SEARCH

SEARCH BY CITATION

Keywords:

  • Auditor choice;
  • auditor quality

Abstract

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

This paper analyzes the auditor choices for a sample of 2,333 predominantly small and mid-sized Finnish firms. Finland requires virtually all commercial enterprises to have a financial statement audit, but allows the smallest firms to choose from four types of audit firms: first tier international firms, first tier national firms, second tier local auditors and non-certified auditors. We find that among the smallest firms, the choice to hire a certified auditor relates to the level of complexity in the organization as measured by size and extent of workforce. For firms that must use a certified auditor, we find that the choice between a first tier and second tier firm is related to size, the extent of debt financing, and complexity associated with being a member of an associated group. Finally, in the upper end of the market, the decision to hire a large international firm relates to size, the need for financing, be it equity or debt, and complexity due to a broad labour force. This pattern is interesting because it indicates that the need for a higher quality auditor is driven first by complexity, then as the firm grows, it is supplemented by the use of debt financing and ultimately by the need to raise equity as well as debt financing.


SUMMARY

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

The benefits of acquiring an audit are multi-faceted and the value of these benefits is likely to vary across firms. In the narrowest sense, the role of auditing is to improve the quality of financial statements since high quality reporting can reduce information asymmetry problems between a firm and providers of financing. Thus, a potential role of auditing is that it reduces the cost of capital. Audits also provide a number of benefits that are internal to the company. For example, the results of an audit can reduce internal agency problems, lead to improvements in process effectiveness, and assist in regulatory compliance. However, high quality auditing and reporting can potentially have adverse effects in the sense that they improve the quality of information that is disclosed to competitors. This paper analyzes the auditor choices for a sample of 2,333 predominantly small and mid-sized Finnish firms. Finland requires virtually all commercial enterprises to have an audit of the year-end financial statements, regardless of public or private ownership, but allows smallest firms to choose from four types of audit firms: first tier international firms (Big Six), first tier national firms (KHT auditors), second tier local auditors (HTM auditors) and non-certified auditors. We find that among the smallest firms, the choice to hire a certified auditor relates to the level of complexity in the organization as measured by size and extent of workforce. For firms that must use a certified auditor, we find that the choice between a first tier and second tier firm is related to size, complexity arising from being part of an associated group and the extent of debt financing. Finally, among firms that choose a top tier firm, the decision to hire a large international firm relates to size, complexity associated with a broad labour force, and the need for financing, be it equity or debt. Taken together, these results indicate that different segments of the market for audit services may be sensitive to different aspects and benefits of the audit process. While the selection of high quality auditors by large companies with extensive financing needs is consistent with evidence in prior market-based studies, smaller companies may be more sensitive to the benefits they obtain from auditors in areas of improving internal operations and obtaining access to expert advice.

I. INTRODUCTION

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

The reasons why an organization chooses a specific auditor are complex and likely to vary across organizations. The benefits of acquiring an audit are multi-faceted (Wallace, 1981; Knechel, 2002), but most research focuses on the reduction in information risk that accrues because an audit leads to more reliable reporting. This is the classic agency theory argument for auditing (Francis & Wilson, 1988), which can apply to both communications between management and outsiders and for agency relationships between management and subordinates. Wallace (1981) notes that there are also other benefits that arise from an audit including improvement in operational efficiency and effectiveness due to auditor evaluation of internal processes, deterrence of management malfeasance, increased compliance with legal and regulatory constraints, and market permission to undertake certain activities (e.g., participate in public capital markets).1

Given the multi-dimensional value of an audit, the optimum mix of different benefits may vary across stakeholders and organizations, potentially influencing the selection of an auditor. For example, management and shareholders in a public company may be most interested in the reduction of agency costs and the cost of capital associated with reliable financial reporting (DeFond, 1992; Francis et al., 1999; Francis & Wilson, 1988; Gul & Tsui, 2001; Krishnan, 2003). Prior research mostly examines public companies. Since these are the firms that are most likely to benefit from a reduction of information risk they would generally place less value on the other benefits of an audit, which are more likely to accrue to small, non-listed firms. More recent research has noted that there are interesting reasons to also study non-public companies (Ball & Shivakumar, 2005, Chaney et al., 2004). Smaller non-public companies may have little separation between ownership and management so an audit may be valuable for other reasons: the resolution of internal agency problems, possible gains in process efficiency, and assurance of compliance with complex regulations. If the value of the audit varies across organizations, then the selection of an auditor may also vary depending on the client's circumstances (e.g., scope of control; see Abdel-Khalik, 1993).

The purpose of this paper is to examine the determinants of auditor choice in a small company market. We use data from Finland because three key characteristics make Finland a unique setting in which to study this issue: (1) virtually all businesses, regardless of size or capital structure, are required to prepare public financial statements that are subject to audit, (2) there are no substitutes for ‘full’ financial statement audits (such as a review), and (3) there exist four distinct categories of auditor from which an organization can select. In general, we find that the smallest businesses select an auditor based on their complexity, mid-size companies select an auditor as a function of their debt levels, and the largest companies select an auditor conditional on whether they are publicly traded and/or their level of unsecured debt. This interesting pattern of results suggests that small firms are most concerned with internal agency and control issues, mid-size firms must cope with banks and other sources of secured financing, and large firms are most focused on external agency costs arising from publicly traded securities (debt or equity).

The remainder of the paper is organized as follows. In Section 2 we discuss the reasons why Finland provides a natural setting for testing auditor choice. In Section 3 we discuss the theoretical and empirical literature on auditor choice and generate a number of hypotheses. This is followed by a discussion of our research method and data and results in the fourth and fifth sections. We close with a summary and conclusion.

2. AUDITING ENVIRONMENT IN FINLAND

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

As a Member State of the European Union, the regulation of auditing in Finland must comply with the Directives of the European Commission. All firms subject to the Fourth and Seventh Directives have to be audited by a certified auditor that has the necessary skills and knowledge as defined in the Eighth Directive. However, differences in national legal systems that arise from cultural differences also affect the auditing environment (Margerison & Moizer, 1996) since auditing (and accounting) standards are issued at the national level.2 The role of a statutory auditor, the process of becoming a certified auditor, and the status of the profession reflect transnational differences that are deeply rooted in the political, historical and economic differences among Member States (European Commission, 1996, 1998, 2000, 2003; Baker et al., 2001). In Finland, auditors not only issue an opinion on the fairness of the financial statements, but they also conduct an administrative audit and issue opinions on the fairness of the proposed profit distribution suggested by the Board of Directors and the company's compliance with applicable laws, regulations and corporate by-laws (Niemi, 2002, 2005). An important aspect of the administrative audit is an examination of the firm's status relative to liquidation rules issued as part of the Finnish Company Law (Sundgren, 1998).

Traditionally, a distinction has been made between code law countries and common law countries (LaPorta et al., 1998). In code law countries, large audit clients are often privately held, auditors face low litigation risk, and the profession is government-regulated. Code law countries can be further grouped into three families of law: French, German, and Scandinavian, the latter including Finland (Jaggi & Low, 2000). Consistent with other code law countries, Finland is considered a low litigation environment (Niemi, 2005; Niemi & Sundgren, 2008). An auditor can be held liable to a client and afflicted third parties for losses arising from a negligent audit, and there is no cap on an auditor's liability. However, there are a number of obstacles to the successful prosecution of a case for liability against an auditor. First, damages are limited to actual losses by the plaintiff. Second, contingent fees and class action suits are not allowed in Finland. Finally, the loser pays the legal fees of the winner, exposing a plaintiff to the defendant's legal costs if the auditor is not found to be liable.

The structure of the market for auditing services in Finland is based on a two-tier system of certification resembling other Nordic countries (Saarikivi, 1999; Loft & Jeppesen, 2001) and Germany (Baker et al., 2001). Auditors approved by the Auditing Board of the Central Chamber of Commerce are known as Keskuskauppakamarin hyväksymä tilintarkastaja (KHT) and are regarded as ‘first tier’ auditors. The prerequisites for taking the KHT exam include three years of work experience under the supervision of a KHT auditor plus a suitable academic degree (e.g., a Master's Degree in economics or business administration). The examination for KHT certification has a low pass rate and is considered a demanding hurdle. The second tier auditors, known as hyväksytty tilimies (HTM), are approved and supervised by the Auditing Committees of regional Chambers of Commerce. HTM auditors need to have three years of auditing experience and a suitable academic degree (e.g., equivalent to a Bachelors Degree). While HTMs also have to pass an exam, it is generally considered to be less demanding than the KHT exam.

The Finnish Auditing Act stipulates that virtually all business enterprises in Finland must be audited, including firms not covered by EU Directives. However, small companies may hire an auditor that does not fulfil the professional requirements of the Eighth Directive (i.e., very small firms are allowed to hire a non-certified auditor). Therefore, in addition to the two tiers of certified auditors, persons holding no certification can audit the smallest firms in Finland. The size criteria for selecting an auditor are summarized in Table 1. Unlike many other countries, there are no alternative assurance services such as compilations, reviews or agreed-upon procedures engagements available for small firms, i.e., they must have a ‘full’ statutory financial statement audit as stipulated by the Auditing Act. All auditors – certified or not – must have a sufficient level of skills and knowledge in accounting and auditing to effectively conduct the audit given its complexity. International audit standards are the basis for professional standards in Finland.3 Also, Finnish accounting firms are subject to peer review every five years conducted under the applicable international standards for quality control (currently ISA 220). Finally, certified Finnish auditors must follow the Code of Ethics of the International Federation of Accountants (Niemi & Sundgren, 2008).

Table 1. Number of client firms in Finland
Category of company and auditor selectionNumber of firmsaPercentage
  • a

    Source: the Committee Report of the Auditing Act Working Group of the Ministry of Trade and Industry. The figures for that report are retrieved from the Statistics Finland databases.

  • b

    For our subsequent analysis, we combine (b) and (c) and describe them as ‘required to have a certified auditor, either HTM or KHT’ since the lead auditor on the engagement can reasonably be expected to be a certified auditor.

(a) No restrictions on auditor selection. Non-certified auditors are allowed.113,83684.2
 Meet 2 of 3 criteria:
  • Total assets < €0.34 million
  • Gross revenues < €0.68 million
  • Number of personnel < 10
(b) Required to hire at least one certified auditor (HTM or KHT) but others may be non-certified.b14,74510.9
 Meet 2 of 3 criteria:
  • Total assets > €0.34 million
  • Gross revenues > €0.68 million
  • Number of personnel > 10
(c) All auditors must be either HTM or KHT certified (no non-certified).b3,6612.7
 Meet 2 of 3 criteria:
  • Total assets > €2.1 million
  • Gross revenues €4.2 million
  • Number of personnel > 50
(d) All auditors must be either HTM or KHT certified (no non-certified), and at least one must be KHT certified.2,9352.2
 Meet 2 of 3 criteria:
  • Total assets > €25 million
  • Gross revenues > €50 million
  • Number of personnel > 300
Total135,177100.0

In 2003, there were over 135,000 operating businesses in Finland that were required to have a financial statement audit, of which 84.2% could choose a non-certified auditor if they wished (Table 1). The larger firms (about 15.8%) are required to hire a certified auditor (HTM or KHT), and 2.2% of the firms are so large that they have to hire a KHT auditor. In 2003, there were 660 individuals holding the KHT certification and 845 holding the HTM certification. There are no reliable statistics available for non-certified auditors. All Big Six accounting firms are considered to be KHT auditors. In 2002, KHTs and HTMs conducted 55,785 and 61,292 financial statement audits, respectively, indicating that many small companies prefer a certified auditor in spite of having the option to use a non-certified auditor. For the purposes of this paper, we examine four types of auditor in our analysis: (1) Big Six auditors, all of whom maintain KHT certification (B6KHT), (2) KHT-certified auditors that are not associated with international firms (KHT), (3) HTM-certified auditors (HTM), and (4) non-certified auditors (NONCERT).

