The Pricing of Industry Specialization

The Pricing of Industry Specialization

by Auditors in New Zealand

David Hay, University of Auckland and

Debra Jeter, Vanderbilt University

February 2008

Abstract

A number of recent research papers examine the impact of industry specialization by auditors on audit fees in a variety of contexts. Most propose that auditors who specialize in an industry are able to perform higher quality audits, which are recognized and valued by financial statement users and thus by managers to such an extent that higher audit fees are warranted and paid to such specialists. Evidence consistent with this proposition has been presented in some studies, while others have generated conflicting findings. To the extent that specialist premiums are supported, this result is a satisfying one for auditors, as it suggests that managers value high quality audits and view audit services as something other than a necessary evil. In this study, we partition New Zealand companies into various market segments to consider the possible benefits of hiring a specialist auditor, and whether (and when) such benefits justify higher audit fees.

We find that there are fee premiums for auditor specialization defined at the city level but not at the national level. The issue of self-selection, whereby clients choose auditors based on the same variables that determine audit fees, does not appear to arise with respect to the choice of a specialist auditor in our setting. Somewhat surprisingly, we find that audit premiums for specialization apply most consistently to larger client firms, and to unlisted companies and low-risk companies. One explanation for this finding is that auditors without specialist expertise in an industry have to offer discounts to attract the most desirable clients, i.e. those that are large, those that have are associated with less danger of liability, and those that exhibit low risk; while clients are not able to negotiate fees as successfully with auditors who have differentiated themselves via industry specialization. The value of this study is that it extends the previous research to a new setting, and obtains similar results; it examines the issue of whether specialist premiums are simply due to auditors and clients self-selecting each other; and it examines subsets of the market to test alternative explanations for specialisation premiums.

DRAFT – PLEASE DO NOT COPY OR QUOTE WITHOUT PERMISSION FROM THE AUTHORS

Acknowledgments:

We appreciate helpful comments by Jeff Casterella, Paul Chaney, David Emanuel, Jere Francis, Carlos Jimenez, Gary McGill, Kevin McMeeking and participants at the University of Florida International Conference on Assurance and Corporate Governance; the Australian National Centre for Auditing and Assurance Research Conference at Australian National University and a workshop at the University of Exeter. We also gratefully acknowledge the contribution made in preliminary work by Jasmine Kwong.

The Pricing of Industry Specialization

by Auditors in New Zealand

1. Introduction

A large body of evidence (see Craswell, Francis & Taylor 1995, Hogan and Jeter 1999, and Knechel, Naiker and Pacheco 2007, for example) suggests that audit firms are desirous of acquiring specialization skills and/or being known as industry specialists. Further, in the U.S., the Government Accounting Office (GAO, 2003) pinpointed “the desire to increase industry expertise” as one of the reasons for the mergers of the Big CPA firms into the current four firms. Possible benefits, from the auditors’ perspective, include the ability to earn higher profits through either (or both) higher fees or lower production costs, reduced risk of audit failures through enhanced understanding of the industries represented, and enhanced growth opportunities.

A number of studies (Francis et al. 2005; Ferguson et al. 2003; DeFond et al. 2000) have presented evidence that industry specialist auditors earn a fee premium, using data from settings as diverse as the U.S., Australia, and Hong Kong. The body of evidence on fee premiums, however, is inconclusive as a number of other studies found little or no evidence of a fee premium (Palmrose, 1986; Ferguson and Stokes 2002), or evidence of such a premium only for certain subsets of firms (Craswell, et al. 1995, only for large clients). The usual assumption in studies finding evidence of a specialist fee premium is that the premium relates to the reputation of the auditor. For example, Craswell, et al. (1995) state: “Industry specialization will increase a Big 8 auditor’s reputation (within that industry) but will require development of expertise . . . additional investment in expertise will require a positive return resulting in an audit fee premium.”

