MASTER THESIS
TRANSPARENCY REPORT DISCLOSURE: A COMPARISON BETWEEN MEMBER FIRMS OF AN AUDIT NETWORK.
PHILIP VELDHUIZEN1 2028875
MSc BUSINESS ADMINISTRATION: ORGANIZATIONAL AND MANAGEMENT CONTROL FACULTY OF ECONOMICS AND BUSINESS
UNIVERSITY OF GRONINGEN JANUARY 2017
Supervisor: Sakshi Girdhar Co-assessor: Martijn van der Steen
Word count: 14.550 (13.028 excluding reference list)
ABSTRACT
Since little less than a decade, audit companies who perform audits for public interest entities (PIEs) in the EU are required to publish an annual transparency report. In these reports audit firms report on their corporate governance and the way high audit quality is ensured. Transparency reports are country specific and multinational audit firms are therefore required by law to publish a report for every country in which it conducts audits for PIEs. This study investigates to what degree multinational audit firms’ disclosure in transparency report is coherent within their network. This is done by comparing information disclosed in transparency reports from three big four audit firms in The Netherlands, The UK and Sweden. Institutional theory and practice variation theory are applied to identify forces that drive convergence and divergent disclosure in transparency reports. Qualitative research is deployed to analyze data. The findings suggest that variation in the transparency reports disclosure exists within audit networks.
Key words: transparency reports, audit, big four, disclosure, practice variation, institutional theory.
1 Diezerkade 2B, 8021 CW, Zwolle. [email protected].
1 INTRODUCTION
In the recent years there has been a growing focus on audit quality (Kinney, 2005; O’Regan, 2010). As a result of some high-profile corporate failures at the beginning of the twenty-first century and the recent financial crisis, there has been increasing pressure for more intervention by regulators in order to increase audit quality (Deumes, Schelleman, Vander Bauwhede & Vanstraelen, 2012). Both audit regulators and audit professionals believe that greater transparency by audit firms can increase confidence in audit quality (Wyman, 2004). In order to enhance transparency, regulators in multiple jurisdictions have mandated transparency reports by audit firms. The European Union was the first to do so, regulators in Australia and Japan have followed a couple of years later (Fu, Carson
& Simnett, 2015)2. In its Eight Directive (2006), the EU requires audit firms who perform audits for Public Interest Entities (PIEs)3 to publish an annual transparency report. The underlying motivation of the EU is that more transparency on audit firm governance is expected to reveal audit quality. When information about transparency and audit quality is publicly available, market participants are able to differentiate among audit firms (International Organization of Securities Commissions, 2009). Transparency reports therefore provide incentives for audit firms to increase their audit quality (Deumes et al., 2012). EU law requires audit firms to disclose information in transparency reports on ten items. Disclosure on these ten items is mandatory but there is no law about the extent to which an item must be reported. Audit firms could just state that certain procedures or policies exist, and not elaborate on it further. Besides the ten legal requirements, audit firms might choose to disclose additional information that could increase the transparency even further. Transparency reports must be published on the website of the audit firm within three months after the financial year has ended, and must remain available for the public for one year4. Audit firms are required to issue transparency reports for each country in which it audits PIEs.
But most audit firms operate only in domestic markets. There are only a few multinational audit companies. The auditing market is dominated by the big four (Deloitte, Ernst & Young, KPMG and PricewaterhouseCoopers) (Francis, Michas & Seavey, 2013). In 2010, these four companies handled 67% of all audits in the United States (Gerakos & Syverson, 2015). Velte and Stiglbauer (2012) concluded that audit markets in developed economies are very likely to exhibit similar concentrations. The big four companies operate in the same business (auditing and consultancy) and are very much alike in a sense of operating procedures, structure and culture (Francis, Michas & Yu, 2013; Knechel, Niemi & Zerni, 2013; Arnold & Cooper, 1999). Member firms from the big four companies are part of the global network. The member firms are striving towards consistency across countries (Barrett,Cooper & Jamal, 2005). This is in line with statements made on ‘one-firm strategy’ by the global networks (table 1). With that knowledge, one could expect that disclosure in the transparency reports should be similar throughout the network. Goal of this study is to gain more insights in the way multinational audit firms disclose national level information in their transparency reports, in order to analyze to what extent consistency is achieved within the audit network.
Although the phenomenon of transparency reports issued by audit firms is relative new, some precursors have already studied the subject. Deumes et al. (2012) studied transparency reports of 103 audit firms in Austria, Germany, The Netherlands and the UK for the financial year 2007/2008. They found variations in the extent and type of governance disclosure in the transparency reports. Their study showed that larger audit firms have a higher transparency report disclosure. Deumes et al. (2012) also concluded that German and Austrian audit firms have higher transparency report disclosure than audit firms from the UK and the Netherlands. They explain these findings due to more extensive guidelines on transparency reporting in Germany and Austria. Fu et al. (2015) did research on 21 transparency reports issued by Australian audit firms in 2013. Their results show that there is diversity in the disclosure between audit firms in relation to the minimum disclosure requirements. All audit firms met the minimum requirements but there is variation in the level of details included in the reports.Petersen and Zwirner (2009) studied German transparency reports. Their findings are in line with the findings of Deumes et al.
(2012) and Fu et al. (2015) that the extent of disclosure varies across audit firms. Oh and Dowling (2014) also studied Australian transparency reports. They found variation in the extent of professionalism and commercialism exhibit in four voluntary issued reports in 2012. 5
2 CPA Act in Japan; The Australian Corporations Legislation Amendment (Audit Enhancement) Act 2012 in Australia.
3 EU Directive Eight Article 2 (13) defines ‘public-interest entities’ as “entities governed by the law of a Member State whose transferable securities are admitted to trading on a regulated market of any Member State. ”
4 The one year period is required by the EU. National regulation may require a longer period. In the UK this period is two years. In the Netherlands and Sweden this period is one year.