3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

What causes a company to choose one auditor over another? Theoretical and empirical research has generally established that an audit has economic value, even in the absence of a mandated audit requirement (Sundem et al., 1996; Chow, 1982). The decision to have an audit, the selection of an audit firm, and the decision to occasionally switch auditors are all complex choices. In general, selection of an auditor may be due to either cost or quality considerations, or both. Prior research has used agency theory to partially explain auditor choice (DeFond, 1992; Carey et al., 2000). In the most narrow sense, the basic role of the audit is to improve the quality of financial statements and extensive literature has found that a high quality audit reduces the incidence of earnings management (Becker et al., 1998; Francis et al., 1999).4 On the other hand, some researchers have suggested that demand for quality auditing is multi-faceted and depends on more than just a cost of capital argument (Knechel, 2001, 2002; Abdel-khalik, 1993; Wallace, 1981). For example, an audit can help managers improve the efficiency of a firm or remove information asymmetries in internal reporting. Behn et al. (1997) observe that companies are more satisfied with their auditor if they are responsive to their needs and involve senior personnel in the engagement. Consequently, the choice of auditor in any given setting may be due to more complex factors than simply the reduction of external agency costs, especially for firms that are small or not publicly traded.

Prior research on audit quality has tended to focus on the differences between Big 8/6/5/4 (Big Six) audit firms and smaller audit firms. DeAngelo (1981) argued that the Big Six provide better quality audits because they have more reputational and legal risk.5 We know from prior research that audits by the Big Six are associated with higher fees (Francis & Simon, 1987; Moizer 1997), lower levels of discretionary accruals (Francis et al., 1999), lower litigation rates (Palmrose, 1988), higher rates of compliance with GAAP (Krishnan & Schauer, 2000), higher earnings response coefficients (Teoh & Wong, 1993), more informative signals of financial distress (Lennox, 1999; Knechel & Vanstraelen, 2007), and less mispricing of IPOs (Beatty, 1989). These attributes are considered to be manifestations of audit quality. In an examination of a specific small-client market, Hay and Davis (2004) were able to find differences in levels of auditor selected by very small clients based on assets, debt and operating expenses.6 Consistent with recent research into non-public companies (Ball & Shivakumar, 2005), these results suggest that small-client markets may be a fruitful area to examine issues of auditor choice.

Evidence from Europe is mixed on the issue of auditor quality. Research from Belgium, a code law country comparable to Finland, suggests that Big Six firms do not constrain earnings management any more than smaller firms (Vander Bauwhede & Willekens, 2004; Sercu et al., 2002), although they may restrain income-decreasing accruals (Vander Bauwhede et al., 2003). There are no differences in audit reporting between Belgian Big Six and non-Big Six firms when problems are readily apparent at a client, although Big Six firms follow more stringent reporting when problems are not evident (Gaeremynck & Willekens, 2003). On the other hand, Vanstraelen (2002) finds that Big Six firms are more likely to modify their opinions due to going concern problems and Willekens and Achmadi (2003) report that Big Six firms generally charge higher fees, both attributes consistent with audit quality.7

Limited evidence exists from Finland as to the effect of different types of auditors on financial and audit reporting: Sundgren and Johansson (2004) report that clients of Big Six firms use shorter depreciation lives for assets, suggesting more conservative reporting than local firms, and Sundgren (1998) finds that non-certified auditors are significantly less likely to modify their opinions than certified auditors. In general, firms that audit small unregulated businesses may not be able to develop economies of scale that allow them to invest in quality improvement, while large firms with a large network of dispersed operations may not be willing to devote resources to the audit of small local businesses (Johnson & Lys, 1990). Additionally, Finnish institutional forces may influence the level of quality delivered by various auditors. For example, non-certified auditors cannot be subject to professional censure such as loss of license for misdeeds. Pricing is used to signal quality for credence goods such as an audit (Dulleck & Kerschbamer, 2006; Hay & Knechel, 2007), and Niemi (2004) finds that hourly billing rates are higher for KHTs than for HTMs. According to the Central Chamber of Commerce, 62% of HTM auditors work part time as sole practitioners, compared to less than 16% of KHT auditors (Niemi & Sundgren, 2008), suggesting that KHT auditors may be able to offer more up-to-date and valuable service to clients since a firm can make investments in quality that an individual cannot (Watts & Zimmerman, 1986).

Hay and Davis (2004) develop a hierarchy of quality that is consistent with the perceived quality of Finnish auditors. They define the first level of quality as education. Since non-certified auditors are not required to meet any specific educational requirements, this characteristic separates NONCERT from HTM, KHT and B6KHT. The second level of quality is professional certification. Since the education, training and testing requirements to achieve KHT status are significantly more stringent than those to obtain an HTM certification, this characteristic separates HTM from KHT and B6KHT. The third and fourth levels of quality are firm size and international reputation, which separates KHT from B6KHT. In sum, the multi-faceted benefits of audits to small clients (Wallace, 1981), the range of factors that influence client satisfaction with an auditor (Behn et al., 1997), and the unique institutional aspects of the Finnish audit market suggest an a priori expectation that there will be perceived quality differences across the four classes of auditors which can influence auditor choice. The remainder of the paper examines three sets of characteristics that might influence auditor choice among those four categories: (1) organizational complexity, (2) financial structure and (3) competitive pressures.

Internal value of an audit: organizational complexity and control

An audit can provide a number of benefits that are internal to a company such as improved process effectiveness and efficiency, increased compliance with laws and regulations, and/or reduced internal information asymmetries. These benefits may arise due to different types of complexity in an organization, e.g., asset, operational, financial and/or transactional complexity (Newton & Ashton, 1989). In a very small company, the owner (manager) can control the operations by direct supervision. However, as a company grows larger it becomes more complex and difficult to control (Kinney & McDaniel, 1989). An increasing number of subordinates, locations and activities may reduce overall efficiency in the organization and give rise to moral hazard problems between the manager/owner and subordinates. For example, subordinates may appropriate wealth from the firm by shirking or stealing, accentuating the need to develop internal control systems. As a result, organizations that are complex are more likely to select a high quality auditor (Simunic & Stein, 1987; Abdel-Khalik, 1993; Hay & Davis, 2004), leading to our first hypothesis:

H1: The choice of auditor is positively associated with the complexity of the firm.

We use four measures to reflect complexity in an organization. First, we define SALARIES as the ratio of salary expense to total operating expenses. SALARIES reflects operational and informational complexity of an organization (Hay & Davis, 2004; Abdel-Khalik, 1993). The more labour intensive an organization is, the broader management's scope of control must be to assure effective operations. Second, we define INVREC as the ratio of inventory and receivables to total assets to reflect asset and transactional complexity (Stice, 1991; Hay et al., 2006). Ge and McVay (2005) observe that the two largest sources of reported material weaknesses in internal control as reported by US companies under Section 404 of the Sarbanes-Oxley Act of 2002 relate to accruals (receivables) and inventory. Third, we define GROUP as a dummy variable that takes the value 1 if the company is the parent company, a subsidiary or a jointly controlled company. Companies that belong to a corporate group are more complex and are likely to use the same audit firm for the whole corporate group (Hay & Knechel, 2005). Ge and McVay (2005) report that a significant number of disclosed material weaknesses in control are due to problems in subsidiaries. Finally, we use SALEGROW, defined as the percentage growth in sales for a specific company, to indicate rapid growth which could lead to stress or breakdowns in internal processes that increase the complexity and riskiness of an organization (Kinney & McDaniel, 1989).8 For all four variables, we expect a positive association between the measure of complexity and the choice of auditor.9

External value of the audit: financial structure

At some point, most small companies need to take on external financing in order to grow. A company that is planning on obtaining new financing will want their financial statements to be credible and may engage a high quality auditor to increase the perceived reliability of financial reports, especially if faced with complex transactions to report. Prior research has shown that the availability (Das & Sengupta, 2001) and success of new financing (Beatty, 1989) depends on the riskiness of the company, and that improved disclosures reduce the risk to investors and the cost of capital to the issuer (Botosan, 1997). These observations lead to our second hypothesis:

H2: The choice of auditor is positively associated with the demand for external financing.

We measure the demand for external financing based on a measure of free cash flow, EXANTEFIN, originally developed by Dechow et al. (1996). This variable is calculated as cash from operations less capital expenditures divided by current assets. The unadjusted ratio for our data is fat-tailed with kurtosis of 112. We transform the measure of free cash flow by taking the square root of the ratio and then restoring the original sign.10 We expect to observe a negative relationship between EXANTEFIN and auditor selection.

Extensive prior research suggests that one major source of demand for auditing is the need to mitigate the client's agency costs arising from conflicting interests among management, owners and creditors (Jensen & Meckling, 1976). This perception has been the basis of a large number of prior studies (Chow, 1982; Beasley & Petroni, 2001). Typically companies in need of financing will first turn to local banks or financial institutions. Evidence suggests that a company can reduce its effective interest rate by engaging a Big Six auditor (Blackwell et al., 1998; Pitman & Fortin, 2004; Mansi et al., 2004; Causholli & Knechel, 2007). Furthermore, external debt financing creates potential agency conflicts between shareholders and creditors (Jensen & Meckling, 1976) and prior research suggests that firms can use income-increasing accounting methods to avoid debt covenant violations (DeFond & Jiambalvo, 1994). The likelihood of violations of covenants (Duke & Hunt, 1990) increases with leverage so an audit facilitates the enforcement of debt covenants that restrict management actions, e.g., changes in leverage, investment decisions and dividend payouts (Smith & Warner, 1979). A creditor's risk is also affected by the extent of collateral available to support a loan and the notional value of collateral may be overstated as a result of aggressive accounting practices by a borrower. Consequently, the borrower's choice of auditor is conditional on the extent of a firm's leverage, leading to our third hypothesis:

H3: The choice of auditor is positively associated with a firm's leverage and the level of unsecured debt.

We use total liabilities in relation to total assets, LIAB, as our measure of leverage. Accounting based covenants are not common among small and mid-sized Finnish firms; however, some rules in the Company Law (CL) serve a similar role as covenants in loan contracts, e.g., requiring that dividends cannot exceed the sum of the current year's earnings and retained earnings from earlier years or that a firm should initiate liquidation if more than half of the stock-capital has been consumed by losses. Also, Bergström et al. (2004) show that there is a strong positive correlation of a firm's real estate and investments with payments to creditors in bankruptcy. We define UNSECURED as the ratio of debt to banks and financial institutions minus land, buildings and investments, in relation to total assets. We expect a positive relationship between LIAB and UNSECURED and the choice of auditor.

As firms continue to grow, they may wish to expand their equity financing or issue shares to the general public, shifting the relative balance of agency costs to potential conflicts between managers and owners. The extent that management can undertake earnings management is a function of the auditor and the selection of an auditor may alleviate the potential agency conflict between majority and minority shareholders. These well-known agency problems give rise to our fourth hypothesis:

H4: The choice of auditor is positively associated with the use of external equity financing.

