The Adherence Level of Sustainability
Disclosures and Firm Value:
MASTER THESIS WITHIN: Business Administration NUMBER OF CREDITS: 15 ETCS
PROGRAMME OF STUDY: International Financial Analysis AUTHOR: Daniela Westerlund
PLACE AND DATE: JÖNKÖPING 05 2021
Empirical Study on the Impact of GRI Report’s
Adherence Level in regard to Firm Value in the
Master Thesis in Business Administration
Title: The Adherence Level of Sustainability Disclosures and Firm Value
Author: Daniela Westerlund Tutor: Haoyong Zhou Date: 2021-05-24
Key terms: GRI, Sustainability Reporting, Sustainability/Environmental Disclosure, Corporate Financial/Sustainability Performance, Firm Value, Tobin’s Q, Stakeholder, Adherence Level.
Abstract
Background: Sustainability reporting has become increasingly important for
firms that want to appease their stakeholders and the society, whilst possibly increasing the corporate financial performance (CFP) of the firm. This is because sustainability disclosures currently work as the main channel for firms to inform their stakeholders of the CSR practices and environmental management carried out by the company. However, there have been various previous studies that examine the relation between corporate social performance (CSP) or the reported CSP, and CFP but not a study that would focus on GRI’s adherence level and its effect on Firm value (FV). The adherence level in the context of a GRI Report refers to the extent to which the GRI Sustainability Reporting Framework and GRI Standards have been applied to a company’s sustainability report (Global Reporting Initiative, n.d.). This study intends to examine if stakeholders can be affected by a sustainability report’s adherence ranking made by GRI, although there necessarily would not be a clear connection to a company’s actual environmental performance.
Purpose: The purpose of this study is to find out if the adherence level affects a
firm’s value and how, although this classification of reports would not say anything about a company’s level of sustainability or a company’s sustainability performance. In short, the study wants to examine if stakeholders or the society surrounding a company are affected by the adherence level of a company’s GRI reporting and if this then can affect the value of the organization in any way.
Aim: The aim of this research is to encourage organizations to become more
transparent or elaborate regarding their sustainability practices if any significance between the adherence level and the FV can be found.
Method: This study was conducted by examining 98 European manufacturing
firms’ GRI adherence levels for the years 2017 to 2019 and comparing them to respective Firm Values (Tobin’s Q) by the usage of panel data regression analysis.
Conclusion: The results show that no significant relationship between the GRI
adherence level and FV can be found in the European manufacturing industry for the period 2017 to 2019.
Acknowledgment
I would like to thank my supervisor Haoyong Zhou for his guidance during this writing process. In addition, I would like to thank the students and friends that made suggestions of improvement to this paper and helped me along the process with any
questions that arose. A special thanks belongs to my partner and my family that supported and encouraged me throughout the extraordinary writing process influenced
by COVID-19.
