The influence of firm-level financial and governance
characteristics on the voluntary CEO-Employee pay ratio
disclosure in European firms
A MASTER THESIS by
Niclas Farrenkopf
Submitted to the
Faculty of Economics and Business at University of Groningen and
Newcastle University Business School at Newcastle University
In partial fulfilment of the requirements for the degrees of M.A. Advanced International Business Management (Newcastle)
and
M.Sc. International Business and Management (Groningen)
Student number: B 50050076 (Newcastle), S 3074129 (Groningen)
Approach/Design/Methodology/Purpose – Building on signalling and legitimacy theory and drawing from literature on corporate transparency and information disclosure, this study uses a multi-industry sample of 128 listed and non-listed European companies and employs binominal logistic regression analysis to explore the influence of firm-level financial and governance characteristics on voluntary CEO-Employee pay ratio disclosure in Europe with the aim to empirically identify significant disclosure determinants.
Implications/Research limitations – This study is limited by its small sample size resulting from the current scarcity of CEO-Employee pay ratio information. In addition, its observations are limited to Europe. It would be interesting for future research to replicate this study with firms in the United States as future disclosure regulation is about to change and more detailed data on the CEO-Employee pay ratio will become available in the United States from 2017 onwards.
Findings/Originality/Value – The developed regression model was unable to find significant results. Consequently, this study found no significant association between the independent variables firm size, industry type, firm age, listing status, CEO gender, audit status and auditor size and the dependent variable of CEO-Employee pay ratio disclosure. Nonetheless it provides a theoretical contribution regarding the potential rational of firms to engage in voluntary disclosure of CEO-Employee pay ratio information and adds to both the literature and the discussion in the field of corporate transparency. It is the first study analysing the influence of firm
characteristics on the disclosure of CEO-Employee pay ratio information. In light of potential regulatory changes, it can provide a ‘before picture’ of firm disclosure determinants and further enhance our understanding of the factors which influence voluntary information disclosure. It also provides an analytical approach for future research on pay ratio disclosure.
Keywords – CEO-Employee pay ratio, voluntary information disclosure,
transparency, corporate governance, disclosure determinants, Europe, financial characteristics, governance characteristics
Abstract ... I List of Tables ... II List of Figures ... III Abbreviations ... IV
1. Introduction ... 1
2. Theoretical Framework ... 4
2.1. CEO-Employee Pay Ratio and its Disclosure ... 4
2.2. Corporate Transparency and Information Disclosure ... 7
2.3. Determinants of Corporate Transparency ... 10
2.4. Theoretical Concepts ... 13
2.4.1. Signalling Theory... 14
2.4.2. Legitimacy Theory ... 18
3. Hypotheses Development and Conceptual Model ... 22
3.1. Firm-level Disclosure Determinants ... 22
3.1.1. Financial Characteristics ... 22
3.1.2. Governance Characteristics... 27
3.2. Control Variables... 30
3.3. Conceptual Model ... 32
4. Methodology ... 33
4.1. Choice of Research Design ... 33
4.4. Study Sample ... 42
4.5. Data Analysis ... 44
4.5.1. Regression Model ... 44
4.5.2. Statistical Data Analysis ... 47
5. Results ... 48
5.1. Descriptive Statistics ... 48
5.2. Multicollinearity ... 51
5.3. Outliers ... 52
5.4. Binomial logistic regression results ... 52
5.4.1. Baseline analysis... 52
5.4.2. Model fit ... 53
5.4.3. Prediction of pay ratio disclosure ... 54
5.4.4. Regression coefficients ... 55
5.4.5. Summary ... 56
5.5. Robustness of the regression model ... 56
5.5.1. Scenario regression analysis ... 57
5.5.2. Summary ... 59
6. Discussion... 60
6.1. Hypotheses Assessment ... 61
7. Conclusion, Limitations and Future Research ... 66
References ... 67
Appendices ... 79
Appendix A: GRI indicator G4-54: Definition and Measurement ... 79
Appendix B: List of firms excluded from the dataset ... 80
Appendix C: List of firms in the dataset ... 81
Appendix F: Suitability of the binominal logistic regression model ... 89
Appendix G: Justification for G4-54 non-compliance ... 95
Appendix H: GRI – Definition of ‘external assurance’ ... 97
Appendix I: NUBS Ethical Approval Form ... 98
Appendix J: Initial study sample ... 103
Appendix K: Additional descriptive statistics ... 104
Appendix L: Scenario 1 – Regression analysis results ... 105
Table 1: Research Hypotheses ... 32
Table 2: Industry sectors ... 39
Table 3: Definition and measurement of variables ... 41
Table 4: Characteristics of the dataset ... 43
Table 5: Descriptive Statistics ... 48
Table 6: Disclosures by firm age group ... 49
Table 7: Disclosures by industry ... 50
Table 8: Disclosures by country / region... 50
Table 9: Disclosures by audit status ... 51
Table 10: Expected influence of variables on disclosure ... 51
Table 11: Regression analysis - Block 0 Classification Table... 53
Table 12: Regression analysis - Omnibus Tests of Model Coefficients ... 53
Table 13: Regression analysis - Hosmer and Lemeshow Test ... 53
Table 14: Regression analysis - Pseudo R2 values ... 54
Table 15: Regression analysis - Block 1 Classification Table... 54
Table 16: Regression analysis - Variables in the equation ... 55
Table 17: Scenarios for the regression analysis ... 57
Table 18: Scenario Analysis - Scenario 1 ... 58
Table 19: Scenario Analysis - Scenario 2 ... 59
Table 20: PCA – Correlation matrix ... 86
Table 21: PCA - KMO and Bartlett's Test ... 87
Table 22: PCA - Individual KMO measures ... 87
Table 23: PCA - Total variance explained ... 87
Table 24: PCA - Component Coefficients ... 88
Table 25: Binominal logistic regression assumptions ... 89
Table 26: Box and Tidwell procedure ... 91
Table 27: Multicollinearity Statistics ... 93
Table 28: Pearson Correlation Matrix ... 94
Table 29: Justifications for G4-54 non-compliance ... 95
Table 30: Disclosures by listing status ... 104
Table 31: Disclosures by auditor size ... 104
Figure 1: Signalling Theory - The Signalling Environment ... 15
Figure 2: Conceptual Research Model ... 32
Figure 3: Saunders’ Research Onion ... 33
Figure 4: Plot of disclosures by firm age group ... 49
BoD Board of Directors CEO Chief Executive Officer CEOs Chief Executive Officers
CSR Corporate Social Responsibility
EC European Commission
EUR Euro
EUR bn In billion Euro EUR m In million Euro EUR tn In trillion Euro
FTE Full-time employees GDP Gross domestic product GRI Global Reporting Initiative PCA Principal component analysis
SEC United States Securities and Exchange Commission SMEs Small and medium-sized enterprises
1 | P a g e
1. Introduction
“Pay ratio disclosure should provide a valuable piece of information” (Stein, 2015) - Kara Stein, Commissioner at the United States Securities and Exchange Commission
“To say that the views on the [new CEO-Employee] pay ratio disclosure requirement
are divided is an obvious understatement” (White, 2015)
- Mary Jo White, SEC Chairwoman
The discussion around CEO-Employee pay ratio disclosure emanates from public and investor concerns regarding the development of executive compensation levels in the last three decades. In particular, the compensation of Chief Executive Officers (CEOs) is subject of public scrutiny and evaluated as far too excessive (Gavett, 2014; Taube, 2014; Mullaney, 2015). Academic research in this area
focused on finding explanations for the rising executive compensations (Boyd, 1994; Sanders and Carpenter, 1998; Frydman and Jenter, 2010) and underlined the
negative ramifications of a divergence between CEO compensation levels and average employee salaries on firm performance or employee satisfaction (Osberg and Smeeding, 2006; Faleye et al., 2010; Faleye et al., 2013; Mohan et al., 2015). However, it is less concerned with exploring the underlying factors that influence the voluntary disclosure of the CEO-Employee pay ratio itself.
