• No results found

The extended audit opinion: creating useful insights or just ticking the box?

N/A
N/A
Protected

Academic year: 2021

Share "The extended audit opinion: creating useful insights or just ticking the box?"

Copied!
39
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

The extended audit opinion: creating useful

insights or just ticking the box?

B.F. Boven

Master Thesis Accountancy

Rijksuniversiteit Groningen

Faculty of Economics & Business

(2)

The extended audit opinion: creating useful insights or just ticking the box? 1

The extended audit opinion: creating useful insight or just ticking

the box?

A study on the effects of institutional isomorphism on the issuing of the extended audit opinions for companies in the financial sector.

Frank Boven S3233804 Esdoornlaan 38 9678RR Westerlee b.f.boven@student.rug.nl 06 83 56 70 09 Rijksuniversiteit Groningen Faculty of Economics & Business Supervisor: Prof. Dr. D.A. de Waard

Co-assessor: Dr. R.C. Trapp

Abstract

This thesis examines whether mechanisms of institutional isomorphism have an influence on the issuing of the extended audit opinion for financial companies. The extended audit opinion has been created to provide more information about the performed audit. The data for this study has been collected from 279 extended audit opinions which have been issued from 2013 up to and including 2016 for 82 companies in the financial sector. Analyses have been performed on whether the size of the audited entity, the year on which the extended audit opinion is issued or the issuing audit firm have a relationship with the diversity of the extended audit opinion. Results showed a statistical significant positive relationship between the size of the audited entity and diversity of the extended audit opinion. A statistical significant negative relationship was observed between the year on which the extended audit opinion is issued and the diversity of the extended audit opinion. No statistical significant relationship was found between the issuing audit firm and diversity of the extended audit opinion. These results indicate that coercive and mimetic isomorphism do have an influence on the issuing of the extended audit opinion and that normative isomorphism does not have an influence. Results of this study should be of interest for investors, regulators and audit firms. Firstly, because the positive influence of coercive isomorphism on the extended audit opinion diversity confirms that the new regulations have the intended effect of providing more useful audit opinions to investors. Secondly, because the results of mimetic isomorphism confirm the concerns that the issuing of the extended audit opinion becomes a standard ticking the box procedure.

(3)

The extended audit opinion: creating useful insights or just ticking the box? 2

Table of Contents

1.

Introduction

4

1.1 The extended audit opinion 4

1.2 Isomorphic pressures on the extended audit opinion 4

1.3 Academic contribution and purpose of the study 6

1.4 Thesis structure 6

2.

Theoretical framework and hypotheses development

7

2.1 Legitimacy theory 7

2.2 Institutional theory 7

2.3 Institutional isomorphism and hypotheses development 8

2.3.1 Coercive isomorphism 8

2.3.2 Mimetic isomorphism 9

2.3.3 Normative isomorphism 10

2.4 The concept of diversity 11

2.5 Conceptual model 12

3.

Research Methodology

13

3.1 Sample 13 3.2 Variables 13 3.2.1 Dependent variable 13 3.2.2 Independent variables 15 3.2.3 Control variables 15 3.3 Statistical model 17

4.

Results

18

4.1 Dataset 18 4.2 Descriptive statistics 18 4.3 Correlation analysis 19 4.4 Regression analysis 20 4.4.1 Control variables 20

4.4.2 Size of the audited entity (H1) 21

4.4.3 Recency of the extended audit opinion (H2) 21

4.4.4 Issuing audit firm (H3) 22

(4)

The extended audit opinion: creating useful insights or just ticking the box? 3

5.

Additional analysis

24

5.1 Interview design 24

5.2 Interview results 24

5.2.1 The extended audit opinion 24

5.2.2 Discussion on hypotheses 25

5.2.3 Discussion on control variables 26

6.

Conclusion and discussion

27

6.1 Conclusion and discussion 27

6.2 Limitations of this study 29

6.3 Ideas for future research 29

References

31

Appendices

35

Appendix 1: Companies included in the sample 35

Appendix 2: Coding sheet for the categorization of KAMs 36

Appendix 3: Operationalization independent variables 37

(5)

The extended audit opinion: creating useful insights or just ticking the box? 4

1. Introduction

In this chapter the master thesis is introduced. The first paragraph of this chapter describes the main subject of this study: the extended audit opinion. In the second paragraph a gap in existing knowledge is identified and the main research question is formulated. The third paragraph elaborates on the purpose and significance of the study. The last paragraph presents how the remainder of the thesis is structured.

1.1 The extended audit opinion

The financial crisis of 2008 sparked a discussion about the effectiveness and usefulness of statutory audits and subsequently one of the products of an audit: the audit opinion. The collapse of the banking sector raised questions about the added value of the audit opinion because financially distressed banks still received an unqualified audit opinion (Sikka, 2009). At the time, the audit opinion had a pass/fail model in which the auditor gives his or her view on whether the financial statements are fairly represented. This view was reported by issuing an unqualified (pass) or qualified (fail) audit opinion. This boilerplate model creates little to no added value because, for example, the audits of banks became increasingly complex while the audit opinion only showed the general and standardized rules of the role and limitations of the audit (Humphrey, Loft and Woods, 2009). Most regulators and stock exchanges require companies to have an unqualified opinion and as a result many publicly traded companies received the same audit opinion with little company-specific information. Investors state that auditors might possess private information about a company which might be useful to them (Lennox, Schmidt and Thompson, 2016).

In response to investors, the Financial Reporting Council (FRC) issued changes in the UK auditing standards concerning the audit opinion. These changes went into effect in October 2012 for financial statements of 2013 and onwards (FRC, 2015). The Royal Netherlands Institute of Chartered Accountants (NBA) issued similar changes that went into effect for financial statements of 2014 and onwards (NBA, 2014). These changes include that auditors have to provide an extended audit opinion that gives an overview of (1) risks of material misstatement; (2) the application of materiality; and, (3) the scope of the audit, including how it responded to risks of material misstatement and the application of materiality (FRC, 2015). The main goal of the extended audit opinion is to give investors company-specific information by providing greater transparency on how the audit is conducted and what risks of material misstatements were identified during the audit. These risks of material misstatement are typically reported as Key Audit Matters (KAMs) which are a substantial part of every extended audit opinion. KAMs enables users of financial statements to understand the matters significance in context of the audit and relate them to the specific circumstances of the audited entity (FRC, 2015). The aim of the KAMs paragraph is to increase the usefulness of the audit opinion for investors. Auditors are mandated to issue an extended audit opinion for listed companies on the UK’s Main Market (FRC, 2015) and for Public Interest Entities (PIEs) in the Netherlands (NBA, 2014).

1.2 Isomorphic pressures on the extended audit opinion

Legitimacy is an important asset for firms to be able to function in society (Power, 2003). This is not different for audit firms and auditors which endure a continuous struggle to gain and/or maintain their legitimacy within society. Changes in the environment of firms creates challenges for firms to maintain their legitimacy (Deegan, 2006). The introduction of the extended audit opinion is such a change in the

(6)

The extended audit opinion: creating useful insights or just ticking the box? 5

environment of audit firms and auditors which poses a challenge for their legitimacy. Audit firms and auditors have to deal with this challenge to maintain their legitimacy, otherwise they are not able to function within society.