We proxy for equity-related agency costs in two ways. First, we use a dummy variable to indicate when a company issues new equity securities. Loughran and Ritter (1997) observe that firms that issue seasoned equity offerings usually show improved operating performance prior to the offering but then show deteriorating performance afterwards. We define SHRGROW to have a value of 1 if a company has issued additional equity in the current year. Second, ownership and control is much more likely to be dispersed among firms that are publicly traded. Therefore, we use a dummy variable that is equal to 1 if a company is publicly traded on the Helsinki Stock Exchange or is a subsidiary of a publicly traded company (PUBLIC).11 We expect a positive association between SHRGROW and PUBLIC and the choice of auditor.

Externalities and disclosure of financial information

A primary objective for any company is to develop and retain a competitive advantage. For a competitor to successfully imitate the strategy of another, it must be aware of the potential for superior profits, have the incentive to compete, be able to diagnose the source of the competitive advantage, and have access to the resources needed (Grant, 1998: 180–185). Businesses have an incentive to avoid revealing proprietary information that could be harmful to the competitive position of the company (Porter, 1980). ‘Conservative’ reporting to protect competitive advantage is also more likely to occur in countries such as Finland that are classified as ‘macro’ countries in their approach to accounting practices (Nobes, 1983; Doupnik & Salter, 1993). One element of sustaining a competitive advantage is to obscure superior performance from competitors and possible entrants (Grant, 1998). Hiring a low quality auditor may increase the opportunities for a company to use income-decreasing accounting methods to disguise the true performance of the firm.12 Furthermore, small, non-traded firms may have more flexibility in disguising their results than large publicly-listed firms. This leads to our fifth hypothesis:

H5: The choice of auditor is negatively associated with the costs of disclosing proprietary information to competitors.

Several studies suggest that the proprietary costs of disclosure are higher in more competitive markets (e.g., Verrechia, 1990).13 Consequently, we study whether there is a relationship between measures of competition in the industry and auditor choice. Our first measure focuses on market concentration. The Herfindahl index of industry concentration (e.g., Harris, 1998; Verrechia & Weber, 2006) is calculated as the sum of (Zi)2 divided by the square of the sum of Zi, where Zi refers to the sales of the ith sample firm in our sample. We compute HERFINDAHL for each industry in our sample at a two-digit level.14 The general assumption in the literature is that competition decreases as market concentration increases (e.g., Harris, 1998; Botosan & Stanford, 2005). A higher value of the ratio means that the industry is more concentrated so we expect HERFINDAHL to have a negative relationship with auditor choice. For supplemental tests, we also compute the concentration ratio (CONCENTRATION) for the four largest firms in the industry (Harris, 1998; Botosan & Stanford, 2005) as inline image, where si is firm i's sales and S is the sum of sales for all firms in the industry (measured at a two-digit level). Regardless of the measure used, the relationship between industry concentration and competition may not be as straightforward as our analysis suggests. Baumol et al. (1988: 217) point out that an effective threat of entrance from new rivals will put downward pressure on profits and rents of a monopolist. Stiglitz (1987) finds that competition may even decrease when there are many firms in a market since search costs increase as it becomes more difficult for customers to identify low cost firms. In this case, increased competition may be indicated by high value of the Herfindahl index and would result in a positive regression coefficient for the concentration measure(s).

Another measure of competition used is the persistence of above-median earnings (Harris, 1998; Botosan & Stanford, 2005). The rationale behind the measure is that the speed at which abnormal earnings are driven towards zero is a function of the competitiveness of the industry (Ohlson, 1995; Dechow et al., 1999). We measure the speed of profit adjustment (ROA-ADJ) by using a firm's abnormal profit rate taken as the difference between the firm's return-on-assets15 and the industry median. We then regress this difference on prior periods and use β2j from the following regression as a measure of competitiveness:

  • image(1)

where Xit is the difference between firm i's return on assets and the industry average in year t, D1 is a dummy variable taking the value 1 if Xit−1 ≤ 0, and D2 is a dummy variable taking the value 1 if Xit−1 > 0. The regression is run separately for each industry using the data in our sample. A smaller β2j (ROA-ADJ) suggests greater competition since an abnormal performance fades away at a faster rate. We expect ROA-ADJ to have a positive relationship with auditor choice since less competitive industries are expected to select more prestigious auditors.

The theoretical association between competition and concentration of the industry is ambiguous so we also use profitability as an alternative measure of competition since consistently high profits for a firm suggests that it may operate in a less competitive industry (e.g., one with high barriers to entry). We measure firm profitability as the firm's return on assets (ROA). We argued above that a company may have an incentive to hire a low quality auditor in order to make it easier to disguise its true profitability. Dedman and Lennox (2007) point out that more profitable firms have stronger incentives to conceal proprietary information, suggesting that profitable firms would also have an incentive to hire low quality auditors. However, managers of unprofitable firms might wish to disguise the poor performance of the firm from shareholders and lenders (Berger & Hann, 2007). We expect that the proprietary cost motive is more likely to dominate among non-public companies that are owned by their managers (cf. Dedman & Lennox, 2007).

4. RESEARCH METHODOLOGY AND SAMPLE SELECTION

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

The data

We use 2,333 firms in our main analyses. The data is taken from the Voitto+ database, which contains the financial statements for almost all firms operating in Finland. Firms in the database have either sent their financial statements directly to Asiakastieto Oy (a major Finnish credit company) which maintains the Voitto+ database, or they have filed the financial statements with the Trade Register at the National Board of Patents and Registration of Finland (NBPR), a department of the Ministry of Trade and Industry. In Finland, all companies regardless of their size must file financial statements with the publicly available Trade Register. While some countries allow very small firms to file abbreviated accounts (e.g., they report the gross margin but not actual sales figures), Finnish companies do not have this choice.

We identified 7,172 manufacturing firms during the period 1994–1999. Information about the auditor's qualifications could be found for 6,558 firms. We use the most recent observation for firms that appear in the sample more than once implying that only companies from the 1996–1999 period were left in the sample. We omit 3,850 firms that lack information for lagged revenues or the prior year auditor, leaving 2,708 firms. We omit a number of companies due to various other data constraints: 27 firms with zero or negative sales, six firms with lagged sales equal to zero, one firm with zero assets, two firms missing salary data, and three cases for which the concentration ratio could not be calculated. Finally, we omit 27 firms audited by Big Six HTM auditors and 27 cases that were audited by an auditor with inadequate credentials given the circumstances of the company. The sample of 2,615 companies is reported in Table 2 for the period 1996 through 1999.16 The sample includes 203 firms (7.8%) audited by a non-certified auditor, 793 firms (30.3%) audited by an HTM auditor, 549 firms (21.0%) audited by a non-Big Six KHT auditor, and 1,070 firms (40.9%) audited by Big Six firms. The Auditing Act stipulates that a company has to hire a HTM or a KHT auditor if its sales, assets or number of employees exceed certain size thresholds as described in Table 1. Information about sales and assets is publicly available but the number of employees is not. Our sample includes a number of companies that met only the sales or asset condition, but not both. Since we do not know the number of employees, we cannot determine which type of auditor would be required by law. These companies are in classes 2 and 4 in Table 2 and are omitted from the sample used for our main analyses, resulting in a final sample of 2,333 companies in class 1, 3 and 5. We examine the firms in classes 2 and 4 in our subsequent sensitivity analysis.

Table 2. Sample firms by size and auditor type
Client size classaAuditor typeTotals
Non-certified auditorHTMKHTB6KHT
  • a

    Client size class. The larger the firm, the less choice the firm has on different auditor types. Three size measures determine the range of auditor choice that the firm has. These measures are total assets, gross revenues, and the number of personnel. When two size thresholds out of three are exceeded, the restrictions for that size class of auditor choice apply to the firm (for specific thresholds figures, see Table 1). For the purposes of our study, we classify our sample firms to the firms that may hire any type of auditor they wish (size class 1, small firms), to the firms that have to choose between a second tier HTM and a first tier KHT (size class 3, mid-sized firms), and to the firms that have to hire a KHT auditor (size class 5, large firms). As our data has been collected from the financial statements, which in Finland do not include information on the number of personnel, we have classified our remaining sample firms into two additional size classes. The firms that may hire any type of auditor they wish if the number of their employees is less than ten (small firms) – but if the number of the employees is larger they have to choose between a HTM and KHT – are classified to size class 2 as we do not know whether they are ‘small’ or ‘medium-sized’. Similarly, we classify the firms that have to hire a KHT auditor if the number of their employees exceeds 300 in size class 4. In our main analyses, we use three ‘clean’ subsamples where firms can be classified correctly as ‘small’, ‘medium-sized’, and ‘large’ without having the data on the number of employees (the size classes 1, 3 and 5). In our tests, we refer to these subsamples as Samples 1, 2 and 3. To assess the sensitivity of our results to the sample selection, we included the size classes 2 and 4 in our Samples 1, 2 and 3 in various ways. The results remain qualitatively the same after the inclusion of size classes 2 and 4 into the subsamples which indicates that our results are not sensitive to the different definitions of ‘small’, ‘medium-sized’ and ‘large’ firms.

11741737986512
229935266240
35274016981,626
424042
515180195
Totals2037935491,0702,615

Descriptive statistics for the final sample are presented in Table 3, classified by size (Panel A) and type of auditor (Panel B). The variables SALEGROW and LIAB had sizable outliers so we ‘winsorized’ the variables 0.5% in the upper and lower tail of the distribution to reduce the impact of these extreme values (i.e., replaced the most extreme values with less extreme ones).17 The smallest firms are more complex based on SALARIES and SALEGROW, similarly complex based on INVREC and less complex based on GROUP. While all three groups have roughly the same level of overall debt (LIAB), the largest firms have the largest need for financing (EXANTEFIN), issue more equity (SHRGROW) and have less secured debt (SECURED).18 The largest firms in the sample operate in more concentrated industries (HERFINDAHL).

Table 3. Descriptive statistics
Panel A: Descriptive statistics by the samples used in Models 1, 2 and 3a
VariablesSample 1 (n = 512) (small firms)Sample 2 (n = 1,626) (mid-sized firms)Sample 3 (n = 195) (large firms)Test statisticb
Mean (Std.Dev.) [25%/50%/75%]Mean (Std.Dev.) [25%/50%/75%]Mean (Std.Dev.) [25%/50%/75%]
SIZE (in million euros)0.1513.36672.897166.07***
(0.126)(4.199)(178.699)
[0.053/0.130/0.215][0.875/1.702/4.048][7.781/29.043/69.266]
SALARIES0.3380.3050.27613.52***
(0.209)(0.139)(0.160)
[0.185/0.345/0.474][0.207/ 0.305/0.395][0.172/0.242/0.358]
INVREC0.3420.3680.3246.18***
(0.243)(0.193)(0.192)
[0.149/0.286/0.508][0.218/0.344/0.492][0.160/0.299/0.468]
GROUPc8.20%25.28%83.08%411.08***
(42)(411)(162)
SALEGROW0.2640.1200.15710.09***
(1.039)(0.435)(0.646)
[−0.098/0.067/0.252][−0.034/0.064/0.183][−0.025/0.062/0.176]
EXANTEFIN0.1890.1270.1972.92*
(0.735)(0.540)(0.617)
[−0.321/0.355/0.650][−0.322/0.289/0.523][−0.217/0.380/0.598]
LIAB0.5920.5850.5810.15
(0.333)(0.256)(0.214)
[0.338/0.585/ 0.806][0.402/0.600/0.784][0.426/0.585/0.763]
UNSECURED0.0420.016−0.09120.93***
(0.289)(0.236)(0.201)
[−0.029/0.000/0.135][−0.131/0.000/0.166][−0.203/−0.050/0.000]
SHRGROWc4.49%7.50%8.21%4.60*
(25)(122)(16)
PUBLICc75.90%
(148)
HERFINDAHL0.0490.0520.0645.84***
(0.054)(0.050)(0.056)
[0.015/0.028/0.054][0.015/0.042/0.054][0.022/0.049/0.083]
ROA-ADJ0.5710.5730.5944.05**
(0.102)(0.098)(0.104)
[0.512/0.598/ 0.647][0.512/0.601/0.623][0.512/0.605/0.647]
Panel B: Descriptive statistics by auditor type
VariableNon-certified auditors (n = 174)HTM auditors (n = 700)(Non-B6) KHT auditors (n = 495)B6 KHT auditors (n = 964)Test statisticb
Mean (Std.Dev.) [25%/50%/75%]Mean (Std.Dev.) [25%/50%/75%]Mean (Std.Dev.) [25%/50%/75%]Mean (Std.Dev.) [25%/50%/75%]
  • a

    Samples consist of the firms in classes 1, 3 and 5 as described in Table 2.