The question of CEO-Employee pay ratio disclosure is part of a bigger discussion about the information that firms want to be transparent on. It is evident that firms with a higher pay ratio are more reluctant to share this information fearing potential negative consequences in form of a decreasing public reputation or overall legitimacy issues in the market. Often the argument of data sensitivity and
confidentiality is hold up against a higher level of transparency, with many firms refusing to disclose information on their compensation policies due to concerns about data protection or simply because they have doubts about the usefulness and sufficiency of a pay ratio indicator (see Appendix G).
2 | P a g e maintain legitimacy within a market and to signal responsibility to both investors and market participants.
In times of rising societal demands for stronger and mandatory disclosure regulations and being aware of the recent regulatory change in the United States regarding the introduction of a mandatory CEO-Employee pay ratio disclosure requirement from 2017 onwards (Lynch, 2015; McGrane and Lublin, 2015), the question arises what characterizes firms that decide to engage in the disclosure of pay ratio information even though they are mandatorily not required to do so?
It seems worthwhile to further explore the existing research gap concerning the factors that influence the decision of corporations to engage in voluntary CEO-Employee pay ratio disclosure as an analysis can provide important insights for policy makers and provide them with an answer to the question on which group of firms they have to focus regulations in order to enhance overall information
disclosure. It can also provide insights on firm factors that drive the voluntary disclosure strategy of firms.
This study applies insights gained from research on corporate transparency and voluntary information disclosure to the context of executive compensation and employee salaries (CEO-Employee pay ratio). Although factors that determine corporate information disclosure have been analysed in general in various studies (Meek et al., 1995; Raffournier, 1995; Inchausti, 1997; Patton and Zelenka, 1997), the number of academic contributions focusing on the determinants of corporate information disclosure related to the CEO-Employee pay ratio is limited (Faleye et
al., 2013). Therefore, this study extends prior research by examining the influence of
firm-level financial and governance characteristics on the disclosure of the CEO-Employee pay ratio for the context of European listed and non-listed firms.
Based on this objective this paper explored the determinants of CEO-Employee pay ratio disclosure using a dataset of 128 European companies that comply with the Global Reporting Initiative (GRI) framework to analyse the following research question:
3 | P a g e By developing and applying a binominal logistic regression model, this study tested the influence of the independent variables firm size, industry type, firm age, listing status, CEO gender, audit status and auditor size on the likeliness of CEO-Employee pay ratio disclosure. Although the developed regression model was unable to find significant results, this study is among the first ones providing a
theoretical contribution regarding the potential rational of firms to engage in voluntary disclosure of CEO-Employee pay ratio information and adds to both the literature and the discussion in the field of corporate transparency.
4 | P a g e
2. Theoretical Framework
2.1. CEO-Employee Pay Ratio and its Disclosure
The discussion around CEO-Employee pay ratio disclosure emanates from public and investor concerns regarding the development of executive compensation levels in the last three decades. In particular, the compensation of Chief Executive Officers (CEOs) is subject of public scrutiny and evaluated as far too excessive (Gavett, 2014; Taube, 2014; Mullaney, 2015).
The debate and the arguments about the disclosure of pay ratios can be divided into two distinct groups; the group of those in favour of a pay ratio disclosure and those who oppose to the disclosure of pay ratio information.
Advocates of pay ratio disclosure point out that rising and excessive CEO compensation levels often come at the expense of shareholders and contribute to an increasing divergence between CEO pay and average employee salaries. High compensation packages have incentivised excessive risk taking by company
executives in the past as cases such as the fraud scandals at Enron and WorldCom demonstrated (Pedrotty, 2011). A more recent example are the malpractices at Lehman Brothers evolving around the structure of executive compensation packages that significantly contributed to the 2008 financial crisis. A disclosure of pay ratio information can provide an instrument to incentivise the internal adjustment of corporate compensation structures (especially considering the relationship between CEO pay and average employee salaries) as it exposes firms to a greater extent to the public eye and thus to issues related to legitimacy and public image. More generally, advocates argue that a pay ratio disclosure is needed as it allows for a comparison of compensation practices across companies.
Critics of an increased transparency regarding pay ratio information argue that investors are misled by CEO-Employee pay ratios as the comparability between companies in different industries, with different employee structures or levels of internationalization varies significantly. As such a firm that operates on a
5 | P a g e different pay ratio compared to companies operating in the construction and
manufacturing industry.
However, while critics provide valid reasons to question the effectiveness and reliability of a one-fits-all pay ratio they fail to give significant consideration to the potential benefits of the disclosure of compensation ratios. While the comparability of pay ratios across industries and countries remains an issue in the absence of a standardized measurement method, pay ratio disclosure has the potential to influence the future development of executive compensation levels.
However, as of 2016, there is no mandatory requirement regarding the
disclosure of firm-specific CEO-Employee pay ratio data. Although regulations in the United States are about to change and, following the implementation of the Dodd-Frank act, will require firms to disclose information on “the median of the annual total compensation of all employees (except the CEO) and the ratio of CEO
compensation to median employee compensation”, the current pay ratio disclosure situation is based on voluntary compliance (Pedrotty, 2011, p. 1; Faleye et al., 2013). The situation in Europe is similar to that with no current mandatory requirement for pay ratio disclosure in place. The European Commission plans to introduce pay ratio disclosure as part of a broader package of measures to improve shareholder rights (Shareholder Rights Directive). The Directive would ask listed firms to disclose not only the pay ratio between the average executive compensation and the
remuneration of the average employee on full-time equivalent terms, but also to report changes in the ratio over the past three years as well as a justification how it contributes to the long-term interest of the firm (Dendrinou, 2014). However, very much like the debate in the United States, the proposed Directive is not yet implemented with critics pointing out that the calculation of a ratio of executive to employee pay is everything but a trivial exercise as it raises questions about ratio comparability and the compensation elements that have to be included (Marsland, 2016).