One way to enhance or maintain legitimacy is to adapt operations towards a template that is perceived as legitimate within a working field or sector. The competitive nature of a working field winnows out organizations that do not adapt their operations towards this legitimate template. This leads to the homogenization of a working field (DiMaggio and Powell, 1983; Lee and Pennings, 2002). A concept that captures this process of homogenization is called institutional isomorphism. Institutional isomorphism is defined as “a constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions” (DiMaggio and Powell, 1983, p. 149). DiMaggio and Powell created a framework that elaborates on three mechanisms through which isomorphism can occur. The framework states that isomorphism occurs in a coercive, mimetic or normative way due to formal and informal pressures, uncertainty or professionalization respectively (DiMaggio & Powell, 1983). A fundamental consequence of isomorphism is organizational legitimacy: the acceptance of an organization by its external environment (Deephouse, 1996).

When linking isomorphism to the extended audit opinion, this study assumes that auditors use the mechanisms of isomorphism to achieve legitimacy in issuing the extended audit opinion. This is based on the fact that the extended audit opinion is a result of formal (FRC/NBA) and informal (investors) pressures which creates uncertainty in the professional field of auditors. The mechanisms of isomorphism explain how auditors increase their legitimacy by adapting their extended audit opinion towards a template which is perceived as legitimate within the auditing working field. For example, an auditor is able to disclose KAMs that are not company-specific but are perceived as legitimate by investors because other auditors disclose similar KAMs for similar companies. If this is done on a large scale it is possible that the extended audit opinion falls back toward a boilerplate model for companies of a similar nature. If this is the case, the intention of the extended audit opinion, providing company-specific information and increasing the usefulness of the audit opinion for investors, has not been fulfilled. This relationship between isomorphism and the extended audit opinion has not yet been examined. This study tried to fill this gap in existing literature. Leading from this reasoning the main research question is as follows:

To what extent does the process of isomorphism influence the issuing of extended audit opinions for audited entities in the financial sector?

As the main research question states, the focus of this study is on extended audit opinions for companies in the financial sector. The reason for this demarcation is that the extended audit opinion is a result of the most recent financial crisis which has been caused by companies in the financial sector (Claessens, 2010). Moreover, the discussion on the usefulness of the audit opinion was sparked by the fact that financially distressed banks still received unqualified audit opinions (Sikka, 2009). This indicates that, especially for financial companies, more transparency and information about the performed audit is needed.

(7)

The extended audit opinion: creating useful insights or just ticking the box? 6

1.3 Academic contribution and purpose of the study

The extended audit opinion is, at the time of writing, mandatory for four and three years in the UK and the Netherlands respectively. The extended audit opinion has been subject to some research in this period. Prior research focuses on: individual investors’ responses to the extended audit opinion (Lennox, Schmidt & Thompson, 2016), the communicative value of KAMs (Köhler, Ratzinger-Sakel & Theis, 2016a; Christensen, Glover & Wolfe 2014) and auditor liability (Brasel, Doxey, Grenier & Reffett, 2016; Backof, Bowlin & Goodson, 2014). Evidence shows that investors do not find the extended audit opinion very informative and use other sources like management disclosure to assess their risks (Lennox et al., 2016). Results on the communicative value of KAMs are mixed. Köhler et al. (2016a) shows evidence that professional investors perceive communicative value of KAMs as positive while nonprofessional investors perceive KAMs as non-communicative. Contrary to this result Christensen et al. (2014) shows evidence that nonprofessional investors actually do change their decisions when receiving an extended audit opinion with a KAMs paragraph. Evidence on auditor liability shows that the extended audit opinion reduces auditor liability but only for certain circumstances like undetected fraudulent misstatements (Brasel et al., 2016) or defining reasonable assurance in the extended audit opinion (Backof et al., 2014). These studies provide a better insight on the effects of the extended audit opinion but there is a need for more research. The discussion paper of Köhler, Quick and Willekens (2016b) expresses this need by presenting several research ideas for the extended audit opinion. One of these ideas is that disclosing KAMs will become a standard “ticking the box procedure” and no longer contains entity-specific informational value. Reasons for this could be that auditors do not disclose certain KAMs because they want to avoid difficult discussions with the audit committee and the management (Köhler et al., 2016b) or because auditors can gain legitimacy by adapting towards a common template. The concerns in the discussion paper of Köhler et al. (2016b) shows the need for research on the risk that the extended audit opinion falls back towards a boilerplate model and becomes a standard ticking the box procedure. Institutional isomorphism can explain this risk and to the researcher’s knowledge a study on the effects of isomorphism on the issuing of the extended audit opinion has not yet been conducted.

This research tries to contribute to existing literature by analyzing extended audit opinions and assess whether audit firms and auditors are influenced by the process of isomorphism. The goal of this study is to provide a better understanding on the isomorphic pressures that influence the issuing of an extended audit opinion. Research findings should be of interest for investors since they expressed the need for more company-specific information in the audit opinion (Lennox et al., 2016). Moreover, findings should be of interest for regulators and audit firms since results gives these parties better insight in the effectiveness of the imposed regulations and the process of issuing extended audit opinions.

1.4 Thesis structure

This chapter introduced the master thesis. The remainder of the thesis is structured as follows: the second chapter discusses the theoretical framework and develops hypotheses based on the literature discussed. The third chapter describes the research methodology. The fourth chapter presents the results of the performed analyses. The fifth chapter provides an additional analysis based on the results presented in chapter four. The sixth and final chapter presents the conclusion, discussion and ideas for future research.

(8)

The extended audit opinion: creating useful insights or just ticking the box? 7

2. Theoretical framework and hypotheses development

This chapter describes the theoretical framework that forms the basis of this study. The first paragraph describes the legitimacy theory. The second paragraph discusses the institutional theory. The third paragraph describes the mechanisms through which institutional isomorphism occurs. Based on these mechanisms, hypotheses are developed. The fourth paragraph describes the concept of diversity and how it applies to this study. The chapter ends with a visual representation of the conceptual model.

2.1 Legitimacy theory

The legitimacy theory poses that organizations need to be perceived as legitimate within their society to be able to function properly. Organizations can only exist when their operations are accepted within the norms and bounds of their communities and environments (An, Davey and Eggleton, 2011). Deegan (2006) argues that an organization has a social contract with the society in which it operates. An organization has to conduct its operations within the expectations of the society at large and not only within the expectations of shareholders. Organizational legitimacy is not fixed and is subject to changes in society attitudes. This poses a challenge for organizations to adapt to the changing requirements of society in order to stay legitimate (Deegan, 2006). Studies about legitimacy are divided in two groups: strategic legitimacy and institutional legitimacy. The strategic group adopts a managerial perspective on how to gain or maintain legitimacy while the institutional group adopts a more detached stance and emphasizes how sector-wide dynamics generate pressure on organizations to gain or maintain legitimacy (Suchman, 1995). This study builds on the institutional view of legitimacy since the extended audit opinion is a result of sector-wide dynamics (FRC, NBA and investors) rather than being a change induced by an organizational/managerial perspective (audit firms that issue audit opinions). Institutional legitimacy is based on institutional theory which is presented in the following paragraph.