A question that has arisen in some of the more recent studies with respect to industry specialization by auditors is whether city-specific leadership alone is a sufficient condition to warrant a fee premium, or whether both national and city leadership are needed. Ferguson et al. (2003) find a fee premium only for firms that are joint national and city-level leaders in their industry (in Australia). Francis, et al. (2005) find that national leaders do not earn a premium, but that joint national and city leaders do, as well as city-specific leaders, and concluded that “auditor reputation for industry expertise is neither strictly local nor strictly national in character” (based on U.S. evidence). Basioudis and Francis (2007) found that there was a premium for city level specialists but not for national leadership and that joint national-city leadership adds nothing beyond the premium for city leadership (in the UK).

We use a dataset of New Zealand firms to explore industry specialization by auditors. Our data provide a unique opportunity to examine the presence of fee premiums for specialist auditors in a smaller country where there is less opportunity to build up extensive experience with a narrowly-defined industry sector, and transferability of office-specific expertise is relatively easy to achieve. Further, many previous studies have not been able to include a complete measure of each audit firm’s industry share, as data for the subsidiaries of overseas companies (such as the international motor companies or oil companies) are not available and thus not included. In New Zealand these companies must disclose their audited financial statements, and we are able to include unlisted as well as listed firms to obtain a more accurate measure of industry market share. We also examine the effects of specialization within a variety of segments of the market for audit services. We provide insight into the following questions related to industry specialization by auditors:

(1)  Do any firms pay a fee premium for national and/or city-specific specialist auditors in our setting?

(2)  Do all firms with specialist auditors pay a fee premium, or only certain subsets of firms?

(3)  If the premiums affect only certain subsets of firms, what does this tell us about the benefits of specialization and the pricing of specialization expertise by auditors?

(4)  Why do client firms often prefer specialist auditors, and is this choice influenced by client size, risk and complexity?

We present evidence that firms in our sample of listed and unlisted New Zealand clients do pay fee premiums to specialist auditors, and that such premiums are observed almost exclusively at the city rather than national level. We find that the fee premiums apply consistently, regardless of our specification, to larger client firms, to unlisted firms, and to low-risk clients. Because we measure specialist premiums relative to non-specialist fees, as is common in the literature, one interpretation of this finding is that non-specialists must offer fee discounts to attract this desirable clientele. We conclude that, even in this relatively simple setting, audit fee determination reflects a complex interplay of demand and supply factors. We find no evidence that selection of a specialist auditor is associated with the variables in the audit fee model, thus lessening the concern that self-selection issues are affecting our interpretation.

The rest of this paper is organized as follows. Section 2 provides an overview of the prior research. Section 3 discusses our motivation and research design. Section 4 describes the sample and method and results, and section 5 summarizes and concludes.

2. Prior Research

A series of papers by Danos and Eichenseher (1981, 1982, 1986) suggests that auditor concentration within industries is related to scale and cost economies, extending even earlier work on auditor concentration by Zeff and Fossum (1967), Rhode et al. (1974), and Schiff and Fried (1976). Eichenseher and Danos (1981) argue that auditor concentration should be higher in industries where the auditor can use scale economies to reduce average production costs, and they expect that scale economies will dominate in industries where the auditor has to make large investments in industry-specific knowledge or expertise. Danos and Eichenseher (1982) argue that cost effectiveness is related to the audit firm's overall market share, the auditor's client- industry market share, and regulation in the client industry. They speculate that in regulated industries, large audit firms can increase market share because of scale-related cost advantages (e.g., related to specialized knowledge), while in non-regulated industries, large audit firms lose market share because of a lack of scale-related cost advantages. Danos and Eichenseher (1986) find that changes in Big 8 market shares are positively related to regulation and negatively related to growth.

More recent evidence on auditor concentration (Hogan and Jeter 1999) reveals that auditor concentration has increased in non-regulated industries over time, while concentration in regulated industries has remained unchanged (though still higher than in non-regulated industries). These findings suggest that industry-specific knowledge requirements extend to non-regulated industries, and that these knowledge requirements increased during Hogan and Jeter’s sample period, 1976-1993. Kwon (1996) takes a slightly different perspective and focuses on auditor dominance in an industry rather than levels of auditor concentration. Kwon contends that firms in concentrated industries will be reluctant to engage the same auditor because of concerns that proprietary information may be inadvertently passed to their competitors. He finds a negative relationship between his measures of auditor dominance and the four-firm concentration ratio in the industry, consistent with an argument that proprietary concerns are important.

