop. but that auditors do not identify more with larger clients. suggesting no support for H4. Coram. Regarding mediation. auditor acquiescence to client-preferred treatment. Willett & Page. the corresponding mediation of the effect of auditor tenure on RAQ acts is not confirmed by the regressions. we examine a broader set of unethical acts. Model 5 indicates a positive relationship between client identification and RAQ acts (p < 0. while Bamber and Iyer (2007) used a sample of auditors employed in Big 4 firms. Coram et al. We also found that auditors who are more experienced are less likely to acquiesce to the client's position. While we found that client identification partially mediates the effect of auditor tenure on auditor acquiescence. The pattern observed in the regressions with auditor acquiescence as the dependent variable is thus partly confirmed in this respect by the regression with RAQ acts as the dependent variable. supporting H3. suggesting that auditors' acquiescence to client-preferred treatment decreases with increasing professional identity. confirming the univariate test.002 (model 2) to 0. Otley & Pierce. Finally. more experienced auditors being . Svanberg & € Ohman. models 2 and 3 indicate a negative relationship between auditor experience and auditor acquiescence (p < 0. Baron & Kenny. does not receive any support from the regressions presented in Table 4. Although the conditions for partial mediation are not completely in place (cf. The control variable TBP is insignificant in both models. Hogg. Discussion and conclusions This study extends previous literature on the effect of auditors' client identification on auditors' client acquiescence (Bamber & Iyer. the present study provides further evidence of the impairment of auditor objectivity that may be caused by relaxing the requirement for an arms-length auditoreclient relationship. indicates that client identification partially mediates the effect of auditor tenure on auditor acquiescence (cf. 2004. We recognize that an auditoreclient relationship could well involve intergroup clashes because the auditor's professional identification is strong and being an auditor entails maintaining a certain distance from the client. Ng. 2002.e. i. and Rast III (2012) argued that identification across organizational boundaries is difficult to achieve if there are intergroup identity clashes. and the significance level decreases as well. In the view of researchers such as Fontaine and Pilote (2011. supporting previous findings (Bamber & Iyer. 1996b. but our expectations were confirmed only in the case of auditor tenure. This relationship. H2 and H7 were tested using the multiple regression analyses presented in models 4 and 5. Glavovic. leaving H7 unsupported. However.. together with the observed decrease in significance levels and in coefficient size between models 2 and 3. 2003. we developed the analysis as a hierarchical regression to detect mediation in the same way as in models 2 and 3. and that client identification is lower than professional identification on average. First. The regression in model 1 does not indicate any re- lationships between client identification and client importance.022 (model 3). We found no apparent relationship between TBP and RAQ acts in the present sample. Finally.. 2007). & Woodliff.05).01) in model 3 indicates a positive relationship between client identification and auditor acquiescence.696 (model 3) and the p-value increasing from 0.952 (model 2) to 0.e. Svanberg. supported by the F-test. 1986). while previous findings regarding the effects of auditors' client identification have been limited to one behaviour. RAQ acts. providing support for H1. we note that there is a relationship between auditor tenure and auditor acquiescence indicated by the significant and positive coefficient for auditor tenure in model 2 (p < 0. 1986). Second. Auditor experience provides some protection against the effects of auditor tenure. between models 2 and 3 indicates that client identification contributes significantly to the explanatory power of the model. for which we found that client identification partially mediates the effect of auditor tenure on auditor acquiescence. Pierce & Sweeney.

