The UK corporate governance quality : a female influence on compliance quality with the corporate governance code

(1)

Amsterdam Business School MSC Accountancy & Control

The UK Corporate Governance Quality

A female influence on compliance quality with the corporate governance code

Name: Basma Moussaoui Student number: 10424725 Date: 19 June 2016

Word count: 13767

Thesis supervisor: Georgios Georgakopoulos

MSc Accountancy & Control, specialization track: Control Faculty of Economics and Business, University of Amsterdam

(2)

2

Statement of originality

This document is written by Basma Moussaoui who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

This study is conducted to examine the effect of gender on corporate governance quality. Inspired and motivated by the upcoming gender quota by 'Davies review’ the study is focused on the female participation in boards and the quality of compliance with the UK Corporate Governance Code. The research is based on the publicly listed top 100 UK firms and the study is based on fully hand-collected data of exclusively observations from 2014 in addition to the latest revision of the UK Corporate Governance code in 2012. The focus on the UK firms was based on the fact that The UK corporate governance code is currently seen as one of the most developed operative codes for best practice provisions, seen the UK had the longest history of principle-based standards (Seidl et al, 2009).The study is conducted by the application of a content analysis based on a hand collection of data, issuing taxonomies for the determination of the quality of compliance.

In line with former gender studies, this study expected a positive relationship between female participation and the quality of corporate governance.

No important significant results found conducting the research, regarding this relationship. And both the suggested positive relationship between either gender and the extent of compliance and gender and the quality of explanations were not found significant, which is in contradiction with mentioned former studies. But additional analyses in the research found that firms that are in non-compliance, always provide a justification for their non-compliance and in 71% of these cases justified their non-compliance by using context specific or also called firm specific justification for their deviation. Firms specific

explanations are seen as efficient information / high quality information. Although this is an positive sign for the monitoring function of the board, no significant evidence was found that female participation in the board lead to an improvement of this monitoring function.

Finally is also found that although the arising interest and government pressure towards a more diversified board, the examined UK firms did not fully implemented this highly requested gender quota.

(4)

4

2.7 Hypothesis ... 17 3 Methodology ... 18 3.1 Sample ... 18 3.2 Research design ... 19 3.2.1 Dependent variables ... 19 3.2.2 Independent variable ... 25 3.2.3 Control variables ... 26 3.3 Statistical model ... 28 4 Results... 29 4.1 Descriptive statistics ... 29 4.2 Hypothesis test ... 34 5 Conclusion ... 37 6 References ... 40

(5)

5

1 Introduction

1.1 Introduction

A lot of prior research has been conducted regarding the influence of female presence on the board of directors and performance quality. Croson and Gneezy (2009) studied the gender differences with respect to professionalism/ethical preferences and concluded that women are more risk averse than men. This study also found significant results for women’s preferences, that woman are adapted more to situation and men are more competitive than women. In their article, Nguyen, Locke and Reddy (2012) have found that the presence of female directors in the board has significant influence on the company’s performance. Prior research, Gul et al. (2007) has already shown that earnings quality is higher for firms with female directors or higher proportion of female directors on the board.

Nguyen, Locke and Reddy (2012) find significant results for the effect of characteristics of board of directors and the influence upon earnings management. The existing gap between the interest of the stakeholders and board of directors can be a driver or motivation of earnings management. This gap is also referred to as the agency problem. Jensen and Meckling (1976) were the first to define the agency problem, with this they gave more body to the relationship between the shareholders and executives (Jensen and Meckling, 1976). The wealth of the shareholders can be harmed or decreased by this misalignment since the executives can be more concerned about the personal wealth, while shareholders aim to increase the value of the firm. So separation of control and ownership is the fundament for this agency issue. Due to the frequent changes in the entities environment and the fact that entities are becoming more complex the conspiracy of these interests is becoming more necessary (Jensen and Meckling, 1976).

These misalignments between parties were not the only driver of the need for

corporate governance and regulations. The past financial crisis and accounting scandals like the Enron case started a discussion among society. These accounting scandals lead to the interference of the government by the application of laws. These accounting scandals resulted in higher expectations of stakeholders, investors and regulation bodies in action against the separation between control and ownership. For instance, the UK corporate governance code and the Sarbanes-Oxley act in 2002 were responses from governments to these scandals. Translating the economic reality with standards instead of rules is the mean purpose of regulating bodies. Standards can be more applicable than rules, since standards can be

(6)

6 adopted to changing operational environments (Arcot et al., 2009). Responding on the

governance and agency failures the UK adopted a code with a new setting and form. The UK corporate governance code is established based on standards instead of rules. Arcot et al. (2009) state that the UK Corporate Governance code stands in contrast for example with the Sarbanes-Oxley act of the US. Since this code enables entities with the comply or explain model to comply with the principles of the code or to give an explanation for the non-compliance with a provision (Corporate Governance Code Committee, 2008). Non-compliance exists when companies deviate from the Code if it can lead to good corporate governance quality. Furthermore (Arcot et al. 2009) state that this approach can lead to more trust by stakeholders since it allows deviations and so this approach should lead to better corporate governance. This trust is created by the flexibility that the code offers to entities. The flexibility of the Code creates more interest in the informativeness of the explanations of deviations from provisions. A substantial disadvantages that can arise from this flexibility is failure in information quality to capital providers. This low quality of information can arise from low quality of the explanations that the firm provides for deviations from the Code. In order to keep this trust by stakeholders the explanations need to be justified and reasoned transparently, otherwise the aim of the code which is enhancing the corporate governance quality could be not achieved. Seidl (2007) went even further and found that when

shareholders feel dissatisfaction with a firm’s decision to deviate from a best practice provision or with the explanations provided for this deviation, they will sell their shares or speak out their discontent. Which highlights the importance of solid explanations.

Prior studies (Seidl et al.; Arcot et al. 2009 and Macneil and Li, 2006) regarding the comply or explain model have focused mainly on the effectiveness of the code and whether companies comply or not. These studies show deviation in the opinion about the effectiveness of the code. (Arcot et al. 2009) found that entities do no use the flexibility of the code to enhance their governance, since in a lot of cases non-compliance was not explained or the explanation were of weak and not a convincing basis.

On the other side ( Seidl et al. 2009) investigated to which extend large public companies complied with the regulations of the UK code. They found that the explanation option that the model provides was more used than research suggested. They also found different drivers for whether entities try to comply or explain.

Macneil and Li (2006) found that the benefits of the flexibility of the code are overstated and that the code could be integrated into regular company regulations.

(7)

7 corporate governance codes like the UK code but also enhanced the role and scope of the audit committee (Audit committee guidance, 2011). This resulted in different studies upon the relationship between audit committees and corporate governance.