Jönköping International Business School, 24th of May 2021
___________________
Daniela WesterlundTable of Contents
... 1
Introduction ... 6
Background ... 6 Problem Discussion ... 6 Research Question ... 7Aim and Purpose... 7
Delimitations ... 8
European Manufacturing Industry ... 8
Frame of Reference and Hypothesis ... 10
Theory ... 10
Description of Theories ... 10
2.2.1 Stakeholders & The Stakeholder Theory ... 10
2.2.2 Legitimacy Theory ... 12
2.2.3 Voluntary Disclosure Theory ... 13
2.2.4 Sustainability Disclosures ... 13
2.2.5 Global Reporting Initiative (GRI) ... 14
Previous Research ... 14
2.3.1 Corporate Financial Performance – CFP ... 14
2.3.2 Corporate Social Performance – CSP ... 15
2.3.3 Relation between CSP and CFP ... 16
2.3.4 Relation between sustainability disclosures and CFP ... 17
2.3.5 Summary of Previous Studies ... 19
Hypothesis ... 21
Research Methodology & Method ... 22
Research Methodology ... 22 3.1.1 Conceptual Framework ... 22 3.1.2 Philosophical Framework... 23 Research Method ... 23 3.2.1 Data Collection ... 23 3.2.2 Independent Variables ... 25 3.2.3 Dependent Variables ... 25 3.2.4 Control Variables ... 26 3.2.5 Analysis Method ... 26
Criticism and Quality Criteria ... 29
3.3.1 Reliability ... 29
3.3.2 Validity ... 30
3.3.3 Generalizability ... 30
Empirical Results ... 31
GRI Adherence Level ... 31
Descriptive Statistics (2017 – 2019) ... 32 4.2.1 2017 ... 32 4.2.2 2018 ... 32 4.2.3 2019 ... 32 Pearson Correlation (2017 - 2019) ... 33 4.3.1 2017 ... 33
4.3.2 2018 ... 33
4.3.3 2019 ... 33
Panel Data Regression Results ... 33
4.4.3 Regression Results ... 34
Data Assumption Test Results ... 35
Summary of Results ... 35
Analysis & Discussion ... 37
Results and the Frame of Reference ... 37
5.1.1 The Stakeholders and the Stakeholder Theory ... 37
5.1.2 Legitimacy Theory ... 38
5.1.3 Voluntary Disclosure Theory ... 39
5.1.4 Previous Studies ... 39
Results and Sustainability Reporting & GRI ... 41
Results and the Data & Delimitations ... 41
5.3.1 The Sample Companies ... 41
5.3.2 Time Period (2017-2019) ... 42
Results and the Problematization ... 42
Conclusion ... 43
Future Studies ... 43
Limitations ... 43
Reference list ... 45
APPENDIX ... 50
Appendix 1 - Sample Companies ... 50
Appendix 2 – Regression Assumptions ... 53
Appendix 3 – Regression Tests ... 53
Abbreviations
BP-LM Test: Breusch and Pagan Lagrange Multiplier Test CFP: Corporate Financial Performance
CSP: Corporate Social Performance CSR: Corporate Social Responsibility ESG: Environmental, Social and Governance
FEM: Fixed Effects Model
FV: Firm Value
GRI: Global Reporting Initiative MNE: Multinational Enterprise OLS: Ordinary Least Squares
REM: Random Effects Model
SME: Small and Medium-sized Enterprise
Definitions
GRI Standards: The latest generation of GRI’s sustainability reporting guidelines. The GRI Standards are being used by organization since July 2018 (Global Reporting Initiative, n.d.).
GRI G4: The 4th generation of the Global Reporting Initiatives (GRI’s) sustainability reporting guidelines (BSR, 2013).
Adherence level: The adherence level in the context of a GRI Report refers to the extent to which the GRI Sustainability Reporting Framework and GRI Standards have been applied to a company’s sustainability report (Global Reporting Initiative, n.d.). I. Comprehensive: “This builds on the Core option by requiring additional
disclosures on the organization’s strategy, ethics and integrity, and
governance. In addition, the organization is required to report more extensively on its impacts by reporting all the topic-specific disclosures for each material topic covered by the GRI Standards” (Global Reporting Initiative, 2016). II. Core: “This option indicates that a report contains the minimum information
needed to understand the nature of the organization, its material topics and related impacts, and how these are managed” (Global Reporting Initiative, 2016).
III. Referenced/Undeclared: No explicit ‘in accordance’ option declared or has not been prepared in accordance with the GRI Standards.
IV. None: No GRI Report.
Stakeholders: Stakeholders are an organizations suppliers, customers, employees, investors, communities, and others (Stakeholder Theory, 2018). A stakeholder, according to Freeman (1984, p.46) is “any group or individual who can affect or is affected by the achievement of the organization’s objectives”. Eden and Ackermann (1998, p.117) define a stakeholder as “People or small groups with the power to respond to, negotiate with, and change the strategic future of the organization”.
Introduction
This chapter will cover this papers background, problematization, research question and hypothesis, aim and purpose, the perspective that will be used in the study, delimitations, and a short description of the European manufacturing industry and its connection to sustainability disclosures.