  • b

    An F-test is used for the continuous variables and a χ2-test for the dichotomous variables

  • *, **, ***

    *, **, ***significant at 10%, 5%, and 1% levels, respectively.

  • c

    For dichotomy variables, a percentage and frequency (in parenthesis) of the cases when the variable takes the value of one is reported.

SIZE (in million euros)0.1181.3663.71817.58115.44***
(0.107)(1.846)(10.597)(84.735)
[0.031/0.079/0.183][0.367/0.807/1.583][0.546/1.373/3.702][1.016/2.900/8.100]
SALARIES0.3000.3350.3000.2988.15***
(0.194)(0.167)(0.146)(0.151)
[0.159/0.284/0.443][0.214/0.337/0.437][0.200/0.302/0.400][0.198/0.288/0.387]
INVREC0.3190.3650.3630.3592.48*
(0.251)(0.203)(0.201)(0.200)
[0.123/0.254/0.481][0.206/0.338/0.493][0.213/0.339/0.492][0.204/0.337/0.495]
GROUPc0%7.00%28.08%44.29%357.93***
(0)(49)(139)(427)
SALEGROW0.1800.1340.1850.1500.71
(0.983)(0.514)(0.735)(0.584)
[−0.135/0.044/0.211][−0.048 /0.056/0.193][−0.027/0.084/0.208][−0.039/0.064/0.187]
EXANTEFIN0.2150.1740.1280.1241.85
(0.731)(0.574)(0.588)(0.586)
[−0.318/0.360/0.674][−0.303/0.342/0.556][−0.314/0.291/0.521][−0.332/0.282/0.544]
LIAB0.5880.5770.5750.5991.21
(0.370)(0.272)(0.264)(0.254)
[0.298/0.563/0.816][0.380/0.583/0.788][0.386/0.580/0.780][0.421/0.617/0.782]
UNSECURED0.0450.0220.015−0.0012.37*
(0.335)(0.238)(0.249)(0.235)
[−0.032/0.000/0.134][−0.127/0.000/0.159][−0.115/−0.001/0.174][−0.131/0.000/0.125]
SHRGROWc6.90%6.00%8.28%7.05%2.34
(12)(42)(41)(68)
PUBLICc2.02%14.32%177.88***
(10)(138)
HERFINDAHL0.0450.0510.0520.0552.67**
(0.052)(0.055)(0.052)(0.049)
[0.015/0.028/0.054][0.015/0.034 /0.054][0.015/0.034/0.054][0.015/0.045/0.058]
ROA-ADJ0.5730.5710.5690.5801.74
(0.092)(0.097)(0.104)(0.101)
[0.512/0.598/0.647][0.512/0.598/0.623][0.512/0.601/ 0.647][0.512/0.605/0.647]
Variable definitions:
Control variable:
SIZETotal assets
Complexity variables:
SALARIESRatio of salaries to operating expenses
INVRECRatio of inventory and receivables to total assets
GROUP1 if a parent company, a subsidiary or a jointly controlled company, otherwise 0
SALEGROWGrowth of sales from previous year divided by sales from previous year
Financial structure variables:
EXANTEFINCash flows from operations less net investments in relation to current assets. A square-root transformation of the ratio was used. If the sign of the unadjusted ratio was negative, then the original sign was restored.
LIABRatio of total liabilities to total assets
UNSECUREDRatio of debt from banks and insurance companies minus buildings, land and shares to total assets
SHRGROW1 if increase in share capital, otherwise 0
PUBLIC1 if a listed company, otherwise 0
Degree of market competitiveness variables:
HERFINDAHLSum of (Zi)2 divided by the square of the sum of Zi, where Zi refers to the sales of the ith sample firm in the industry. Industry is measured at a two-digit level. The value of one (max) would indicate that only one firm exists in the industry whereas the index approaches zero in the industry with a large number of firms of equal size.
ROA-ADJSpeed of ROA adjustment from year t− 1 to t. Measured at a two-digit industry level

Panel B presents descriptive statistics by auditor choice. There are univariate differences across auditor types regarding SIZE, SALARIES, INVREC, GROUP, UNSECURED, PUBLIC,19 and HERFINDAHL. The Big Six have many more clients traded on the Helsinki Stock Exchange than the non-Big Six KHT auditors (14.3%/2%). Over 25% of the firms are part of an affiliated group and they are always audited by a certified auditor, mostly KHT auditors. The cell means are not equal across the four categories of auditor for SALARIES and INVREC but neither is the relationship among the cells monotonically increasing or decreasing. Also of interest are the relatively larger values for SHRGROW for the two KHT categories versus the other two categories (although the difference across the four categories is not statistically significant).

The estimation models

We use three logistical regression models to examine different auditor choice decisions: (1) certified versus non-certified, (2) HTM versus KHT, and (3) Big Six versus non-Big Six KHT. This formulation allows us to test whether client attributes have a differential impact as auditor choice moves from small to large accounting firms, as well as from the smallest to largest companies. One would expect that the choice of auditor in large firms is more complex in the sense that more factors may influence the choice of auditor.20 Since the choice of alternative auditors varies across size categories, the sample for each model depends on the alternatives actually available to a company.

Model 1 represents the choice between a certified and non-certified auditor. Only the smallest companies in our sample are allowed to choose a non-certified auditor. Specifically, a company must use a certified auditor if two of the following three conditions are met: assets are over €0.34 million, revenues are over €0.68 million and/or the number of employees is over 10 (see Class 1 in Table 2).21 We estimate the following logistical regression:

  • image(2)

where CHOICE1 is 1 for a certified auditor, and 0 otherwise. We use size, measured as the natural log of assets (LNSIZE), as a control variable since larger companies can be expected to use larger audit firms. PUBLIC is not used in this model because none of these firms are publicly listed. GROUP is not used since all firms that belong to a group were audited by certified auditors. Since the sample used for Model 1 constitutes the smallest companies, we expect that the dominant influence on auditor choice will be complexity (H1), followed by the need for financing (H2) and debt-related agency problems (H3).

Model 2 represents the choice between a KHT and an HTM auditor for any firm that is too big to qualify for a non-certified auditor, but is not publicly listed or large enough to be required to use a KHT auditor. Consequently, the sample for Model 2 consists of firms that must have a certified auditor but are not required to have a KHT auditor (see Class 3 in Table 2). We estimate the following logistical regression:

  • image(3)

where CHOICE2 is 1 for a KHT auditor, and 0 for an HTM auditor. PUBLIC is not used in this model because none of these firms are publicly listed. Since the firms in this sample are on the cusp between very small and moderately large, we expect that the need for financing (H2), especially debt financing (H3), and concerns about proprietary costs of disclosure (H5) will be the predominant drivers of auditor choice.

Model 3 reflects the choice between a non-Big Six KHT auditor and Big Six firm among companies that must have a KHT auditor (see Class 5 in Table 2). For those firms, we estimate the following logistical regression:

  • image(4)

where CHOICE3 is 1 for a Big Six auditor, and 0 otherwise. The firms in this sample can be publicly listed (PUBLIC). Since these companies are the largest in our sample, we expect that the need for equity financing (H4) and concerns about proprietary costs of disclosure (H5) will be the predominant drivers of auditor choice.

5. RESULTS

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

The correlation results for the variables used in each model are presented in Table 4 and the main results are presented in Table 5. We indicate the expected sign for each independent variable and report the coefficients and one-tailed significance levels for each variable in each model. We also report the model significance and the pseudo R2. We will discuss each model separately and then compare the results across models in the Discussion and Conclusions section that follows.

Table 4. Pearson correlation matrices
Panel A: Sample 1 (small audit client firms), n = 512a
VariablesbLNSIZESALARIESINVRECGROUPSALE-GROWEXANTEFINLIABUN-SECUREDSHR-GROWHERFIN-DAHLROA-ADJ
LNSIZE1.0000          
SALARIES0.05771.0000         
INVREC−0.0239−0.01981.0000        
GROUP0.1761−0.0060−0.00221.0000       
SALEGROW0.0674−0.05910.05130.09221.0000      
EXANTEFIN−0.14310.0460−0.1672−0.0541−0.06391.0000     
LIAB−0.0461−0.04790.2840−0.02990.0260−0.18651.0000    
UNSECURED−0.0580−0.07180.2293−0.03940.0633−0.17080.49711.0000   
SHRGROW0.1049−0.1150−0.02460.03130.0004−0.11320.03010.02401.0000  
HERFINDAHL0.0580−0.07710.0347−0.04760.0308−0.09540.0046−0.00230.02151.0000 
ROA-ADJ−0.04760.1130−0.00760.0116−0.00270.01580.02010.0162−0.0376−0.41881.0000
Panel B: Sample 2 (medium-sized audit client firms), n = 1626a
VariablesbLNSIZESALARIESINVRECGROUPSALE-GROWEXANTEFINLIABUN-SECUREDSHR-GROWHERFIN-DAHLROA-ADJ
LNSIZE1.0000          
SALARIES−0.19141.0000         
INVREC−0.0468−0.19831.0000        
GROUP0.2940−0.0525−0.05001.0000       
SALEGROW0.0963−0.12090.03680.03281.0000      
EXANTEFIN−0.06200.0311−0.12860.0112−0.00311.0000     
LIAB−0.0764−0.01390.2127−0.01640.0794−0.26941.0000    
UNSECURED−0.0621−0.07140.3096−0.08630.0520−0.17290.53261.0000   
SHRGROW0.0596−0.03340.00490.03850.0810−0.05150.03980.01381.0000  
HERFINDAHL0.0773−0.04870.1735−0.01100.0622−0.01350.01420.08490.01511.0000 
ROA-ADJ−0.04460.0042−0.2066−0.0045−0.05770.0449−0.0367−0.08160.0393−0.40811.0000
Panel C: Sample 3 (large audit client firms), n = 195a
VariablesbLNSIZESALARIESINVRECGROUPSALE-GROWEXANTEFINLIABUN-SECUREDSHR-GROWPUBLICHERFIN-DAHLROA-ADJ
  • a

    Sample 1 consists of audit client firms that due to their small size are allowed to hire any type of auditor including non-certified ones. Sample 2 consists of firms that due to their larger size have to choose at least one certified auditor (i.e. between first tier KHT auditors and second tier HTM auditors). Sample 3 consists of listed companies and largest privately held firms which have to choose at least one first tier KHT auditor (i.e. between non-Big N KHT and Big N KHT). For specific monetary size thresholds, see Table 1.