The reasons given by the European Commission for considering a mandatory implementation of pay ratio disclosure are in line with the arguments of most
6 | P a g e to shift the focus of management from short-term risk taking to a long-term
orientation towards the interest of the company and to help firms making conscious decisions and reflect on the relative value of good management (EC, 2014).
Certain European countries have mandatory requirements regarding the disclosure of executive compensation. The United Kingdom, Ireland and France for instance have mandatory disclosure requirements regarding executive remuneration through individual level reporting, while Portugal or Denmark demand compensation disclosure only on an aggregated level with a limited individual breakdown for
directors (Ferrarini and Moloney, 2004; Right2Info, 2016). However, these compensation disclosure requirements do not include information on employee compensation policies.
The Global Reporting Initiatives’ G4 framework which was introduced in 2013 provides a platform for firms to engage in the voluntary disclosure of information related to Corporate Social Responsibility (CSR), including an indicator on the CEO-Employee pay ratio in its comprehensive reporting standard. While the indicator itself suffers from similar shortcomings that critics point out for pay ratio disclosure in general (e.g. a limited comparability across industries and countries) it provides a standard definition for pay ratio calculation (see Appendix A) and thus enables investors at least to a certain extent to compare compensation practices between companies within the same industry and the same country.
With the majority of firms refusing to disclose pay ratio information often based on the arguments of an insufficient ratio comparability and frequently also by referring to the confidentiality of corporate remuneration policies (see Appendix G) the question arises why and what kind of firms engage in voluntary disclosure of pay ratio data. With CSR becoming a central part of the overall strategy of modern corporations, companies already became more conscious about the importance of transparency and accountability to shareholders in order to maintain legitimacy in a market and to signal socially responsible behaviours and business practices
(Cheers, 2011; Kühn et al., 2014). Building on this idea, this study aims to find firm-level characteristics and factors that influence pay ratio disclosure decisions.
7 | P a g e decision-making on transparency and will add to the academic debate about pay ratio disclosure.
Thus, aiming to determine the characteristics of firms that are more likely to disclose CEO-Employee pay ratio information on a voluntary basis, this study will draw from the literature in the field of corporate transparency and the wider field of corporate information disclosure to identify firm-level financial and governance factors that influence a firm’s disclosure decision. Building on this theoretical framework the hypotheses will be developed.
2.2. Corporate Transparency and Information Disclosure
To determine key characteristics of firms disclosing CEO-Employee pay ratio information it is paramount to understand the fundamental concepts of voluntary information disclosure and thus corporate transparency in general.
Corporate transparency can be understood as the “availability of firm-specific information to those [who are] outside [of] publicly traded firms” (Bushman et al., 2004, p. 207) or as the “perceived quality of intentionally shared information from a sender” (Schnackenberg and Tomlinson, 2014, p. 5). As pointed out by
Schnackenberg and Tomlinson (2014), a definition of corporate transparency needs to be broad enough to cover for the fact that the concept of transparency is one that exists across contexts, in various domains of research and encompasses different dimensions. As such corporate transparency can be seen as consisting of the three key dimensions information disclosure, clarity and accuracy (Schnackenberg and Tomlinson, 2014).
Information disclosure is an essential element of corporate transparency as it
enables different stakeholder groups to gain better insights into firm operations, management and policies (Patel et al., 2002; Bushman et al., 2004; Nicolaou and McKnight, 2006). As such it refers to the availability and accessibility of information that enables an audience to interpret and determine relevant facts and data
8 | P a g e reports or a firm’s articles of association are examples for this type of information disclosure. Firms can also decide to disclose information on a voluntary basis without being legally required to do so (Ho and Wong, 2001). This kind of behaviour is often observed with firms that follow Corporate Social Responsibility standards or frameworks (Belkaoui and Karpik, 1989; Roberts, 1992), but can also be the result of economic-driven strategies (Jensen and Meckling, 1976; Diamond and Verrecchia, 1991). Voluntary disclosure reduces information asymmetries and companies can benefit from making information publicly available as it “enhance[s] the performance of firms by reducing agency costs […] [but also by helping] corporate insiders to make optimal decisions” in favour of long-term firm interests (J.J. Choi et al., 2012, p. 40).
However, an increased corporate disclosure and information transparency is not unanimously positive in nature. Especially when the shared information is unfavourable for the company, the detrimental effects of voluntary information transparency can harm firms public image or market legitimacy (see Chapter 2.4).
In case of the CEO-Employee pay ratio, information disclosure is achieved by granting stakeholders access to firm-specific information regarding the difference between the compensation level of the Chief Excusive Officer and the average employee of a firm. Given this information different interest groups are able to evaluate and determine a firm’s relative divergence in remuneration structures in relation to others.
For the purposes of this research voluntary information disclosure is particularly interesting. Although previous research came to the conclusion that a “disclosure of pay ratio improves […] performance” on both a firm and individual level and demanded an obligatory implementation of pay ratios in remuneration reports, most regulatory industry requirements or national laws do not yet enforce a
mandatory disclosure (Lacmanović, 2013, p. 165).
The recent regulatory changes in the United States leading towards a mandatory disclosure of CEO-Employee pay ratios for listed companies from 2017 onwards1 can be seen as a step in the direction of higher corporate transparency and information disclosure standards (Lynch, 2015). Though with “companies […]
9 | P a g e not [being] required to disclose non-executive employee compensation [or pay ratio] data in publicly available sources” the situation at the moment is one of
non-mandatory pay ratio disclosure (Faleye et al., 2013, p. 3261). Thus, firms which make ratios publicly available are doing this on a voluntary basis.
Clarity as a second dimension is defined by Schnackenberg and Tomlinson
(2014) as the “comprehensibility of information received from a sender”
(Schnackenberg and Tomlinson, 2014, p. 9). While clarity of information can be limited by elements such as the unfamiliarity with the language in which it is provided, a high level of complexity or through the use of industry- or firm-specific vocabulary and terms, it is generally seen as the degree to which shared information is understandable for an audience outside the firm environment (Larsson et al., 1998; Nicolaou and McKnight, 2006; Granados et al., 2010). As such the meaning of a message has to be clear and understandable without the potential of
misinterpretation on the way from receiver to sender (Schnackenberg and Tomlinson, 2014). Among academic researchers this “perceived quality of information” (Schnackenberg and Tomlinson, 2014, p. 10) is seen as a central aspect of corporate transparency to which literature also refers to by using
alternative terms such as ‘understandability’, ‘coherence’ or ‘interpretability’ (Miller, 1996; McGaughey, 2002; Nicolaou and McKnight, 2006). For the research question explored in this paper clarity can be understood as the clear interpretation of the pay ratio between workers and the Chief Executive Officer as a wage gap and
divergence indicator within a firm.