Prior studies presenting evidence that specialist auditors provide higher quality auditors include Solomon et al. (1999), who suggest that industry-specialist auditors have greater knowledge within their specialist industry; Carcello and Nagy (2004), who find a negative association between auditor industry specialization and client financial fraud; and Balsam, Krishnan and Yang (2003), who document higher earnings response coefficients for clients of specialist auditors. Gramling and Stone (2001) provide an extensive review of the literature in this area, remarking that the evidence is quite mixed, and that “many critical issues related to audit firm industry expertise remain under- or un-investigated.”

O’Keefe et al. (1994) suggest that the provision of audit services to a client requires investments in general knowledge, industry-specific knowledge, and client-specific knowledge. We posit in this paper that one of the benefits to the client of an auditor with superior levels of industry-specific knowledge is that such knowledge facilitates the efficiency of the audit process. In other words, the client is not called upon to explain to the auditor the nature of aspects unique to the client’s industry as the industry specialist auditor is already well-versed in those aspects.

Demand side: Do client firms pay more for an audit by a specialist; and if so, why?

Recent papers examining audit fees find some evidence of a premium for auditor industry specialization (at either the city or national level, or in some cases both), generally assumed to indicate that clients value an auditor with greater expertise in their industry enough to pay more for such services. Craswell et al. (1995) state: “Some accounting firms voluntarily invest in expertise beyond the minimum required by professional standards and therefore have incentives to maintain their reputations by producing higher-quality audits.” Craswell et al. (1995) present evidence of specialist premiums measured at the national level in Australia in the 1980s. However, their results are quite sensitive to the percentage market share used for the cut-off in their definition of specialization, with statistical significance levels varying widely.

Cullinan (1998) proposes a framework in which the use of cost-based versus market-based pricing for audit services determines whether premiums or discounts are associated with greater perceived industry expertise. DeFond et al. (2000) finds that Big 5 industry specialists in Hong Kong earn a fee premium, while specialist non-Big 5 firms charge discounted fees to clients in their specialist industries.

Although most of the earlier specialization studies identified specialists based on national market share, more recent studies have turned to local or city leadership measures. Ferguson and Stokes (2002) find no evidence of specialist premiums measured at the national level in Australia in the 1990s. In contrast, Ferguson et al. (2003), also using Australian data, find that specialist premiums are earned by auditors who were identify as specialists at both the city and national levels. Francis, Reichelt and Wang (2005) also find evidence of a specialist premium where the auditor is a specialist at both the national and city level in the U. S. Francis, et al. (2005) further suggest that city-specific leadership alone is likely to be a sufficient condition, whereas Ferguson et al. (2003) argue that leadership at both city and national levels is needed to warrant a fee premium.

However, the city-alone premium documented in Francis, et al. (2005) is not supported in two of their robustness tests, and the authors conclude that their results are mixed and inconclusive on this point. They find no evidence of a fee premium for firms that were leaders at the national level only, and they conclude that there is “at least weak evidence of transferability of expertise and firm-wide reputation benefits,” but not strong transferability effects. Using UK data, Basioudis and Francis (2007) find specialization premiums at the city level but not in national market share measures, suggesting that industry specialization benefits are more likely to occur in narrower market segments than in national market share measures.

Mayhew and Wilkins (2003) draw on Porter (1985) to suggest that auditor specialization is a form of differentiation, leading not only to greater efficiency and lower costs, but also (potentially) to greater value to the client. Where audit firms are successful in differentiating themselves from their competitors, then specialization will lead to fee premiums. When they are unsuccessful in achieving differentiation, they must offer a discount to attract clients, where “achieving differentiation” means obtaining a significantly greater market share than one’s competitors (at least 10% more).