Background
Green is today’s new black when it comes to sustainability reporting for companies, and this can be seen in the worldwide sustainability reporting rate which is a staggering 80%. The reason for the increased reporting of a company’s, not only economic, but also environmental and social practices is not only due to new regulations and laws but also because it has an impact on a firm’s corporate value and financial performance (KPMG, 2020). Stakeholders, an organizations suppliers, customers, employees, investors, communities, and others (Stakeholder Theory, 2018), currently value sustainability highly and as a result the company increases their focus on it as they want to attract the goodwill of their stakeholders. Therefore, companies that want to retain their value must invest in the best possible Corporate Social Responsibility (CSR) and report it visibly to their stakeholders (Tapaninaho & Kujala, 2019).
The Global Reporting Initiative (GRI) currently remains the dominant standard for sustainability reporting (KPMG, 2020). GRI reporting help organizations better understand their impacts on the economy, society, and environment. The goal of GRI Reporting is to make companies more transparent and increase the accountability in organizations’ sustainable development (Global Reporting Initiative, n.d.).
Problem Discussion
Sustainability is important, and organizations want to appease their stakeholders to a large extent while simultaneously gaining a higher corporate value. Also, the Investor Responsibility Research Center Institute (IRRCi) suggests that stakeholders involve sustainability as a factor to investment decisions (IRRCi, 2016). Hence, this study will tackle the question of the correlation between sustainability reporting and FV from a different angle where the adherence level of the GRI Report will be considered instead of a subjective definition of its quality or quantity. Adherence level refers to GRI’s own categorization of reports. The GRI G4 and the GRI Standards report type allows reports to be conducted in accordance with either the Core or the Comprehensive level. If the
report is not in accordance with the GRI Standards its adherence level is either Undeclared, Referenced, or None depending on if it is a GRI G4 Report or a GRI Standards Report, respectively (Global Reporting Initiative, n.d.). Also, this study looks at Europe’s Large and Multinational Enterprises (MNE’s) which covers a new region for this type of study. In short, the previous literature has not covered how the adherence level of sustainability reports may or may not affect financial performance of manufacturing firms in Europe and that is why this study will contribute with a new perspective to the existing literature.
Research Question
The research question of this study is the following:
How does the adherence level of the sustainability report affect European manufacturing companies’ firm value?
Specifically, this paper will test the following hypotheses:
H0: The adherence level on a company’s GRI Report will have no effect on a company’s firm value.
H1: The adherence level on a company’s GRI Report will have a negative effect on a company’s firm value.
H2: The adherence level on a company’s GRI Report will have a positive effect on a company’s firm value.
This study will look at the GRI adherence levels in the manufacturing industry in Europe. The FV will be quantified by using Tobin’s Q.
Aim and Purpose
This study will by no means imply that a company using a more extensive application of the GRI sustainability reporting framework and GRI Standards would be more sustainable than a company using a less detailed report. Instead, the purpose is to find out if the adherence level affects the firm’s value and how, although this classification of reports would not say anything about the company’s level of sustainability or the company’s sustainability performance. In short, the study wants to examine if stakeholders or the society surrounding a company are affected by the adherence level of a company’s GRI reporting and if this then can affect the value of the organization in any way. The aim of this research is to encourage organizations to become more transparent
and elaborate regarding their sustainability practices if any significance between the adherence level and the FV can be found.
Delimitations
This study will make use of both the GRI Standards and GRI G4 Reports since they are the most recent generations of the GRI reporting guidelines. However, the focus lies within the adherence levels of these reports and both GRI Standards and G4 are measured similarly with the core, comprehensive, undeclared/referenced and ‘none’ options. Hence, the use of both GRI Reporting generations should fulfil the purpose of this paper. However, this research does not represent all generations of the GRI Reports. Also, the study considers only one region and one industry (several industry sectors however), Europe, and the manufacturing industry, hence this study does not represent any other region or industry. Also, the sample companies are Large or Multinational Enterprises thus the finding of this research cannot be considered for Small and Medium-sized Enterprises (SME’s).