Accuracy as third dimension of corporate transparency is understood as “the
perception that information is correct to the extent possible given the relationship between sender and receiver” (Schnackenberg and Tomlinson, 2014, p. 10). As such it is the degree to which provided information is valid, comparable and reliable. As pointed out by Fung (2014) it can also be seen as the “truthfulness” of information (Fung, 2014, p. 75). In the context of CEO-Employee pay ratio disclosure accuracy poses a major problem. With no mandatory pay ratio disclosure requirements present, there is no standardized definition of how to measure the CEO-Employee pay ratio. This clearly affects the degree of exactness and precision of most reported pay ratios. This paper hopes to sufficiently account for the mentioned issue of
10 | P a g e data on the CEO-Employee pay ratio, which lays out a standardized definition of measurement in its implementation guidelines (see Appendix A). Due to the use of the GRI pay ratio indicator accuracy will also be interpreted as a degree of
compliance with the given framework definition.
Furthermore, Fung (2014) suggests the use of timeliness, completeness and
materiality of information as additional elements of corporate transparency.
Understanding timeliness as the disclosure of information in a timely and quick manner, completeness is seen as the disclosure of information in a way that is “sufficient to enable investors to make informed decisions” (Fung, 2014, p. 75). Materiality of information refers to the idea that disclosed information should be of relevance for the decision making process and as such consisting of material that can “influence investment decisions” (Fung, 2014, p. 75).
In sum, corporate transparency is therefore interpreted as the function of the six dimensions information disclosure, clarity, accuracy, timeliness, completeness and materiality of information with each of them contributing “uniquely to [the] overall […] level of transparency by increasing stakeholder confidence in the quality of information received” from a firm (Schnackenberg and Tomlinson, 2014, p. 11).
2.3. Determinants of Corporate Transparency
A variety of academic papers has explored potential variables that determine corporate transparency, which can generally be distinguished in two main
categories: firm-level and country-level characteristics. Both categories can be further broken down for instance into financial and governance factors. The
distinction in these two subcategories of corporate transparency variables was firstly introduced by Bushman et al. (2004) who used a factor-analysis approach to initially capture firm-specific information environments for a number of countries (Cheung et
al., 2007). In doing so, the author found evidence for the existence of financial and
11 | P a g e Khlif and Souissi (2010) used a meta-analysis approach to determine the significance of corporate characteristics such as ownership dispersion, analysts following, audit firm size, leverage, corporate size, profitability, and the multi‐
nationality of a firm as determinants of corporate disclosure and found evidence for a positive relationship between disclosure and audit firm size. Research by Yao et al. (2012) on voluntary disclosure of internet financial reporting2 analysed variables such as profitability, industry type, company size, ownership diffusion and the existence of an audit committee as determinants of information disclosure for emerging market firms. However, for their sample they found that variables such as age, size, profitability, industry and the existence of an audit committee were not significant while non-family ownership was determined as a significant explanatorily variable to the levels of voluntary disclosure (Yao et al., 2012). On the contrary, research by Bokpin (2013) confirmed the significance of “firm size, financial
leverage, audit quality, age and profitability” as relevant firm-level characteristics to determine corporate disclosure (Bokpin, 2013, p. 127). Furthermore, work by
Archambault and Archambault (2003), who empirically tested a sample of companies from 33 countries, indicates that disclosure is “influenced by culture, national […] and corporate systems” (Archambault and Archambault, 2003, p. 173). As such their empirical model includes firm-level factors as for instance ownership structure, firm size and industry, but also country-level factors like the economic situation
(measured via inflation and GDP development) and the political environment
(measured through components such as the legal system in place, freedom of press as well as the degree of political freedom) in a country (Archambault and
Archambault, 2003).
All the mentioned studies make valuable contributions to the field of corporate transparency. However, one criticism of much of the literature in this field is that there is no clear set of predictor variables available, painting a picture of a rather contradicting set of studies that sometimes see factors such as firm size, firm age or industry type as significant variables to explain corporate transparency decisions but nearly as often oppose to their relevance. Based on these findings, this study
proposes to use the general idea of both financial and governance characteristics on
13 | P a g e 2.4. Theoretical Concepts
Having elaborated on the basic understanding of corporate transparency with its six dimensions and different determinants, the next section turns towards the fundamental theories in the field of information disclosure.
As determined by Shehata (2014) academic research in the area of voluntary information disclosure builds most frequently on the theoretical concepts of “agency theory, signalling theory, capital need theory, and legitimacy theory” (Shehata, 2014, p. 18). Other researchers come to the same conclusion regarding the importance of agency, signalling and legitimacy theory in the context of corporate transparency (Tower and Rusmin, 2012; Uyar et al., 2013).
Originally rooted in information economics, agency theory is referring to the problems arising from principal-agent relationships within business environments and based on the assumptions of existing information asymmetries and incomplete contracts between principals and agents as well as different objectives and
possibilities of controlling each other (Eisenhardt, 1989). Based on the three central problems of adverse selection, hold-up and moral hazard a principal-agent approach to corporate transparency aims to resolve issues by increasing the level of
information disclosure with the aim to reduce potential agency costs (Eisenhardt, 1989; Schmitz, 2001).
Capital need theory argues that the level of corporate transparency can
significantly influence corporate capital structures with higher levels of transparency enabling firms to potentially reduce “owner-manager agency costs as well as [costs associated with] high levels of accounting disclosures, financial governance […] [and] audit intensity” (Oxelheim, 2006, p. 337). Assuming that a firm’s costs of capital include a price premium “for investors’ uncertainty about the adequacy and accuracy of the information available about the company” it suggests that voluntary
information disclosure can support firms in obtaining access to cheaper sources of capital as potential investors are better able to obtain information about the activities and prospects of a firm (Frederick Choi, 1973; Shehata, 2014, p. 20).
Signalling theory explains voluntary information disclosure of corporations as
14 | P a g e mandatory reporting requirements by law can be used to signal superiority over other firms in the market (David Campbell et al., 2001; Shehata, 2014).
Lastly, legitimacy theory as a concept of social interaction between information senders and receivers views corporate information disclosure as an instrument to gain legitimation in the market by providing external stakeholders with insights on firm processes and policies and as a mechanism to influence and alter external stakeholder’ perceptions and beliefs about a firm (Cormier and Gordon, 2001).
Consequently, it is apparent that the topic of voluntary information disclosure can be analysed through a variety of different theoretical lenses. This study sees the aim of firms to gain legitimacy and the process of signalling their quality as the basis for decisions on voluntary information disclosure such as the publication of pay ratio information. While other theories can provide additional insights and highlight other aspects of the disclosure discussion, legitimacy is seen as the core requirement of companies in order to operate successfully in a market. However, an analysis of all the potential theoretical angles is out of the scope of this paper and given the set amount of time and space restrictions, this study sees the last two theoretical frameworks of signalling and legitimacy theory as most relevant and influential for the analysis of voluntary CEO-Employee pay ratio disclosure determinants and will therefore focus on both for the hypotheses development process.