2.2 Institutional theory

When discussing the institutional theory it should be noted that there are two streams of institutionalism, namely old and new institutionalism. Old institutionalism focuses on the institutionalization of organizations self. Selznick (1996) defines institutionalization as “the emergence of orderly, stable, socially integrating patterns out of unstable, loosely organized, or narrowly technical activities” (p. 271). New institutionalisms focuses on legitimation as a driving force behind organizational actors. New institutionalism sees legitimacy as an organizational need to justify particular forms and practices. This justification encourages institutional mimicry which means that the organization is highly sensitive to the environment within it lives (Selznick, 1996). Suchman (1995) adopts a similar view by stating that new institutional theory poses that organizations adapt their operations on pressures created by their environment to be perceived as legitimate. The difference between new and old institutionalism is that in new institutionalism, organizational forms, structural components and rules are institutionalized on a general level between more organizations while in old institutionalism organizations are institutionalized on an individual level (Selznick, 1996).

Meyer and Rowan (1977) argue that institutional rules have effects on organizational structures and the implementation of their actual activities. Formal/bureaucratic organizational structures best reflect the concepts of institutionalism and how actual activities are institutionalized. Meyer and Rowan (1977)

(9)

The extended audit opinion: creating useful insights or just ticking the box? 8

describe the formal structure of an organization as a “blueprint for activities which includes the table of organization: a listing of offices, departments, positions and programs which are linked to explicit goals that make up a rational theory of how, and to what end, activities are to be fitted together.” (p. 341-342). Organizations that incorporate formal structures gain legitimacy, resources, stability and better survival prospects (Meyer and Rowan, 1977). DiMaggio and Powell (1983) build on this by stating that the bureaucratization of organizations is no longer driven by competitiveness or the need for efficiency. They contend that bureaucratization is the effect of institutional change that makes organizations more legitimate but not necessarily more competitive or effective. Many firms converge to a “common template” in their organizational field to enhance their legitimacy. The competitive environment to which firms are exposed winnows out firms that do not converge to this “common template”. This will eventually lead to the homogenization of an organizational field (Lee and Pennings, 2002). A concept that captures this process of homogenization of an organizational field is called institutional isomorphism which is described in the following paragraph.

2.3 Institutional isomorphism and hypotheses development

Institutional isomorphism is defined by DiMaggio and Powell (1983) as: “a constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions” (p. 149). DiMaggio and Powell created a framework on institutional isomorphism to obtain a better understanding of the institutional theory. Their paper identifies three different mechanisms through which isomorphic change occurs: coercive, mimetic and normative isomorphism. DiMaggio and Powell state that this typology is an analytical one. The mechanisms are not always empirically distinct, they tend to derive from different conditions and they may lead to different outcomes. The following subparagraphs discuss the mechanisms of institutional isomorphism and based on these mechanisms hypotheses are developed.

2.3.1 Coercive isomorphism

The first mechanism is coercive isomorphism which stems from political influence. Coercive isomorphism results from both formal and informal pressures exerted on organizations, by organizations upon which they are dependent. Formal pressures can occur in the form of governmental mandates in which organizational change could be a direct reaction to these mandates (DiMaggio and Powell, 1983). Verbruggen, Christiaens and Milis (2011) shows that large Belgian non-profit organizations have a higher level of compliance towards reporting standards when their financial statements are subject to an external audit. This is an example of a coercive isomorphism: the Belgian government enforces external audits for large organizations with the aim to create compliance with reporting standards. This in turn leads to the homogenization of financial statements since all large non-profit organizations comply to the same reporting standards. The fact that Verbruggen et al. (2011) find higher compliance by larger firms could be because larger firms endure increased pressures imposed by the Belgian government or because media and society monitor larger firms more. Chizema and Kim (2010) argue that governments pressurize larger firms more than smaller firms because larger firms have an important role in a government’s economy and because larger firms fill in an exemplary role for smaller firms. Larger firms are also pressurized more by media and society. Media monitors larger firms more than smaller firms, placing them in a public view that is monitored by society (Fang and Peress, 2009). From this view it can be stated that larger firms feel more formal (government) and informal (media and society) pressures which in turn

(10)

The extended audit opinion: creating useful insights or just ticking the box? 9

creates more incentive to engage in coercive isomorphism. Coercive isomorphism could influence the issuing of the extended audit opinion since it was created from both formal and informal pressures. Formal pressures stem from the fact that the extended audit opinion has been mandated by the FRC and NBA for listed companies in the UK and PIEs in the Netherlands respectively (FRC, 2015; NBA, 2014). Informal pressures stem from investors that demand a more informative and useful audit opinion. As stated, larger firms feel more formal and informal pressures by government, media and society. From this reasoning, it is expected that auditors feel more formal and informal pressures when issuing extended audit opinions for larger firms. This in turn creates more incentive to engage in coercive isomorphism when issuing the extended audit opinion. The general assumption behind coercive isomorphism is that organizations become less diverse because they comply to the same set of rules and/or expectations. Taking this point of view, one could expect that extended audit opinions for larger firms are less diverse due to increased coercive pressures. However, this assumption is arguable since formal (FRC/NBA) and informal (investors) pressures are exerted to create more diverse extended audit opinions. Therefore, this study expects that extended audit opinions for larger firms are more diverse because pressures to engage in coercive isomorphism are stronger for larger firms. The following hypothesis is formulated:

Hypothesis 1: Extended audit opinions for larger firms are subject to stronger pressures of

coercive isomorphism, resulting in more diverse audit opinions. 2.3.2 Mimetic isomorphism

The second mechanism through which isomorphic change could occur is mimetic isomorphism. DiMaggio and Powell (1983) state that uncertainty is a strong force that encourages imitation. Organizations that use difficult technologies, have ambiguous goals and/or are active in a volatile environment tend to model themselves on other organizations. This response to uncertainty is defined by DiMaggio and Powell (1983) as ‘modeling’. Organizations that face problems with ambiguous causes or unclear solutions use modeling to find a viable solution with little expense. Haveman (1993) adds to this view by stating that: “organizations imitate organizations within their population, as the actions of these organizations tend to be more salient than the actions of organizations in other populations” (p. 596). Haveman (1993) assumes that organizations in one industry constitute a population and therefore organizations will mimic other organizations within their industry. Organizations in the same population are seen as more important competitors (Porac and Thomas, 1990) and are to be monitored more closely than organizations outside the population. When linking this to the extended audit opinion it can be stated that audit firms operate in a volatile environment due to changing audit mandates which creates uncertainty within the audit profession. The introduction of the extended audit opinion brings this uncertainty within the audit profession. To cope with this uncertainty auditors could model their extended audit opinions on other extended audit opinions that has been issued within the auditing sector. Thus, the uncertainty created by the extended audit opinion could create pressures to engage in mimetic isomorphism when issuing the extended audit opinion.