Cianci and Bierstaker (2009) found that the effect of TBP on RAQ acts is explained by cognitive bias. social desirability bias. because we expected that auditors who identified with their clients would tend to look more favourably on the client's in- ternal control. The first reason was that previous studies have found that auditors under pressure alter their judgment in favour of client-preferred alternatives in response to efficiency concerns (Bierstaker et al. and that this assessment by the auditor in turn causes the auditor to inappropriately believe that foregoing some substantive testing. though not conclusive ones. since this study provides evidence that client identification has an effect on RAQ acts. The present study. Some previous research has examined the impact of such factors as the audit firm's leadership (Otley & Pierce. we confirm previous findings that auditors identify with their clients but. combined with the findings of Cianci and Bierstaker (2009). Appendix. Francis. 2012). We examined the effects of auditors' client identification on five different RAQ acts. although previous research has associated RAQ acts more or less exclusively with TBP. € J. 1996a. in contrast to Bamber and Iyer (2007) who examined Big 5 firms. Previous evidence indicates that experienced auditors are less susceptible to the negative effects of TBP and that junior auditors perform most of the audit field work. suggests that auditor tenure is an antecedent of client identification and that the effect of auditor tenure is partially mediated by client identification.. A possible reason for the lower level of professional identification observed in the present study is that non-Big 4 auditors are somewhat more distant from the audit profession than are Big 4 auditors (cf. 2010). disagree. we have found that the auditor objectivity problem due to client identification exists in non-Big 4 firms.g. These factors taken together may explain why TBP has less in- fluence on the behaviour of the current sample of auditors. The use of the control variable office size provides some assurance that differences in the amount of field work between auditors do not distort the analysis of the relationship between client identification and RAQ acts. provides evidence suggesting that the underlying psychological mechanism in situations in which TBP causes RAQ acts is the same as in situations in which client identification causes RAQ acts. Further studies could empirically examine potential differences between Big 4 and non-Big 4 auditors with respect to professional identification. Furthermore. and the mea- surement of auditor acquiescence using a constructed scenario. P. However. the nature of the findings presented meant we could not formally conclude that client identification partially mediates the effect of auditor tenure on RAQ acts. in order to save time. according to much previous research (e. they are vulnerable to the negative effects on objectivity and more likely to commit RAQ acts. strongly agree) When someone praises this client. whereas previous studies of the impact of TBP on RAQ acts have surveyed junior auditors whose time budgets are imposed upon them. We suggest that the cognitive bias affecting auditors who identify with their clients is similar to the bias affecting auditors who are affected by TBP. and found evidence of the expected relationship between auditors' client identification and RAQ acts. The totality of the evidence. Our sample of experienced auditors on average conducts RAQ acts almost as frequently as previously reported for junior auditors (Otley & Pierce. it is unsur- prising to find these negative effects of client identification. Moreover. will have less adverse effects. Measures used in the study Client identification (Scale: strongly disagree. produce lower-quality audits than do Big 4 auditors. We found that increasing client identification increases not only the auditor's tendency to acquiesce to the client-preferred treatment of a material accounting position.. Arnold. Ohman € & Wallerstedt. future research could continue to explore how client identification affects behaviours other than those intimately related to auditor objectivity. Nevertheless. and more important. the pattern that emerged provides some indication of the expected mediation. Svanberg. but the likelihood of identifying with clients appears unrelated to auditor experience. The other. Glover et al. The fact that we did not find that TBP affects RAQ acts can be reconciled with the fact that previous research has used low ranking auditors with little experience while our sample contains experienced auditors who also have management roles in Swedish non-Big 4 firms. We found some indications. This assures us that the sampled auditors are suitable for the purpose of the study despite the fact that their roles differ from those of the junior auditors examined in most previous research. neutral. & Pierce. 2005). reason was suggested by recent developments in TBP research. the auditors examined here have more influence over their time budgeting and for this reason perceive less TBP. When non-Big 4 auditors identify with their clients. agree. but that the auditor is more likely to commit a range of RAQ acts. Ohman / The British Accounting Review 47 (2015) 395e408 405 less likely than their less experienced colleagues to acquiesce to client-preferred treatment. In conclusion. that client identification mediates the effect of auditor tenure on RAQ acts. it feels like a personal compliment. 1996a) and € ethical culture (Svanberg & Ohman. 1999. 1996b). . This finding supports the expectation that the effects of auditors' client identification extend to areas previously regarded as related mostly to TBP. They argued that TBP causes the auditor to look more favourably on the client's internal control.. 2013. These limitations indicate that the results should be inter- preted cautiously. indicating that the problem is more severe than previous research suggests. but our impression is that most research argues that TBP explains most of these detrimental effects. There were two reasons to expect that client identification would have an impact on RAQ acts. Our finding that client identification is associated with RAQ acts suggests that research on the link between TBP and RAQ acts need to be addressed using complex models. Sweeney. Since non-Big 4 auditors. we did not find that professional identification in non-Big 4 firms counteracts the effects of client identification. however. taking the regressions for auditor acquiescence and RAQ acts together. 2004). Limitations of this study include the modest response rate and non-response bias. We expected auditors' client identification to have the same detrimental effect on the frequency of RAQ acts.