5 Two other studies on transparency reports can be identified. These studies are not on variation in disclosure though. Pheijffer (2010) studied transparency reports in the Netherlands. He concluded that the reports meet the minimum legal requirements, but disclosure in transparency reports could be increased further if regulations are further specified. Pott et al. (2008) studied the practitioners’ perceptions of transparency reports. They found that the information that is demanded and valued by the different users (accountants and auditors) varies.
2 Table 1
Presentations on ‘one-firm’ strategy, as stated on website of the network.
Firm Statement
Deloitte “With over 150 years of hard work and commitment to making a real difference, our organization has grown in scale and diversity, yet our shared culture remains the same.”6 EY “We’re not merely a loose collection of national practices – we are a global organization, unified
in our approach.”7
KPMG “Our member firms commit themselves to a common set of KPMG values. Firms must abide by quality standards governing how they operate and how they provide services to clients.”8 PwC “Member firms are bound to abide by certain common policies and to maintain the standards of
the PwC network as put forward by PwCIL.”9
These studies all have one conclusion in common; there is variation in transparency report disclosure.
However, there has been no study that focusses on variation in disclosure by member firms within an audit network. This leads to the central question of this study: To what extent is there variation in transparency report disclosure between member firms of an audit organization? In order to answer this question, two research question need to be answered first. These research questions are:
RQ1: Is there variation in transparency report disclosure between member firms of an audit network?
If yes, another research questions arises;
RQ2: What are the sources of variation in transparency report disclosure?
From a theoretical perspective this study aims to: (a) contribute to the literature gap on variation in transparency report disclosure within an audit network and (b) contribute to the literature regarding national level practices by big four audit firms. As mentioned earlier, Deumes et al. (2012), Fu et al. (2015), Petersen and Zwirner (2009) and Oh and Dowling (2014) studied the subject of transparency reports, yet no study focused on variation in disclosure by member firms of a network. DeFond and Zhang (2014) call for further research regarding characteristics of audit firms, such as ownership structure, governance systems and compensation systems. Audit firms elaborate on these characteristics in their transparency report, so studying these reports can contribute to the existing literature regarding national level audit governance.
From a practical perspective this research has contributions for society, the government and audit firms.
First, society benefits from this research in the sense that potential clients of big four firms can get a better view of the organization they would like to do business with. Second, governments and markets can get more insights in the corporate governance of the big four organizations, and the way they ensure a high audit quality. In fact, governmental institutions wanting more transparency in the way audits are performed was the main reason for the introduction of transparency report. Third, audit firms can use this study in order to further improve their transparency reports. This is not only useful for big four firms, but also for smaller audit firms who perform audits for PIEs.
This study is organized as followed; the next section discusses institutional theory and practice variation theory. Section three outlines the methodology of this study. Section four presents the results and section five concludes this study.
LITERATURE REVIEW
This study uses the concepts of institutional theory and practice variation theory. Institutional theory provides arguments why transparency report disclosure by member firms of an audit network is expected to be similar.
Practice variation theory argues why transparency reports are expected to be convergent throughout the audit network.
Institutional theory explores the impact of rules, social norms and rationalized concepts on formal organizational structures, practices and procedures (Meyer and Rowan, 1977; Scott; 2008). Institutional theory is used to explain similarities between organizations in a specific field (DiMaggio & Powell, 1983). DiMaggio and Powell (1983) state that organizations show considerable diversity in approach and form in the earlier stages of their lifecycle. Over time, a field becomes embedded and the organizations in the field will push towards divergent practices and procedures, which leads to homogenization among organizations (DiMaggio & Powell, 1983).
6 https://www2.deloitte.com/global/en/pages/about-deloitte/articles/about-deloitte.html
7 http://www.ey.com/gl/en/about-us/our-global-approach/2020-vision_our-global-approach
8 http://home.kpmg.com/xx/en/home/about/governance/structure.html
9 http://www.pwc.com/gx/en/about/corporate-governance/network-structure.html
3 Organizations show such behavior because they strive to gain legitimacy (Björkman, Fey & Parks, 2007). Suchman (1995) defines legitimacy as “a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions” (p.574).
Homogenization among organizations arises due to the existence of isomorphism. Isomorphism is “a constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions” (Hawley, 1968, p. 334). DiMaggio and Powell (1983) identify three types of isomorphism as the main drivers of convergence between organizations; (a) coercive isomorphism, (b) mimetic isomorphism and (c) normative isomorphism.
Coercive isomorphism refers to formal and informal pressures exerted on an organization (DiMaggio &
Powell, 1983; Scott, 2008). Sources of coercive pressures can be, for example, governmental organizations or powerful international organizations (DiMaggio & Powell, 1983). Björkman et al. (2007) identify international legislation as the main driver of coercive isomorphism. Teo, Wei and Benbasat (2003) concluded that subsidiary firms are exposed to coercive pressures by the parent company. In the context of transparency reports, member firms of an audit network could be exposed to coercive pressures imposed by the network. Member firms of an audit network need to reply to procedures and practices set out by the global network (Zimmerman & Volckmer, 2015). When the network has rules and guidelines on transparency reports, the transparency reports issued by the member firms of the audit network should be similar. Mimetic isomorphism arises when organizations imitate the practices, processes or structures employed by successful organizations with the goal of countering uncertainty (Brown, 2011). Mimetic isomorphism occurs through the adoption of best practices, set out by a leading organization (DiMaggio & Powell, 1983). The underlying assumption is that managers, in time of uncertainty, assume that leading organizations have already adopted appropriate practices (Benders, Delsen & Smits, 2006).
Since transparency reports are a relative new phenomenon, it is not unlikely that audit firms are uncertain about what information to disclose in these reports. Therefore, an audit firm could imitate other member firms from the network in the way they disclose information in their transparency report. Normative isomorphism results primarily from pressures by professional groups (Tsamenyi,Cullen & Gonzalez, 2006). DiMaggio and Powell (1983) state that normative isomorphism arises since members of an occupation collectively struggle to define the conditions and methods of their work. By developing shared values and beliefs, professional groups enhance the collective expectations regarding appropriate practices (DiMaggio & Powell, 1983). The EU has been striving towards harmonization in accounting and auditing standards since the 1970s (Combarros, 2000). As a result, professional bodies have developed standards for accounting (Combarros, 2000). One might expect professional bodies, such as the IOSCO, IASC or IFAC, to contribute towards harmonization of audit firms as well. These normative pressures could therefore increase convergence in transparency report disclosure between member firms of an audit network.