  • b

    For variable definitions, see Table 3.

LNSIZE1.0000           
SALARIES−0.20881.0000          
INVREC−0.1110−0.12011.0000         
GROUP−0.01260.13570.05071.0000        
SALEGROW−0.0488−0.01240.02160.09511.0000       
EXANTEFIN−0.01560.1467−0.01990.0754−0.10471.0000      
LIAB−0.13400.08390.36260.09230.1357−0.12691.0000     
UNSECURED−0.0560−0.05090.3287−0.07390.0997−0.13560.33711.0000    
SHRGROW0.1543−0.0975−0.00950.08510.0107−0.1313−0.1470−0.00141.0000   
PUBLIC−0.20000.23330.0068−0.03050.03990.06640.0618−0.00480.03741.0000  
HERFINDAHL0.2213−0.20130.0831−0.1012−0.0497−0.07790.16450.0359−0.0296−0.15061.0000 
ROA-ADJ0.0201−0.0786−0.2940−0.1371−0.1030−0.0327−0.14290.0223−0.05480.2244−0.24611.0000
Table 5. Logistic regression results: Auditor choicea = f (LNSIZE, SALARIES, INVREC, GROUP, SALEGROW, EXANTEFIN, LIAB, UNSECURED, SHRGROW, PUBLIC, HERFINDAHL, ROA-ADJ)
VariablebPredicted signModel 1Model 2Model 3
Coefficient Z-value p-valuecCoefficient Z-value p-valuecCoefficient Z-value p-valuec
  • a

    Auditor choice in Model 1: 1=certified auditor (i.e. HTM, KHT or B6KHT), 0=non-certified auditor (NONCERT); Model 2: 1=KHT or B6KHT, 0=HTM; Model 3: 1=B6KHT, 0=KHT. For description of auditor types (NONCERT, HTM, KHT, and B6KHT), see Table 1.

  • b

    LNSIZE is natural log of assets. Other variables are defined in Table 3.

  • c

    Significance levels (p-values) are based on one-tail tests.

LNSIZE+0.414 4.50 0.0000.652 9.48 0.0000.534 2.15 0.016
SALARIES+1.027 2.12 0.017−0.128 −0.29 0.3843.778 1.54 0.062
INVREC+0.386 0.90 0.1860.094 0.28 0.389−2.324 −1.15 0.125
GROUP+     1.443 8.03 0.0000.226 0.28 0.389
SALEGROW+0.142 1.27 0.103−0.041 −0.26 0.3970.448 0.48 0.317
EXANTEFIN−0.035 −0.26 0.4000.025 0.22 0.4120.894 1.59 0.056
LIAB+0.033 0.09 0.4630.761 2.79 0.0031.365 1.64 0.051
UNSECURED+−0.017 −0.04 0.483−0.255 −0.85 0.1990.830 0.49 0.311
SHRGROW+−0.547 −1.24 0.108−0.179 −0.80 0.2115.038 3.03 0.001
PUBLIC+          0.194 0.16 0.438
HERFINDAHL+2.149 1.03 0.1531.538 1.21 0.1145.551 0.77 0.222
ROA-ADJ+−0.012 −0.01 0.4960.0518 0.08 0.4690.747 0.18 0.427
Constantnone−1.796 −2.24 0.013−4.898 −6.62 0.000−4.639 −1.14 0.128
Chi-square/sign.  39.73 0.000  259.42 0.000  20.27 0.062 
Pseudo R2  0.0605    0.1266    0.1916   
Sample size 512    1,626    195    

Model 1

The sample used for Model 1 includes the firms that have the option of selecting a non-certified auditor, i.e., Class 1, Table 2. The sample for Model 1 consists of 512 companies, of which 174 were audited by non-certified auditors, 173 by HTM auditors, 79 by KHT auditors and 86 by Big Six auditors. The mean (median) assets of the firms is €0.15 million (€0.13 million). As can be seen from Table 4, the correlations between the independent variables in Model 1 are generally low. LIAB and UNSECURED are the highest correlated variables with a Pearson correlation of 0.50, while HERFINDAHL and ROA-ADJ have a correlation of −0.42. The results for Model 1 are reported in Table 5. Model 1 has a χ2 of 39.73 which is significant at p < 0.001. As expected, complexity issues dominate the choice of auditor for the smallest firms since potentially important benefits of the audit include objective evaluation of internal control and assessment of compliance with laws and regulations. The complexity variable SALARIES has the predicted positive sign with a p-value of 0.017. LNSIZE has the predicted positive sign and it is significant at the p < 0.001 level.

As a robustness check, we re-estimated Model 1 using the sub-sample of 512 companies that clearly had the option of hiring non-certified auditors plus one-third of class 2 with the smallest salaries. The companies with the smallest salaries are the ones most likely to have fewer than 10 employees and consequently may belong to the group of firms that can be audited by non-certified auditors. This results in a sample with 592 companies. The results are qualitatively similar to those reported in Table 5. Both SALARIES and LNSIZE have a positive coefficient significant at p < 0.029 and p < 0.001 respectively in one-tailed tests. Also, given the negative correlation between HERFINDAHL and ROA-ADJ, we re-ran the analysis with one or the other variables but not both. The results are essentially the same as the model with both variables included simultaneously.

To examine whether the influence of complexity differs across the types of certified auditors, we estimated a multinomial regression using the sample from Model 1. For this analysis, the firms audited by each of the four auditor types were assigned to separate groups.22 These analyses (not reported) show that SALARIES are highly significant in regressions comparing the non-certified auditors and HTM auditors, but are mostly insignificant in regressions comparing HTM and KHT auditors and HTM and Big Six auditors. Taken together, these results support H1 for the smallest firms that have the option of using a non-certified auditor. Financing, agency and competition issues are less important for these companies.

Model 2

Model 2 considers the choice between a KHT and an HTM auditor. The sample for Model 2 includes all firms in size class 3 in Table 2. The sample size is 1,626 of which 527 companies have an HTM auditor, 401 companies a non-Big Six KHT auditor, and 698 companies a Big Six auditor. The mean (median) assets of these firms is €3.4 million (€1.7 million). The correlations among the independent variables in Model 2 are generally low, e.g., LIAB and UNSECURED have a correlation of 0.53, HERFINDAHL and ROA-ADJ have a correlation of −0.41, and UNSECURED and INVREC have a correlation of 0.31 (see Table 4). The results for Model 2 are reported in Table 5. The model is highly significant with a χ2 value of 259.42 (p < 0.001). LIAB, GROUP and LNSIZE have the expected positive coefficients and are highly significant in the model (p < 0.003 in all cases). These results support H1 and H3.

Model 2 was also re-run as a robustness check using class 3 plus one-third of class 2 with the highest salaries and one-third of class 4 with the lowest salaries. This increases the sample to 1,714 firms and the results (not reported) are qualitatively similar to those reported in Table 5.23 Given the correlation between LIAB and UNSECURED, we also performed a sensitivity analysis if one or the other variables is excluded from the model. LIAB is still significant in Model 2 (p < 0.01) if UNSECURED is dropped, and UNSECURED remains insignificant if LIAB is dropped. Finally, since HERFINDAHL and ROA-ADJ are negatively correlated, we re-ran the analysis with one of the other variables but not both. The results are essentially the same as the model with both variables included simultaneously.

To examine whether the influence of complexity differs across the types of certified auditors, we estimated a multinomial regression using the sample from Model 2. For this analysis, the firms audited by each of the three certified auditor types were assigned to separate groups. These analyses (not reported) show that LNSIZE, LIAB and GROUP have higher coefficients in the regressions comparing Big Six auditors to HTM auditors than in regressions comparing non-Big Six KHT auditors to HTM auditors, indicating that higher value of the variables are associated with higher quality auditors.24

Model 3

The sample for Model 3 includes class 5 in Table 2, i.e., firms that had to choose between a non-Big Six KHT and a Big Six auditor. The sample size is 195, of which 15 were audited by a non-Big Six KHT auditor. The mean (median) assets of these firms is €72.9 million (€20.7 million). The correlations among the independent variables in Model 3 are again generally low, e.g., LIAB and UNSECURED have a correlation of 0.34 and HERFINDAHL and ROA-ADJ have a correlation of −0.25.25 The results are reported in Table 5. The model's χ2 is 20.3 (p < 0.063). LNSIZE (p < 0.016), SALARIES (p < 0.062), LIAB (p < 0.051) and SHRGROW (p < 0.001) have the predicted positive coefficients and are significant. EXANTEFIN is also significant (p < 0.056) but the sign is opposite to what we expected. Thus, these results provide partial support for H1, H3 and H4, but counter to H2.

To extend the analysis of the choice of Big Six, we expand the sample for Model 3 to include all companies that were actually audited by a KHT auditor in the size classes 1, 3 or 5 in Table 2 (not reported). This sample reflects the firms already predisposed to having a KHT audit so we can test the differences for companies that chose Big Six firms rather than non-Big Six KHT firms. The sample consists of 1,459 firms, of which 964 were audited by a Big Six KHT auditor. The mean (median) assets of these companies is €12.9 million (€2.3 million). PUBLIC, LNSIZE, GROUP and LIAB have the expected positive coefficients (p < 0.05). EXANTEFIN is no longer significant. Thus, among all companies that hire a KHT auditor, larger publicly-traded firms are, as one would surmise, more likely to hire a Big Six firm.26

Supplemental tests

Given the potential measurement error in some of our test variables, we conducted a number of supplemental tests. Table 5 shows that there is a positive relationship between LIAB and the choice of a higher quality auditor in some models. Slightly more than the half of the firms in the sub-sample used for Model 1 had no bank debt so we re-estimated Model 1 using only the 264 firms having bank debt. We observe that the coefficient of LIAB is 1.27 (p < 0.03, one-tail test) which is considerably higher than in Model 1. If lenders appreciate a high quality auditor, one would expect that firms whose debt is increasing would be more likely to switch to a higher quality auditor. The average change in liabilities is 3.2% for the 2,234 firms that did not switch auditors and 8.8% for the 99 firms that switched to a higher quality auditor (p < 0.20).27 Furthermore, approximately 10%(236/2,333) of the companies had liabilities that increased by more than 30%; of these, 6.8% (16/236) switched to a higher quality auditor. The switching rate for firms that did not have a rapid increase in liabilities was 4% (83/2,097). A Pearson χ2 test shows that the difference in the fractions is significant (p < 0.041).

Using SALARIES as a measure of complexity, we found support for H1 among the smallest firms in the sample. To test whether the result is sensitive to the way we measure our proxy, we re-estimated the models using salaries scaled by total assets. This measure is positive and significant at the 1% level in Models 1 and 2. The coefficient is not significantly different from zero in Model 3. Thus, this measure provides stronger support for H1 in Models 1 and 2, but weaker support in Model 3, than the measure used in the main analyses and is more in line with our expectation that complexity as measured by SALARIES is most important for smaller firms.

The results reported in Table 5 for equity financing do not strongly support H4 concerning equity-related agency problems. To study whether the results are sensitive to our measure of equity growth, we replaced SHRGROW with a dummy variable taking the value 1 if share-capital increased during the two years after the year of data used in this study (suggesting preparation for a planned equity offering). This variable could be calculated for 2,018 firms and is insignificant in all models. Another possibility is to test whether firms that actually issue equity are more likely to switch to higher quality auditors. However, the rate of switches between the categories is very similar: 3.1% of the firms that issued equity switched to a higher quality auditor versus 4.3% of firms that did not issue equity. The corresponding percentages for companies issuing equity in the next two years are 3.8% and 4%. A possible explanation is that a high proportion of share offerings in our sample are not public offerings, but are made to parties that are already well informed about the firm's quality and prospects (e.g., family members or existing business associates).