C. Boone. D. F. it is unnecessary to make adjusting entries in the financial statements. V. In the current year's audit. M. This client's successes are my successes. This client is considered one of the best companies to work for. Bedard. were the time budgets for the jobs you worked on during the last year (Scale: very easy to attain. and statistical con- siderations.. L.. Simunic. & Kenny. 21(2). L. 21e38. rarely.. Auditing: A Journal of Practice and Theory. L. Signed off an audit-programme step without completing the work or noting the omission. Made superficial reviews of client documents. 25(1).. Drieenhuizen.. Reduced the amount of work performed on an audit step below what you consider reasonable. agree. (1982). C. Moderatoremediator variables distinction in social psychological research: conceptual. (2002). Auditing: A Journal of Practice and Theory. B.. I.. impossible to attain? How often do you achieve your time budgets? (Scale: nearly always. C. disagree. & Robson. 1(2). sometimes. it feels like a personal insult. When someone criticizes this client. 1e24. sometimes. 51(6). The client's CFO argues that the total amount of unrecorded liabilities is immaterial and. As the auditor. K. very tight/practically unattainable. sometimes. & Biggs. Ashforth. for the purpose of our study we ask that you respond (1) based on the limited information provided and (2) assuming that the case involves your largest client referred to above. N. D. Aranya. Auditors' client acquiescence Please respond to the following short audit case. Blokdijk. D. Client importance (Scale: no importance. neutral. Big 5 auditors' professional and organizational identification: consistency or conflict? Auditing: A Journal of Practice and Theory. R. K. a dispute has arisen between you and the management of your largest client over the mate- riality of certain unrecorded liabilities discovered by you during the audit. Armstrong. References Alderman. The effect of audit quality on earning management. considerable importance. disagree. M. 271e280. K.. Organizations and Society. 330e352. My profession's successes are my successes. . When someone praises my profession. (2006). Journal of Personality and Social Psychology. However. (1984). strongly agree) When someone criticizes my profession. J. Accounting. 26(2). Aranya. attainable with considerable effort. 14(1). neutral.. 27e48. & Iyer. Grey. RAQ acts During the last year. & Mael. 99e122. Jiambalvo. Professional and firm guidelines do not provide a definitive answer on the materiality of the amount involved. K. Accounting Horizons. Social identity theory and the organization. Estimating non-response bias in mail surveys. M. Bamber. The role of problem shifts on auditors' decision processes in analytical procedures. F. M. A reexamination of accountants' organizationaleprofessional conflict. 18e36. P.never. 20e39. Auditing: A Journal of Practice & Theory. (2012). Organizations and Society. Auditors' perceptions of time budget pressures and premature sign-offs: a replication and extension. Becker.. J. (2007). 396e402. I am very interested in what others think about my profession. how likely is it that you will not require these liabilities to be recorded? Please indicate your response as a likelihood between 0 (very low likelihood) and 100 (very high likelihood). therefore.. (1986). (1998). never.. & Subramanyam. little importance. C. Auditors' identification with their clients and its effect on auditors' objectivity. rarely. Accounting. A. Academy of Management Review. & Iyer. Khurana. how often would you attain your time budget? (Scale: nearly always. strongly agree) This client does not have a good reputation in the business community. rarely. (2001). 