The other theory used in this study is practice variation theory. Originating from institutional theory, practice variation theory aims to explain organizational variety (Lounsbury, 2008). Practice variation theory draws on the concept of multiple embeddedness (Arena & Jeppesen, 2016). Multiple embeddedness arises when organizations are influenced by multiple institutional logics (Lounsbury, 2008). Institutional logics are the organizing principles that affect cognition and guide decision making in a field (Friedland & Alford, 1991). Cruz, Major and Scapens (2009) state that firms that are embedded in multiple organizational fields with divergent institutional logics create the need to vary practice. This is the result of tensions between institutional logics (Edgley, Sharma & Anderson-Gough, 2016). The concept of multiple institutional logics has been studied in order to gain a better understanding of practice variation among organizations. Meyer, Mudambi and Narula (2011) claim that subsidiaries of a multinational organization are subject to multiple embeddedness. Lukka (2007) studied sources of practice variation in a Finnish multinational. He concluded that a source of practice variation was the tension between standard operating rules that the parent company wanted to imply and the sub-units’ technical (internal) logic. The research by Cruz et al. (2009) suggests that practice variation could also be the result of exercise of agency in the dependent organization. Actors in the dependent organization might adapt external processes when integrating these systems into their organization (Cruz et al., 2009). An comparable claim is made by Kostova and Roth (2002); “ when confronted with internal organizational pressure from their parent company to adopt a practice, foreign subsidiaries act as active agents and may adopt the practice to varying degrees ” (p.217). Ratzinger-Sakel and Schönberger (2015) studied national level regulation for audit firms. They concluded that there is significant variation in national legal requirements and professional standards for audit firms. National regulation could be a source of practice variation since audit firms are required to comply with rules and guidelines in their domestic country. Rosenzweig and Singh (1991) found that although subsidiaries of a multinational face pressures for consistency, they are also subject to local institutional logics such as legal systems. With this knowledge, it is likely that member firms of an audit network are subject to multiple institutional logics. This
4 would lead to practice variation among member firms of a network, and as result member firms would disclose divergent information in their transparency reports
In conclusion, transparency report disclosure by member firms of a network could either be convergent or divergent. Institutional theory provides arguments for why transparency reports are expected to be convergent.
Coercive, mimetic and normative pressures are expected to drive similarities in transparency reports by member firms. On the other hand, practice variation theory supports arguments for divergent transparency reports. Two potential sources of variation can be identified; national regulation and the exercise of agency.
METHODOLOGY 1. Research method
Since there is a lack of research regarding transparency report disclosure by member firms of a network, this study adopts a qualitative research method. Qualitative research is useful for studying new phenomena (Arino, LeBaron
& Milliken, 2016). Data is collected through a content analysis. A content analysis is a useful method for analyzing text data (Cavanagh, 1997). A content analysis can be performed either through (a) computational method or (b) manual method. This study is the latter, a manual content analysis. The aim of a manual content analysis is to find meaning and information from context (Viney, 1983). In addition, a content analysis suits this study for multiple reasons. First, content analysis allows an assessment on the disclosure in transparency reports, because the policies and practices effective in an audit firm have different motivations. Second, content analysis enhances controllability of a research since this method is repeatable, and therefore reduces bias (Cascio & Aguinis, 2008).
Third, a content analysis is flexible (Cascio & Aguinis, 2008). Initial categories might be leading at first, other (sub)categories are expected to emerge during the analytic process (Haines-Saah, Bell & Dennis, 2015).
2. The selection of audit firms and countries.
A total of nine transparency reports are included in this study; reports from three audit firms in three countries.
Transparency reports from big four companies are selected for this study because of their dominance in the audit world and due to the fact that bigger companies are likely to disclose more information (Deumes et al., 2012). The transparency reports of Deloitte, Ernst and Young (EY) and PricewaterhouseCoopers (PwC) are compared for the member firms in the Netherlands (NL), the United Kingdom (UK) and Sweden10. Transparency reports are selected based on two criteria; (a) the firm needs to publish transparency reports in all three countries and (b) the transparency report needs to be a stand-alone report. From the big four audit firms, Deloitte, EY and PwC all meet these criteria. The transparency reports issued by KPMG do meet criteria (a), but criteria (b) is not met. KPMG publishes the transparency reports in its integrated report, and not as stand-alone report. It is impossible to tell what information in their integrated report is part of their transparency report and what information is not.
Transparency reports by KPMG are therefore not taken into account in this study.
The Netherlands, the UK and Sweden are selected on the base of their legal system. These three countries have different legal systems in place, which leads to diversity in national regulation for audit firms (Doupnik &
Perera, 2015). Diversity in national regulation may result in more or less rules and guidelines regarding transparency reports. Studying countries with different legal systems could therefore be interesting in the light of multiple embeddedness, since member firms must comply with both national regulations as pressures by the network. The UK is typified as a common law system. This type of legal system is characterized by strict accounting rules that are issued by professional or nongovernmental bodies (Doupnik & Perera, 2015; Iatridis, 2012; Frost & Ramin, 1996). Doupnik and Perera (2015) further state that financial statements in common law countries generally must be published in accordance with a prescribed format. In code law countries, such as the Netherlands, are accounting laws rather general and issued by the government (Doupnik & Perera, 2015;
Hesselink, 2006). Sweden is part of the Scandinavian (or Nordic) legal system (Zweigert & Kötz, 1998). This legal systems falls between common law and code law (La Porta,Lopez-de-Silanes, Shleifer & Vishny, 2000), although is it mostly influenced by code law systems (Elling, 1993)11.