European Manufacturing Industry
The sample firms used in this paper are European since the degree of sustainability development strategies are similar, which means that in the region the same philosophy is applied when it comes to sustainability practices (Boiral, 2013). Whenever a firm is said to be European in this study it is referred to the location of the firm’s headquarters. In the case of subsidiaries European means that the company reports to this region (Global Reporting Initiative, n.d.).
The manufacturing industry in Europe generated €1820 billion value added in 2017 and employed 28.5 million people. The industry is composed of 24 subsectors whereas a sample of 98 firms of the five largest subsectors are included in this study (automotive, chemicals, equipment, food and beverage products, and metals products) (European Union, 2020).
Many of the European countries belong to the European Union (EU) which means that listed companies with more than 500 employees from these states must comply with EU regulations regarding disclosures on social and environmental challenges. The
information that must be disclosed contains environmental and social matters, treatment of employees, human rights, bribery, anti-corruption, and diversity of company boards (European Commission, 2020). These matters regulated by the EU are included in the GRI Standards (Global Reporting Initiative, 2016).
Frame of Reference and Hypothesis
This chapter will provide the theoretical background to Stakeholder Theory, Legitimacy Theory, Voluntary Disclosure Theory, sustainability disclosures, GRI and, finally, on previous studies within the topic. The hypothesis will be formulated with the help of the theories and previous studies and will be presented in the end of this chapter.
The sources utilized for this theoretical framework have been gathered using primo and google scholar. The key words used in the search process were GRI, sustainability reports, sustainability/environmental disclosure, firm value, adherence level, stakeholders, CSR, CFP, CSP, and different combinations of all these. All journals considering previous research will have the ABS level of 2 or higher, are peer-reviewed and have a considerable number of citations to ensure the relevance of the studies.
The previous studies will be presented descriptively with a categorisation of CFP and CSP and finally the found relationship between these two concepts. The reason for this categorisation is that there seem to be a wide range of different definitions and ways to measure both CFP and CSP in the previous studies.
Theory
The previous literature findings regarding CSP and sustainability disclosures in relation to CFP and FV are mixed to some degree and the most used theories are the Stakeholder Theory and the Legitimacy Theory. The Voluntary Disclosure Theory is mentioned as a contrast to the Legitimacy Theory (Boiral, 2013). Hence, the focus of the theories used in this paper will be both Stakeholder Theory and Legitimacy Theory, however the Voluntary Disclosure Theory will be taken into consideration since it is focused on sustainability reporting, and thus it may contribute to this research. Additionally,
sustainability disclosures and more specifically GRI reporting will be discussed in more detail in this theory section.
Description of Theories
2.2.1 Stakeholders & The Stakeholder Theory
Analysing previous research that considers CSR and its effect on corporate financial performance (CFP), it is apparent that Stakeholder Theory is widely used, and for a good reason. According to Montiel and Delgado-Ceballos (2014) the Stakeholder Theory is one of the most common approaches used in sustainability, social and environmental management research. The reason for the frequent use of Stakeholder Theory in papers regarding sustainability reporting is that the theory offers an explanation for why companies would engage in what would otherwise be considered unconventional, or unprofitable practices (Fernando & Lawrence, 2014). Also, according to Donaldson and Preston (1995) there are distinctions between different stakeholder theories. In this study the theory applied will be the original version, the normative or also called the integrative
Stakeholder Theory that was brought forward by Freeman and his colleagues in 1984 and further developed in 2010 (Hörisch, Freeman & Schaltegger, 2014). However, the groundwork for the framework was made by Igor Ansoff already in 1965 (Roberts, 1992). The reason for using the integrative Stakeholder Theory is that this theory is closely linked to sustainability management (Hörisch et al., 2014) hence it is relevant for this study where the purpose is to find out if the adherence level of a sustainability reports affects FV. Firstly, a stakeholder will be defined and secondly the Stakeholder Theory will be explained further.