29. agree. An examination of professional commitment in public accounting. V. R. (1981). M. An analysis of cross-sectional differences in Big and non-Big public accounting firms' audit programs. Tests of time: organisational time reckoning and the making of accountants in two multi-national accounting firms. J. 1e15. DeFond. (1989). Professional identification (Scale: strongly disagree. M. Auditing: A Journal of Practice and Theory. 26(2). (1977).. S.. strategic. the amount is material. 15(1). 239e264. attainable with reasonable effort. S. Journal of Marketing Research. Svanberg. We appreciate that normally you would require more information. J. J. & Overton. I usually say “we” rather than “they”. Dalton. 18(1). H. Bagley. E. M.. The public thinks highly of this client. & Stein. how often did you act in the following manner when carrying out an audit? (Scale: nearly always. W. often. E. the client management strongly disagrees.. & Amernic. S. T. The Accounting Review. However.. never) Accepted weak client explanations. Contemporary Accounting Research. Client image (Scale: strongly disagree. Do the Big 4 and the second-tier firms provide audits of similar quality? Journal of Accounting and Public Policy. I usually say “we” rather than “they”.. Bierstaker. If you did not under-report time. The factors that affect accountants' decisions to seek careers with Big 4 versus Non-Big 4 accounting firms. it feels like a personal insult. often. Bamber. (1999). J. A. moderate importance. K. (2010).. W. When I talk about my profession.406 € J. & Ferris. F. Pollock. P. P. high importance) Please estimate the importance of this client to your firm. 1173e1182. L.. J. M. N. & Ortegren. & Raman.. Anderson-Gough. R. 6(4). Failed to research an accounting principle. 54e68. 14(3). it feels like a personal compliment. 26(2). J.. K. TBP In general. 1e24. Baron. often.. In your opinion. 59(1). Ohman / The British Accounting Review 47 (2015) 395e408 When I talk about this client. & Deitrick. F. E.

(2009). The International Journal of Auditing. 239e263. The Accounting Review. (2008). do?uri¼COM:2010:0561:FIN: EN:PDF Accessed 11. 103e123. F. (1999). (2001). & Hogg. Informal mentoring relationships and the career processes of public accountants. Academy of Management Review. MA: The MIT Press. P. Auditing: A Journal of Practice and Theory. M.. Michas. 205e224. Ditto. Kunda. (2012).. 931e956. M. Applied Psychology: An International Review. (2003). self and identity: a review and research agenda. Journal of Accounting Research. D. 30(1). Johnstone... (1997). R. Wilks. 6(1). (1994). & J. & Pattison. D. & Prislin. P. A. Lundgren. DeZoort. Gleason. T. Auditing: A Journal of Practice & Theory. (2004). 15(4). M.. Managerial Auditing Journal. K. Gibbins. Chung.. Glover. 48(4). (2004). & Barefield. Available on the internet at http://www. J. 25(1). P. Haslam. D. & McDaniel.. van Knippenberg. D. 927e944. 15(2). NJ: Hampton Press.. 27(1). Does audit market concentration harm the quality of audited earnings? Evidence from audit markets in 42 countries. 221e233. Hollingsworth. (1997). & Lord. Audit reporting of Big 4 versus non-Big 4 auditors: the case of ex-Andersen clients. Hogg.. 29(7).. Auditing: A Journal of Practice & Theory. British Accounting Review. 193e208. SFS 1995:665 [Decree on accountants. R. Accounting and Business Research. G. C.. € J. 39(2). (2008). 