Transparency reports of the financial year 2015/2016 are studied. This financial year is selected since this is the most recent financial year. Audit firms have published transparency reports for almost a decade, so they have some experience with disclosing information in their reports. In addition, as mentioned in the literature review Deumes et al. (2012) stated that there are fewer disclosure guidelines in the Netherlands and the UK in comparison with Germany and Austria. However, their research was on transparency reports issued in 2008. Since 2008,
10 A list with abbreviations can be found in the appendix
11 Since there a more countries with the same legal system, the further selection of countries is based on the language. Transparency reports are often issued in the domestic language. The author masters the Dutch and English language. Swedish transparency reports are usually published in English, however PwC Sweden did publish their report in Swedish.
5 professional and governmental bodies in the Netherlands and in the UK have increased regulations for audit firms.
More regulations for audit firms is expected to result in more disclosure in transparency reports. This in line with the findings of Deumes et al. (2012) that transparency report disclosure increases over time. Transparency reports are therefore expected to have the highest disclosure in the most recent financial year, 2015/2016.
3. Coding
This paper compares the transparency reports on all ten mandatory items, as listed in Article 40 of EU Directive Eight. These items are (a) a description of the legal structure and ownership; (b) where the audit firm belongs to a network, a description of the network and the legal and structural arrangements in the network; (c) a description of the governance structure of the audit firm; (d) a description of the internal quality control system of the audit firm and a statement by the administrative or management body on the effectiveness of its functioning; (e) an indication of when the last quality assurance review took place; (f) a list of public-interest entities for which the audit firm has carried out statutory audits during the preceding financial year; (g) a statement concerning the audit firm's independence practices which also confirms that an internal review of independence compliance has been conducted; (h) a statement on the policy followed by the audit firm concerning the continuing education of statutory auditors (i) financial information showing the importance of the audit firm, such as the total turnover divided into fees from the statutory audit of annual and consolidated accounts, and fees charged for other assurance services, tax advisory services and other non-audit services; (j) information concerning the basis for the partners' remuneration (EU Directive Eight, 2006).
The coding progress was carried out in the following manner. First, the nine transparency reports were obtained through the websites of the audit member firms. Second, in line with Krippendorff (2004) and Weber (1990), categories were identified. The categories are based on Article 40 of EU Directive Eight, as mentioned above. Each item from Article 40 is a category. The studies by Deumes et al. (2012) and Fu et al. (2015) are used as a basis for creating subcategories. However, some subcategories are added by the author as well. The third step in the coding process is the manual reading of all nine transparency reports. Finally, the data on the different categories was retrieved from the reports by reading the transparency reports. Notes and headings were written in the text, as prescribed by Elo and Kyngäs (2008). After all the data was collected, transparency reports were read again to make sure that there is no data missed.
Reliability in content analyses consist of three elements; (a) stability, (b) reproducibility and (c) accuracy (Krippendorff, 2004). Krippendorff (2004) defines stability as the degree to which a process is invariant or unchanging over time. A test-retest procedure can be used to obtain stability. The transparency reports in this study are therefore read twice. Once before the data collection, and once after all the data was collected. Reproducibility is the degree to which a process can be recreated under varying circumstances, at different locations, using different coders (Krippendorff, 2004). Since this study is conducted by only one researcher it is impossible to achieve inter- rated reliability. However, the data studied is all online available for the public. Other researchers should therefore be able to reproduce this study. Accuracy is the degree to which a process conforms to a known standard (Krippendorff, 2004). To achieve accuracy, the coding process for a content analysis as indicated by Elo and Kyngäs (2008) is adopted. The content analysis process consists of (1) open coding, (2) creating categories and (3) abstraction (Elo and Kyngäs, 2008). The study follows this process and the steps are explained earlier in this section. Steps (2) was executed before step (1) since the coding in this study is based on earlier studies.
Validity in content analyses consists of two steps; (a) developing a coding scheme and (b) assessing the coding scheme to a standard (Potter and Levine‐Donnerstein, 1999). This requires the existence of a standard set of codes. The coding scheme used in this study is consisted with the coding used by Deumes et al. (2012) and Fu et al. (2015), and the coding scheme in those studies can be seen as the standard set of codes. The coding scheme used in this study can be found in the appendix.
RESULTS12 1. (a) Legal structure and ownership (table 2)
1.1 Deloitte. All three Deloitte firms state that they operate in a limited liability partnership, and all three firms mention that these partnerships are owned by partners of the member firms. Number of partner is disclosed by all three firms as well. In addition, Deloitte NL and Deloitte Sweden state the names of national subsidiaries in their report. Deloitte UK does not include this information.
1.2 EY. All three EY firms disclose that they operate in a limited liability partnership/company. EY NL and EY Sweden state that the firms are owned by its partners, but EY Sweden also mentions EY Europe as owner. EY UK
12 Tables with all data are published in the appendix.
6 does not explicitly mention ownership in their report. All EY firms state that EY Europe has voting control in the member firms. EY Sweden and EY UK disclose the number of partners in their report. EY NL does not include this. No EY firm mentions the names of national subsidiaries.
1.3 PwC. All PwC firms state that they operate in limited liability partnership, and that partners own the firm.
Number of partners and names of national subsidiaries are disclosed by all PwC firms.
1.4 Variations in disclosure. All three network show consistency among its member firms. Some minor differences exists but overall the disclosure is very similar in all networks.
2. (b) The network legal and structural arrangements (table 3).
2.1 Deloitte. All three Deloitte firms mention that they are part of the DTTL network. Deloitte NL and Deloitte UK disclose characteristics of the network (number of employees, countries and total revenues of the network).
Notable here is a small difference in total employees of the network disclosed in the Deloitte NL and Deloitte UK report (225,000 and 225,400 respectively). This could be a sign of this information not being supplied by the network. Deloitte NL also include number of offices. Deloitte Sweden does not include any characteristics of the Deloitte network.
2.2 EY. All three EY firms disclose that they are part of the EYG network and all firms disclose information on number of employees and number of countries in which EYG operates.
2.3 PwC. All three PwC firms state that they are part of the PwCIL network. PwC NL and PwC UK disclose information on network characteristics as well (number of employees, countries, offices and total revenues of the network). PwC Sweden does not include characteristics of the network.