2.4.1. Signalling Theory
Signalling theory is a transaction-specific communication model in
15 | P a g e Figure 1: Signalling Theory - The Signalling Environment
Source: Connelly et al. (2011, p. 44)
While signallers are seen as corporate insiders that obtain “private information [which] provides […] [them] with a privileged perspective regarding the underlying quality of […] [the] organization”, receivers are corporate “outsiders who lack
information about the organization in question” (Connelly et al., 2011, p. 44-45). The
signal itself is seen as communicated “positive and negative private information” to
the outside world and the feedback as the reaction following the receiver’s interpretation of these information (Connelly et al., 2011, p. 44).
A common pattern that can be observed in the literature and the studies evolving around signalling theory is the use of quality as central element to distinguish firms. The majority of studies distinguishes two groups of firms: high-quality and low-high-quality firms. Existing information asymmetries between actors
create a situation in which firms have better knowledge of their own performance and quality, while corporate outsiders such as shareholders, investors or customers are less able to determine the true quality or value of a firm (Connelly et al., 2011). Signalling theory suggests that as both groups have the opportunity to signal their true quality to the market, two situations can occur: a separating or a pooling
equilibrium. While in the first situation only high-quality firms engage in signalling and
16 | P a g e Signalling theory in the field of voluntary information disclosure postulates that firms with good performance and positive information are more willing to share these with corporate outsiders as doing so helps them to be recognized as superior
compared to other firms in the market (Spence, 1973; Healy and Palepu, 1993). Consequently, management is motivated to engage in voluntary information
disclosure to influence market perceptions and to correct an over- or undervaluation of the firm by the market. On the contrary, firms with below average performance or non-standard business practices are less motivated to share information voluntarily as they anticipate potential negative ramifications connected to the disclosure of information. However, non-transparency can be interpreted as a negative signal, especially when information disclosure decisions of firms are not in line with the standards of the business environment or industry in which it is operating.
Transferring this knowledge to the context of CEO-Employee pay ratio disclosure it is apparent that firms with higher pay ratios are less likely to disclose this information compared to those with a lower pay ratio. However, attributing the disclosure decision solely to the level of the pay ratio is rather simplistic and does not take into account that there are other factors influencing the rational of firms to
engage in information disclosure, as multiple studies related to voluntary information disclosure have shown before (see Chapter 2.3). Industry pressures, national
differences in regulatory environments, or firm-specific differences in size, age or governance structures can influence disclosure decisions.
17 | P a g e (Farvaque et al., 2011). This can result in a competitive disadvantage for smaller firms.
One criticism of much of the literature on signalling theory is its strong
orientation on quality as the driving factor to distinguish firms. While the assumption of quality differences between companies seems logical and is applied by the majority of studies in the field, the term quality is rather general and can be interpreted in a variety of ways (e.g. financial superiority, higher levels of social responsibility, better management). Furthermore, the assumption that firms with a better performance are more willing to engage in signalling through information disclosure is inconclusive with some studies in support and others opposing to it (Bini et al., 2010). It is difficult to isolate a single signalling effect as information disclosure decisions are not solely influenced by firm quality and signalling costs, but depend on a variety of contextual factors (Ahmed and Courtis, 1999; Bini et al., 2010).
In sum, there are two main conclusions that can be drawn from the analysis of the present literature on signalling theory: Firstly, the main element of distinction used in signalling theory is quality, with most studies distinguishing high-quality from low quality firms. Secondly, there are costs connected to the signalling process between firms and the market which limit its benefits. The implementation of
18 | P a g e
2.4.2. Legitimacy Theory
Legitimacy is a key resource required by firms to operate successfully. Thus, it comes as no surprise that legitimacy theory is among the most frequently used concepts to explain information disclosure behaviour of firms with the majority of studies in the field focusing on the connection between legitimacy issues and corporate social responsibility reporting as form of voluntary information disclosure (Guthrie and Parker, 1989; Dennis M Patten, 1992; Lightstone and Driscoll, 2008a). Other studies build on legitimacy theory to find determinants of voluntary information disclosure for different settings (Hackston and Milne, 1996; Gamerschlag et al., 2011; Shehata, 2014).
In general, legitimation is a “process whereby an organization justifies to a peer or subordinate system its right to exist” (Maurer, 1971, p. 361). Other authors describe legitimacy as “the extent to which the array of established cultural accounts provide explanations for [the] existence [of organizations]” (Meyer and Scott, 1983, p. 201) and the pivotal paper of Suchman (1995) defines it as “the generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions” (Suchman, 1995, p. 574).
A review of the key literature in the field shows that it has become
commonplace to distinguish legitimacy in two types: Organizational Legitimacy and
Institutional Legitimacy Theory.
19 | P a g e Acknowledging the critique of this typology by Suchman (1995) who points out that the divergent assumptions of organizational and institutional legitimacy “about agency and cultural embeddedness often lead them to talk past one another”, this paper adopts his proposed distinction into the three types of pragmatic, moral and
cognitive legitimacy (Suchman, 1995, p. 572).
Pragmatic legitimacy can be broken down into exchange, influence and
dispositional legitimacy. While Suchman (1995) explains exchange legitimacy as the
support of firm behaviour by stakeholders considering the direct effects of
organizational actions on their wellbeing and relative to their expectations, he sees influence legitimacy resulting from the responsiveness of an organization to
stakeholder’ suggestions as an indicator for ongoing commitment. Furthermore, the author sees dispositional legitimacy attributed to firms that have dispositional
characteristics such as trustworthiness or honesty which are of importance to stakeholders.
In contrast to pragmatic legitimacy, moral legitimacy evaluates firm actions not based on the potential benefits for individual stakeholders but rather on the basis that corporate behaviour and action is in line with moral standards and promotes overall social welfare (Suchman, 1995). The evaluation of firm actions as socially responsible is however not objective in nature but based on the subjectively
constructed social values system of the audience. In general, moral legitimacy can be present in four forms: consequential, based on what an organization
accomplishes in terms of quality and value, procedural by using morally acceptable techniques and production procedures to achieve outcomes, structural by providing general organizational features that foster trust such as internal control systems and lastly personal, based on the charisma and attitude of individual leaders in a
company (Suchman, 1995).
Lastly, cognitive legitimacy poses a third type of legitimacy which can be broken down in the two classes of comprehensibility and taken-for-grantedness. While the first class of legitimacy arising from comprehensibility “stems mainly from the availability of cultural models that furnish plausible explanations for the
20 | P a g e influence and sees the behaviours and actions of companies as legitimate due to the taken-for-grantedness status of these actions in society (Suchman, 1995, p. 582).
Issues of legitimacy can be of major influence for the disclosure of CEO-Employee pay ratio information. In fact, with the rise of CSR as a central element of the overall strategy of the modern corporation, companies are more aware of the importance of transparency and accountability to shareholders in order to maintain legitimacy in a market (Cheers, 2011; Kühn et al., 2014). Especially through the three forms of pragmatic legitimacy (exchange, influence and dispositional
legitimacy) markets can exercise influence over firm actions and act as a corrective for socially unacceptable behaviour. Research by Kelly and Seow (2016) found evidence that “if companies are concerned about negative public perceptions […] pay ratio disclosures may be better able than current CEO pay disclosures at shaming companies into restraining CEO pay” (Kelly and Seow, 2016, p. 107).