Our results for the general need for financing also did not support our expectations (H2). Using EXANTEFIN as the measure, we found weak support for the prediction that firms with a higher demand for external financing would be audited by higher quality auditors in Model 3 but no support in Models 1 and 2. The variable used was based on Dechow et al. (1996); however, we used a continuous variable and they used a dummy variable taking the value 1 if the cash flows from operations less net investments in relation to current assets was less than −0.5. Investments were calculated as an average over three years. The dummy variable used by Dechow et al. (1996) could be calculated for 1,844 firms. The variable is insignificant in Models 1 and 2, while Model 3 could not be estimated since the variable took the value zero for all companies audited by non-Big Six KHT auditors. The ratio was insignificant when the cut-off was adjusted upward to 0. In conclusion, the supplemental results give little support for the prediction that the demand for financing affects auditor choice.

We found no support for H5 that competition in an industry would affect auditor choice as measured using the Herfindahl index for concentration and ROA-ADJ as an empirical measure of the sustainability of abnormal profits.28 A potential reason for the lack of association between the Herfindahl index and the choice of auditor is that the effect of concentration on competition is ambiguous as discussed in Section 3. The fact that ROA-ADJ and HERFINDAHL are negatively correlated in all samples may support this ambiguity, i.e., the correlations in Models 1, 2 and 3 are −0.419, −0.408 and −0.246, respectively. If both variables capture the degree of competition, one would expect that they would be positively correlated. One possible interpretation of the correlation is that competition on average increases with higher industry concentration as suggested by Stiglitz (1987) or the metrics lack construct validity as suggested by Dedman and Lennox (2007).

As a further test of the proprietary cost hypothesis, we replaced HERFINDAHL with a concentration ratio (CONCENTRATION) as specified in Harris (1998) and Botosan and Stanford (2005). The results using CONCENTRATION generally support H5 since the ratio was positive in all three models and significant at the 0.10 level in Model 1, at the 0.01 level in Model 2, and at the 0.05 level in Model 3 (all one-tailed tests). We also replaced ROA-ADJ with the firm's actual return on assets (ROA). If more profitable firms have stronger incentives to conceal information about the true profitability of the firm and a low quality auditor gives them better possibilities to do so, one would expect a negative association between the choice of a high quality auditor and the profitability. ROA is negative in Model 1 (p < 0.10, two-tailed), insignificant in Model 2, and positive and significant in Model 3 (p < 0.10, two-tailed). Overall, these results provide some support for H5 in the sub-sample of the smallest firms. A possible reason for the positive sign of ROA in Model 3 is that agency concerns are relatively more important than proprietary costs among large firms (cf. Berger & Hann, 2007).

6. DISCUSSION AND CONCLUSIONS

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

The benefits of acquiring an audit are multi-faceted and the value of these benefits is likely to vary with firm size. In the narrowest sense, the role of auditing is to improve the quality of financial statements since high quality reporting can reduce information asymmetry problems between the firm and providers of financing. Thus, a potential role of auditing is that it reduces the cost of capital and improves the availability of capital (Blackwell et al., 1998; Pitman & Fortin, 2004). Audits also provide a number of benefits that are internal to the company. For example, the results of an audit can reduce internal agency problems, lead to improvements in process effectiveness, and assist in regulatory compliance. However, high quality auditing and reporting can potentially have adverse effects in the sense that they improve the quality of information that is disclosed to competitors.

In this paper we set forth two hypotheses related to the internal benefits of auditing (H1, H2), two hypotheses related to external agency conflicts (H3, H4), and one hypothesis related to the cost of disclosing high quality information to potential competitors (H5). The hypotheses were tested on three samples: (1) very small companies that could be audited by non-professional firms, (2) firms that had to use a certified auditor (HTM or KHT) auditor, and (3) large firms that had to use a KHT auditor (Big Six or non-Big Six KHT). Among the smallest firms, it was found that LNSIZE and SALARIES were significant. Larger firms and firms with higher salaries in relation to the total operating expenses were more likely to hire a certified auditor. Firms with more employees would need access to expertise in order to manage internal agency problems, comply with regulations, and obtain feedback on control systems. We also found that there were no significant differences in salaries between the different types of certified auditors (i.e., HTM, KHT and Big Six). This result indicates that the decision whether to hire a certified or non-certified auditor is the most important choice made by small firms. We did not find any support for the hypotheses related to the role of auditing in transactions with debt or equity holders among the smallest firms.

For the slightly larger firms that had to choose between an HTM and KHT auditor, the results support the complexity hypothesis as measured by GROUP (H1) and leverage hypothesis (H3). We found that the significance of leverage is increasing with the size of the business. This may indicate that banks and other providers of debt rely more on financial statements, and value a high quality auditor, as the size of the business grows because they may be less likely to require personal guarantees and collateral from the owners. We did not find any support for the hypotheses that firms planning to issue equity securities or firms with a high demand for external financing would be more likely to hire high quality auditors. A possible reason for this is that the issuance of equity to outsiders is a relatively remote occurrence among the class of firms considered in Model 2.

Among the largest firms that had to choose between a KHT and Big Six auditor, we found support for the complexity hypothesis (H1), leverage hypothesis (H3) and the equity hypothesis (H4). While the choice of auditor by larger companies continues to be influenced by leverage, the primary difference for the largest companies in the sample is that public firms are more likely to be audited by Big Six auditors. A possible reason for this is that the ownership of public firms is more dispersed than that of private firms. While we have no data on company ownership, smaller privately-held firms are probably family-run and have little separation of ownership and control. However, other factors might contribute to the higher probability of having a Big Six firm audit listed companies: (1) the expectations and requirements for financial reporting are more demanding for listed firms, necessitating greater levels of consultation among professionals, (2) firms may also have operations outside Finland, resulting in demand for services from the international network of Big Six auditors, and (3) IPO organizers might require that the firm choose a Big Six auditor.29

Thus, the overall pattern of these results taken together suggests that firm complexity affects all levels of auditor decisions in one way or another. As firms grow, complexity is supplemented first by the need for debt financing, and eventually for the largest companies, the need for equity.

However, the results reported in this paper are susceptible to a number of important limitations. First, the classification of some of the firms in the sample may depend on the number of employees in a firm which is unobservable for our study. Second, an important issue that might influence our result is endogeneity. This might arise if the pairing of a client and auditor has as much to do with the preferences of the audit firm as it does with the attributes of the client. We are unable to directly test this possibility but argue that the sequence of steps in selecting an auditor may alleviate the problem a bit. Given that auditors decide which markets/segments/industries in which to specialize, the auditor selection process is then in the hands of the clients, who presumably choose the best of the available firms. Since Finland is a low litigation country, and most of the firms in the sample are small, we do not feel that auditors will avoid specific clients for the purposes of managing their own portfolio risk. A further limitation of the paper is the difficulties with measuring the degree of competition and the incentive to hire low quality auditors in order to improve the possibilities of withholding information from competitors. Thus, even if the results in the paper gave little support for H5, it is possible that more refined measures would give other results, suggesting the need for further research.

In summary, the results of our analysis of auditor choice across four types of auditors and three different sizes of companies reveals some very interesting overall patterns. First, the choice of auditor by the smallest companies is driven most by issues of complexity and internal operations. Second, mid-size companies choose auditors based on a need for debt financing. Finally, the largest firms are also influenced by leverage (particularly unsecured debt) but the dominant driver of auditor choice is the public listing of equity securities. In general, these results are consistent with our expectations and empirical evidence in prior studies that examined these issues in isolation. The contribution of this paper is to demonstrate that the auditor choice follows hypothesized patterns of behaviour across a range of companies, auditors and conditions.

ACKNOWLEDGMENTS

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

The authors would like to thank Allen Blay, Juha Kinnunen, Peter Moizer, Stuart Turley, Risto Walden and participants at the workshops in the Fisher School of Accounting, University of Florida and at the Department of Accounting and Finance at Helsinki School of Economics, the EARNet Symposium in Manchester 2003, the EAA conference in Prague 2004, and the AAA meeting in San Francisco 2005 for helpful comments.

NOTES
  • 1

    Another example of auditing facilitating market permissions is assurance that firms are in compliance with ISO 9000 which allows international trading relationships to develop.

  • 2

    In Finland, the requirements of the Eighth Directive regarding regulation of statutory auditors are implemented through the Auditing Act (936/1994). EU Member States are moving from national regulation to the adoption of international standards. By 2005, listed companies in the EU have to adopt the use of IFRSs (International Financial Reporting Standard). In May 2003, the European Commission issued a proposal that ISAs (International Standards on Auditing) issued by IAASB (International Auditing and Assurance Standards Boards) be adopted as the auditing standards in the EU (European Commission, 2003).

  • 3

    Non-certified auditors are not required to follow international auditing standards but they are required to follow the rules of the Auditing Act, which are generally considered to be less substantive than international standards.

  • 4

    As is discussed later in the paper, the evidence that Big Six auditors reduce earnings management is mixed in Europe (Vander Bauwhede & Willekens, 2004).

  • 5

    There is also some evidence that size correlates with audit quality even among smaller accounting firms (Deis & Giroux, 1992; Colbert & Murray, 1998; Casterella et al., 2003).

  • 6

    Hay and Davis (2004) examined a sample of incorporated, non-profit societies in New Zealand for which an audit was voluntary.

  • 7

    Some of the results observed in a similar code law country such as Belgium may not transfer to Finland because the threshold for mandatory auditing is much lower than in Belgium (€3.125 million vs. €0.34 million).

  • 8

    Also, Ge and McVay (2005) report revenue recognition as the fourth most common source of material weaknesses in internal control over financial reporting.

  • 9

    Another potential measure of complexity suggested by the extensive literature on audit fees is the existence and extent of multiple divisions, subsidiaries and business units. However, in a small client market, the existence of such organizational components is rare and is not generally disclosed in the types of reports used in this study.

  • 10

    Dechow et al. (1996) used the average capital expenditures during three years. To avoid losing observations we use the capital expenditures during one year in the main analyses; however, we test a variable based on the three-year average of capital expenditures as a robustness check. Further, Dechow et al. (1996) used a dummy variable taking the value 1 if the ratio was less than 0.5 and 0 otherwise. We use the continuous variable since there was not sufficient variation in the variable to estimate Model 3 in Table 5. In our supplemental tests we report results using the dummy variable in Dechow et al. (1996). We also tested a rank transformation of our measure of free cash flow and the results are qualitatively consistent with the square root transformation used in the main analysis of the paper.

  • 11

    Data about the actual dispersion of ownership is not available in the database used for this study. Since the vast majority of organizations included in the study is small and family owned, the population should be relatively homogeneous on this dimension, except for the publicly traded companies.

  • 12

    Theoretical studies show that the decision to disclose information depends on the actual context, including whether good or bad news are about to be disclosed and the type of competition (Cournot or Bertrand). The results in the empirical literature are also mixed (see Ackert et al., 2000). Based on the expectation that excess profits are more likely in less competitive markets, Harris (1998) predicted that firms in more competitive industries disclose less segment information.

  • 13

    Contrary to Verrechia (1990), Darrough and Stoughton (1990) suggest that competition encourages firms to disclose information. The difference between the studies is that Darrough and Stoughton (1990) focus on the threat of entry, that is, potential competition rather than actual competition. The empirical measures used in this paper are aimed to measure the actual competition.