DiMaggio and Powell (1983) argue that “there is much to be gained in terms of our understanding of the process of isomorphic change by attending to similarity as well as to diversity among organizations and, in particular, to change in the degree of homogeneity or variation over time” (p. 158). Slack and Hinings (1994) address this issue of change in the degree of homogeneity over time. They use national-level sport

(11)

The extended audit opinion: creating useful insights or just ticking the box? 10

organizations that are subject to environmental pressures to adopt a more bureaucratic organizational design, to examine mimetic isomorphism. Their results show that, over time, the sport organizations begin to mimic each other which results in a decrease of organizational diversity. This result is backed up by other studies that also confirm that organizations become less diverse when time progresses (Carolan, 2008; Ashworth, Boyne and Delbridge, 2007). Based on this information the expectation is that, when time progresses, extended audit opinions become less diverse due to mimetic isomorphism. When the extended audit opinion was introduced in 2013, auditors could not engage in mimetic isomorphism because examples of extended audit opinions were not available. However, in the following year auditors can engage in mimetic isomorphism and mimic extended audit opinions which are perceived as legitimate. As time progresses more extended audit opinions are issued which makes it easier for auditors to engage in mimetic isomorphism since there are more examples. Therefore, it is expected that more recent extended audit opinions are less diverse, because auditors engage in mimetic isomorphism faster. Based on this information the following hypothesis is formulated:

Hypothesis 2: More recent extended audit opinions are subject to stronger pressures of mimetic

isomorphism, resulting in less diverse audit opinions. 2.3.3 Normative isomorphism

The third mechanism through which institutional isomorphism occurs is normative isomorphism. This stems primarily from professionalization which is defined by DiMaggio and Powell (1983) as “the collective struggle of members of an occupation to define the conditions and methods of their work, to control the production of producers and to establish a cognitive base and legitimation for their occupational autonomy” (p. 152). Two aspects of professionalization are important sources of normative isomorphism. The first aspect is formal education and the cognitive base produced by universities. For example, auditors that attended the same university are exposed to a similar source of socialization and create a similar view on the world and their profession. The second source is the growth and elaboration of professional networks across organizations which creates new models within the profession (DiMaggio and Powell, 1983). Socialization within a profession leads to routines and a “taken-for-granted” use of new models. Professional networks helps diffusing new models and contribute to the homogenization within a profession (Beckert, 2010). Normative isomorphism could influence the issuing of the extended audit opinion since it is created by auditors that have a similar educational background and use their professional network to elaborate on the extended audit opinion. This could lead to a taken-for-granted model when issuing an extended audit opinion which in turn results in less diverse extended audit opinions.

The professional field of auditors is dominated by the Big 4 audit firms PwC, KPMG, Deloitte and EY. A big part of the extended audit opinions is issued by Big 4 firms. A survey of the FRC consisting of 278 extended audit opinions in the UK shows that 97% is issued by Big 4 firms (FRC, 2016). In the Netherlands a similar survey has been conducted which shows that 76 out of the 77 extended audit opinions are issued by Big 4 auditors (NBA, 2015). Big 4 audit firms have a professional network in which auditors elaborate on job specific issues like the extended audit opinion. Elaboration on these networks could lead to the creation and diffusion of a model of the extended audit opinion. This in turn could lead to a taken-for-granted use of this model which contributes to the homogenization of the extended audit opinion within an audit firm.

(12)

The extended audit opinion: creating useful insights or just ticking the box? 11

Subsequently, the expectation is that extended audit opinions of the same audit firm are less diverse than extended audit opinions of other audit firms due to the diffusion of a model within the audit firm. Based on this information the following hypotheses is formulated:

Hypothesis 3: Extended audit opinions that are issued by auditors of the same audit firm are

subject to normative isomorphism, resulting in less diverse audit opinions.

2.4 The concept of diversity

The proposed hypotheses emphasize the diversity of the extended audit opinions. An explicit specification of diversity is important to enhance the understandability of the hypotheses (Harrison and Klein, 2007). Prior research has struggled to conceptualize and study diversity effectively because varied theoretical perspectives are used. One perspective draws on similarity attraction (Byrne, 1971) and social categorization theories (Turner 1985, Tajfel, 1981). A second perspective focuses on the principle of variety to highlight heterogeneity benefits (Ashby, 1956). A third perspective focuses on differences in hierarchy or status characteristics theories (Blau, 1960; Berger, Fisek, Norman and Zelditch, 1977). Based on these perspectives Harrison and Klein (2007) provided three different types of diversity. These are separation, variety and disparity. Separation encompasses that, within units, members differ from one another in their position along a single continuous attribute (e.g., perception of leader charisma, organizational commitment). Variety involves that, within units, members are different from one another on a categorical attribute (e.g., gender of board member, functional background). Disparity is that, within units, members differ to which they hold or receive a share of a certain attribute (e.g., power, pay, status). The following figure gives a visual representation of the different types of diversity:

(13)

The extended audit opinion: creating useful insights or just ticking the box? 12

The diversity of the extended audit opinion can be influenced by the KAMs paragraph. The KAMs paragraph enables users of financial statements to understand the matters significance in context of the audit and relate them to the specific circumstances of the audited entity (FRC, 2015). Consequently the KAMs paragraph reflects the professional judgement of the auditor on whether issues during the audit have to be reported. The disclosure of KAMs can reflect coercive isomorphism because, for example, an auditor will report more KAMs for larger audited entities than for smaller entities with the argument to better be safe than sorry. The KAMs can also reflect mimetic and normative isomorphism because auditors mimic other KAMs due to uncertainty or because the auditors elaborate within their firm on which KAMs have to be reported. When linking the different measures of diversity to the KAMs, the variety measure is applicable. The KAMs can be categorized which enables measurement of the variety of the KAMs, and consequently the diversity of the extended audit opinion.

2.5 Conceptual model

Based on the proposed hypotheses the following conceptual model is established:

Figure 2: Conceptual model

Diversity of the extended

audit opinions

(Variety of KAMS)

Coercive isomorphism

(Firm size of audited entity)

Mimetic isomorphism

(Recency of the extended

audit opinion)

Normative isomorphism

(Extended audit opinions

issued by the same audit

(14)

The extended audit opinion: creating useful insights or just ticking the box? 13

3. Research Methodology

This chapter describes the research methodology used to test the proposed hypotheses. The first paragraph discusses how the sample is selected and what size it is. The second paragraph describes the dependent, independent and control variables used in this study and how these has been measured. The third and last paragraph provides an oversight of the used variables and sets out the statistical model.

3.1 Sample

The data that has been used in this study is collected manually from the annual reports of listed companies in the UK and the Netherlands. The data is collected by a research group consisting of students writing their master thesis about the extended audit opinion. The UK sample consists of 251 companies linked to the FTSE 100 and FTSE 250. The Dutch sample consists of 73 companies that are listed on the AEX, AMX or AScX. This results in a total sample size of 324 companies with an extended audit opinion. This is not the final sample size since several requirements have to be met. Firstly, companies with an extended audit opinion that does not contain KAMs are excluded from the sample. Secondly, companies that received an extended audit opinion from a non-Big 4 auditor are excluded. Finally, as stated in paragraph 1.2, this study focuses on extended audit opinions for companies in the financial sector. This means that only companies active in the financial sector are selected in the sample. To distinguish the companies that are active in the financial sector the Standard Industrial Classification (SIC) code in the Orbis database is used. When applying these filters the total sample size consists of 82 companies. For these companies 279 extended audit opinions are issued between 2013 up to and including 2016. Appendix one shows all companies included in the sample.

3.2 Variables

This section describes the dependent, independent and control variables that are used in this study and how these variables were measured.

3.2.1 Dependent variable

The goal of this thesis is to determine whether the process of isomorphism influences the issuing of extended audit opinions. The different mechanisms of isomorphism could influence the diversity of extended audit opinions. Therefore, the dependent variable of this study is the diversity of the extended audit opinion (DIV). In this study the diversity of the extended audit opinions is measured by the variety of the KAMs disclosed in the extended audit opinion.