2.4 Variations in disclosure. Minor variations exists between the Deloitte and PwC firms. EY firms disclose the same information.
3. (c) Description of the governance structure of the audit firm (table 4).
3.1 Deloitte. From the Deloitte firms, only Deloitte NL discloses a report by the Public Interest Committee (PIC).
Deloitte UK includes a report from its Independent Non-Executives (INEs). Management teams are included by all three Deloitte firms, and all three firms mention a board of directors. Deloitte NL is the only Deloitte firm that discloses the presence of a supervisory board. Deloitte UK discloses further information on management teams as well (for example, meeting attendance by board members). This information is not disclose by Deloitte NL and Deloitte Sweden. Names of sub-committees are only disclosed by Deloitte NL and Deloitte UK.
3.2 EY. No EY firms includes a PIC report, but EY UK discloses a report from its INEs. Management teams are included by all three EY firms. The board is mentioned by all firms. EY NL is the only EY firm to disclose that a supervisory board is also in place. EY UK discloses further information on management teams, EY NL and EY Sweden do not include this information. In addition, EY UK is the only firm to disclose information on sub- committees.
3.3 PwC. A PIC report is included by all three PwC firms. All three firms also include management teams in their transparency report. In addition to all three PwC firms having a board, PwC NL and PwC UK have a supervisory board as well. From the PwC firms, only PwC UK discloses further information on management teams. Sub- committees are only disclosed by PwC NL and PwC UK.
3.4 Variations in disclosure. There is variation in disclosure by member firms of all three networks. For example, all three Dutch firms mention a supervisory board while firms from the Sweden and the UK do not mention this, except PwC UK. Another example is UK firms disclosing extensive information on management teams, while Dutch and Swedish firms do not elaborate on management teams. These variations are the result of national regulations. In 2012, the Dutch NBA issued the Code for Audit Firms. This Code requires audit firms who perform audits for PIEs to install a Public Interest Committee. In 2014, the NBA published the ‘In het publiek belang’
(English: In the public interest) report. This led to two new governance principles in the Netherlands; (a) audit firms who perform audits for PIEs are required to install a Supervisory Board and (b) the tasks and functions of the Public Interest Committees will be integrated with those of the Supervisory Board. UK firms disclosing extensive information on management teams is due to the Audit Firm Governance Code in the UK (FRC, 2016).
Provision A.1.3. of this Code requires firms to disclose the following information on management teams and governance structures; (a) names and job titles, (b) how they are elected or appointed, (c) their terms, length of services, (d) meeting attendance in the year and (e) relevant biographical details. The Code also requires audit firms to disclose a report from its independent non-executives (INEs) and/or PIC. National regulation could therefore be identified as a source of practice variation. Other differences, such as the disclosure on sub-committees are likely to be the effect of exercise of agency.
7 4. (d) Description of the internal quality control system of the audit firm and a statement by the administrative or management body on the effectiveness of its functioning (table 5).
In 2009, the IAASB issued a report regarding the Internal Standard on Quality Control 1 (ISQC 1). In this report, the IAASB states the following: “ISQC 1 deals with a firm’s responsibilities for its system of quality control for audits and reviews of financial statements, and other assurance and related services engagements” (p. 38). “The objective of ISQC 1 is for a firm to establish and maintain a system of quality control to provide it with reasonable assurance that: (a) the firm and its personnel comply with professional standards and applicable legal and regulatory requirements, and (b) reports issued by the firm or engagement partners are appropriate in the circumstances” (p. 39) .ISQC 1 consists of six elements; (a) Leadership responsibilities for quality within the firm, (b) Relevant ethical requirements, (c) Acceptance and continuance of client relationships and specific engagements, (d) Human resources, (e) Engagement performance, (f) Monitoring (IAASB, 2009). Firms that follow ISQC 1 disclose information on all six elements.
4.1 Deloitte. Only Deloitte Sweden mention that they follow ISQC 1, and thus include all six elements of ISQC 1 in their report. Deloitte NL and Deloitte UK disclose information on: (a) leadership, (b) ethics, (e) engagement performance and (f) monitoring. In addition, Deloitte UK also states information on (d) human resources. A statement on the effectiveness of the internal quality control system is disclosed by all three Deloitte firms. These statements are signed by the management in the reports of Deloitte NL and Deloitte Sweden.
4.2 EY. There is no EY firm that mentions following ISQC 1. All EY firms do disclose information on: (a) leadership, (b) ethics, (c) client relationships, (e) engagement performance and (f) monitoring. EY NL does also include a (d) human resource policy. All three firms disclose a statement on the effectiveness of the internal quality control systems and these are all signed by the management.
4.3 PwC. All three PwC firms state that they follow ISQC 1, and therefore mention all the six elements in their transparency reports. In addition, all three PwC firms included a statement by the management on the effectiveness of the internal quality control system. However, this statement is not signed by the management in the reports of PwC Sweden and PwC UK.
4.4 Variations in disclosure. There is variation in disclosure between the Deloitte firms, with only Deloitte Sweden mentioning ISQC 1. EY and PwC disclose similar information throughout their networks. Furthermore, it is interesting to see that EY Sweden does not mention ISQC 1. The Revisornämnden (Swedish Supervisory Board of Public Accountants) state on their website that Swedish audit firms are required to comply with ISQC 113. Deloitte Sweden and PwC Sweden disclose that they have adopted ISQC 1. The fact that EY Sweden does not include information on ISQC 1 could be a sign that EY has set global standards for transparency reports, since the information on internal quality control systems disclosed by EY Sweden is the same as EY NL and EY UK.
5.(e) External monitoring (table 6).
5.1 Deloitte. Deloitte NL (6) is inspected by most bodies in comparison to Deloitte Sweden (1) and Deloitte UK (3). Deloitte NL is also the only Deloitte firm that includes findings from external reviews in their transparency reports. Deloitte Sweden and Deloitte UK mention that findings can be found on the websites of the Revisorsnämnden and the FRC respectively.
5.2 EY. EY NL (6) is subject to inspection by most bodies in comparison to EY Sweden (2) and EY UK (3). EY NL and EY UK disclose findings from external reviews in their transparency report. EY Sweden states that these findings can be found on the website of the Revisorsnämnden and the PCAOB.