13(1). T. (2004). M. M. Ng. S. M. (2000). Apanovitch. (2003). R. & Margheim. Kennedy. & Melendrez. R. M. & Newton. Client employment of previous auditors: shareholders' views on auditor independence. 357e371.. disclosure. M. 45(2). Kelley. V. & Pilote. Green paper: Audit policy: Lessons from the crisis. Cresskill. E. E. Investors' perceptions of auditors' economic dependence on the client: post-SOX evidence. McCracken. J. 462e479). (1987). S. Lembke. 43e59. W. M.. Auditing: A Journal of Practice and Theory. The Accounting Review. Dukerich. 53e69.. J. J. 1995:665]. K. 171e194. E.. A. Administrative Science Quarterly.. Gibbins. L. Psychology in organisations: The social identity approach. transforming business: Building responsive and responsible workplaces. 197e233. 369e393. 51(7). J. The case for motivated reasoning. C.. In M. C. Motivated cognitive processing and attitude change. 15. F. Audit quality and earnings management for Taiwan IPO firms. The congruency of audit partner and chief financial officer recalls. The Leadership Quarterly. & Zhou. Svanberg. 117e127. (2003). & Ohman. J. M. Jenkins. Survey on the differential effects of time deadline pressure versus time budget pressure on audit behavior. Client importance. European Commission. 37(2).. M. (1998). Organizations and Society. D. F. The relationship between client advocacy and audit experience: an exploratory analysis. 871e893. Cambridge.. The Accounting Review. 24(Suppl.. (1995). Call for evidence: Auditors: Market concentration and their role. C. F.14. R. 715e726. J. Social cognition: Making sense of people. Hogg. Heitzman. E. D. van Knippenberg. (1999). Putting the “team” into teamwork: alternative theoretical contributions for contemporary management practice. (1981). non-audit services. A. & Nutt. P Tangney (Eds. (2012). Kelley. L. 21(1). M. Knapp.. Y. 49(3). G. & Kelley.. Personality and Social Psychology Bulletin. J. House of Lords. Work motivation and performance: a social identity perspective. (1992). S. (2005). 71(1). Auditing: A Journal of Practice & Theory. (2003). (2013). S. E. 759e778. K. and abnormal accruals. (2011). (2013).. J. & Lowe.uk/business/committees/ Accessed 11. An empirical study of Canadian companies to determine clients' preferred relationship approach with their financial auditor. R. De Cremer.. Auditor fees and cost of debt. 3(3). M. J. 100e122. A. 28e85. (1999). J. (1999). 1e6. B. M. K. D. & Verrecchia. Social identification effects in social dilemmas: a transformation of motives. Auditing: A Journal of Practice & Theory. Select Committee on Economic Affairs. A. The effect of audit billing arrangement on under-reporting of time and audit quality reduction acts. T. 75(1). 38e45.eu/LexUriServ/LexUriServ. R. & Van Vugt. Herbohn.. and the cost of capital. 23(1). Ohman / The British Accounting Review 47 (2015) 395e408 407 € Carrington. J. van Knippenberg. A. Kunda. Psychological Bulletin.. G. (1998). Fontaine. Johed.. & Wilson. Hogg. (1998). (1990). 325e355. (2005). European Journal of Social Psychology.. A survey of time budget pressure and reduced audit quality among Australian auditors. C. & Bierstaker. A. S. D. S. Leuz.europa. Jenkins.. Auditor commitment to privately held clients and its effect on value-added audit service. Haynes. 32(2). & Nelson. Journal of Marketing Trends.. R.. M. R. A. Coram. Auditing & Finance. Mael. 202e211. Accounting information.. Cianci. 20e27. Dalton. 24(Suppl. Leary. & Seavey. M. van Knippenberg. A.. Social identity and self-categorization processes in organizational contexts. A. Lai... Scepansky.). R. L. A. J. A. Deetz. 