From a firm point of view and building on the given definition of legitimacy as a concept of social interaction between information senders and receivers, information disclosure can provide an instrument to influence and alter external stakeholder’ perceptions and beliefs about a firm (Cormier and Gordon, 2001). As such
legitimation can be achieved both through mandatory disclosures – as for instance provided through annual financial reports because of regulations, and voluntary disclosures provided in form of voluntary social reporting by following Corporate Social Responsivity frameworks such as the Global Reporting Initiative (Magness, 2006; Lightstone and Driscoll, 2008b; Shehata, 2014).
What can be seen as the main point of critique with legitimacy theory as identified by Suchman (1995) is its failure of to adequately define the exact meaning of the term legitimacy, with it being “more often invoked than described and […] more often described than defined” (Suchman, 1995, p. 573). Nonetheless, legitimacy theory offers an influential theoretical framework to analyse and
22 | P a g e
3. Hypotheses Development and Conceptual Model
3.1. Firm-level Disclosure Determinants3.1.1. Financial Characteristics
1. Firm Size
A number of researchers argue that firm size is an essential element of voluntary information disclosure. In fact, a review of previous academic contributions paints a rather inconclusive picture regarding the influence of firm size with a number of studies providing evidence for a positive relationship between firm size and
information disclosure (Chow and Wong-Boren, 1987; Depoers, 2000; Abd-Elsalam and Weetman, 2003; Abraham and Cox, 2007; Jonas Oliveira et al., 2011) and others seeing a negative relationship between both variables (Mak, 1996; Aljifri and Hussainey, 2007; Kou and Hussain, 2007; Aljifri, 2008). However, when considering the theoretical arguments based on signalling and legitimacy theory this rather inconclusive influence of firm size is not surprising.
Legitimacy theory suggests that larger firms are exposed to a higher level of disclosure requirements as they are more visible to the public eye and the media and thus under greater pressure to respond in line with stakeholder and social
expectations (Alsaeed, 2006). In addition, larger firms often have more complex corporate structures and complexity requires a higher degree of disclosure in order to maintain market legitimacy (Cooke, 1989). Previous studies confirm the
assumption that larger firms are more exposed to pressure due to their visibility and are thus more likely to disclose voluntary information to gain pragmatic and moral legitimacy (John Patten, 1991; Asquer et al., 2016).
23 | P a g e companies are more concerned with the potential effect of it on their competitive position in the market (Uyar et al., 2013).
On the contrary, larger firms often have bigger executive compensation levels and thus higher CEO-Employee pay ratios. For these firms a disclosure of pay ratio information sends a negative signal to the market and implies a lower firm quality. As markets are less likely to attribute pragmatic and moral legitimacy to these firms as their remuneration policies are not in line with stakeholder interests or social
expectations, larger firms can also be expected to have a reduced willingness of CEO-Employee pay ratio disclosure.
While being aware of the potential that both a positive and a negative association between firm size and CEO-Employee pay ratio disclosure is possible, the author of this paper sees more arguments speaking in favour of a positive relationship
between both variables and thus will test the following hypothesis regarding firm size:
H1: There is a positive relationship between firm size and the disclosure of CEO-Employee pay ratio information.
2. Industry Type
A variety of academic studies noted that the type of industry has an influence on voluntary information disclosure. Voluntary information disclosure has been found to be more frequent and comprehensive in industries such as financial services or manufacturing with literature pointing out a variety of explanations for this industry effect (Barako, 2007; Krusberge, 2012). Some studies see voluntary disclosure differences explained by industry variations of proprietary costs (Meek et al., 1995), general pressure to release industry-related information exercised by investors in order to assess the value and position of a firm in the industry (Dye, 1985),
differences in regulatory industry controls (Owusu-Ansah, 1998) or in the intensity of international activities of particular industries (Raffournier, 1995).
24 | P a g e Raffournier, 1995; Lídia Oliveira et al., 2006; Burgwal and Vieira, 2014) nearly as high as the number of papers which conclude that there is no significant association between both variables (Wallace et al., 1994; Owusu-Ansah, 1998; An et al., 2011).
However, the assumption of an association between industry type and the disclosure of pay ratio information as form of voluntary disclosure is consistent with legitimacy theory which states that as some industries are exposed to a higher degree of societal and stakeholder pressure to provide additional information, firms in those industries are more willing to engage in voluntary disclosure in order to avoid “legitimacy gap[s] between the society and […] corporate operations” (Deegan et al., 2002; Burgwal and Vieira, 2014, p. 66). For example, firms in the highly
regulated banking sector are under pressure to inform corporate outsiders about key financial performance indicators and thus might fear to loose legitimacy by not
disclosing pay ratio information.
The type of industry also affects the level of employee salaries as well as the compensation of company executives (Dickens and Katz, 1986). As signalling theory postulates, market participants interpret differences in pay ratio levels as differences in firm quality which provides an explanation why firms in certain industries are more likely than others to disclose CEO-Employee pay ratio information. Especially in industry sectors with typically higher executive compensations but only moderate average employee wage levels firms want to avoid sending a signal of low quality to the market and pay ratio non-disclosure can be expected.
Therefore, the following hypothesis will be tested regarding industry type:
H2: The type of industry is significantly associated with the likeliness of voluntary CEO-Employee pay ratio disclosure.
3. Firm Age
25 | P a g e corporate voluntary disclosure, Uyar et al. (2013) identified listing age as a significant explanatory variable for the level of information disclosure among listed Turkish firms and Leuz and Wysocki (2008), Prencipe (2004) and Owusu-Ansah (1998) confirm that company age can be associated with enhanced disclosure. In general, a review of previous literature regarding the influence of firm age on voluntary information disclosure provides inconsistent results, with some studies seeing a positive (Hossain and Hammami, 2009), others a negative (Ibrahim, 2014) and also with some that do not see any significant relationship between both variables (Bukh et al., 2005; Alsaeed, 2006; Uyar et al., 2013).
Company age can be seen as a proxy for risk (Cerbioni and Parbonetti, 2007). Consequently, older firms can be expected to have more developed internal control systems while younger firms have no existing market reputation and as such low levels of legitimacy. At the same time, Owusu-Ansah (1998) points out that younger firms are suffering from a competitive disadvantage as the process of collecting, certifying and disclosing information is relatively costlier for them and they cannot rely on an existing disclosure record. While younger firms have higher signalling costs they are also in higher need to engage in signalling to proof their quality to the market and to substitute for their lack of reputation. Previous studies noted that younger companies are in fact more willing to engage in voluntary information disclosure “to increase the level of communication [with the market in order] to reduce scepticism and amplify the trust of investors” (Kateb, 2014, p. 632).