5.3 PwC. PwC NL (4) is inspected by most bodies in comparison to PwC Sweden (1) and PwC UK (3). PwC UK is the only PwC firm that discloses findings from external reviews in their transparency report. PwC NL does not mention anything about the findings. PwC Sweden states that findings can be found on the website of the Revisorsnämnden.
5.4 Variations in disclosure. There is variation in external monitoring disclosure in all three audit networks.
However, firms are subject to inspections by national reviews bodies so the source of the variation in this case is practice variation. On the other hand, the fact that some firms disclose findings in their report while other member firms do not include findings is likely due to the exercise of agency.
6. (f) A list of public-interest entities for which the audit firm has carried out statutory audits during the preceding financial year (table 7).
6.1 Deloitte. While all three Deloitte firms disclose a list with PIEs audited, Deloitte NL and Deloitte UK provide additional information. Deloitte NL mentions mutations in the list of PIEs from 2014/2015 as well. Deloitte UK includes UK company numbers from the PIEs in their report.
13 http://www.revisorsnamnden.se/rn/english/laws_rules_and_standards.html
8 6.2 EY. All three EY firms disclose a list with PIEs audited. No EY firms disclose additional information.
6.3 PwC. All three PwC firms include a list with PIEs audited. In addition, PwC NL discloses mutations on the list of PIEs from 2015/2015 also. PwC Sweden mentions the regulated markets that the PIEs are transferred in as well. For example, the Scandinavian exchange companies Nasdaq OMX and NGM Stockholm. PwC UK does not disclose additional information.
6.4 Variations in disclosure. There is consistency between the EY firms. Some small differences exists in the Deloitte and PwC network but the overall disclosure is very similar.
7. (g) A statement concerning the audit firm's independence practices which also confirms that an internal review of independence compliance has been conducted (table 8).
7.1 Deloitte. Deloitte NL and Deloitte Sweden disclose Deloitte network policies on independence. Deloitte UK does not mention these network policies. Deloitte Entity and Compliance System (DESC) and Global Independence Monitoring System (GIMS) are both mentioned by Deloitte NL and Deloitte Sweden. Deloitte NL also mentions that there is a global network of Independence Directors. Deloitte NL (Independence Center) and Deloitte Sweden (Director of Independence) both further disclose only one national independence policy in place.
Deloitte UK mentions seven national independence policies. A confirmation that an internal review of independence has been conducted is disclosed by all three Deloitte firms.
7.2 EY. All EY firms state two global policies on independence: Global Independence System (GIS) and Global Monitoring System (GMS). The three member firms of EY also disclose the same national policies on independence; EY independence compliance. EY independence compliance consists of: (a) Independence confirmation, (b) Global independence compliance reviews, (c) Personal independence compliance testing, (d) Non- audit services (e) Global independence learning, (f) Service Offering Reference Tool, (g) Business Relationship Evaluation Tool and (h) Audit committees and oversight of independence. EY NL also disclose an Independence Desk to be in place. EY NL and EY Sweden disclose a confirmation that an internal review of independence has been conducted. EY UK does not explicitly mention such a confirmation.
7.3 PwC. All three PwC firms disclose global policies on independence. However, there is variation between the number of global policies included; PwC NL (1), PwC Sweden (4) and PwC UK (3). Authorisation for services (AFS) is mentioned by all three firms. PwC Sweden and PwC UK also mention Central Entity Service (CES) and Independence Checkpoint as global policies. PwC Sweden further state that the PwC network has a Global Breaches Reporting System. National policies are also included by all three PwC firms, yet the national policies show no consistency among the PwC firms. A confirmation that an internal review of independence has been conducted is included by all PwC firms.
7.4 Variations in disclosure. There is variation among Deloitte and PwC member firms. The EY firms disclose similar information.
8. (h) Continuing education of statutory auditors (table 9).
8.1 Deloitte. All three Deloitte firms state minimum requirements of its auditors in the light of education. Deloitte NL (40 per year) and Deloitte Sweden (120 over a three-year period) state a minimum number of hours that must be spent on education. These two firms also disclose education programs that auditors can participate in. Deloitte UK only mentions that auditors are required to attend an annual technical update, a course covering audit, accounting and other matters such as professional skepticism, audit tools and project management. Deloitte NL and Deloitte Sweden disclose that they monitor staff compliance regarding education. No Deloitte firm state that they take employee participation in education programs into account in the process of promoting employees.
8.2 EY. All three EY firms mention that its auditors are required to spent 120 hours over a three year period, with a minimum of 20 per year. The content of education programs is disclosed by all three firms and is similar throughout the network, although EY NL mentions other education programs as well. EY NL state that they monitor employee compliance regarding education. EY Sweden and EY UK do no mention this in their report.
There is no EY firm that disclose that they take employee participation in education programs into account in the process of promoting employees.
8.3 PwC. None of the PwC firms disclose a minimum number of hours that must be spent on education by its auditors. PwC Sweden and PwC UK mention that auditors are required to complete an annual risk and quality update training spanning matters relating to compliance, independence and ethics. Although PwC NL does not mention anything an minimum requirements, they are the only PwC firm to disclose the content of education programs. All PwC firms state that they monitor employee compliance regarding education. PwC NL and PwC UK also include that they take employee participation into account in the process of promoting employees.
8.4 Variations in disclosure. There is variation among the Deloitte and PwC networks. EY firms disclose similar information in their transparency reports. Some variations that exist in the Deloitte and PwC networks are the
9 result of national regulations. Dutch auditors must spent 120 hours over a three-year cycle (with a minimum of 20 hours per year) on education (NBA, 2014), and Swedish auditors are required to spent 100 hours over a five-year cycle 14. There appears to be no guidelines for auditors from the UK. Other differences in disclosure are likely to be the result of exercise of agency.
9. (i) Financial information (table 10)
9.1 Deloitte. Financial information is disclosed by all three Deloitte firms. All three Deloitte firms divide revenues into revenues from the different service lines. Deloitte UK discloses operating profit from audit services as well.