A stakeholder, according to Freeman (1984, p.46) is “any group or individual who can affect or is affected by the achievement of the organization’s objectives”. Eden and Ackermann (1998, p.117) define a stakeholder as “People or small groups with the power to respond to, negotiate with, and change the strategic future of the organization”. The term seems to have been first implemented in the early 1960’s where the intension was for the term to sound reminiscent to a stockholder and to imply that there may be other individuals or groups that have an interest in an organization and its management (Goodpaster, 2015). Also, according to Kenneth Goodpaster who is a Professor emeritus in Ethics and Business Law, a stakeholder can be for example a shareholder, employee, supplier, creditor, competitor, government, customer, or a community (Goodpaster, 2015).
If we compare stakeholders and their importance in the Europe and the US, we can see from Jurgens, Berthon, Papania and Shabbir’s (2010) study that in the US, companies steer businesses themselves whereas in Europe the stakeholders steer companies, to fulfil stakeholder interests, to a wide extent. In short, stakeholders in Europe play a greater part in company decision making in comparison to stakeholders in the US. However, there seem to be differences also in the European stakeholder model where the Northern European (Scandinavia, Austria, and Germany) model suggests the use of a corporate governance approach and the British and Irish models seem to be a mixture of the European and the US model. The Northern European, also called the Rhineland, approach is a system of co-management between stakeholders where all strive for the same goal, to keep the organization running (Jurgens et al., 2010). The Stakeholder Theory that will be used in this research, just as in the study made by Hörisch et al (2014), is the Rhineland approach (Jurgens et al., 2010) where the focus lies in managing stakeholder relationships
rather than managing stakeholders. This approach has a normative core and is integrative (Hörisch et al., 2014) and it works well with the region studied in this paper.
The main purpose of the Stakeholder Theory is to encourage organizations to become connected closely with their stakeholders to ensure the company’s future. This can be done by aligning operations with stakeholder desires and keep stakeholders informed of the said operations (Fernando & Lawrence, 2014). Freeman (2010) says that the purpose of business is to create value for all stakeholders but the easiest way to do so is when all stakeholders come together and collectively improve the situation for everyone involved. In short, the company will bloom when the company’s purpose aligns with the stakeholder preferences, this being the basis to the Stakeholder Theory (Fernando & Lawrence, 2014).
To create accountability between stakeholders and an organization, stakeholders desire a trustworthy way to see if their preferences have been met or not, usually by asking for an annual report or a sustainability report depending on the stakeholder’s inclination towards the organization (Logsdon & Lewellyn, 2000). Environmental and social values are increasingly important for stakeholders (KPMG, 2020) and with the help of sustainability reports organizations can communicate that they fulfil stakeholder desires and their duty to the society (Karagiorgos, 2010). Hence, the Stakeholder Theory can be used as a tool by managers to better understand how to create value to stakeholders whilst simultaneously combine business and ethics (Freeman, Rusconi, Signori & Strudler, 2012).
The Stakeholder Theory is overlapping with the Legitimacy Theory when considering the relationship between society and a company, but the Stakeholder Theory is much narrower because it considers only stakeholders and not the whole society (Schaltegger & Burritt, 2010).
2.2.2 Legitimacy Theory
According to Suchman (1995, p. 574) “Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions.” Lindblom (1993) says that legitimacy from an organization’s perspective can be achieved when an entity’s
value system aligns with the value system of the society. A threat to the company’s legitimacy, which is also a threat to its existence, arises if its values differ from the larger social system. An illegitimate organization will face difficulties in its attempt to attract stakeholders (Deegan, 2006) since the society provides the organization with resources and employees according to how much the entity is wanted by the society (Deegan, 2004). Traditionally, in economics, profit maximisation was viewed as a measure of corporate performance but according to the Legitimacy Theory profit is a metric of the company’s level of legitimacy (Ramanathan, 1976). In short, a company’s legitimacy is a resource it depends on for its survival (Deegan, 2006).