Contrary to that, some studies find that older firms disclose more information as they already operate for a longer period of time and are thus more likely to have integrated the idea of legitimacy in their corporate value systems in order to persist in the market (Deegan, 2006).
In sum, firm age as an independent variable finds support in both legitimacy and signalling theory and has been applied in previous studies. Therefore, it can be expected that firm age has an influence on the voluntary disclosure of
26 | P a g e
H3: Younger firms are more likely to engage in CEO-Employee pay ratio disclosure than older firms.
4. Listing Status
A variety of previous studies analysed the relationship between the listing status of a firm and voluntary information disclosure with some studies in support of a positive relationship (Wallace et al., 1994) and others seeing no significant association (Buzby, 1975) between both variables.
Firms have to apply for listing at the stock exchange and the success of such an application is often connected to the fulfilment of disclosure requirements that demand firms to publish information on their internal corporate activities as well as on key performance indicators (Buzby, 1975). Thus, it can be expected that listing status has an influence on information disclosure in general.
In line with legitimacy theory, Tian and Chen (2009) argues that society has higher expectations for listed companies’ information disclosure and Zhang and Zhang (2014) state that “mandatory […] information disclosure alone can no longer satisfy the diverse needs of investors” for information from listed firms (Zhang and Zhang, 2014, p. 143). Consequently, listed firms see themselves exposed to increased social pressures and demands to disclose more information voluntarily.
Therefore, this study will differentiate listed from non-listed firms and test the following hypothesis regarding listing status:
27 | P a g e
3.1.2. Governance Characteristics
1. CEO Gender
While there are no previous studies which link the gender of a firm’s Chief Executive Officer to the disclosure of CEO-Employee pay ratio information in particular, previous academic work has already analysed the effects of gender on corporate risk taking and information disclosure in general (Nalikka, 2009; Larkin et
al., 2013; Xiang et al., 2014; Loukil and Yousfi, 2016; Pucheta‐Martínez et al., 2016). In particular, a study by Nalikka (2009) regarding the impact of gender diversity on voluntary disclosure found interesting results, concluding that a female Chief Financial Officer is positively associated with a higher level of voluntary disclosure in annual reports. However, the same study found that in cases of female CEOs there is no statistically significant impact on voluntary information disclosure (Nalikka, 2009). Research by Gul et al. (2011) found that “gender-diverse boards improve the quality of public disclosure through better monitoring”(Gul et al., 2011, p. 315). In addition, some studies argue that female directors provide greater supervision and monitoring and thus actively contribute to a reduction of uncertainty (Adams and Ferreira, 2009; Terjesen et al., 2015). Female directors do not only have better attendance rates in board meetings compared to their male colleagues, but are also more likely to be found in “monitoring positions on audit, nominating, and corporate governance committees”(Habib and Hossain, 2013, p. 95). As “audit and corporate governance committees are directly involved in increasing [corporate] transparency”, gender can be seen as an important explanatory variable of information disclosure (Habib and Hossain, 2013, p. 95). Consequently, female directors have most likely an influence on the perceived market legitimacy of firms as they enhance
trustworthiness and the quality of internal monitoring measures.
Even though the majority of previous studies analysed the influence of female directors on a board as a whole, this study will adopt CEO gender as an explanatory variable for CEO-Employee pay ratio disclosure.
28 | P a g e
H5: Firms with a female Chief Executive Officer disclose their CEO-Employee pay ratio to a greater extent than those firms with a male Chief Executive Officer.
2. Auditor Size
Previous studies have related the size of an audit firm to voluntary information disclosure providing inconclusive results as some of them are in support of
(Inchausti, 1997; Uyar et al., 2013) and others opposing to a significant association (Wallace et al., 1994; Alsaeed, 2006) between both variables. The argument for an association between both variables is often based on the higher level of experience and internationalization in bigger auditing firms such as the Big Four3. Consequently, these auditing firms are seen as more concerned about their own reputation in the market and thus might require clients to disclose more information (Uyar et al., 2013).
While audit in general is most often seen as the screening process of reports that are connected to mandatory reporting requirements, studies such as Wallace et al. (1994) argue that auditors do not only audit but also influence the level of disclosure. Inchausti (1997) suggests that audit firms use the information disclosed by their clients to signal their own quality to the market. Consequently, clients of Big Four audit firms are expected to disclose higher levels of information.
However, the author is aware of a potential size effect originating from the assumption that bigger firms are more interested in being audited by a Big Four auditor due to the potential positive signalling effects to the market. Referring to Titman and Trueman (1986), firms hiring a big prestigious and known auditor such as a Big Four company could increase the trustworthiness of audited reports and signals quality of the disclosed information to the market which influences the firm’s legitimacy. Therefore, it has to be pointed out that the influence of auditor size for the analytical model has to be interpreted carefully and to be accounted for at the stage of data analysis.
29 | P a g e This study will test the following hypothesis regarding auditor size:
H6: Firms whose GRI report was audited by a Big Four auditor are more likely to engage in the voluntary disclosure of CEO-Employee pay ratio information compared to those firms that were audited by a non-Big Four auditor.
3. Audit Status
Building on the arguments in favour of including the variable auditor size in the research model, this study also tests the influence of GRI report audit status. While most studies focus on the existence of an audit committee or auditor size as
variables to explain voluntary information disclosure, none of them includes the audit status of voluntary disclosed information as a variable. This comes as no surprise as the audit of company reports is taken for granted due to existing regulatory
requirements and also a process that follows already taken disclosure decisions. However, audit requirements are only extending to mandatory financial reporting but not to voluntarily disclosed information and as other studies concluded, auditors are likely to influence the level of information disclosure during the audit process
(Wallace et al., 1994).
The disclosure of CEO-Employee pay ratio information is a form of voluntary disclosure which is not covered by most Corporate Social Responsibility reporting standards. In addition, only a limited number countries in Europe such as France even require mandatory CSR reporting and none of them demands for a mandatory CSR report audit (Kühn et al., 2014).
This study tests audit status based on whether a firm’s GRI report has been externally assured or not using the GRI’s definition of external assurance as
“activities designed to result in published conclusions on the quality of the report and the information (whether it be qualitative or quantitative) contained within it” (see Appendix H for further details) (GRI, 2016b, p. 13).
30 | P a g e an increased value of reporting following from external assurance (GRI, 2016b). Resulting from that, the market is more willing to attribute pragmatic and moral legitimacy to these firms.
However, while the inclusion of audit status finds support in both legitimacy and signalling theory, it has to be pointed out that results regarding the significance of the variable have to be interpreted carefully. Firstly, audit quality varies from firm to firm as companies that voluntarily hire auditors can determine the level and intensity of the auditing process. Secondly, the signal of audit as quality assurance indicator has to be questioned in general as there are issues of conflict of interest present.