165e186. 108(3). Journal of Accounting. L.. Clients' preferred relationship approach with their financial statement auditor. Organizational image and member identification.. Glavovic. 22(3e4). Behavioral Research in Accounting. D. 47e65. & Chandler. Aggressive client reporting: factors affecting auditors' generation of financial reporting alternatives. Transforming communication.. De Cremer. 315e336. Social identity. 86e104.04. E. III (2012). 24(7). T. E. 5. 61e69. . C. Contemporary Accounting Research. Auditors as advocates for their clients: perceptions of the auditoreclient relationship. 43(3). Herda. What do we know about audit quality? British Accounting Review. C. A comparison of dysfunctional behaviours by tax accountants and auditors under time budget pressure. 127e149. Intergroup leadership in organizations: leading across group and organizational boundaries. D.).. Journal of Personality and Social Psychology. P. 17(2). Francis. Chen. R. 36(4). D. J. V. H. 121e140. (2013). 73e78. R. Z. & Ashforth. 1(7). Alumni and their alma mater: a partial test of the reformulated model of organizational identification. J. 385e420. Australian Accounting Review. DeAngelo. Iyer. 480e498. S. An empirical exploration of complex accountability in public accounting. T. L. Handbook of self and identity (pp. S.. Auditors' incentives and their application of financial accounting standards. T. Lambert. S. D.. Fontaine. C. (2013). (1998). D. Dutton.parliament. 16. (2000). & Harquail. Hackenbrack. (2007).. Munro. 25(2). & Lockhart.... D.. H. Journal of Applied Business Research. Coram. S. & Rast. D. D. M. Johansson. 1e20. T. A. The moral intensity of reduced audit quality acts. 32(1). Fo Francis. E. Identification of accounting firm alumni with their former firm: antecedents and outcome. T. Kadous. Journal of Accounting. C. Human Relations. Why do auditors over-rely on weak analytical procedures? The role of outcome and precision. A review and synthesis of pressure effects research in accounting. 13(2). Audit conflict: an empirical study of the perceived conflict of auditors to resist management pressure. B. & Lavelle. Advances in Ac- counting. & Terry. Bamber.. Dart. P. S. 78(3). & Biggs.. D. Auditing & Finance. & Peecher. Journal of Applied Business Research. E. (1997). Leadership. (2010). Available on the internet at http://eur-lex. T.. & Woodliff. Accounting.. N. Ng. Bedard. J... 232e255. 1e22. The Accounting Review. C. (1994). London: Sage Publications. Advances in Accounting. An empirical test of the hierarchical construct of professionalism and managerialism in the accounting profession.04. (2010). P. 825e856. & Kallapur. Advances in Accounting... S. (2002). New York: The Guilford Press. 78(4). Margheim. Academy of Management Review. 36(4). 27(1). G. Motivated sensitivity to preference-inconsistent information. 183e199. & Salterio. 25... (2005). Journal of Accounting and Economics. Journal of Accounting Literature. K. E. Dhaliwal. Journal of Accounting Research. J. Prawitt. Journal of Organizational Behavior. 495e524.). & Pilote. 15. K. K. M. & Li. 20(1). S. 88e104. Auditors' efficiency motivated evaluation. The effect of quality assessment and directional goal commitment on auditors' acceptance of client- preferred accounting methods. J. (1996). Incorporating Advances in International Accounting. 345e368.. 60(2). K.. (2013). & Woodliff.14. Lin. 113e137. R. €rordning om revisorer. D. G. L.. Auditor size and audit quality. S. Current Issues in Auditing. (1985). M... Z..