Deloitte NL and Deloitte Sweden do not disclose such information.
9.2 EY. Financial information is disclosed by all three EY firms. All three EY firms divide revenues into revenues from the different service lines. EY UK discloses operating profit from audit services as well. EY NL and EY Sweden do not include such information.
9.3 PwC. Financial information is disclosed by all three PwC firms. All three PwC firms divide revenues into revenues from the different service lines. PwC UK discloses operating profit from audit services as well. PwC NL and PwC Sweden do not include such information.
9.4 Variations in disclosure. There is small variation in all three networks due to the fact that all three UK firms disclose operating profits from audit services and Dutch and Swedish firms do not include this information. UK firms are required to publish the operating profit from audit services as well, as described in the Voluntary Code of Practice on Disclosures of Audit Profitability (Consultative Committee of Accountancy Bodies, 2009). This is line with the Doupnik and Perera (2015) who state that financial statements in common law countries must be published according to a prescribed format. In the Netherlands and Sweden there are no national rules on disclosing financial information. National regulations therefore drive variation in financial information disclosure.
10. (j) Information concerning the basis for the partners' remuneration (table 11).
10.1 Deloitte. From the Deloitte firms, only Deloitte UK includes a statement that audit partners are not rewarded for the sale of non-audit services to audit clients. All three firms disclose that partners' remuneration consists of a performance related bonus. Deloitte Sweden and Deloitte UK further disclose factors that determine the performance of partners, but the factors disclosed by Deloitte Sweden differ from those disclosed by Deloitte UK.
Deloitte NL does not include factors that determine the performance of partners.
10.2 EY. All three EY firms disclose a statement that audit partners are not rewarded for the sale of non-audit services to audit clients. All three EY firms further state that partners' remunerations consist of a performance related bonus. The factors that determine the performance of partners are disclosed by all EY firms and are the same for each firm.
10.3 PwC. From the PwC firms, only PwC UK discloses a statement that audit partners are not rewarded for the sale of non-audit services to audit clients. All three PwC firms state that partners' remuneration consists of a performance related bonus. In addition, PwC Sweden also mention that partners receive a base salary. PwC NL and PwC Sweden disclose factors that determine the performance of partners. The factors disclosed are the same in both reports. PwC UK does not include factors that determine the performance of partner.
10.4 Variations in disclosure. There is variation in disclosure between member firms of Deloitte and PwC. EY firms are consistent in disclosing information on partners' remuneration.
DISCUSSION AND CONCLUSION
The goal of this study is to gain more insights in the way big four audit firms disclose national level information in their transparency report. The central question of this is study is therefore; To what extent is there variation in transparency report disclosure between member firms of an audit organization? Two additional research question were formed to answer the central question. RQ1: Is there variation in transparency report disclosure between member firms of an audit network? If yes, RQ2 arises: What are the sources of variation in transparency report disclosure? The 2015/2016 transparency reports from Deloitte, Ernst and Young and PricewaterhouseCoopers are studied in three countries; the Netherlands, the United Kingdom and Sweden.
The results show that there is variation in transparency report disclosure between member firms of an audit network. However, not every audit network shows the same degree of variation among its member firms.
EY member firms are most consistent in disclose information in transparency reports. For example, all three member firms of EY disclose similar information on; the EY network (item b), internal quality control systems
14 EY Sweden mention in their report that the Revisornämnden requires auditors to spent 100 hours over a five-year cycle. Revisornämnden publishes almost all regulations in Swedish. This made it impossible to back up the statement by EY Sweden. The claim by EY Sweden is therefore leading in this study.
10 (item d), independence practices (item g), continuing education (item h) and partners’ remuneration (item j).
Coercive isomorphism and mimetic isomorphism could be the source of this homogenization among EY member firms. Coercive isomorphism could relate to the EY global network pressuring its member firms to adopt a certain set of standards for transparency reports. This is in line with the findings by Teo et al. (2003) that subsidiaries of a multinational are required to comply with the rules of the parent company. Mimetic isomorphism implies that EY member firms imitate a leading member firm in the way they disclose information in transparency reports.
This is consistent with the findings of DiMaggio and Powell (1983) that organizations copy ‘best practices’ by other organizations. The findings by Barrett et al. (2005) support mimetic isomorphism. They concluded that member firms strive towards consistency. The results do not reveal evidence that normative isomorphism contribute towards homogenization since international professional bodies, such as the IFAC of IOSCO, do not have guidelines on transparency reports. If such international professional bodies would have guidelines on transparency reports, the reports by Deloitte and PwC member firms would have been more homogenous.
Furthermore, none of the nine member firms disclose international professional bodies influencing their report. A possible explanation could be that international professional bodies leave room for national professional bodies (such as the FRC and the NBA) to imply their own rules and guidelines.
Furthermore, the disclosure by EY Sweden on (d) internal quality control systems is very interesting.
Even though Swedish regulation requires EY Sweden to adopt ISQC 1, EY Sweden does not mention ISQC 1 in their transparency report. In fact, EY Sweden discloses the same information on internal quality control systems as EY NL and EY UK. Yet, since EY Sweden need to comply with Swedish rules, it is likely that they have adopted ISQC 1 but choose not to disclose this in their transparency report. It appears that EY Sweden rather complies with disclosure guidelines set out by the EY network than to disclose country specific information.
The Deloitte and PwC member firms are more divergent in disclosing information in transparency reports.