Auditors should on one hand be a neutral third party assuring the correctness of the disclosed information, while on the other hand they are interested in maintaining clients and thus might be inclined to overlook shortcomings in voluntary information disclosure. Lastly, the GRI’s definition of external assurance only assures that reports are audited – not verified – what significantly limits the quality of the assurance and the audit process itself.
Being aware of these issues, this study tests the influence of a firm’s GRI report audit status on CEO-Employee pay ratio disclosure using the following hypotheses:
H7: Firms whose GRI report was audited have a higher likeliness of disclosing CEO-Employee pay ratio information.
3.2. Control Variables 1. Economic Development
Bushman et al. (2004) suggest that transparency factors vary depending on the specific economic situation of a firm’s host country environment. Ball (2001) states that the disclosing infrastructure of a country “evolves as a complementary
component of their economic, legal and political infrastructures” (Bushman et al., 2004, p. 209).
31 | P a g e compared to firms in countries that have a negative economic development. As the latter might fear the risk of losing reputation or legitimacy and with it the access to crucial sources of low-cost financing they can be expected to take a non-disclosure position when it comes to the disclosure of their CEO-Employee pay ratio.
This study will use the economic development as control variable for
CEO-Employee pay ratio disclosure which will be measured through three proxy variables: real GDP growth in percent, the inflation rate in percent and the unemployment rate in percent.
2. Political and Legal Environment
Previous research by Bushman et al. (2004) identified the political and legal environment of firms as an important factor influencing corporate transparency. Morris and Gray (2007) points out that the degree of corporate transparency and information disclosure can be influenced and limited “because of political and legal barriers to their implementation at the local (country) level, including [the] quality of [the] legal system and [the] extent of political involvement in the economy” which can result in “cross-country financial reporting differences” (Morris and Gray, 2007, p. 3). Additionally, countries can vary significantly in their legal environments with some of them having very advanced disclosure regulations and strong enforcement
mechanisms (e.g. France requires mandatory CSR reporting from listed firms), while other countries in Europe lack strong regulatory disclosure requirements (Hail and Leuz, 2006).
It is therefore that this study will use the political and legal environment as a control variable.
32 | P a g e Table 1: Research Hypotheses
Variable Hypothesis
H1 Firm Size There is a positive relationship between firm size and the disclosure of CEO-Employee pay ratio information.
H2 Industry Type The type of industry is significantly associated with the likeliness of CEO-Employee pay ratio disclosure.
H3 Firm Age Younger firms are more likely to engage in CEO-Employee pay ratio disclosure than older firms.
H4 Listing Status Listed firms have a higher likeliness to engage in the voluntary disclosure of CEO-Employee pay ratio information.
H5 CEO Gender Firms with a female Chief Executive Officer disclose their CEO-Employee pay ratio to a greater extent than those firms with a male Chief Executive Officer.
H6 Audit Firms whose GRI report was audited by a Big Four auditor are more likely to engage in the voluntary disclosure of CEO-Employee pay ratio information compared to those firms that were audited by a non-Big Four auditor.
H7 Auditor Size Firms whose GRI report was audited have a higher likeliness of disclosing CEO-Employee pay ratio information.
3.3. Conceptual Model
Based on the previous elaborations this study will use the following conceptual model (see Figure 2):
33 | P a g e
4. Methodology
4.1. Choice of Research Design
This study applies the research onion concept (see Figure 3) introduced by Saunders et al. (2009) as a general research framework since it provides an
established and commonly used guideline for the structure of an academic research process including key elements such as the research philosophy, approach,
strategy, choice of method, time horizon as well as data collection and analysis (Blaxter, 2010; Gray, 2013).
Figure 3: Saunders’ Research Onion
Source: Saunders et al. (2009)
As pointed out by Bryman and Bell (2015), a deductive research approach builds on an existing theoretical background for the hypotheses development process and advances from general theory to specific knowledge.
34 | P a g e transparency and information disclosure, the study follows a deductive research
approach and a positivist paradigm as research philosophy to test the specifically
developed hypotheses regarding the disclosure of the CEO-Employee pay ratio. While the chosen approach of positivism can be seen as a partially inflexible concept that does not give much room for adaption during the research process and which is limited in its ability to determine the meanings that people attach to social phenomena, it is still seen as a feasible choice for this study since it allows for a statistical analysis and thus for a comparison of quantitative data to determine
potential underlying patterns and relations between factors (Cupchik, 2001; Creswell, 2013).
The research strategy was based on a secondary data collection process that combined firm-level financial and governance information from well-established and reputable business databases in the field of corporate financial disclosure and corporate social responsibility reporting. Measurement and data sources for the variables used in this study are elaborated in Chapter 4.2 and 4.3 respectively.
In terms of methodological choice this study collected secondary quantitative data through a mono-method analysis of relevant financial and governance factors (Azorín and Cameron, 2010). Primary data collection through interviews or surveys was not seen as a feasible option, as the large-scale, cross-national and multi-industry European sample approach of measurement that this study pursued would have made it difficult to gather the relevant data under the given time and resource restrictions. Furthermore, key financial data on firm size and age as well as on the additional country-level control variables was obtainable through secondary financial databases.
This study used a cross-sectional time horizon, collecting data for the financial year 2014. This is due to both reasons of feasibility and reliability. A larger
35 | P a g e 4.2. Sampling
This paper explores the influence of financial and governance factors on the CEO-Employee pay ratio disclosure for the greater population of listed and non-listed European firms. The specified sampling frame for the study consists of listed and non-listed European companies that complied with the Global Reporting Initiatives’ G4 comprehensive reporting standard for the financial year 2014. The
comprehensive version of the report includes the indicator G4-54 (see Appendix A for the definition and measurement of the indicator as laid out by the GRI) that refers to the disclosure of a firm’s CEO-Employee pay ratio, which is the main focus of this paper. The choice of Europe as geographical setting for this study is due to two considerations: Firstly, data on CEO-Employee pay ratio disclosure is more readily available for European firms compared to companies that are for instance located in North America or Asia. Secondly, the focus on European firms helps to limit potential differences in the legal, social and institutional environment. Although this study controls for differences between different geographical regions within Europe, a larger sample frame including non-European countries was not seen as feasible under the given time and resource restrictions and especially due to the necessary one-by-one screening process for GRI reports that is required to obtain the data.
The study sample includes both financial and non-financial firms. Financial firms are in many European countries (e.g. France, Germany or Switzerland) subject to different disclosure requirements (Glaum and Street, 2003; Berglöf and Pajuste, 2005; Brown and Tarca, 2005). Being aware of this situation, and following the example of Faleye et al. (2013), Eng and Mak (2003) and other studies in the field of corporate information disclosure, this research performs a separate analysis for financial firms to mitigate concerns that the results are driven by companies operating in the financial industry.
Due to the limited number of firms that include the G4-54 indicator in their GRI report for 2014 compared to the overall population size, a non-random convenience
sampling method was chosen. Although non-probability sampling can be criticized