Walker (2004) finds that the performance in quality of audit committees depends on several characteristics, like their involvement, their knowledge and skills.

O’Sullivan (2000) investigated the impact of the board composition on the recommended financials aspects of corporate governance. The researcher found that non-executive directors as part of the board tend to prefer intensive audits while reduction of intensive auditing could result in the reduction of agency costs expected from the separation of control and ownership.

All the insights described prior lead to the definition of corporate governance as a dynamic concept. The past few years were in the focus of enhancing the corporate governance by implementing new rules and regulations regarding board structures,

independence and composition. Jensen and Meckling (1976) state that the corporate boards are one of the most important internal governance mechanisms.

Campbell & Minguez-Vera (2008) found different characteristics of board of directors and audit committees that influence the level of earnings management and corporate

governance quality. The emphasized on independence and experience of board composition as characterises that could be of influence.

Lately in certain countries the importance of board composition was highlighted based on diversity. Several studies like (Campbell & Minguez-Vera, 2008 and Adams & Ferreira 2009) combined gender with financial performance and found significant positive results for this relation. To participate in this emphasize of diversity, the Netherlands for example tried to enhance the role of diversity by incorporating a quota. This woman-quota obliged board to incorporate at least a deviation of 30/70% between woman and man. As stated above prior research found a positive relation between woman’s presence and performance quality. So inspired by the increasing interest in woman incorporation in boards and the on-going discussion about which characteristics of boards effect corporate

governance quality, this paper will tend to combine these dynamics and examine the effect of gender on the corporate governance quality. Quality is be tested on the level of

informativeness that companies provide regarding their corporate behaviour.

No important significant results found conducting the research, regarding this relationship. And both the suggested positive relationship between either gender and the

(8)

8 extent of compliance and gender and the quality of explanations were not found significant, which is in contradiction with mentioned former studies.

1.2 Research Question

As stated above this paper will focus on the relationship between gender and compliance or non-compliance for corporate governance quality. Prior literature (Seidl et al.; Arcot et al. 2009 Macneil and Li, 2006) and Akkermans et al., 2007) created a suggestion that the comply or explain method can provide the best results regarding corporate governance quality, since the corporate governance code sets out standards of good practice for boards in relation to shareholders. This also creates the suggestion that full compliance will lead to better corporate governance quality. So, Responding to the prior literature limitations (Hooghiemstra, 2012 and Drost, 2104) and tending to combine these two worlds of gender and corporate governance quality the following research question will be adopted in this paper.

“What is the effect of gender on the quality of compliance with corporate governance codes?

The following steps needs to be carried out in order to investigate the above research question. First, to analyse whether companies comply or not comply with the provisions of the UK corporate governance code. Second, to analyse the quality of explanations for compliance and non-compliance. Third, to analyse the influence of gender as a characteristics of corporate boards to compliance quality.

1.3 Contribution

This part addresses the motivation that underlies this study. Contribution of this study is together withscientific especially from a societal point of view. Gul et al (2007) also argues that woman demonstrate greater risk aversion and are less aggressive in decision taking. This study can help enhance the role of woman within the accounting field. And if there is any effect found this study can help us respond on this effect by better understanding of the characteristics that are of mean influence. In scientific literature (Dalton et al., 2010 and Ding et al., 2013) there is research performed with respect to the influence of female participation in corporate boards and the effect on performance quality, however limited research has been conducted in connection with the comply or explain method for corporate governance

(9)

9 First of all a lot of influencers of corporate governance quality are already tested, for example the expertise, independency and activity (Hooghiemstra, 2012 and Ujunwa, 2012) but not a lot of research is conducted on the effect of gender differences on corporate governance quality. So, this research will complement the existing corporate governance literature by studying information quality of corporate behaviour.

Secondly, this study is trying to contribute to the existing literature by among others focusing on a different corporate governance code with a different setting. Earlier research (Drost, 2014) that combined the effect of gender with corporate governance quality was focused on the Dutch gender quota with the Dutch corporate governance code. The researcher found in her study that the government was not successful with this quota, since a lot of entities did not follow the rule of the 30/70% deviation. This study is trying to enhance the reliability of previous research upon gender and corporate governance quality by focusing on firms with a more preferred deviation of gender. And since the UK does not have any strict rules regarding the gender structure of the boards this research can show us the effect of this approach on the behaviour of entities regarding this aspect.

Furthermore, Hooghiemstra (2012) stated in his research that by examining the non-compliancy within a sample of more EU countries could enhance the generalizability of result of corporate behaviour. He also stated that it would be interesting based on previous results to investigate whether explanations for non-compliance are also affected by certain firm specific characteristics. Especially this interest and gap in literature is seen as inspiration for this study. Finally, relating gender ‘as a firm characteristic and based on previous gender and corporate firm relations’ to this specific corporate governance comply-or-explain mechanism makes this thesis unique and is therefore contributing to the existing scientific literature regarding gender.

1.4 Structure

This thesis is structured as follows. Section 2 provides an overview of the prior literature build upon the basis of the agency theory. Furthermore this section ends with the formulation of the hypothesis. Section 3 introduces the methodology used in this thesis. Section 4 presents and discusses the results and section 5 forms the conclusion and gives insight for future research.

(10)

10

2 Literature review

This paragraph provides an overview of the theories that will be built upon in this study for better understanding of the variables and to result in a hypothesis.

2.1 Agency theory

Agency theory is till now the most commonly used theory in corporate governance research. This theory is particularly useful for understanding the creation of the misalignment between boards and stakeholders. So the agency theory focuses on the behavior of listed firms. Jensen and Meckling (1976) define agency theory as

“an relationship in which one or more persons (the principal) engages another person ( the agent) to perform some service on their behalf which involves delegating some decision-making authority to the agent”.

So the basis of this theory is that the interest of the principles and the agents do differ. The agency relationship develops when the shareholders hire a manager to create firm value on their behalf. To be able to operate on behalf of the shareholders decision-making power is delegated to managers (Jensen and Meckling 1976). The delegation of this control can cause information asymmetry and different goals. These two concepts are the basis for the agency problem. But the real driver for the agency problem is the existence of different risk

preferences, which is created and reinforced by information asymmetry (Jensen and Meckling 1976). Risk preferences is illustrated by the level of risk aversion of managers and

shareholders. The question that arises is which risk are shareholders desiring to take relative to operations and investments and which risk are managers willing to take relative to their agenda. So both the shareholders as the executives can have different risk taking behaviors since their tolerance towards risk differs and so both may take different actions in practice. Agency theory is also used to explain the decisions of firms relative to disclosures of corporate information. Larker and Tayan (2011), state that the disclosure of information reduces information asymmetry between the manager and the shareholders and also the corresponding risks.