10(11). Auditing: A Journal of Practice & Theory. Houston.. collective self. Psychology of intergroup relations (pp.. 49e70. 6(1). 203e221. 95e120. Mayhew.. J. Otley. M. Perceived audit quality. C.. The effects of precedents and client position on auditors' financial accounting policy judgment. & Sweeney. B.. 65e84. The impact of perceived ethical culture of the firm and demographic variables on auditors' ethical evaluation and intention to act decisions. 13(3). modified audit opinions and the likelihood of liquidating bankruptcy among financially weak firms. C. J. R. Individual self.). D. Arnold.. European Accounting Review. Audit time budget pressure: consequences and antecedents. B. M. 31(1). Audit quality and earnings management by seasoned equity offering firms. W. & Liao. J. 477e499. 7e24). Shafer. A survey of time budget pressure and irregular auditing practices among newly qualified UK chartered accountants. Auditing & Accountability Journal. Accounting. € Ohman. E. J. 805e817. Stefaniak. Rediscovering the social group: A self categorization theory. D. S. 15(2). T. M. McNair. & Page. Audit regulation and the development of the auditing profession: the case of Sweden. Philadelphia. B. F. (2012). The social identity theory of intergroup behavior. S. Shafer.. W. 16(7). (2013). D. W. O'Keefe. . IL: Nelson-Hall.. Client satisfaction and usefulness to external stakeholders from an audit client perspective. 241e257. Auditors' time pressure: does ethical culture support audit quality? Managerial Auditing Journal. 13(1). Organizations and Society. & Cornell.. Supervisory actions and job satisfaction: an analysis of differences between large and small public accounting firms. 93(4). M. 299e337. Worchel. Willett. Ohman / The British Accounting Review 47 (2015) 395e408 Markus. M. Oakes. T. Annual Review of Psychology. British Accounting Review. L. V.. S. G. (1996b). R. Cost-quality conflict in audit firms: an empirical investigation. R.. 467e486. G.. 11(2). Wan-Huggins. 724e749. D.. C... relational self. The development and longitudinal test of a model of organizational identification. Conformance to GAAS reporting standards in municipal audits and the economics of auditing: the effect of audit firm size. Pierce. Asia-Pacific Journal of Accounting and Economics. Inter- national Journal of Auditing. (2010). (1995). H.. & Pierce. The effects of employer and client identification on internal and external auditors' evaluations of internal control deficiencies. R.. (2012). 31e58. (1992). In S. & Sevcik. W. J. (2002). 101e120. B. Schatzberg. Ha€ckner. (1998). J. Patten. P. (2012). N. B. Effects of ethical context on conflict and commitment among Chinese accountants. (2010). Proper compromises: the management control dilemma in public accounting and its impact on auditor behavior. D. 9(2). & Pierce. 39e56. (2001). Orga- nizations and Society. Managerial Auditing Journal. H. Randall. 572e591. Managerial Auditing Journal. P. Journal of Business Ethics. J. & So €rbom. & Brewer. J. M. Accounting. € Ohman. Svanberg. 635e653. Sedikides. Sweeney. E.408 € J. C. 15(1).. & Wallerstedt. 531e551. Turner. & Elder. (1987). C. (1991). & Ohman. (1987). 46e68. W. E. W. & Turner. (1985). M. Auditing: A Journal of Practice & Theory. (2001). (1996). M. 27(5). (1996). Tajfel. Salterio. R. & Fernandes. & Wang. 17e28. (1996a). & Westort. C. Oxford: Blackwell.. Accounting History. Journal of Business Ethics. B. (2008). & Wurf. 17(2). & Griffeth. B. J. C. 9(1). organizational-professional conflict and work outcomes: a study of certified management accountants. Accounting. E. W. PA: Psychology Press. Otley. Auditing & Accountability Journal.. S. CPA examination performance and competition. T. A. P. Hogg. 28(8).. 377e400. Park. Zhou. 28(2).. M. The effect of accounting uncertainty and auditor reputation on auditor objectivity. 20(2). 25(4). D. Accounting. (1991). € Svanberg. B. (2004). The operation of control systems in large audit firms. 28(7).. Chicago. Auditing: A Journal of Practice & Theory. M.. 21(5). Riordan.. & Pierce. & W. 415e441. The social desirability bias in ethics research. 39e77. The dynamic self-concept: a social psychological perspective. Research in Accounting Regulation. Accounting Horizons. P.. Austin (Eds. P. D... J. Z. & Wetherell. Sundgren. Journal of Applied Social Psychology. E. M. 38. Reicher. (2009).. P. Professionalism.