Variety is, as discussed in paragraph 2.4, diversity within an unit based on the fact that they differ from each other on categorical attributes. To be able to measure the diversity of the extended audit opinions, the KAMs are coded in nineteen categories. Appendix two shows which categories are used to operationalize the dependent variable. The categorization of the KAMs is done by seven researchers of a research group of which the researcher is part of. This means that KAMs are categorized by different group members which could lead to inconsistencies between the categorizations of the KAMs. To ensure that the KAMs are categorized in a consistent manner the categorization is reviewed and supervised by another member of the research group. This member collects the data from all the seven researchers and maintains the data-set. The categorization of the KAMs for the years 2013-2015 is supervised and reviewed by this group member. The categorization of the KAMs of 2016 is not supervised and reviewed by this group member due to the timing of the data collection and the completion of the master thesis.

(15)

The extended audit opinion: creating useful insights or just ticking the box? 14

The variety of the KAMs needs to be operationalized to be able to make statistical inferences about the diversity of the extended audit opinion (Harrison and Klein, 2007). To operationalize the variety of KAMs, the Blau’s index of heterogeneity (Blau, 1977) has been used. This is a commonly used index which has been suggested as an optimal measure of diversity to capture variations within a group (Harrison and Klein, 2007; Miller and del Carmen Triana, 2009). The Blau index assigns a numeric value to each extended audit opinion which enables statistical analysis. The following formula is used to calculate the Blau index:

𝐵𝑙𝑎𝑢′𝑠 𝑖𝑛𝑑𝑒𝑥 𝑜𝑓 ℎ𝑒𝑡𝑒𝑟𝑜𝑔𝑒𝑛𝑒𝑖𝑡𝑦:

1 −

(𝑝

𝑘 2

)

In this formula variety can take k = 1, 2, …, K possible categories and p represents the proportion of KAMs in the kth category. The Blau index always has a value between 0 – 1 and the higher the Blau index, the more variety in the KAMs is observed. This in turn translates to a more diverse extended audit opinion. The Blau index of each extended audit opinion has been calculated as follows: for example, an extended audit opinion has 5 KAMs from which 2 focus on goodwill, 1 on current assets and 2 on implicit fraud. The calculation for the Blau index is: (1 - (2/5)2 + (1/5)2 + (2/5)2) = .64. A statistical interpretation of this number is that the chance that two randomly selected KAMs are different is 64%.

Important to note is that the Blau index is influenced by the amount of categories available within an unit (Harrison and Klein, 2007). This is the case in this study because not all extended audit opinions have the same amount of KAMs. The maximum possible variety of KAMs increases when an extended audit opinion contains more KAMs. Audit opinions with more KAMs, have a “richer” set of possible categories to draw from and thus have a higher theoretical maximum of diversity. One way to mitigate this issue is by standardizing the Blau index. The Blau index is standardized by dividing the unstandardized Blau index by its theoretical maximum which yields the Index of Quality Variation (IQV) (Agresti and Agresti, 1978). The theoretical maximum of diversity is computed by (K – 1) / K (Harrison and Klein, 2007). When applying this to the example the theoretical maximum is: (5-4)/5 = .80, and the IQV is: .64/.80 = .80.

This study uses both the Blau Index as the IQV to measure the diversity of the extended audit opinions. The reason for this is explained by elaborating two points of view. On the one hand the Blau index gives good insights in the diversity of the extended audit opinion because it takes into account the amount of KAMs disclosed. More KAMs means that the Blau index can reach a higher value and thus the diversity of the extended audit opinion is higher. For example, an extended audit opinion with six KAMs is more diverse than an extended audit opinion with three KAMs. On the other hand the IQV gives good insights in the diversity of the extended audit opinion because it mitigates the influence of the amount of KAMs disclosed. The IQV measures only the variety of the extended audit opinion in terms of categories of KAMs disclosed in the extended audit opinion. An example of this is that an extended audit opinion with nine KAMs of which seven are in the same category is less diverse than an extended audit opinion with five KAMs that all are in a different category.

To summarize, the variety of the KAMs reflects the diversity of the extended audit opinion which is the dependent variable of this study. The variety of the KAMs is operationalized in two ways. Firstly by the Blau index which yields the unstandardized diversity of the extended audit opinion (DIV) and secondly by the IQV which yields the standardized diversity of the extended audit opinion (STD_DIV).

(16)

The extended audit opinion: creating useful insights or just ticking the box? 15 3.2.2 Independent variables

This study used three different independent variables to test the hypotheses. These are the size of the audited entity (SIZE), the financial year on which the audit opinion is formed (FY) and which Big 4 audit firm issued the extended audit opinion (BIG4).

The first independent variable, the size of the audited entity, is measured by the natural logarithm of the total assets. The annual reports in the sample are issued in pounds, euros or dollars. To be able to compare firm size properly, annual reports in euros and dollars are converted to pounds since most annual reports are issued in pounds. The conversion rate of 31 December of the year of the annual report is used. So, for the annual reports of 2013 the conversion rate of 31 December 2013 is used, for the annual reports of 2014 the conversion rate of 31 December 2014 is used and so on. The conversion rates are displayed in appendix three.

The second independent variable, the year on which the extended audit opinion is formed, is measured by the end of the financial year of the audited entity. This is also the case for entities with a split financial year. So, for example, an extended audit opinion issued for an entity with a split financial year of 2015-2016 was measured as an extended audit opinion that is issued in 2015-2016.

The third independent variable is the audit firm that issued the extended audit opinion. The sample exists only of extended audit opinions that are issued by Big 4 audit firms which are PwC, KPMG, Deloitte and EY. For these audit firms a dummy variable is created, with PwC as point of reference. Appendix three shows how the dummy-coding is applied.

3.2.3 Control variables

The goal of this study is to establish whether the process of isomorphism influences the issuing of extended audit opinions. This has been tested by using the variables described in previous paragraphs. However, there are more variables which might be of influence on the diversity of the extended audit opinions but are not in the scope of this study. To ensure that only the effect of the independent variables is measured, control variables are added to the analyses. The control variables used in this study are: the applied materiality during the audit (MAT), the frequency of audit committee meetings of the audited entity (FREQ_ACM) and the profitability of the audited entity (PROF).

With the introduction of the extended audit opinion auditors have to disclose how they applied the concept of materiality during the audit (FRC, 2015). Materiality can be described as the risks of misstatements in the financial statements that could influence the economic decisions of the users of the financial statements (IAASB, 2016). Thus, auditors that apply a lower materiality are more critical on financial misstatements than auditors that apply a higher materiality. Consequently, lower materiality leads to more risks of material misstatement. Empirical evidence has shown that the level of materiality is positively related to the amount of KAMs disclosed. The reasoning behind this is that higher materiality indicates less auditor effort which leads to more KAMs (Sierra-Garcia, Gambetta, García-Benau and Orta-Pérez, 2017). The amount of KAMs disclosed can influence the variety of KAMs and thus the applied materiality is controlled for when conducting the study. The variable materiality has been measured by the natural logarithm of the materiality disclosed in the extended audit opinion. The disclosed materiality is conversed into pounds in a similar way as the size of the audited entity.