  • 14

    We used the sample data for 1998 before any omissions because of missing auditor data or missing variables for the calculations. This data includes 6,608 observations. The year 1998 was selected since it is in the middle of our sampling period and because the data includes the largest number of observations for this year. Industries with fewer than ten companies were omitted.

  • 15

    Return on assets is defined as the operating profit plus interest and dividend income, in relation to total assets.

  • 16

    The data for four of the companies is from 1996, 101 from 1997, 1,844 from 1998 and 666 from 1999.

  • 17

    SALEGROW was replaced with 7.70 if its value exceeded 7.70 and with −0.76 if the value was below −0.76. LIAB was replaced with 1.57 if its value exceeded 1.57 and with the value 0.022 if the value was below 0.022.

  • 18

    Only the largest firms can be PUBLIC given the manner in which the subsamples are constructed.

  • 19

    All public companies must have a KHT auditor so the test statistic is between two groups, KHT and B6KHT.

  • 20

    The rationality of the decision of a small firm is likely to be ‘bounded’ in the sense that with relatively small costs and benefits at stake, it is not efficient to spend large resources on considering advantages and disadvantages with an auditor choice (for a review of literature on bounded rationality, see Conlisk, 1996).

  • 21

    We report the results from a number of supplemental tests using different sample specifications later in the paper. In general, the main results are not dependent on the sample selection.

  • 22

    In this regression, the dependent variable took different values depending on whether the firm was audited by a non-certified, HTM, KHT or Big Six auditor. We assigned firms audited by HTM auditors to the base category. Thus, the regression studies differences between firms audited by HTM auditors and the other auditor categories. See Hosmer and Lemeshow (1989: 216–38) for a discussion of the multinominal logistic regression model.

  • 23

    Six firms in size class 2 that were audited by non-professional auditors were excluded.

  • 24

    The coefficient (p-value) of LIAB was 0.32 (0.32) in the regression comparing non-Big 6 KHT auditors to HTM auditors and 1.08 (<0.01) in the regression comparing Big 6 auditors to HTM auditors. The corresponding coefficients (p-values) for GROUP are 1.28 (<0.01) and 1.54 (<0.01). The coefficients and p-values for LNSIZE are 0.42 (<0.01) and 0.81 (<0.01). p-values are here for two-tailed tests.

  • 25

    As with Models 1 and 2, running models with just HERFINDAHL or ROA-ADJ does not change the results.

  • 26

    The results were qualitatively similar when the companies assigned to class 2 or class 4 in Table 2 were added to the sample (not reported in tables).

  • 27

    A case was coded as a switch to a higher quality auditor if it switched from a non-certified auditor to a certified auditor, if it switched from an HTM to a KHT or from a non-Big Six KHT to a Big Six auditor. The change in liabilities was calculated as the change in total liabilities scaled by lagged total assets.

  • 28

    We also tested whether the results for the Herfindahl index for the concentration of sales is driven by the concentration in industries with very few companies by excluding those industries in which there were fewer than 100 companies. This resulted in an omission of 202 companies. The index was insignificant in Models 1, 2 and 3 as in the main analysis.

  • 29

    There are eight companies in the sample that are either the parent company or a subsidiary to a company that went public during the year that the company is included in the data. All these are from 1998. There were altogether 36 IPOs in Finland in the period 1996–1999 (Spohr, 2004).

