Attributes Influencing Clients' Auditor Choices: The Expectation Gaps between Auditors and Board Members

This paper reports the results of a survey of board members about the importance of various auditor attributes when making auditor choices,¹ as well as audit partners and (senior) managers about the importance they expect board members to assign to the same attributes. Surveying board members and auditors allows us to identify “expectation gaps” between the actual importance of attributes for board members in auditor choice decisions and the importance as expected by auditors. Auditor choices are driven by the interplay between board members and audit firms. Board members evaluate the proposals handed in by audit firms participating in audit tenders, and the evaluations will be influenced by the weight that board members assign to various audit firm attributes. Proposals and presentations during audit tenders provide audit firms with the opportunity to “fine tune” their offer by highlighting certain attributes (Taminiau and Heusinkveld 2017). Audit firms will highlight those attributes during audit tenders that they expect to be important for board members. However, it is unclear whether auditors correctly estimate the importance of those attributes for board members.

Audit tenders and auditor choice scenarios have become more frequent recently. In Europe, mandatory audit firm rotation (European Commission 2014b) has shortened audit firm tenure. Audit firm rotation is not mandatory in the U.S. (PCAOB 2011). Auditor tenure has, however, been under scrutiny, leading to the obligation of its disclosure (PCAOB 2017, AS 3101.10b). This disclosure could increase awareness of potentially undue long auditor tenure, which might cause audit tenders to also become more frequent in the U.S.

A vast body of archival literature draws on economic models of auditor choice to explain the impact of various demand-side factors on clients' choice of audit quality level (Dopuch and Simunic 1980). Prior archival studies used clients' audit firm tier choices (e.g., a Big N audit firm) to proxy for different audit quality levels demanded by clients. Demand-side factors examined in prior archival literature include client ownership structure (e.g., management, family, and foreign ownership), audit committee and board of director characteristics (e.g., board independence, audit committee size, and meeting frequency), other client-level determinants (e.g., client size, complexity, leverage, and political connectedness), as well as client country-level determinants (e.g., legal environment).²

However, while economic models and demand-side factors can explain the client's choice of audit firm tier, they are less able to explain the choice of the specific audit firm within the tier. Extant archival research has documented that companies are more likely to choose a specific Big 4/Big 5 audit firm if an alumna/alumnus of that audit firm is present in management. Independent audit committees mitigate this effect (Lennox and Park 2007; Dhaliwal, Lamoreaux, Lennox, and Mauler 2015). Most prior research on auditor choices beyond audit firm tiers is based on surveys and interviews. The working relationship and audit fees have been consistently identified as the most important attributes driving auditor choices (e.g., Eichenseher and Shields 1983; Butcher, Harrison, and Ross 2013). The two most important reasons for retaining the incumbent auditor in Beattie and Fearnley (1995) are the offering of a reduced fee by the incumbent auditor and avoidance of disruption and loss of management time, both economic reasons. However, when asked to rate the importance of various audit firm characteristics independent from their current audit firm, the three most important characteristics were integrity of the audit firm, technical competence of the audit firm, and the quality of the working relationship with the audit partner (Beattie and Fearnley 1995). Similarly, financial managers deem audit value and the working relationship as key determinants of auditor changes (Fontaine, Letaifa, and Herda 2013). However, all of these prior studies have used (executive) managers as survey participants.³ Changes in corporate governance systems have shifted the responsibility of audit matters away from executive management to board members and audit committees. Management, however, seems to continue to exercise influence in auditor choice decisions (e.g., Cohen, Krishnamoorthy, and Wright 2010; Dhaliwal et al. 2015). Almer, Philbrick, and Rupley (2014) found management to be an important information source for audit committees in order to assess timeliness, ability to liaise with the auditor's offices, as well as the audit firm's technical and industry expertise. Their study has, however, not assessed the importance of audit quality attributes and independence to board members.

Board members have tasks and incentives in firms' corporate governance that differ from those of executive management. From an agency perspective, the primary task of a board member is to monitor and control management (e.g., Cohen, Krishnamoorthy, and Wright 2008). Hence, board members will likely consider other attributes to be important than executive management. Because no audit firm can be expected to outrank competing audit firms on all attributes, the relative importance of attributes to board members is important. Indeed, prior archival literature has shown that independent audit committee members demand high audit quality and are more likely to choose specialist auditors (Abbott and Parker 2000). Higher audit fees are paid by more independent (Abbott, Parker, Peters, and Raghunandan 2003) and more active (Bratten, Causholli, and Sulcaj 2019) audit committees.

We examine three research questions:

We survey board members about the importance they assign to different auditor choice attributes, as well as auditors about the importance they expect board members to assign to the same attributes. We recruited both board members and auditors in Germany and Luxembourg, who have additional work experience outside their resident countries. The board members also held board mandates in other European countries such as the U.K., The Netherlands, and France. In the European Union, the audit committee is responsible for audit tenders.⁴ However, depending on the EU member state and whether an audit committee has been established, the board⁵ itself can be directly responsible for audit tenders.⁶ Of the 42 board members that were asked to participate in this study, 37 currently are or have been (vice) chairperson of a board or audit committee members.⁷

Table 1 presents demographic data.