(17)

The extended audit opinion: creating useful insights or just ticking the box? 16

The frequency of audit committee meetings, which is the second control variable, could influence the amount of KAMs disclosed. Empirical evidence has shown that there is a positive relation between the frequency of audit committee meetings and the quality of the financial statements (De Vlaminck and Sarens, 2015; Li, Mangena and Pike, 2012). This could influence the diversity of the extended audit opinion because financial statements of better quality can lead to less KAMs. However, the study of Stewart and Munro (2007) indicate that when an audit committee meets on a more frequent basis, auditors spend more hours to the audit. This in turn could lead to more KAMs disclosed because the auditor spends more time to discuss issues with the audit committee. Thus the frequency of audit committee meetings could either lead to more or less KAMs disclosed which in turn influences the variety of the KAMs. Therefore, the frequency of audit committee meetings is controlled for in this study. The control variable is measured by the amount of the audit committee meetings disclosed in the annual report of the audited entity. The third control variable used in this study is the profitability of the audited entity. The profitability of the audited entity influences the amount of KAMs disclosed in the extended audit opinion. Firms that are more profitable use less creative accounting and have less discussion with the auditor. Consequently, more profitable firms tend to receive an unqualified audit opinion (Beasley, Carcello & Hermanson, 1999; Laitinen & Laitinen, 1998). When linking this evidence to the amount of disclosed KAMs it can be stated that more profitable firms receive less KAMs because more profitable firms use less creative accounting and have less discussions with auditors. This relationship is confirmed by the study of Morais and Pinto (2017). Thus, the more profitable the audited entity, the less KAMs are disclosed. This in turn influences the diversity of the extended audit opinion. This influence has been controlled for by creating a control variable. The profitability of the audited entity is measured by the Return On Assets (ROA) which is a widely used measure for firm profitability (Aupperle, Carrol & Hatfield, 1985; Griffin & Mahon, 1997). ROA is calculated with the following formula:

𝑅𝑂𝐴 = 𝑁𝑒𝑡 𝐼𝑛𝑐𝑜𝑚𝑒 𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠

The ROA ratios have been collected from the Orbis database or are calculated by hand when not available in Orbis. When calculated by hand the net income of the firm has been obtained using Compustat. The total assets of the firms were collected manually by the research group.

(18)

The extended audit opinion: creating useful insights or just ticking the box? 17

3.3 Statistical model

This study is conducted in a quantitative way by performing statistical tests. A multiple regression analysis is used to test the influence of coercive isomorphism (the size of the audited entity), mimetic isomorphism (the recency of the extended audit opinion) and normative isomorphism (the issuing audit firm) on the diversity of the extended audit opinions. The following regression models are used:

𝐷𝐼𝑉 = 𝛽0+ 𝛽1𝑆𝐼𝑍𝐸 + 𝛽2𝐹𝑌 + 𝛽3𝐵𝐼𝐺4 + 𝛽4𝑀𝐴𝑇 + 𝛽5𝐹𝑅𝐸𝑄_ACM+ 𝛽6𝑃𝑅𝑂𝐹 + 𝜀𝑖

𝑆𝑇𝐷_𝐷𝐼𝑉 = 𝛽0+ 𝛽1𝑆𝐼𝑍𝐸 + 𝛽2𝐹𝑌 + 𝛽3𝐵𝐼𝐺4 + 𝛽4𝑀𝐴𝑇 + 𝛽5𝐹𝑅𝐸𝑄_ACM+ 𝛽6𝑃𝑅𝑂𝐹 + 𝜀𝑖

In this formula the constant is β0, βi represents the variables used in this study and ɛi is the distortion term. The following table summarizes the variables used in the regression models:

Table 1: Description of variables

Variable Measure

Dependent (DIV) (STD_DIV)

Blau index of heterogeneity Index of Quality Variation Independent

(SIZE) (FY) (BIG4)

The natural logarithm of the total assets

The financial year on which the extended audit opinion is issued (2013-2016) The Big 4 audit firm that issued the extended audit opinion (PwC, KPMG, Deloitte, EY)

Control (MAT) (FREQ_ACM)

(PROF)

The natural logarithm of the materiality disclosed in the extended audit opinion The amount of audit committee meetings disclosed in the annual report of the audited entity

Return On Assets obtained from the Orbis database or calculated by hand: net income/total assets

(19)

The extended audit opinion: creating useful insights or just ticking the box? 18

4. Results

In this chapter the results of the conducted research is presented. The first paragraph describes the transformation of outliers in the dataset. The second paragraph consists of the descriptive statistics of the different variables. Paragraph three discusses the correlation analysis that checks for multicollinearity. The fourth paragraph shows the results of the performed regression analyses that test the proposed hypotheses.

4.1 Dataset

Before performing the regression analyses the data is checked for outliers. Outliers skew the normal distribution of data which can lead to distorted results (Field, 2013). Outliers are identified and replaced by a process called winsorizing. Outliers are identified by creating boundaries that are calculated by the mean of a variable plus/minus the standard deviation times three (Field, 2013). All the data that falls outside these boundaries were revalued by the mean plus/minus three times the standard deviation. This resulted in the upward adjustment of 4 values to 15,4889 in the independent variable SIZE and the downward adjustment of 3 values to 26,17% in the control variable PROF.

4.2 Descriptive statistics

Table 2 shows the descriptive statistics of the variables used in this study. The table shows the sample size, mean, standard deviation, minimum and maximum of the different variables. In total 279 extended audit opinions are used to test the proposed hypotheses. The unstandardized diversity (Blau index) of the extended audit opinion is 0,5597 on average with a standard deviation of 0,2372. This means that the chance that two randomly selected KAMs from the entire sample are different is 55,97%. The standardized diversity (IQV) of the extended audit opinion shows a mean of 0,8161 with a standard deviation of 0,3261. The chance that two KAMs are different when measuring the extended audit opinion diversity in a standardized way is 81,61%. The natural logarithm of the size of the audited entity is on average 22,85 with a standard deviation of 2,40. The natural logarithm of materiality shows a mean of 16,94 and has a standard deviation of 1,27. On average all audit committees had 5,16 meetings in a year with a minimum of 0 meetings and a maximum of 13 meetings. The profitability of all the audited entities in the sample is on average 5,50% and ranges from -9,09% to 26,17%. The standard deviation is 6,81% which is higher than the mean. This indicates that the profitability of the firms is quite spread out.

Table 2: Descriptive statistics

Variable N Mean Std. Deviation Minimum Maximum

DIV 279 0,5597 0,2372 0,0000 0,8571 STD_DIV 279 0,8161 0,3261 0,0000 1,0000 SIZE 279 22,85 2,40 15,49 28,16 MAT 279 16,94 1,27 12,90 20,25 FREQ_ACM 279 5,16 2,10 0 13 PROF 279 5,50% 6,81% -9,09% 26,17%

In table 2 the independent variables years and audit firm are not presented because the mean, standard deviation, minimum and maximum do not provide useful information about these variables. Table 3 shows the descriptive statistics of each year included in the study and table 4 shows the descriptive of the Big 4 audit firms used in the study. The most (80) extended audit opinions are issued for the year 2016

(20)

The extended audit opinion: creating useful insights or just ticking the box? 19

and the least (45) for 2013. For the years 2014 and 2015 75 and 79 audit opinions have been issued respectively. The reason why less extended audit opinions were issued for the year 2013 is because not all UK companies received an extended audit opinion and because the extended audit opinion was not mandatory for Dutch firms (FRC, 2015; NBA, 2014). The descriptive statistics for the independent variable audit firm are as follows: PwC issued the most extended audit opinions with 107 audit opinions. KPMG issued 68 audit opinions and both Deloitte and EY issued 52 audit opinions.