REFERENCES

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES
  • Abdel-Khalik, A. R. (1993), ‘Why do private companies demand auditing? A case for organizational loss of control’, Journal of Accounting, Auditing and Finance, Vol. 8, No. 1, pp. 3152.
  • Ackert, L. F., Church, B. K. & Sankar, M. R. (2000), ‘Voluntary disclosure under imperfect competition: Experimental evidence’, International Journal of Industrial Organization, Vol. 18, No. 1, pp. 81105.
  • Baker, C. R., Mikol, A. & Quick, R. (2001), ‘Regulation of the statutory auditor in the European Union: A comparative survey of the United Kingdom, France and Germany’, The European Accounting Review, Vol. 10, No. 4, pp. 76386.
  • Ball, R. & Shivakumar, L. (2005), ‘Earnings quality in UK private firms: Comparative loss recognition timeliness’, Journal of Accounting and Economics, Vol. 39, No. 1, pp. 83128.
  • Baumol, W., Panzar, J. & Willig, R. (1988), Contestable Markets and the Theory of Industry Structure, Revised Edn, San Diego, CA: Harcourt Brace Jovanovich.
  • Beasley, M. S. & Petroni, K. R. (2001), ‘Board independence and audit-firm type’, Auditing: A Journal of Practice and Theory, Vol. 20, No. 1, pp. 97114.
  • Beatty, R. (1989), ‘Auditor reputation and the pricing of initial public offerings’, The Accounting Review, Vol. 64, October, pp. 693709.
  • Becker, C., DeFond, M., Jiambalvo, J. & Subramanyam, K. (1998), ‘The effect of audit quality on earnings management’, Contemporary Accounting Research, Vol. 15, pp. 124.
  • Behn, B. K., Carcello, J. V., Hermanson, D. R. & Hermanson, R. H. (1997), ‘The determinants of audit client satisfaction among clients of Big Six firms’, Accounting Horizons, Vol. 11, pp. 724.
  • Berger, P. & Hann, R. (2007), ‘Segment profitability and the proprietary and agency costs of disclosure. The Accounting Review, Vol. 82, No. 4, pp. 869906.
  • Bergström, C., Eisenberg, T. & Sundgren, S. (2004), ‘On the design of efficient priority rules for secured creditors: Empirical evidence from a change in law’, European Journal of Law and Economics, Vol. 18, No. 3, pp. 27397.
  • Blackwell, D. W., Noland, T. R. & Winters, D. B. (1998), ‘The value of auditor assurance: Evidence from loan pricing’, Journal of Accounting Research, Vol. 36, No. 1, pp. 5770.
  • Botosan, C. (1997), ‘Disclosure level and the cost of equity capital’, Accounting Review, Vol. 72, No. 3, pp. 32349.
  • Botosan, C. & Stanford, M. (2005), ‘Managers' motives to withhold segment disclosures and the effect of SFAS No. 131 on analysts' information environment’, The Accounting Review, Vol. 80, pp. 75172.
  • Carey, P., Simnett, R. & Tanewski, G. (2000), ‘Voluntary demand for internal and external auditing by family businesses’, Auditing: A Journal of Practice and Theory, Vol. 19, Suppl., pp. 3751.
  • Casterella, J., Jensen, K. & Knechel, W. R. (2003), ‘Audit litigation in regional and local audit firms: An empirical study of professional malpractice claims’, Working Paper, University of Florida.
  • Causholli, M. & Knechel, W. R. (2007), ‘Lending relationships, IPOs, and the effect of auditor quality on the cost of debt’, Working paper, University of Florida.
  • Chaney, P., Jeter, D. & Shivakumar, L. (2004), ‘Self-selection of auditors and audit pricing in private firms’, The Accounting Review, Vol. 79, No. 1, pp. 5172.
  • Chow, C. W. (1982), ‘The demand for external auditing: Size, debt and ownership influences’, The Accounting Review, Vol. 57, No. 2, pp. 27291.
  • Colbert, G. & Murray, D. (1998), ‘The association between auditor quality and auditor size: An analysis of small CPA firms’, Journal of Accounting, Auditing and Finance, Vol. 13, No. 2, pp. 13550.
  • Conlisk, J. (1996), ‘Why bounded rationality? Journal of Economic Literature, Vol. 34, June, pp. 669700.
  • Darrough, M. N. & Stoughton, N. M. (1990), ‘Financial disclosure policy in an entry game’, Journal of Accounting and Economics, Vol. 12, Nos. 1–3, pp. 21943.
  • Das, S. & Sengupta, S. (2001), ‘Asymmetric information, bargaining, and international mergers’, Journal of Economics and Management Strategy, Vol. 10, No. 4, pp. 56590.
  • DeAngelo, L. E. (1981), ‘Auditor size and audit quality’, Journal of Accounting and Economics, Vol. 3, No. 3, pp. 18399.
  • Dechow, P., Sloan, R. & Sweeney, A. (1996), ‘Causes and consequences of earnings management: An analysis of firms subject to enforcement actions by SEC’, Contemporary Accounting Research, Vol. 13, No. 1, pp. 136.
  • Dechow, P., Hutton, A. &. Sloan, R. (1999), ‘An empirical assessment of the residual income valuation model’, Journal of Accounting and Economics, Vol. 26, Nos. 1–3, pp. 134.
  • Dedman, E. & Lennox, C. (2007), ‘Managers' perceptions of product market competition, profitability and firms' voluntary disclosures of sales', Working paper, Hong Kong University of Science and Technology.
  • DeFond, M. L. (1992), ‘The association between changes in client firm agency costs and auditor switching’, Auditing: A Journal of Practice and Theory, Vol. 11, No. 1, pp. 1631.
  • DeFond, M. L. & Jiambalvo, J. (1994), ‘Debt covenant violation and manipulation of accruals’, Journal of Accounting and Economics, Vol. 17, Nos. 1–2, pp. 14576.
  • Deis, D. R. Jr. & Giroux, G. A. (1992), ‘Determinants of audit quality in the public sector’, The Accounting Review, Vol. 67, July, pp. 46279.
  • Doupnik, T. S. & Salter, S. B. (1993), ‘An empirical test of a judgmental international classification of financial reporting practices’, Journal of International Business Studies, Vol. 24, No. 1, pp. 4160.
  • Duke, J. & Hunt, H. (1990), ‘An empirical investigation of debt covenant restrictions and accounting related debt proxies’, Journal of Accounting and Economics, Vol. 12, Nos. 1–3, pp. 4563.
  • Dulleck, W. & Kerschbamer, R. (2006), ‘On doctors, mechanics and computer specialists’, Journal of Economic Literature, Vol. 44, No. 1, pp. 542.
  • European Commission (1996), Green Paper: The Role, the Position and the Liability of the Statutory Auditor within the European Union, Brussels: Commision of the European Communities.
  • European Commission (1998), Communication from the Commission: The Statutory Audit in the European Union: the Way Forward, Brussels: Commision of the European Communities.
  • European Commission (2000), Commission Recommedation: Quality Assurance for the Statutory Audit in the EU: Minimum, Brussels: Commision of the European Communities.
  • European Commission (2003), Communication from the Commission: Reinforcing the Statutory Audit in the European Union, Brussels: Commision of the European Communities.
  • Francis, J. R. & Simon, D. T. (1987), ‘A test of audit pricing in the small-client segment of the US audit markets’, The Accounting Review, Vol. 62, January, pp. 14557.
  • Francis, J. R. & Wilson, E. R. (1988), ‘Auditor changes: A joint test of theories relating to agency costs and auditor differentiation’, The Accounting Review, Vol. 63, No. 4, pp. 66382.
  • Francis, J. R., Maydew, E. L. & Sparks, H. C. (1999), ‘The role of Big Six auditors in the credible reporting of accruals’, Auditing: A Journal of Practice and Theory, Vol. 18, No. 2, pp. 1734.
  • Gaeremynck, A. & Willekens, M. (2003), ‘The endogenous relationship between audit-report type and business termination: Evidence on private firms in a non-litigious environment’, Accounting and Business Research, Vol. 33, No. 1, pp. 6579.
  • Ge, W. & McVay, S. (2005), ‘The disclosure of material weaknesses in internal control after the Sarbanes-Oxley Act’, Accounting Horizons, Vol. 19, September, pp. 13758.
  • Grant, R. (1998), Contemporary Strategy Analysis – Concepts, Techniques, Applications, (3rd edn), Malden, MA: Blackwell Publishers.
  • Gul, F. A. & Tsui, J. S. L. (2001), ‘Free cash flow, debt monitoring, and audit pricing: Further evidence on the role of director equity ownership’, Auditing: A Journal of Practice and Theory, Vol. 20, No. 2, pp. 7184.
  • Harris, M. (1998), ‘The association between competition and managers' business segment reporting decision’, Journal of Accounting Research, Vol. 36, No. 1, pp. 11128.
  • Hay, D. & Davis, D. (2004), ‘The voluntary choice of an audit of any level of quality’, Auditing: A Journal of Practice and Theory, Vol. 23, September, pp. 3753.
  • Hay, D. & Knechel, W. R. (2005), ‘Evidence on the association among elements of control and external assurance’, Unpublished working paper, University of Auckland.
  • Hay, D. & Knechel, W. R. (2007), ‘The effects of advertising and solicitation on audit pricing’, Unpublished working paper, University of Auckland.
  • Hay, D., Knechel, W. R. & Wong, N. (2006), ‘Audit fees: A meta-analysis of the effect of supply and demand attributes’, Contemporary Accounting Research, Vol. 23, Spring, pp. 14192.
  • Hosmer, D. & Lemeshow, S. (1989), Applied Logistic Regression, New York: John Wiley & Sons.
  • Jaggi, B. & Low, P. Y. (2000), ‘Impact of culture, market forces, and legal system on financial disclosures’, The International Journal of Accounting, Vol. 35, No. 4, pp. 495519.
  • Jensen, C. M. & Meckling, H. W. (1976), ‘Theory of the firm: Managerial behavior, agency costs and ownership structure’, Journal of Financial Economics, Vol. 3, No. 4, pp. 30560.
  • Johnson, W. B. & Lys, T. (1990), ‘The market for audit services: Evidence from voluntary auditor changes’, Journal of Accounting and Economics, Vol. 12, Nos. 1–3, pp. 281308.
  • Kinney, J. R. Jr. & McDaniel, L. S. (1989), ‘Characteristics of firms correcting previously reported quarterly earnings’, Journal of Accounting and Economics, Vol. 11, No. 1, pp. 7193.
  • Knechel, W. R. (2001), Auditing: Assurance and Risk, 2nd edn, Cincinnati, OH: SouthWestern College Publishing.
  • Knechel, W. R. (2002), ‘The role of the independent accountant in effective risk management’, Tijdschrift voor Economie en Management, Vol. XI, pp. 6586.
  • Knechel, W. R. & Vanstraelen, A. (2007), ‘The relationship between auditor tenure and audit quality implied by going concern opinions’, Auditing: A Journal of Practice and Theory, Vol. 26, No. 1, pp. 11331.
  • Krishnan, G. V. (2003), ‘Audit quality and the pricing of discretionary accruals’, Auditing: A Journal of Practice and Theory, Vol. 22, No. 1, pp. 10926.
  • Krishnan, J. & Schauer, P. C. (2000), ‘The differentiation of quality among auditors: Evidence from the not-for-profit sector’, Auditing: A Journal of Practice and Theory, Vol. 19, No. 2, pp. 925.
  • LaPorta, R., Lopez-de-Silanes, F., Shleifer, A. & Vishny, R. W. (1998), ‘Law and finance’, Journal of Political Economy, Vol. 106, pp. 111355.
  • Lennox, C. (1999), ‘Are large auditors more accurate than small auditors? Accounting and Business Research, Vol. 29, No. 3, pp. 21727.
  • Loft, A. & Jeppesen, K. K. (2001), ‘Harmonizing the education and qualification of auditors? Implementing the EU's Eighth Directive in Denmark’, Proceedings of First European Auditing Research Network Symposium, University of Wuppertal and University of Essen, pp. 5581.
  • Loughran, T. & Ritter, J. R. (1997), ‘The operating performance of firms conducting seasoned equity offerings’, Journal of Finance, Vol. 52, December, pp. 182350.
  • Mansi, S. A., Maxwell, W. F. & Miller, D. P. (2004), ‘Does auditor quality and tenure matter to investors? Evidence from the bond market’, Journal of Accounting Research, Vol. 42, No. 4, pp. 75593.
  • Margerison, J. & Moizer, P. (1996), ‘Auditor licensing in The European Union: A comparative study based on cultural differences’, The European Accounting Review, Vol. 5, No. 1, pp. 2956.
  • Moizer, P. (1997), ‘Auditor reputation: The international empirical evidence’, International Journal of Auditing, Vol. 1, No. 1, pp. 6174.
    Direct Link:
  • Newton, J. D. & Ashton, R. H. (1989), ‘The association between audit report technology and audit delay’, Auditing: A Journal of Practice and Theory, Vol. 8, Suppl., pp. 2237.
  • Niemi, L. (2002), ‘Do firms pay for audit risk? Evidence on risk premiums in audit fees after direct control for audit effort’, International Journal of Auditing, Vol. 6, No. 1, pp. 3751.
  • Niemi, L. (2004), ‘Auditor size and audit pricing: Evidence from small audit firms’, European Accounting Review, Vol. 13, No. 3, pp. 54160.
  • Niemi, L. (2005), ‘Audit effort and fees under concentrated client ownership: Evidence from four international audit firms’, The International Journal of Accounting, Vol. 40, No. 4, pp. 30323.
  • Niemi, L. & Sundgren, S. (2008), Auditing, Trust & Governance – Regulation in Europe (eds. Quick, R., S.Turley, and M.Willekens), London: Routledge.
  • Nobes, C. W. (1983), ‘A judgmental international classification of financial reporting practices’, Journal of Business, Finance & Accounting, Vol. 10, No. 1, pp. 119.
  • Ohlson, J. (1995), ‘Earnings, book values and dividends in equity valuation’, Contemporary Accounting Research, Vol. 11, No. 2, pp. 66187.
  • Palmrose, Z.-V. (1988), ‘An analysis of auditor litigation and audit service quality’, The Accounting Review, Vol. 63, No. 1, pp. 5573.
  • Pitman, J. A. & Fortin, S. (2004), ‘Auditor choice and the cost of debt capital for newly public firms’, Journal of Accounting and Economics, Vol. 37, No. 1, pp. 11336.
  • Porter, M. E. (1980), Competitive Strategy: Techniques for Analyzing Industries and Competitors. New York: Free Press.
  • Saarikivi, M.-L. (1999), ‘Tilintarkastajan riippumattomuus’ (Auditor independence, in Finnish). Doctoral thesis, No A-155. Helsinki School of Economics.
  • Sercu, P., Vander Bauwhede, H. & Willekens, M. (2002), ‘Earnings quality in privately held firms: The roles of audits, stakeholders and governance mechanisms’, unpublished working paper, KULeuven.
  • Simunic, D. A. & Stein, M. T. (1987), Product Differentiation in Auditing: Auditor Choice in the Market for Unseasoned New Issues, Vancouver, BC: The Canadian Certified General Accountants' Research Foundation.
  • Smith, C. & Warner, J. (1979), ‘On financial contracting: An analysis of bond covenants’, Journal of Financial Economics, Vol. 7, No. 2, pp. 11761.
  • Spohr, J. (2004), ‘Earnings management and IPOs – evidence from Finland’, The Finnish Journal of Business Economics, Vol. 53, No. 2, pp. 15772.
  • Stice, J. D. (1991), ‘Using financial and market information to identify pre-engagement factors associated with lawsuits against auditors’, The Accounting Review, Vol. 66, No. 3, pp. 51634.
  • Stiglitz, J. (1987), ‘Competition and the number of firms in a market: Are duopolies more competitive than atomistic markets? Journal of Political Economy, Vol. 95, No. 5, pp. 104161.
  • Sundem, G. L., Dukes, R. E. & Elliott, J. A. (1996), The Value of Information and Audits, New York: Coopers & Lybrand.
  • Sundgren, S. (1998), ‘Auditor choices and auditor reporting practices: Evidence from Finnish small firms’, European Accounting Review, Vol. 7, No. 3, pp. 44165.
  • Sundgren, S. & Johansson, C. (2004), ‘The effects of the auditor's professional qualification and the firm's financial health on depreciation in Finland’, Accounting and Business Research, Vol. 34, No. 2, pp. 12543.
  • Teoh, S. H. & Wong, J. J. (1993), ‘Perceived auditor quality and the earnings response coefficient’, The Accounting Review, Vol. 68, April, pp. 34666.
  • Vander Bauwhede, H. & Willekens, M. (2004), ‘Evidence on (the lack of) audit-quality differentiation in the private client segment of the Belgian audit market’, European Accounting Review, Vol. 13, No. 3, pp. 50122.
  • Vander Bauwhede, H., Willekens, M. & Gaeremynck, A. (2003), ‘Audit firm size, public ownership, and firms' discretionary accruals management’, The International Journal of Accounting, Vol. 38, No. 1, pp. 122.
  • Vanstraelen, A. (2002), ‘Auditor economic incentives and going-concern opinions in a limited litigious Continental European business environment: Empirical evidence from Belgium’, Accounting and Business Research, Vol. 32, No. 3, pp. 17186.
  • Verrechia, R. E. (1990), ‘Endogeneous proprietary costs through firm interdependence’, Journal of Accounting and Economics, Vol. 12, No. 4, pp. 24550.
  • Verrechia, R. E. & Weber, J. (2006), ‘Redacted disclosure’, Journal of Accounting Research, Vol. 44, No. 4, pp. 791814.
  • Wallace, W. (1981), The Economic Role of the Audit in Free and Regulated Markets, http://raw.rutgers.edu/raw/wallace/homepage.html
  • Watts, R. L. & Zimmerman, J. L. (1986). Positive Accounting Theory, Englewood Cliffs, NJ: Prentice-Hall.
  • Willekens, M. & Achmadi, C. (2003), ‘Pricing and supplier concentration in the private client segment of the audit market: Market power or competition? The International Journal of Accounting, Vol. 38, No. 4, pp. 43155.

AUTHOR PROFILES

  1. Top of page
  2. Abstract
  3. SUMMARY
  4. I. INTRODUCTION
  5. 2. AUDITING ENVIRONMENT IN FINLAND
  6. 3. THEORETICAL BACKGROUND AND PRIOR LITERATURE ON AUDITOR CHOICE
  7. 4. RESEARCH METHODOLOGY AND SAMPLE SELECTION
  8. 5. RESULTS
  9. 6. DISCUSSION AND CONCLUSIONS
  10. ACKNOWLEDGMENTS
  11. REFERENCES
  12. AUTHOR PROFILES

W. Robert Knechel is the Director, International Center for Research in Accounting and Auditing and the Ernst and Young LLP Professor of Accounting in the Fisher School of Accounting at the University of Florida. Professor Knechel has written extensively on auditing issues in research and education for over 25 years.

Lasse Niemi is an assistant professor in accounting at the Helsinki School of Economics with research and teaching interests in auditing and financial reporting. He currently serves as an academic member on the Auditing Board for Chartered Public Finance Auditors operating under the auspices of the Ministry of Finance and a deputy member of the Auditing Board of the Central Chamber of Commerce operating under the auspices of the Ministry of Trade and Industries.

Stefan Sundgren is Öhrlings PricewaterhouseCoopers professor in Accounting and Auditing at Umeå School of Business at Umeå Univesity, Sweden. His research interests are financial accounting and auditing.