The invitation to participate in the survey was sent by the researchers via email directly to 42 named board members, and to contact persons at Big 4 audit firms, and one non-Big 4 audit firm (a “next 10” audit firm with an international network), who distributed the invitation by email to eligible participants (i.e., partners, senior managers, and managers) in the respective local offices. We defined partners, senior managers, and managers as eligible participants for the auditor perspective, because these ranks have been found to be commonly involved in audit tenders (Taminiau and Heusinkveld 2017).

The survey was accessible between March 23 and October 10, 2018. We received 73 usable responses. We received 32 usable responses by board members and 41 by auditors, corresponding to response rates of 76 percent (board members) and 31 percent (auditors).⁸ The high response rate regarding the board members can be attributed to the fact that all board members were addressed personally and asked for their approval before taking part in the survey prior to the distribution. No systematic difference was apparent in the responses from early and late respondents.

In Sections 1−3 of the survey, respondents were asked to rate the importance of various attributes in auditor choice decisions. Responses were collected using four-point Likert scales (1 = “unimportant”; 2 = “less important”; 3 = “important”; 4 = “very important”). Attributes were selected based on prior auditor choice research (e.g., Beattie and Fearnley 1995; Behn, Carcello, D. Hermanson, and R. Hermanson 1997; Carcello, Hermanson, and McGrath 1992; Eichenseher and Shields 1983; Fontaine et al. 2013; Schroeder, Solomon, and Vickrey 1986). Additionally, certain attributes were considered separately at the level of the audit firm, audit partner, and engagement team, as prior research did not make this differentiation. This led to 38 attributes (items).

In order to structure the attributes for survey respondents, and as a basis for our willingness-to-pay analysis (Sections 4 and 5), we organized the attributes into three groups. Following Ruhnke and Schmidt (2016), we grouped attributes that prior studies (Knapp 1991; Carcello et al. 1992; Carpenter and Reimers 2013) have found to be associated with audit quality into beneficial audit attributes (Section 1) such as expertise and independence. Attributes that enhance the benefit of the audit for the client, but are not directly related to audit quality, such as quality of management letter recommendations or advice, as well as audit fees, are grouped into beneficial non-audit attributes (Section 2).⁹ The third group comprises attributes that are related to the quality of the auditor-client working relationship (Section 3). In Sections 4 and 5, we asked respondents to indicate the required monetary increase (decrease) in audit fees for a 30 percent increase (decrease) for each of the three groups of attributes, using a medium level of audit fees and a medium level of each of the groups as a benchmark.

Three board members with a minimum of ten years of work experience as board members reviewed the research instrument. Two also have substantial previous work experience as auditors (one was a former partner and the other was a former manager at a Big 4 firm). Hence, these three reviewers had a similar background as our survey respondents. Our reviewers confirmed the grouping of the attributes for Sections 4 and 5.

We use a Wilcoxon test for independent samples to compare the responses of the two respondent groups because a normal distribution cannot be assumed, and the response data are ordinally scaled.

The board members deemed the survey to be understandable (mean = 3.7) and complete (mean = 2.1).¹⁰ The two means are significantly different from the midpoint of the scale in a one-sample t-test (understandability: t₃₂ = 23.68, p < 0.001, two-tailed; completeness: t₃₂ = 15.1, p < 0.001, two-tailed). The auditors also deemed the survey understandable (mean = 3.3), but did not find that the survey covered all attributes that are important to clients in auditor choice decisions (mean = 2.7). The two means are significantly different from the midpoint of the scale in a one-sample t-test (understandability: t₄₁ = 22.8, p < 0.001, two-tailed; completeness: t₄₁ = 20.2, p < 0.001, two-tailed). The difference in the completeness ratings between the two groups of participants already suggests an overall “expectation gap.” Board members deem our selection of attributes as complete, but auditors deem attributes to be relevant for clients that clients themselves do not consider to be important.

Table 2, Panels A–C present the importance ratings by the two respondent groups.

The five attributes rated most important by board members were “industry expertise of the audit partner,” “business know-how of the audit firm,” “industry expertise of the audit firm,” followed by “professional skepticism and independence of mind of the audit partner” and “useful management letter recommendations after the audit” (with equal importance). We had grouped four of these attributes into Beneficial Audit Attributes (Panel A). Lower audit fees vis-à-vis direct competitors does not seem to be an important attribute for board members.

Auditors, on the other hand, expect board members to assign the highest importance to “meeting previously agreed schedules and deadlines,” “accessibility of the audit partner,” “industry expertise of the audit firm,” “working relationship with the audit partner,” and “industry expertise of the audit partner.” Three out of these five attributes were grouped into Working Relationship Attributes (Panel C).