Furthermore agency theory suggests that the principal can limit opportunistic action by the agent by providing the agent with appropriate incentives. Opportunistic actions are actions that could harm the principals in firm value (Jensen and Meckling 1976). Larker and Tayan (2011) describe these actions as executives tending to increase their own wealth while

(11)

11 the possibility exists that shareholders wealth could be affected and so harmed. To refer to this as the agency problem and the creation of agency costs as a result from this. Studies like (Deumes and Knecel 2008) and (Vander bauwhede and Willekens 2008) conclude that agency costs and disclosure costs are associated with certain firm characteristics.

Hooghiemstra (2012) identified that informativeness of firms explanation for deviations from a corporate governance code can be seen as a form of disclosure.

2.2 Gender diversity

General understanding of gender differences as a characteristic of corporate board structure. is important to be able to understand the effect of these differences in accounting

perspectives. The overall arises of women in the top increased the attention for diversity of gender within corporate boards. This increased voice of women let many researches study the relationship between gender and board performance. This studies aim to understand the effect of the differences in personal characteristics on corporate performance.

A study of (Campbell and Mı´nguez-Vera, 2007) argues that diversity within a certain group can enhance the creativity and innovation. Gul et al (2011) develops on this and found that women communicate more directly than men, since they tend to discuss issues that are unusual. Overall they found that female board members undertake a more constructive manner of speech during meetings.

The study of Carter et al. (2003) investigated the link between board diversity and firm value in the context of agency and stakeholder theory and their study suggests that greater diversity can increase the independence of the board, since woman have a more aggressive approach and show more courage in asking questions in contrast to male directors. Main past literature like (Johnson and Powell 1994) suggests that women are more risk averse, discrete and ethical in contrast to man. As described in the previous paragraph risk aversion is associated with the misalignment between shareholders and executives. Since women have a greater desire for security while man have a greater desire for high returns they tend to be more risk averse than man.

Finally Smits (2011) argue that diversity can enhance problem-solving thinking. This study also argues that gender diversity in the board may also improve a firms competitive advantage if the diversity improves the image of the firm.

The benefits from gender diversity on corporate governance is examined in two directions. First, there are several studies who directly examined the relationship between

(12)

12 gender and financial performances and the relationship between gender and overall boards effectiveness. Adams and Ferreira (2009), Huse et al (2009) and Nielsen and Huse (2010) confirmed the above statements about female behavior and all showed positive results

towards this relationship and state that female participation eventually leaded to improvement of the quality and effectiveness of decision making process, so the boards effectiveness and the overall achievement of boards objectives. However the relationship between gender and financial performances showed contradictory results. Erhardt et al (2003) showed positive results towards gender and financial performances. This study finds that a more presence of female board members contributed to higher financial ratio’s. Burke (2000) found a positive relationship between the number of females in board and revenue assets and profit margins. Adams and Ferreira (2009) also found positive results for female participation and financial performance, however the contrary aspects lies in the finding that female membership only lead to better performance when females enters firms with low performance. But when they enter firms with good performance the firm values seems to decrease. Another contrary result to the expected is found in the study of Broome et al (2011). They introduced a concept which is referred to as the ‘critical mass’. Which illustrates a border on the positive effect of woman on firm performance. This border means that adding more female members will not constantly lead to better firm performance. Because when a certain level of female

participation is achieved, women will no longer be seen as outsiders that could have

additional value. According to certain studies like Broome et al (2008) the critical mass for women participation is three or more.

The second direction into which the relationship of gender and corporate governance is examined, is with the understanding of the ethical and economic arguments behind the concept of gender diversity (Broome 2011). Ethical arguments regarding gender diversity in boards emphasize that it is unacceptable from the viewpoint of several members of society to exclude certain groups based on gender or any other characteristic from the corporate class (Broome 2011). Economic arguments emphasize that the systematic ignorance and non-selection of able female candidates can eventually harm firm financial performance (Burke 2000).

This study relies on the viewpoint that firms are continually doing stakeholder management to create stakeholder satisfaction. Giving more voice to woman is an example here for in this context. This pressure to firms regarding a diversified gender composition in a board is put by various stakeholders, like shareholders, governmental bodies and politics (Francoeur et al., 2008).

(13)

13 2.3 Corporate governance

Corporate governance has a broad and sometimes vague scope within the accounting world. According to Gillan (2006) corporate governance is a system of laws, rules and factors that control operations at a company. This same study argues that corporate governance is relevant in the understanding of corporate behaviour and organizational performance. The existence of self-interested executives created the need for corporate governance. Control and monitoring systems for these agency problems are called corporate governance (Larker & Tayan, 2011). These authors also define corporate governance as:

“the collection of control mechanism that an organization adopts to prevent or dissuade potentially self-interested managers from engaging in activities to the welfare of shareholders and stakeholders”.

Larcker et al (2007) argues that corporate governance is “a set of mechanisms that influence the decisions made by managers and directors when there is a separation of ownership and control”. These governance mechanisms ensure managers to take actions that are in line with the interest of shareholders. Gillan (2006) distinguishes internal mechanisms and external mechanisms. These mechanism could be seen as elements that could influence the corporate governance of a firm. Internal mechanisms include i.e. board strength, capital structure and internal control systems. So, these are factors that can be directly influenced by the firm itself. External mechanisms include i.e. law and regulation and monitoring by financial analysts. These mechanisms will be highlighted more in the following sections.

2.4 Corporate boards

Within the agency perspective corporate boards are one of the most important internal mechanism for corporate governance. The boards role is important and can be divided into two streams. Board of directors have first of all an obligation towards shareholders and secondly, the board is responsible for the formulation of strategic objectives and the direction and monitoring of these objectives. Therefore the boards have an important role in resolving the agency problem.

The corporate board consists of two subsidiaries, the executive board and the non-executive board. According to Huse et al. (2009) the non-executive board deals with the

development and decision processes of strategy. The non-executive board on the other hand fulfils more the control and monitoring task and is so more focused on the alignment with shareholders. This board furthermore consists of several committees, like the audit and

(14)

14 compensation committee. (Larcker & Tayan, 2011) merged the board tasks into two overall functions. So they recognize the advisory function and the oversight function.

During the past several studies (Hooghiemstra, 2012 and Ujunwa, 2012) were conducted to elaborate on the characteristics of boards and firm performance. Especially the independence of the board is questioned in different cases (Hooghiemstra, 2012 and Seidl et al., 2009) after several popular scandals. Byrd and Hickman (1992) suggest that effective governance and firm performance increase with board independence. Larcker and Tayan (2011) describe independence as:

“The degree to which a director is free from conflicts of interest that might compromise his or her ability to act solely in the interest of the firm.”