Deegan (2004) argues that quite a few previous studies use Legitimacy Theory to study social and environmental reports hence proposing a relationship between corporate disclosures and the expectations of the community. However, Legitimacy Theory itself argues that companies would improve the perception that the public has of their sustainability practices using sustainability disclosures (Deegan, 2002). It is also said that the relationship between a company’s performance within sustainability and its sustainability disclosure would be negative if we look at the Legitimacy Theory although the relationship between these variables remains somewhat ambiguous in both theory and empirics (Hummel & Schlick, 2016).
2.2.3 Voluntary Disclosure Theory
Controversially to the Legitimacy Theory the Voluntary Disclosure Theory suggests that the relationship between an organizations performance in sustainability has a positive correlation to its sustainability reporting practices (Hummel & Schlick, 2016). This theory was originally applied only to financial voluntary disclosure but later has been used also for non-financial voluntary disclosures. The reasoning behind this is that a company with higher-level CSP would voluntarily reveal its true performance with the hope of it leading to higher market value (Clarkson, Li, Richardson & Vasvari, 2008). Also, public reputation and a relevant sustainability image keeps companies on their toes when it comes to untruthful sustainability reporting (Verrecchia, 2001).
2.2.4 Sustainability Disclosures
Sustainability reporting has become increasingly important, especially for large manufacturing firms since they have a significant impact on the environment (KPMG,
2020). These disclosures’ purpose is to provide a company’s stakeholders with an accurate and reasonable representation of a firms CSP including both negative and positive contributions (Global Reporting Initiative, 2016). However, some scholars argue that sustainability reports can camouflage, or can be used to camouflage, the actual sustainability related issues in a company (Boiral, 2013). Contrarily, according to previous research and the Voluntary Disclosure Theory there is a positive correlation between CSP and sustainability disclosures (Hummel & Schlick, 2016).
2.2.5 Global Reporting Initiative (GRI)
To strengthen the accuracy and transparency of sustainability reporting to stakeholders is the main goal of the GRI. To be able to achieve this goal GRI assesses various indicators such as stakeholder inclusiveness, sustainability context, completeness balance, materiality, clarity, comparability, and reliability (Boiral, 2013). In addition, GRI reports provide an alternative way to investigate the relationship between sustainability reports of CSP and CFP (Chen, Feldman & Tang, 2015).
Previous Research
2.3.1 Corporate Financial Performance – CFP
In the previous research about the relationship between CSP and CFP two main categories of economic performance, or CFP measures, can be found that are being used: market-based measures (e.g., stock market performance, market return) and accounting-market-based measures such as ROA or asset turnover (van Beurden & Gössling, 2008; Brooks and Oikonomou, 2018).
Rhou, Singal and Koh (2016) mention in their research that stock-market or market-based performance measures would be less subjective compared to accounting-based measures when considering managerial manipulation for example (Branch, 1983). Thus, Tobin’s Q ratio, which is a market-based measure, defines the CFP in the study made by Rhou et al. (2016). Fatemi, Glaum, and Kaiser (2018) and Li, Gong, Zhang, and Koh (2018) quantify FV also with Tobin’s Q. Also, Klassen and McLaughlin (1996) refer in their study to the stock return when discussing CFP. The authors chose this metric due to public events, such as environmental crises, being highly related to a firm’s stock price (Klassen &
McLaughlin, 1996). Additional market-based measure is market value of equity that defines the FV in the study made by Hassan (2018).
Plumlee, Brown, Hayes, and Marshall (2015) measure FV using a company’s cost of equity components and cash flow and López, Garcia, and Rodriguez (2007) used different accounting-based performance indicators (e.g., ROE, ROA, cost of capital) to measure the corporate performance of the European sample firms in their study. Also, Chen et al. (2015) use ROE as a metric for CFP.