Table 3: Descriptive statistics per year Table 4: Descriptive statistics per audit firm

Year Frequency Percentage Cumulative Accountant Frequency Percentage Cumulative

2013 45 16,1% 16,1% Deloitte 52 18,6% 18,6%

2014 75 26,9% 43% EY 52 18,6% 37,3%

2015 79 28,3% 71,3% KPMG 68 24,4% 61,6%

2016 80 28,7% 100% PwC 107 38,4% 100%

Total 279 100% Total 279 100%

As stated in paragraph 3.2.2 a dummy variable for the issuing audit firms is created. When making use of a dummy variable a rule of thumb is to work with a reference group which is left out of the rest of the analyses. It is usual to choose the variable that is the largest when selecting a reference group. In this study PwC issued the most extended audit opinions and thus, is taken as the point of reference and left out in the rest of the analyses.

4.3 Correlation analysis

Before the results of the regression analyses are presented a correlation analysis is performed. The correlation analysis checks for multicollinearity between the variables used in this study. If multicollinearity exists it is hard to obtain unique estimates of the regression coefficients. Multicollinearity is assumed when the correlation between two variables is higher than 0.8 or lower than -0.8 (Field, 2013). Table 5 shows the results of the correlation analysis.

Table 5: Correlation matrix

Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 1. DIV 1 2. STD_DIV ,907** 1 3. SIZE ,326** ,162** 1 4. FY -0,102 -0,047 -0,051 1 5. KPMG -,207** -,187** 0,015 -0,026 1 6. Deloitte 0,001 -0,035 -,136* -0,036 -,272** 1 7. EY 0,025 0,075 0,071 0,034 -,272** -,229** 1 8. MAT ,152* 0,038 ,696** -0,051 0,010 -0,037 -0,004 1 9. FREQ_ACM ,291** ,155** ,576** -0,036 0,104 -,138* -0,085 ,383** 1 10. PROF -,242** -,212** -,436** -0,016 -0,095 0,001 -0,016 -,157** -,310** 1 ** Correlation is significant at the 0.01 level (2-tailed)

(21)

The extended audit opinion: creating useful insights or just ticking the box? 20

From the results in table 5 it can be stated that multicollinearity exists between the unstandardized diversity and the standardized diversity of the extended audit opinion (.907). This is not a problem for the regression analyses since both variables are used as dependent variables and are not placed into the same regression models. Based on the correlation matrix two significant relationships exist between the dependent and independent variables. Firstly, a positive relationship between the diversity of the extended audit opinion and the size of the audited entity is identified (.326 and .162). This indicates that when the size of the audited entity increases, the diversity of the extended audit opinion increases. Secondly a negative relationship between the diversity and audit firm KPMG is identified (-.207 and 187). This indicates that KPMG issues less diverse audit opinions in comparison with PwC which is the reference group. Important to note is that these results are only indicative. The conclusive results are presented in the following paragraph in which the regression analyses are discussed.

4.4 Regression analysis

This paragraph presents the results of the regression analyses. Since the dependent variable is measured in two different ways each regression is performed twice with the only change being the dependent variable. In the models labeled ‘A’ the unstandardized diversity of the extended audit opinion is the dependent variable. In the models labeled ‘B’, the standardized diversity of the extended audit opinion is the dependent variable. Five different models are used to perform the regressions. The first model includes the dependent variable and all control variables used in the study. In the second, third and fourth model an independent variable is added to the first model to test the proposed hypotheses. The fifth model includes all variables used in this study.

4.4.1 Control variables

The first model of the regression analysis shows the relationship between the dependent variables and the control variables. The results of this analysis are presented in table 6:

Table 6: Results regression analysis model 1

Variables Model 1A Model 1B

Constant ,334* ,920***

MAT ,008 -,009

FREQ_ACM ,025*** ,017*

PROF -,006*** -,009***

Adjusted R-Square ,102 ,044

***Significant on a 1% level, **Significant on a 5% level, *Significant on a 10% level

The adjusted R-square of model 1A is ,102 which indicates that the variation in the diversity of the extended audit opinion is explained for 10,2% by the control variables. In model 1B the control variables explain the variation in the standardized diversity of the extended audit opinions for 4,4%. Model 1A indicates a significant positive relation (β = 0,025; p < 0,01) between the frequency of audit committee meetings and the unstandardized diversity of the extended audit opinion. The model also indicates a negative significant relationship (β = -0,006; p < 0,01) between the profitability of the audited entity and the diversity of the extended audit opinion. Similar relationships are recorded in model 1B although the relationship between audit committee meetings and the standardized diversity of the extended audit opinion is weaker and less significant (β = -0,017; p < 0,10). The relationship between profitability and the standardized diversity of the extended audit opinion is a little bit stronger and on the same level of significance (β = -0,009; p < 0,01). No significant relationship (β = 0,008; p > 0,10 / β = -0.009; p > 0,10)

(22)

The extended audit opinion: creating useful insights or just ticking the box? 21

between the materiality of the performed audit and the diversity of the extended audit opinion was observed. Thus, both models indicate that when the firm’s audit committee has more meetings, the diversity of the extended audit opinions increases and when firms are more profitable, the diversity of the extended audit opinion decreases.

4.4.2 Size of the audited entity (H1)

The second model is used to test the first hypothesis which proposes that: extended audit opinions for larger firms are subject to stronger pressures of coercive isomorphism which results in more diverse audit opinions. This means that a positive relationship between firm size and the diversity of the extended audit opinion is expected. Table 7 shows the results of the performed regression to test hypothesis 1:

Table 7: Results regression analysis model 2

Variables Model 2A Model 2B

Constant ,214 ,849*** SIZE ,028*** ,016 MAT -,021 -,025 FREQ_ACM ,016** ,012 PROF -,003 -,007** Adjusted R-Square ,126 ,046

***Significant on a 1% level, **Significant on a 5% level, *Significant on a 10% level

When including the independent variable size in the regression model, the explanatory power increases from 10,2% in model 1A to 12,6% in model 2A. The explanatory power of model 2B increases from 4,4% to 4,6%. Model 2A shows a significant positive relationship (β = 0,028; p < 0,01) between the size of the audited entity and the diversity of the extended audit opinion. However, as presented by model B, when the diversity of the audited entity is standardized this relationship is not significant (β = 0,016; p >0,10). Thus, hypothesis 1 can be accepted when using the unstandardized diversity of the extended audit opinion but is rejected when using the standardized diversity of the extended audit opinion as dependent variable. 4.4.3 Recency of the extended audit opinion (H2)

The third model tests the second hypothesis which proposes that: more recent extended audit opinions are subject to stronger pressures of mimetic isomorphism, resulting in less diverse audit opinions. A negative relationship is expected between the year on which the extended audit opinion is issued and the diversity of the extended audit opinion. Table 8 presents the results of the performed regression analysis to test hypothesis 2:

Table 8: Results regression analysis model 3

Variables Model 3A Model 3B

Constant 43,527* 30,514 FY -,021* -,015 MAT ,007 -,009 FREQ_ACM ,025*** ,017* PROF -,006*** -,009*** Adjusted R-Square ,107 ,043

(23)

The extended audit opinion: creating useful insights or just ticking the box? 22

When including the year on which the extended audit opinion is issued in the regression model the explanatory power of model A increases from 10,2% to 10,7%. For model B a decrease in explanatory power is observed from 4,4% to 4,3%. Model 3A shows a negative significant relationship (β = -0,028; p < 0,10) between the year on which the extended audit opinion is issued and the diversity of the extended audit opinion. When standardizing the diversity of the extended audit opinion this relation is still negative but not significant (β = -0,015; p > 0,10). This means that, similarly to hypothesis 1, hypothesis 2 can be accepted when using the unstandardized diversity of the extended audit opinion but is rejected when using the standardized diversity of the extended audit opinions as dependent variable.