Klein (2000) examines the relationship between audit committee and board characteristics to earnings management. This study found for both the audit committee and the board

independency a negative relation in accordance to abnormal accruals. These results suggests that more independence will lead to a more reliable corporate governance. The link between diversity and independence is set by (carter et al., 2013). This study found that any form of diversity will lead to more perspectives and so will expand the scope of boards activities and innovation. Rose (2007) went even further and made a direct link stating that board diversity reduces the agency conflict by improving the effectiveness of corporate leadership.

2.5 Corporate governance codes

Seidl et al (2009) defines corporate governance codes as a non-obligatory set of principles and standards for best practices. These codes can also be categorized as an external

mechanism for corporate governance. Corporate governance codes furthermore tend to enhance the integrity, transparency and quality of boards (Akkermans et al., 2007). Since corporate codes are highly recommended but not obligatory, there are no globally accepted corporate governance standards that when complied to imply good governance. Therefor countries took the responsibilities to individually determine codes for best practices. But since these codes all focus on the same topics and despite the differences in national law and regulations we find similar content on a universal level ( Akkermans et al., 2007)

Corporate governance codes can be divided into two categories. Literature distinguishes rules-based standards and principle-based standards. The rules based

approaches are a set of detailed rules for the accounting standards. While the principle-based approach is more flexible and offers room for own professional judgement, instead of

(15)

15 inflexible regulations. The rules based approach is sometimes attacked from the viewpoint whether ’one size fits all’ (Larcker & Tayan, 2011). Certain companies like the US follow the rule-based approach but many environments prefer the principle-based approach which offers flexibility in the practice of the standards that are set by the Corporate governance code. The purpose of practice-based codes is that companies are expected to take their unique

organizational setting into account and apply what fits their particular firm setting ( Seidl et al, 2009).

Corporate governance codes were developed in order to ensure the alignment of ownership and control and to restore stakeholders trust. Several codes were developed in history, for example the Sarbanex-Oxley Act of (2002) and the European Parliament and Council (Directive/2006/43). But the first corporate governance code was developed in 1992 by the Cadbury Committee in the UK, which is referred to as the “the Cadbury Report’ (Akkermans et al., 2007). The purpose of this code was to create a set of rules and best practices to direct and control the behaviour of corporate management from the agency viewpoint.

This paper will zoom in on the UK corporate governance code, by developing on the latest revision of the UK corporate Governance Code (2012). Since this code emphasizes more attention on the characteristics of the board. And the UK is seen as a pioneer in the field of corporate governance. The UK corporate governance code is currently seen as one of the most developed operative codes for best practice provisions, seen the UK had the longest history of principle-based standards (Seidl et al, 2009).

A main characteristic that is derived from the principle based approach is the ‘comply or explain’ principle (Seidl et al, 2009). This principle is based on the idea that firms should have more space for firm specific application of the standards for good practice and whereby the explanation for deviation should justify the unique circumstances of the firm ( Arcot & Bruno, 2006). Broad insight in this latter concept is provided in the following section.

2.6 Comply or explain principle

A model which is derived from corporate governance codes is the comply-or-explain principle. As stated in the corporate Governance Code Committee (2008) this principle includes that entities should state in their annual reports how the firm applied the principles of the model and handled the best practice provisions of the code and if the firm failed to comply with the provisions should explain why the provision was not applied. Although

(16)

16 corporate governance codes encourage compliance, any compliance with the code is not mandatory, since the only legal obligation that listed entities have is reporting on their compliance with corporate governance code (Hooghiemstra, 2012). Faure-grimaud et al (2005) investigated whether the UK corporate governance code led to more compliance and to more explanations. They found that the code advanced compliance more than the initial code ‘the Cadbury Code’. As described above, compliance with the codes is seen as an element of good governance. But non-compliance is not necessary an implication of bad governance. The comply-or-explain principle is only considering bad governance when the explanations of deviations from the provisions are limited or not existing. Furthermore Seidl et al. (2013) states that when firms apply the principle of comply-or-explain, they need to put more emphasis on the quality of the explanations for deviations from the code. Since a reasoned explanation can fully give justice to the non-compliance and still satisfy for example shareholders. Compliance with the code is in essence seen as an element of good governance, but as explained above there is possibility given for non-compliance which not directly is seen as bad governance. Hooghiemstra (2012) stated that bad governance is only concluded when non-or limited explanation for deviation from the code is given. Seidl et al (2013) even states that good explanation for non-compliance can even give evidence for compliance with the code, when it promotes delivery of business objectives.

So, firms are expected to provide information to shareholders that will enable them to judge whether deviation from the UK corporate governance code is for valid reasons.

Macneil and Li (2006) state that in order to enable shareholders, firms should provide the shareholders with explanation for deviation. This explanation should be based on firm

specific circumstances that caused the firm to deviate from certain provision in order to fulfill the shareholders disclosure desire. But in this study Macneil and Li found that firms often provide concise, general and uninformative explanations for deviation. Arcot et al (2010) also concluded that companies provide uninformative explanations, which do not give any serious indicate for commitment to good corporate governance. The audit Committee Institute (2008) also stated that the ‘Comply or explain” principle adds value to corporate governance due to its flexibility as mentioned before. However this institute also raised questions towards this principle with respect to the sustainability. They foresee a rising demand of shareholders to comply with all the provisions since the explanations that are given for non-compliance are overall just general statements. And this eventually could lead firms to behave in a non-sustainable manner, so comply with provisions just for the compliance sake. And this will surpass the purpose of this principle.

(17)

17 2.7 Hypothesis

Empirical studies involving gender are mostly dividable into two categories. Some studies have a more feminist approach attempting to understand gender discrimination and develop social contribution (Campbell & Minguez-Vera, 2008 and Adams & Ferreira 2009). Other studies use gender as an independent variable to explain variations in corporate boards effectiveness (Drost, 2014). This study will be more focused on the second stream. Since prior literature suggested the effects of different aspect of corporate boards, this study is tended to use gender as a characteristic of boards to explain differences in behavior.

As concluded from the above, gender studies showed a positive relationship between woman in corporate boards and firms performance (Campbell & Minguez-Vera, 2008; Adams & Ferreira 2009 and Drost, 2014). Looking at the fact that female board members tend to be more strict in monitoring and compliant with rules and regulations this study expects a positive relation between female participation in corporate boards and the quality of compliance with the UK Corporate Governance Code.