Table 2, Panel D shows the board members' actual and the auditors' expected willingness-to-pay. Specifically, we asked for the willingness-to-pay (acceptable audit fee increase) for a 30 percent increase in a group of attributes. The willingness-to-pay is highest for an increase in beneficial audit attributes (+11.94 percent), and considerably lower for an increase in beneficial non-audit attributes (+7.58 percent) and working relationship attributes (+6.77 percent). The rank order is similar for a 30 percent decrease in the groups of attributes. The fee reduction required by board members is highest for beneficial audit attributes (−41.33 percent) and lower for working relationship (−25.67 percent) and beneficial non-audit attributes (−25.17 percent).

The auditors expected the acceptable audit fee increase for a 30 percent increase in attributes to be highest for an increase in working relationship attributes (+13.88 percent), second highest for an increase in beneficial non-audit attributes (+12.75 percent), and lowest for an increase in beneficial audit attributes (+11.93 percent). The required decrease in audit fees compensating a 30 percent decrease in attributes was expected to be highest for beneficial audit attributes (−16.23 percent), second highest for a decrease in beneficial non-audit attributes (−15.40 percent), and lowest for a decrease in working relationship attributes (−15.05 percent).

We find several “expectation gaps” between the importance as assigned by board members and the importance that auditors expect board members to assign to these attributes for several attributes. Table 2, Panels A–C, present Z-statistics and associated p-values when comparing the importance rankings (means) assigned by the two respondent groups (Wilcoxon test for independent samples).

We find that several attributes are more important to board members than auditors expected: “degree to which audit standards are met,”¹¹ “technical competence, industry expertise, professional skepticism, and independence of mind of the audit partner,” “professional skepticism and independence of mind of the engagement team,” and “business know-how of the audit firm.” Conversely, auditors overestimated the importance to board members for attributes such as “low total audit fee in comparison with direct competitors,” “retaining the same audit partner,” the “accessibility of the audit partner,” and “meeting previously agreed schedules and deadlines.”

Taken together, the findings suggest that auditors underestimate the importance of attributes that signal audit quality (beneficial audit attributes, Panel A), and overestimate the importance of audit fees as well as working relationship attributes (Panel C). Notably, two of those working relationship attributes focus on the audit partner.

The findings from our willingness-to-pay analysis (Panel D) support these findings. We find that the board members' actual willingness-to-pay is significantly lower than the willingness-to-pay as expected by auditors for an increase in the working relationship. A potential explanation for this overestimation is that auditors think that working relationship attributes can more easily be influenced in the short term, and they overestimate board members' willingness-to-pay for changes. Additionally, auditors underestimated board members' demand for audit fee reductions for decreases in all three groups of attributes. The difference is greatest for a decrease in beneficial audit attributes, followed by beneficial non-audit attributes, and working relationship attributes. The willingness-to-pay analysis shows that board members exhibit high price elasticity for increases in attributes, and auditors underestimate this elasticity. Conversely, board members exhibit lower price elasticity for decreases in attributes, and auditors overestimate this elasticity.

Our findings have implications for auditors and board members. Knowing the importance of attributes to board members in auditor choice decisions is crucial, especially for auditors in audit tenders. Auditors underestimate the importance of audit quality attributes and overestimate the importance of the working relationship to board members. However, our findings suggest that, when the tender process is in the hands of board members, auditors should focus more on attributes that signal audit quality to the client, such as professional skepticism, independence, technical competence, and business know-how, because board members deem these attributes to be the most important. Auditors also engage in lowballing (e.g., Desir, Casterella, and Kokina 2014). Engaging in lowballing to win audit tenders is consistent with our finding that auditors expect clients to assign a high importance to audit fees.

The results of the willingness-to-pay analysis is consistent with the importance ratings for the attributes, in that board members assign the highest importance and value to attributes that signal audit quality. This is plausible, because board members have a self-serving interest in high audit quality when they use the audit results while fulfilling their own supervisory responsibilities.

Although this study's research context is Europe, we believe our results should generalize to other jurisdictions, including the U.S. The responsibility of audit committees in audit tenders, as well as the composition and role within corporate governance are comparable: In the U.S., the audit committee is responsible for the auditor appointment process (PCAOB 2012). In the EU, although the final auditor choice is made by the shareholders, the audit committee (or board) has a significant influence on the selection process.¹² In the EU, the audit committee is a subcommittee of the board, and the majority of the audit committee members must be independent. If no audit committee has been established, the board is responsible for audit tender, and the chairperson of the board needs to be independent. The same applies in the U.S., where the audit committee is a subcommittee of the board of directors and is composed of board members and shall otherwise be independent (SOX, Section 301). Hence, U.S. audit committees are composed of the same individuals as the respondents in our survey.¹³

Future research could build on our findings and examine the interplay between different attributes with other research designs. In particular, understanding the importance of attributes in auditor choice decisions for different levels of engagement risk might be interesting. The willingness-to-pay in general and the interplay between audit fees, audit quality, and the working relationship in particular might be an interesting area for future research as well.