4.4.4 Issuing audit firm (H3)

The fourth model tests the third hypothesis which proposes that: extended audit opinions that are issued by auditors of the same audit firm are subject to normative isomorphism, resulting in less diverse audit opinions. A negative relationship is expected between all audit opinions of one issuing audit firm and the diversity of the extended audit opinion. Table 9 presents the results of the performed regression analysis to test hypothesis 3:

Table 9: Results regression analysis model 4

Variables Model 4A Model 4B

Constant ,405** 1,011*** KPMG -,152*** -,185*** Deloitte -,031 -,073 EY -,027 -,004 MAT ,006 -,011 FREQ_ACM ,027*** ,019* PROF -,007*** -,010*** Adjusted R-Square ,158 ,089

***Significant on a 1% level, **Significant on a 5% level, *Significant on a 10% level

When including the issuing audit firms into the regression model the explanatory power of model A increases to 15,8% and of model B to 8,9%. Both model 4A (β = -0,152; p < 0,01) as 4B (β = -0,185; p < 0,01) shows a significant negative relationship between dummy variable KPMG and the diversity of the extended audit opinion. Both model A and B indicate an insignificant negative relationship for the dummy variables Deloitte (β = -0,031; p > 0,10 / β = -0,073; p > 0,10) and EY (β = -0,027; p > 0,10 / β = -0,004; p > 0,10). This result indicates that the extended audit opinions of KPMG are less diverse than the audit opinions of PwC. This however, is not the case for the extended audit opinions issued by Deloitte and EY. Because of the insignificant relationships observed for Deloitte and EY the third hypothesis is rejected.

(24)

The extended audit opinion: creating useful insights or just ticking the box? 23 4.4.5 All variables

The fifth model includes all variables used in the study. This model mainly serves to analyze the explanatory power of all variables. Table 10 presents the results of the regression analysis which includes all the independent and control variables:

Table 10: Results regression analysis model 5

Variables Model 5A Model 5B

Constant 44,558* 35,776 SIZE ,025*** ,010 FY -,022* -,017 KPMG -,148*** -,184*** Deloitte -,022 -,071 EY -,036 -,007 MAT -,020 -,021 FREQ_ACM ,018** ,015 PROF -,004** -,009*** Adjusted R-Square ,184 ,087

***Significant on a 1% level, **Significant on a 5% level, *Significant on a 10% level

Model 5A shows that 18,4% of the variation in the diversity of the extended audit opinion is explained by all the independent and control variables used in this study. The explanatory power of all independent and control variables in model 5B is 8,7%. This indicates that variables used in this study have higher explanatory power when diversity of the extended audit opinion is unstandardized. The direction (positive or negative) and significance of the relationships between the dependent and independent variables observed in model 5A and 5B are similar to the relationships observed in models 2A, 2B, 3A, 3B, 4A and 4B.

(25)

The extended audit opinion: creating useful insights or just ticking the box? 24

5. Additional analysis

The results presented in the previous chapter provided useful insights on the effect of institutional isomorphism on the issuing of the extended audit opinion. In order to test the validity of these results, interviews were held with professionals from one of the Big 4 audit firms. Two interviews were held in which the results of the previous paragraph were discussed. The first paragraph of this chapter describes the design of the interviews held. In paragraph two the results of the interviews are presented and discussed.

5.1 Interview design

To increase the validity of the results presented in the previous chapter two interviews were held. The interviews were held in a semi-structured way. In this format a series of questions is asked and additional questions based on the answers from the interviewee can be asked (Smith, 2015). Open questions are asked to give the interviewees the best chance to give their view on the results presented in chapter four and to encourage discussion about the presented results. Interviews were held with two professionals working at one of the Big 4 audit firms. Both professionals have substantial experience with the issuing of the extended audit opinion. The first professional issued extended audit opinions for four different insurance companies. The other professional issued extended audit opinions for four listed companies. With this interview, professionals can place the results of this study into context with their experience. This places the results of the previous chapter in a more practical point of view when compared with the theoretical approach adopted in the previous chapters. The questions asked during the interviews are presented in appendix four.

5.2 Interview results

In this paragraph the results of the conducted interviews are presented. The first subparagraph shows the general view of the interviewed professionals on the extended audit opinion. The second subparagraph shows the discussion regarding the results of the tested hypotheses. The third subparagraph presents the discussion regarding the significant results of the control variables.

5.2.1 The extended audit opinion

In general, professionals argue that the extended audit opinion is a positive development within the audit profession. The transparency provided by the extend audit opinion is embraced by the professionals since external parties like funders, investors and employees expressed their the need for more transparency. However, within the audit profession there is a discussion about what and how much information needs to be disclosed. Auditors are still searching for a good balance regarding the level of disclosure in the extended audit opinion. There are two sides captivated on this disclosure issue. On the one hand, auditors state that a maximum amount of information should be disclosed in order to increase information transparency. On the other hand, auditors state that the disclosure of information should be within bounds because the interpretation of information could harm the audited entity.

The intention of the extended audit opinion is clear since both professionals state that the extended audit opinion should be a reflection about the performed audit. Both professionals argue that the extended audit opinion should at least provide information about which issues occurred during the audit and how these issues were resolved. Moreover, professional judgement should be used when composing the extended audit opinion.

Referenties

GERELATEERDE DOCUMENTEN

The aim of this chapter is to find a suitable hedg- ing strategy such that the risk of the difference of the hedging portfolio and the claim is minimized under a simple spectral

Methods/Design: The MESSI trial (Mechanochemical Endovenous ablation versus radiofrequency ablation in the treatment of primary Small Saphenous vein Insufficiency) is a

In the previous chapter, we looked into the results of my empirical study to answer the question of how first-in-family Students of Colour experience higher

Tiago Filipe Montes de Jesus University of

The mega-sporting events taken into account within this paper will be the summer and winter Olympics, the FIFA World Championships, and the UEFA European Championships

By analysing 350 cases of conflict situations spread over the world, we explore how the governance systems (i.e. community’s concentration of power and contestation

Again, large connectivity changes occurred across the first stimulation period, and much smaller changes upon subsequent periods, indicating that the network also

Here, we combine Species Distribution Models (SDMs), field surveys and molecular analyses to shed light on the presence of a cryptic long-eared bat species from North Africa,