The second aspect of this study is the effectiveness of this code and its corresponding “comply or explain principle”. As Macneil and Li (2006) found that disclosures from the comply or explain principle are often doubtful, since the explanations for deviations could be too concise and uninformative. So third parties, like shareholders and investors have to base their decisions upon unjustified deviations. Another expectation in attempt to develop a hypothesis is the research of Hooghiemstra (2012) the results of this research suggested that certain firm characteristics are associated with a firms choice to provide either general and uninformative explanations or more firm specific and informative explanations. He

furthermore argues that firms with a weaker board tend to approach comply-or-explain more symbolically than substantively. So, firm that seek compliance just for compliance sake. Based on the above and concluding from the theoretical analyses the following hypothesis is formulated:

H1a: female members in the board of directors positively influence the extent of compliance with the UK corporate governance code.

H1b: female members in the board of directors positively influence the quality of explanations provided for compliance with the UK corporate governance code. H1c: female members in the board of directors positively influence the quality of explanations provided for non-compliance with the UK corporate governance code.

(18)

18

3 Methodology

A quantitative archival database research is assumed appropriate for this research question and therefor used in this thesis to test the hypothesis. The total research process will be explained in this chapter. This process consist of an explanation of the used sample, and exploration of the used variables and a discussion of the used research model.

3.1 Sample

The sample in this research is mainly based on hand-collected data and partially on archival data. The study consists of mainly hand colleting and coding since a database for the disclosures on comply-or-explain principles is lacking. This study will use a UK based sample. This paper is focused on the UKFTSE100 (this stock market includes different industries) in order to achieve some form of comparability within the sample. These companies are obliged to report under the same economic and political circumstances. The focus on the UK firms was based on the fact that The UK corporate governance code is currently seen as one of the most developed operative codes for best practice provisions, seen the UK had the longest history of principle-based standards (Seidl et al, 2009). Furthermore the implementation of the gender quota in the UK. The quota is still not an obligation but a challenge for UK listed firms to maintain a diverse composition of gender in the corporate boards. The challenge is to create a 70/30 distribution. This target is set based on the ‘Davies review’ and currently limited to the FTSE 100.

The hand-collected and archival data only includes observation from 2014, due to the last revision in the UK Corporate Governance Code of 2012 and in order to examine the recent emphasis on board characteristics.

Table 1:Sample selection

Listed firms UK FTSE 100 per 31-12-2014 100

Less: firms operating in the financial sector -16

Total sample of UKFTSE 100 active firms 84

The above table shows the detailed selection of applicable firms. The sample consists of non-financial companies since the regulatory environment of these companies may well interact with the provisions of the UK corporate governance code as suggested by (Fraure-Grimaud et al. 2005).

(19)

19 3.2 Research design

In this research the quality of compliance is measured based on the level of alignment with the provisions of the UK Corporate Governance code. Moreover it is expected that when boards consist with a higher number of female members, firm will show more compliance with the Corporate Governance Code. And if no compliance is found the reasons for compliance will be extendedly argued referring to firm-specific issues.

To test for the compliance level there are 12 provisions used of the UK Corporate Governance code, derived from the code of 2012 itself. The attention is focused on this code since it is more focused on the board characteristics. Furthermore these 12 provisions are chosen since these are focused on the board characteristics also. After the determination of the level of compliance with the provisions, explanations for compliance as well

non-compliance will be analysed. Two types of explanations need to be analysed for this research. Conducting a content analyses can help to classify these explanations on the basis of their informativeness. The two explanation types that need to be distinguished are,

non-compliance, so explanations for deviating of the code and non-compliance, explanations for complying with the code. So summarized that means that in first order the extent of compliance is examined based on the 12 provisions. And after this a content analyses is conducted to investigate the informativeness of the disclosed explanations. This part of the research was fully hand collected (Drost, 2014 and Stap, 2015). All the information is collected from annual reports and corporate governance statements

The following sections give a description of the used independent, dependent and control variables.

3.2.1 Dependent variables

To determine the operationalization of the dependent variable two analyses are conducted, which are fully described below, namely the analyses for the extent of compliance and the analyses for the quality of explanations. These two analyses are divided into three proxies. Three different proxies were used to calculate the dependent variable in order to be able to test for the three hypotheses described in the previous chapter. These three proxies are, the extent of compliance, the disclosed explanations for compliance and the disclosed

(20)

20 3.2.1.1 Extent of compliance

In first order the extent of compliance is investigated to examine to which degree UK firms are compliant with the UK Corporate Governance Code per 2014. This analyses was

conducted by manually reading the Corporate governance statements and /or annual report to determine the extent of compliance for the 12 applicable provisions from the code. This variable is operationalized by a dummy variable. Which can have the value of 1 of 0. The value of 1 is given when a firm deviates with the provision from the code and 0 when the firm complies with certain provision of the code and this is done for all the applicable

provisions and all sample firms. This variable is measured by expressing it in a percentage by summarizing the compliant provisions per firm and dividing this by the total number (12) of applicable best practice provision. So summarized this means that for each applicable and tested provision of the UK Corporate Governance Code the firm is tested using the before mentioned documentation, if the firm complies with the code or deviates from the code. And this test is coded as stated above and is expressed finally into a percentage indicating the extent of compliance with the code per sample firm.

3.2.1.2 Quality of explanations

A content analysis is conducted to analyze the quality of the informativeness of the explanations. This research method enables researchers to make qualitative inferences based on content documents. This analyse is done based on the models of (Seidl et al., 2013) and (Hooghiemstra 2012). These models give the opportunity to classify the explanations regarding the best-practice provisions of the UK Corporate Governance Code on the basis of their informativeness. Two different types of explanations are measured based on these models. First the explanations for deviations are hand-collected from all the statements as mentioned before and measured and later on the explanations for compliance are hand-collected and measured.

3.2.1.2.1 Quality of explanations for compliance

In first order the explanations for compliance were measured. Less informative explanations for compliance with provisions are expected, since corporations under this law are not obligated to explain or discuss in which manner they do comply with a provision. One of the first studies to focus on the quality of the content of explanations were (Arcot and Bruno, 2006). They used a content analysis to rank explanations on their basis of their level of informativeness using six categories to investigate the quality of corporate governance. The categories were: no

(21)

21 explanation, general explanation, inline explanation, limited explanation, transitional explanation, genuine explanation. A marginal difference is found between the categories therefore this study will research by combining the categories that in fact capture similar explanations, resulting in three categories. This categorisation is also used by former research of Hooghiemstra (2012) that elaborated further on the work of Arcot and Bruno (2006). Table 2a below shows the final and used categories to determine the quality given explanations.