Rezaee and Tuo’s (2019) study examines the relationship between the quantity/quality of sustainability disclosures and innate/discretionary earnings quality. Innate earnings quality refers to the earnings quality derived from natural traits of a company such as the business model. Discretionary earnings quality refers to earnings quality that stems from non-natural company traits such as managerial short-term decisions. Summarized, the CFP in this study is measured through a percentage of income resulting from higher sales or lower costs i.e., earnings quality (Rezaee & Tuo, 2019).
2.3.2 Corporate Social Performance – CSP
CSP has been measured in various ways in the previous literature (Brooks & Oikonomou, 2018; van Beurden & Gössling, 2008) hence CSP can be divided into three categories according to van Beurden and Gössling (2008). First parameter indicates the level of a firm’s sustainability disclosures which can be measured by a content analysis for example. The second level of CSP is about corporate actions and the third one concerns the corporate reputation ratings such as Fortune or Business Ethics.
Chen et al. (2015) and Plumlee et al. (2015) measure CSP by looking at voluntary disclosure quality (e.g., GRI indicators) which refers to the first CSP parameter mentioned by van Beurden and Gössling (2008). Similarly, Rezaee and Tuo (2019) use GRI database in their study to determine both the quantity and quality of the sustainability disclosures for the chosen sample companies. Rhou et al. (2016) uses CSR and CSR awareness to describe CSP in their study and these variables are measured by KLD STATS which falls into the third category (third party social rating agency) of CSP if we follow Beurden and Gössling’s (2008) categorization of CSP. Similarly, Fatemi et al. (2018) uses KLD Research and Analytics to measure CSP but also the Bloomberg’s
disclosure score (DISC), which also Hassan (2018) and Li et al. (2018) uses, when measuring ESG disclosures.
Brooks and Oikonomou (2018) examine both sustainability disclosures and CSR performance in relation to CFP studies in their literature review which can be advantageous since Klassen and McLaughlin (1996) mentions that voluntary environmental self-disclosures by firms cannot necessarily be associated with the actual environmental performance of the organization. Klassen and McLaughlin (1996) also argue that there is a difference between environmental performance in the long-term and the short-term. To avoid these potential weaknesses in the study Klassen and McLaughlin (1996) use environmental events such as third-party award announcements or environmental crises to measure environmental performance since these events usually inform the public of strong or respectively weak historical performance within environmental management. Alike Klassen and McLaughlin (1996) are López et al. (2007) discussing the short- and long-term effects of sustainability practices being taken by a firm. To avoid these differences the authors, turn to conduct the study in the same region (Europe) where the development of sustainability practices is similar. In this study sustainability practices are measured by the usage of the Dow Jones Sustainability Group Index (DJSI) since the metric is widely used in European firms to disclose their sustainability, similarly to GRI or the Global Compact (López et al., 2007).
2.3.3 Relation between CSP and CFP
Klassen & McLaughlin (1996) conducted their study on the linkage between environmental management within a firm to its financial performance. The authors find that environmental crises are significantly related to a firm’s negative returns, thus supporting the notion of a linkage between financial performance of a firm and its environmental management practices. Also, positive unpredictable stock-returns seem to follow environmental events that are of a positive nature according to the study (Klassen & McLaughlin, 1996). López et al. (2007) argue that both investors and businesses entrust that the firm can gain long-term value when sustainability practices are implemented. However, the study finds a negative impact of CSR practices on business performance in the short-term. Since van Beurden and Gössling (2008) find most of the literature review studies they used having a positive correlation between CSP and CFP they conclude that there is clear empirical evidence of a positive relationship. They also mention that the
studies that found a significant negative relationship would use outdated material for their studies (van Beurden & Gössling, 2008).
Another research perspective that supports the Stakeholder Theory by Rhou et al. (2016) finds that sustainability initiatives can result in financial gain depending on the level of stakeholders’ CSR awareness. In short, if stakeholders are aware of different sustainability practices it can lead to a benefit for a firm with CSR practices which can lead to possible improvement of the company’s financial performance. Fatemi et al. (2018) supports the notion that strengths within a firm’s environmental practices increases FV but also that the opposite is true when it comes to weaknesses within environmental practices.