This is in line with the findings by Deumes et al. (2012) and Fu et al. (2015) that there is variation in transparency report disclosure. In the case of Deloitte, member firms disclose different information on the following items;
governance structure (item c), internal quality control systems (item d), last external quality review (item e), independence practices (item g), continuing education (item h) and partners’ remuneration (item j). In addition, the fact that Deloitte NL and Deloitte UK disclose slightly different information on the number of employees in item (b) could be a sign that this information is not provided by the global network. Member firms of PwC disclose different information on the following items; governance structure (item c), last external quality review (item e), independence practices (item g), continuing education (item h) and partners’ remuneration (item j). Both Deloitte and PwC member firms disclosure different information on legal structure and ownership (item a), the network (item b), list of PIEs (item f) and financial information (item i) as well, but the differences in these items are significant smaller than the aforementioned items. The divergent transparency report disclosure by member firms of Deloitte and PwC could be due to member firms being subject to multiple embeddedness. Multiple embeddedness creates the need to vary practices (Cruz et al., 2009), due to the existence of multiple institutional logics (Lounsbury, 2008). The source of practice variation in this study indicates national regulation. Member firms need to reply to national rules and guidelines for audit firms. For example, the Consultative Committee of Accounting Bodies requires UK audit firms to disclose profits from audit services when disclosing financial information. As a result, UK member firms disclose this information in their transparency report, while Dutch and Swedish member firms do not. Other examples of national regulation that lead to practice variation are; (a) Dutch regulation that audit firms who perform audits for PIEs are required to have a supervisory board, (b) UK firms needing to disclose extensive information on management teams, (c) UK firms needing to disclose a report by the PIC or INEs, (d) the fact that member firms are subject to national external quality reviews; in the Netherlands (6) are more external monitoring bodies than Sweden (2) and the UK (3), and (e) the fact that Dutch and Swedish regulators have set a standard for continuing education. However, not all practice variation is driven by national regulation. As Cruz et al. (2009) pointed out, the exercise of agency can also be identified as a source of practice variation. This seems to be the case in this study; member firms of Deloitte and PwC disclose information differently, and this is not always the result of national regulations.
To summarize, big four audit firms appear to have different policies on transparency reports. The transparency reports issued by EY member firms are homogeneous due to coercive and mimetic isomorphism. The transparency reports by Deloitte and PwC member firms are divergent due to national regulations and the exercise of agency. In answering the research question; To what extent is there variation in transparency report disclosure between member firms of an audit organization?, this study shows there is variation in transparency report disclosure between member firms of an audit organization, but not every audit network shows the same degree of variation among its member firms. It is interesting to see how multinational audit firms deal with transparency reports. While scholars such as Francis et al. (2013) and Knechel et al. (2013) argue that big four audit firms are similar in the sense of operating procedures, structure and culture, this study shows that these firms have different
11 practices and procedures as well. These findings contribute to the literature on big four audit firms. The study also contributes to the transparency report literature in the sense of a cross-country analysis of member firms in the same network.
The findings suggest implications for practitioners as well. Although big four firms strive towards consistency among member firms (Barrett et al., 2005) this study does not support these findings. To increase consistency, policy makers could create internal rules and guidelines on transparency reports in an audit network.
These rules could be on a standard format that must be used by all member firms, but could also imply information that must be disclosed regarding the ten legal requirements. Since transparency can be useful tool to differentiate a market participant among audit firms (IOSCO, 2009), having global rules on transparency reports could benefit the audit network. Variations in the exercise of agency can be countered with global rules. However, when creating these global rules policy makers must remind that member firms are required to comply with national regulations.
This study has the following limitations. First, the transparency report by PwC Sweden was published in Swedish. While older reports issued by PwC Sweden were in English, the transparency report from 2015/2016 was in the domestic language. After contacting PwC Sweden about this it turned out that there were no plans to issue an English version as well. Google Translate is issued to translate the transparency report in English. Second, this study only examines the information disclosed in transparency reports. In the light of economic theory, audit firms will consider the costs and benefits of disclosing information in their transparency reports (Deumes et al., 2012). When certain policies are not disclosed in transparency reports, this does not necessarily mean that these policies do not exists with a firm. The conclusions found in this study are therefore only applicable to transparency reports, and can not be used for drawing conclusions on policies and practices set out by big four firms. Third, this study focusses only on three countries and three audit firms which makes is hard to draw global conclusions.
Further research could therefore focus on transparency reports of other audit firms or other countries. An increasing number of countries worldwide require audit firms who perform audits for PIEs to issue a transparency report. It would be interesting to see how member firms from different continents disclose information in their transparency report. Findings from such a study could help develop global conclusions. Another opportunity for further research could be interviewing policy makers from audit firms studied in this research. This method is useful for identifying the underlying motivation by audit firms to disclose certain information in their transparency reports. Since the EY network seems to have adopted a policy on transparency report disclosure it would be interesting to find out the motivation by EY to do so. Deloitte and PwC are more divergent in their transparency report disclosure by their member firms. Interviewing policy makers from these organizations could help create an understanding of why these firms have such a policy on transparency reports. Further research could also focus on the differences in national regulation since the rules and guidelines for audit firms are different across countries.
To conclude, this study is in line with Deumes al. (2012), Pheijffer (2010) and Fu et al. (2015) in their conclusion that current regulation on transparency report needs to be reconsidered. The ten mandatory items on which audit firms have to elaborate on in their transparency report are not specific enough to force audit firms to disclose specific information. For instance, subcategories could be implemented to make sure audit firms elaborate more extensively on their policies and procedures.
REFERENCES
Arena, M., & Jeppesen, K. (2016). Practice Variation in Public Sector Internal Auditing: An Institutional Analysis. European Accounting Review, 25 (2): 319-345.
Arino, A., LeBaron, C., & Milliken, F. J. (2016). Publishing qualitative research in academy of management discoveries. Academy of Management Discoveries, 2 (2): 109-113.
Arnold, P. J., & Cooper, C. (1999). A tale of two classes: The privatization of Medway Ports. Critical Perspectives on Accounting, 10 (2): 127–152.
Barrett, M., Cooper, D, & Jamal, K. (2005). Globalization and the coordinating of work in multinational audits.
Accounting, Organizations & Society, 30 (1): 1-24.
Benders, J., Delsen, L., & Smits, J. (2006). Bikes versus lease cars: the adoption, design and use of cafeteria systems in the Netherlands. International Journal Of Human Resource Management, 17(6): 1115-1128.
Björkman, I., Fey, C. F., & Parks, H. J. (2007). Institutional theory and MNC subsidiary HRM practices:
evidence from a three-country study. Journal Of International Business Studies, 38(3): 430-446.