Table 2a: Categorisation of Explanations for compliance (Hooghiemstra, 2012)

Category Definition

1: No explanation Company provides only the statement of

compliance but no further detail. 2: General explanation Company provides statements using

standard phrases without given any specific details regarding the firms circumstances. 3: Firm-specific explanation Company provides detailed and firm

specific information in order to justify the compliance.

3.2.1.2.2 Quality of explanations for non-compliance

In second order the quality of explanations for deviations from the best practice provisions was given. The classification of these explanations for non-compliance is based on the created hierarchy by (Seidl et al., 2013). In this paper they critically analysed earlier studies of Macnail and Li (2006); Akkermans et al. (2007); Arcot et al (2010) and Hooghiemstra (2011) to combine their methodologies and create a redefined and tightened taxonomy. This thesis will measure the quality of explanations for non-compliance based on this created taxonomy since it is based on previous methodologies and papers that tested the equality of corporate governance by focussing on the compliance quality. Therefore it enhances the reliability and usefulness for this thesis as well.

This taxonomy consists of three general categories with each subcategories. The first general category is ( 1 ) deficient justification, this is also the ‘no explanation’ category. So these type

(22)

22 of justifications are aligned to firms that are in non-compliance without given any arguments for their deviation. The second general category is ( 2 ) context specific justification, this is also the firm-specific explanation. The explanations that are given for non-compliance under this category are in reference to the specific situation of the firm and provide firm specific reasons for deviation. The third general category is ( 3 ) principled justification, this is also the general explanation for deviation. This type of justifications arise when a firm contends that a provision does not reflect the best practice. Firms that give principle justification refer to for example general implementation problems and or conflicts with law and regulations when reasoning their deviation. As explained before all three general categories have their corresponding subcategories, which elaborated more on the specifications of the general categories for explanations. In the following table (2b) al the general categories with their corresponding subcategories are explained in detail, obtained of (Seidl et al., 2013).

Table 2b Categorization of explanations for non-compliance

(Seidl et al. 2013)

Categories of explanation Sub-categories of explanation

Definition

1: Deficient justification also referred to as the ‘no explanation’.

- Pure disclosure - Description of

alternative practice

- Empty justification

- Company only declares that it deviates from the code provision.

- Company presents an alternative solution to the governance problem that the code provision addresses.

- Company provides an explanation that seems like a justification for its deviation but which does not possess an

explanatory power

2: Context-specific justification

Also referred to as the ‘firm specific explanation’. - Size of company or board - Company structure - International context of company - Company justifies deviation with regard to the size of the company. - Company justifies

deviation by regarding the code provision as inappropriate. - Company justifies

deviation with regard to specific aspects of its international operations.

(23)

23 - Other company specific reasons - Industry specificities - Transitional justification - Company justifies deviation with regard to the particular situation of the company other than size.

- Company justifies deviation with regard to the specificities of the industry.

- Company justifies deviation with regard to either novelty of the code provision or the fact that the company is a new entrant.

3: Principled justification also referred to as the ‘general explanation’.

- Effectiveness

- General

implementation problems

- Conflicts with laws or societal norms

- Company justifies deviation by pointing out that an application of the code provision will be sub-optimal generally not just for its own operations.

- Company justifies deviation by pointing out some general problems in implementing the code provisions. - Company justifies

deviation by pointing out that the code provision conflicts with societal norms, values or laws.

For the purpose of the total analysis, these categories and explanations need to be

operationalized. In this thesis the classification of the categories per explanation has been done by the use of a dummy variable in order to distinguish different treatment groups of the analysis. If a firm assigned in the statements an explanation to a particular category then this dummy variable will have the value of 1 and 0 for all other categories. After this the analysis will provide a total sum of explanations per category and per firm. The following table (2c) provides a detailed overview of the measurement of the dependent variable. Which is obtained based on the paper of (Drost, 2014) and (Hooghiemstra, 2012).

(24)

24 Table 2c Measurement of the dependent variable

(Drost, 2014)

Dependent variables Proxy Measurement

Compliance Extent of compliance Dummy variable with a value of 1 if a firm deviates from a provisions followed by dividing the total with all the 12 examined provisions.

Compliance general explanation

Quality of explanations for compliance

Dummy variable with a value of 1 if an explanation is assigned toward this category and 0 for all other categories

Compliance firm specific explanation

Quality of explanations for compliance

Dummy variable with a value of 1 if an explanation is assigned toward this category and 0 for all other categories.

Non-compliance deficient justification

Quality of explanations for non-compliance

Non-compliance context specific justification

Non-compliance principled justification

In order to further operationalize the dependent variable regarding a statistical regression analysis an overall score is calculated. Which is referred to as the ISCORE. This score represents the informativeness of the explanations that firms provided in their annual reports and statements, and is therefore used as a proxy for the dependent variable for both the

(25)

25 explanations for compliance as the explanations for non-compliance. The ISCORE is adapted from Hooghiemstra (2012) and represents a weighted average score. The higher the ISCORE the more informative the explanations for compliance and non-compliance are. The ISCORE for both compliance and non-compliance is calculates as follow (Hooghiemstra, 2012):

ISCORE_C = 4 * firm-specific explanations + 2 * general explanation + 1 * no explanation Total number of explanations for compliance

ISCORE_NC =

4 * context-specific explanation + 2 * deficient explanations + 1 * principled explanation Total number of explanations for compliance

The overall ISCORE is calculated by assigning points per category, divided by the total number of explanations for compliance and non-compliance. The weights are given in such order that firm-specific/context-specific explanations weighs more in total than the

general/deficient explanations and this latter weighs more in total than the no/principled explanations. This represents the development of the scores 4, 2 and 1 in the formula for the ISCORE. Firm-specific, context-specific are weighted a 4, since these explanations give a better view of the firms context/situation and could be therefore more preferred and so on. Therefore this overall score indicates the level of informativeness of the explanations for compliance and non-compliance with the best practice provisions of the UK Corporate Governance Code.

The data for both the extent of the compliance as the quality of compliance was collected by manually checking the mentioned sources ( annual reports, corporate governance statement and compliance statements). The hand-collection process consisted of first

analysing the extent of compliance as described before per firm and per provision of the UK Corporate Governance code. Secondly the quality of explanations was analysed by again manually reading all the sources of all the sample’s firms and assigning explanations towards categories per firm in order to test the compliance level.

3.2.2 Independent variable

As mentioned before this thesis is testing the relationship between gender in corporate boards and the quality of corporate governance. So therefore the independent variable of this thesis

(26)

26 is gender. And in this analysis there is distinction made between the relationship between gender in boards and the extent of compliance and between gender in corporate boards and the quality of explanations. In other words, do firms with more females in board comply with more provisions so have a greater extent of compliance. And provide qualitative more and better explanations so provide more informativeness. The independent variable gender is determined by manually checking all the annual reports for the UKFTSE100 firms and first determining the total board members and determine the total amount of female participants in the board. A percentage of females per board is determine out of this analyse, by dividing the total number of females by the total amount of board members. Table 3 below shows the measurement of the independent variable Board gender.

Table 3 Measurement of the independent variable.

Independent variable Measurement

BoardGender Total number of board participants divided

by the total amount of board members, expressed in a percentage.

3.2.3 Control variables

Since other factors could influence the relationship between gender in corporate boards and the quality of corporate governance, the addition of control variables is necessary. In first order the total assets of the applicable firms are determined and used in the model to control for size differences between the firms (Klein, 2006). Firm size is included as a control

variable of the specific firms, since larger firms are more viewable and more likely to operate in shareholders’ interests and demands from the agency viewpoint (Li et al., 2008).

Continuing on the agency perspective, larger firms are expected to disclose higher quality information (Ahmed and Courtis, 1999), since these type of firms have greater agency problems ( Dey, 2008). From this perspective and regarding the level of compliance,

explanations towards compliance and non-compliance are sometimes seen as a form of corporate disclosure. And corporate disclosure is in relation to the quality of financial

reporting. According to Beyer et al. (2010) and Kim and Shi (2011) cost of equity is affected by corporate disclosure (agency problem) and firms that are actively operating on equity markets often have higher quality of information disclosure. Therefore participation on the

(27)

27 equity market (issued share capital) is enclosed as a control variable. Furthermore also the link between the level of disclosure and company debt is issued in several studies (Fernandes, 2008; Leuz, 2004). This researches related the firms leverage to the agency problem, since the degree of reliance on debt can influence the corporate disclosures. Therefore leverage is added as an control variable. Also board size is added as an control variable, since research found that larger boards often provide financial reporting of higher quality (Arcot et al., 2012) And the last included control variable is board strength. Referring to Hooghiemstra (2012) several board characteristics influence the level of quality and extent of provided information disclosure. For example, as well as Hoitash et al. (2009) concluded that certain characteristics defining the board strength could influence the information disclosure. Summarising this control variable is a mixed score measuring and control for the effect of board strength in this relationship by combining, board size; board independence; frequency of board meetings and the number of subcommittees. Adding the total effect of these factors yields the board strength measure which can result in a certain value indicating a 0 for ‘weak board’ and a 4 for a ‘strong board’.

Again the data for these control variables is fully hand collected by issuing the annual reports of the applicable corporations (Hooghiemstra, 2012 and Drost, 2014).

Table 4 Measurement of control variables

Control variable Measurement

Agency-Problem This is a dummy variable with a value of 1

if a firm issued share capital in 2014 and otherwise the value is 0

Size Total assets per firm in 2014, (in absolute

numbers).

Leverage This is a leverage formula obtained by

dividing total debt by total equity per firm in 2014.

Board-size This is the total number of board members

in absolute numbers.

Board-Strength This is a mixed score per firm consisting of several factors between 0 and 4. Indicating the strength of the board. 0 indicates a weak board and 4 indicates a strong board.

(28)

28 3.3 Statistical model

As mentioned before this thesis is testing the relationship between gender in corporate boards and the quality of corporate governance. To test for the relationship between the two a

statistical model is created, in form of a multivariate regression analysis.

Compliance = β0 + β1 BoardGender + β2 AgencyProblem + β3 Size + β4 Leverage + β5

BoardSize + β6 BoardStrength + ei

This above regression was obtained to run a test for the first hypothesis (H1a) in order to test for the relationship between female board members and their relationship with the extent of compliance.

The statistical progression of the ISCORE_C:

Quality of explanation _C = β0 + β1 BoardGender + β2 AgencyProblem + β3 Size + β4

Leverage + β5 BoardSize + β6 BoardStrength + 𝜀𝑖

The statistical progression of the ISCORE_NC:

Quality of explanation _NC = β0 + β1 BoardGender + β2 AgencyProblem + β3 Size + β4

Leverage + β5 BoardSize + β6 BoardStrength + 𝜀𝑖

These above two regressions were obtained to run a test for the second (H1b) and third (H1c) hypothesis in order to test for the relationship between female board members and their relationship with both the explanations for compliance and explanations for non-compliance. Summarized, the first model measures the extent of compliance, second model the quality of compliance given for compliance (C) and the third model given for

non-compliance (NC). Where βi stands for the coefficients and 𝜀𝑖 for the error term all the other

(29)

29

4 Results

4.1 Descriptive statistics

The descriptive statistics provided in table 5 provide insights in the characteristics of the dependent, independent and control variables. The total compliance rate (ISCORE_C) and the total non-compliance rate (ISCORE_NC) have a mean of respectively 2,8718 and 2,3494. The standard deviations are respectively 0,78650 and 1,81749. This indicates that this sample consist of slightly more companies that provide more explanations towards compliance than non-compliance with the UK Corporate Governance Code. The dependent variable extent of compliance, have a mean of 0,9066 and standard deviation of 0,06687. This means that almost 91% of the sample firms were in full compliance with the tested provisions of the UK Corporate Governance code. This is in line with the research of Drost (2014) which found a 91% score for the full compliance. Furthermore the mean of the independent variable, Percentage_F, which indicates the total number of females in the board is 0,2186 and standard deviation is 0,08349 which indicates that on average in this sample the total percentage of females on board is 22%, this is in contrary with the research of Drost (2014) who found an average of 12%. Furthermore the control variable, board size has a mean of 10,52 and a standard deviation of 2,396. Which indicates that on average the corporate boards consisted of 11 members. Finally the control variable board strength has a mean of 2,36 and standard deviation of 0,932. This indicates that on average the firms in the sample scored 2,36 on strength between 0 and 4. This is in line with the study of Drost (2014) which scored a 2,00 on average and slightly higher than (Hooghiemstra, 2012) scoring an average of 1.440 on board strength.

(30)

30 Table 5: Descriptive statistics

Variables Obs Mean Std Min Max Skewness

Dependent variables ISCORE_C 83 2,8718 ,78650 2.00 4,00 ,168 ISCORE_NC 83 2,3494 1,81749 0,00 6,00 -,176 ExtentofCompliance 83 0,9066 ,06687 ,83 1,00 ,223 Independent Variables PercentageF 83 0,2186 ,08349 0,00 ,44 -,020 Control Variables Sharecapital 83 0,6900 ,467 0,00 1 -,820 Totalassets 83 22326,52 41890,027 1373 284305 4,066 Leverage 83 2,7396 4,65209 0,03 40,72 6,867 Boardsize 83 10,52 2,396 6,00 17 ,667 Boardstrenght 83 2,36 ,932 1 4 ,231

A breakdown of quality of compliance and the corresponding explanations is presented in table 6a and b. The review of the corporate governance statements, annual reports and other corporate governance statements showed that UK listed companies are compliant with the UK Corporate Governance code for 91% on average. As explained before the distinguished explanations for compliance are ‘General explanations’ and ‘firms specific explanations’. The examination of the sources lead to the conclusion that there is no

substantial variance between the given general and firm specific explanations for compliance. So, firms that complied with the Code which are 91% on average provided 48% of general explanations and 52% of firm specific explanations justifying their compliance.

Table 6a: Quality of Compliance

Industry Size % Compliance Total number of explanations for compliance Mining 9 94 32 Construction 3 90 27 Manufacturing 29 94 126 Transportation 17 90 41 Wholesale Trade 2 87 4 Retail Trade 11 90 46 Services 12 92 39 Total 83 91 315

(31)

31 Table 6b: Quality of Compliance explanations

Industry Size General

explanations % General explanations Firm specific explanations % Firm specific explanations Mining 9 12 8 20 12 Construction 3 6 4 21 13 Manufacturing 29 57 37 70 43 Transportation 17 21 14 20 13 Wholesale Trade 2 4 3 0 0 Retail Trade 11 24 15 22 13 Services 12 29 19 10 6 Total 83 153 48 163 52

A breakdown of quality of compliance and the corresponding explanations is presented in table 7a and b. The review of the corporate governance statements, annual reports and other corporate governance statements showed that UK listed companies are non-compliant with any certain provision of the UK Corporate Governance code for 71% on average. As explained before the distinguished explanations for non-compliance are ‘deficient’, ‘context specific’ and ‘principled’. The examination of the sources lead to the conclusion that there is a substantial variance between the given deficient, context specific and principled explanations for non-compliance. So, firms that did not fully comply with the Code which are 71% on average provided 18% deficient, 71% context specific and 15% principled explanations. Concluding that all the companies that deviated from the Code justified this by at least stating the fact that they are not fully compliant with one or more provisions. These results show that firms make considerable use of the flexibility of the comply-or-explain principle as assumed before, since 71% firms on average are not fully complaint with the code but hence justify this by providing any of the before mentioned explanations. There is an obvious variance between the categories since on average 71% of the corporations that do not fully comply with the code decided to justify their

non-compliance by providing context specific justification. This means that 71% of the specific firms decided to provided disclosure for deviations of the provisions that was of firm specific and internal nature.

(32)

32 Table 7a: Quality of Non-Compliance

Industry Size % Compliance Total number of explanations for compliance Mining 9 70 8 Construction 3 75 1 Manufacturing 29 68 41 Transportation 17 67 19 Wholesale Trade 2 70 2 Retail Trade 11 72 7 Services 12 69 15 Total 83 71 93

Table 7b: Quality of Non-Compliance explanations Industry Size Deficient Deficient

% Context specific Context specific % Principled Principled %

Mining 9 1 6 6 9 3 30 Construction 3 0 0 1 2 0 0 Manufacturing 29 6 35 30 45 5 50 Transportation 17 3 8 14 21 2 20 Wholesale Trade 2 0 0 2 3 0 0 Retail Trade 11 1 6 4 6 0 0 Services 12 6 35 9 14 0 0 Total 83 17 18 66 71 10 15

The dependent, independent and control variables are checked for normality and in addition for outliers, which is of influence on the regressions analysis. In order to do this the kurtosis and the skewness was tested. Also the distribution of the mentioned variables was checked in plots for normality and outliers. According to literature the skewness needs to be between -1 and +1. All the variables met these conditions expect for the control variables leverage and total assets (firm size) which had a very high skewness. In addition to this test the distribution did not seem to be normally distributed. Both variables showed a distribution that is skewed to the right. A transformation is mostly applied to data to more closely meet the assumptions of a statistical procedure. Therefore in order to use these variables in the regression test a transformation was needed. In order to transform a variable a regular logarithmic or LN function is used. In this study both were checked and in the end firm size

(33)

33 expressed in total assets showed a skewness of 0,714 after using the regular logarithmic function and leverage showed a skewness of -0,592 after the regular logarithmic

transformation. The transformation of firm size was consistent with former research (Hooghiemstra, 2012) and (Drost, 2014).

Table 8: Pearson Correlation

Variables 1 2 3 4 5 6 7 8 9 1. ISCORE_C * 2. ISCORE_NC -,050 3. ExtentofCompliance ,146 -.648** 4. PercentageF -,050 -,112 ,118 5. sharecapital ,125 ,145 -,102 ,102 6. Totalassets (LOGTA) ,058 ,052 -,071 ,038 ,059 7. Leverage (LOGLEV) -,188 -,181 ,120 ,125 -,060 ,146 8. Boardsize -,082 ,174 -,100 ,071 -,071 .296** _,024 9. Boardstrenght .238* _,083 _,026 _,096 _-,073 _-,094 _-,043 _,068 _*

*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).

Table 8 shows the Pearson correlations between the variables. This Pearson correlation is a univariate analysis since this matrix determines the separate correlation coefficients between the used variables. The coefficients stated in bold are significant. This matrix is applied for examination since issues regarding multicollinearity could arises when there is too much correlation between variables. The line is set at a coefficient of 0,8. Multicollinearity could arise when a variable coefficient exceeds 0,8 this could lead to unreliable inferences regarding the data set.

The correlation matrix did not show many correlation between the variables which could otherwise indicate multicollinearity. The variable extent of compliance and

ISCORE_NC show a significant quit high correlation of -0,648. However this correlation can be explained by the fact that a higher extent of compliance is linked to a lower ISCORE_NC. Since firms that comply fully with the code or with most provisions would possibly disclose less or no explanations for compliance. This variables are checked for multicollinearity since both are used in the regression analysis. Furthermore board strength and ISCORE_C show a significant low correlation of 0,238. This can be explained by the possibility that a strong board has correlation and a link with a higher ISCORE_C, so more explanations for

compliance. This is reasonable because strong boards show more compliance with the code. Lastly, a significant low correlation is found between board size and firm size (total assets). A multicollinearity check is done by checking the variance inflation factor (VIF). A VIF test determines whether an variable could be considered as a linear combination of

The UK corporate governance quality : a female influence on compliance quality with the corporate governance code