2.3.4 Relation between sustainability disclosures and CFP
Chen et al. (2015) looked at the relationship between CSR disclosures and financial performance. The authors argue that there is not a lot of previous studies that would use GRI or another internationally accepted CSR reporting standard to explore this relationship. The use of 75 sample companies in the manufacturing industry resulted in a significant positive correlation between the GRI categories Human Rights, Society and Product responsibility and ROE. This conclusion holds for various CSR indicators according to the authors (Chen et al., 2015). Another study in 2015 by Plumlee et al. (2015) find that voluntary environmental disclosure quality is positively related to a company’s cost of equity components and cash flow (i.e., FV). The study uses both industry sectors that are sensitive to environmental issues and industry sectors that are not, within the manufacturing industry. Similarly, Brooks and Oikonomou (2018) find that social disclosures are in general correlated with better CSR performance, which supports the Voluntary Disclosure Theory, as well as better CFP and that there is a positive significant relation between CSP and CFP, but controversially with an unassuming economical linkage. Furthermore, Fatemi et al. (2018) looked at ESG disclosures in isolation and concludes that they can be found to decrease FV whilst also in 2018 Hassan (2018) argues that enhancement of a sustainability report would improve organizational visibility and attract investors. Li et al. (2018) argues there to be a positive association between ESG disclosure level and FV but also that interaction between a sustainability disclosure and CEO power relates to FV positively. However, Rezaee and Tuo (2019) find a positive correlation both on quality and quantity of a sustainability
disclosure with innate earnings quality. Simply put, a company’s financial well-being would be positively correlated to the quantity and quality of a sustainability report conducted. Interestingly, the study finds as well that, for example, managerial short-term decisions would not affect the quality or quantity of a CSR disclosure (Razaee & Tuo, 2019).
2.3.5 Summary of Previous Studies
Table 1: Previous Studies
Author(s) Year Title Country/ Industry
Main research question
Empirical basis Theory applied Main findings Klassen McLaughlin 1996 The Impact of Environmental Management on Firm Performance Manufacturing Industry Examines if strong environmental performance positively affects the financial performance of the firm, and conversely if weak environmental performance has a negative effect. NEXIS database. 1985-1991. Listed firms NYSE or AMEX. 96 manufacturing firms. Efficient Market Theory Environmental and perceived future financial performance are linked. Positive abnormal stock returns from positive environmental events. López Garcia Rodriguez 2007 Sustainable Development and Corporate Performance: A Study Based on the Dow Jones Sustainability Index
Europe Examines whether business
performance is affected by the adoption of practices included under the term CSR.
2 groups of 55 firms for the period 1998-2004.
Legitimation Theory, Triple Bottom Line
Differences in CFP exist between firms (DJSI and DJGI). These differences exist due to CSR practices. A negative short-term impact on CFP was found. Argues for long-term value when it comes to CSP strategies. van Beurden Gössling 2008 The Worth of Values: A Literature Review on the Relation between Corporate Social and Financial Performance
Not specified What is, according to the literature, the relationship between Corporate Social Performance and Corporate Financial Performance, and which factors have an influence upon it? 34 studies (literature review) Many different theories (literature review) Majority of previous studies find positive correlation between CSP and CFP. States that there is clear empirical evidence for a positive correlation between CSR and financial performance (voices that state the opposite refer to outdated material). Chen Feldmann Tang 2015 The relationship between disclosures of corporate social performance and financial performance: Evidences from GRI reports in manufacturing industry Manufacturing Industry Examines the relationship between CSR disclosures and CFP.
GRI report as data sources, 75 sample companies, financial data from annual reports of the sample companies
Stakeholder theory and Institutional Theory
The results indicate that some categories (Human Rights, Society and Product Responsibility) have a positive
correlation with ROE. Same conclusion is argued to be true for many CSR indicators. Plumlee Brown Hayes Marshall 2015 Voluntary environmental disclosure quality and firm value: Further evidence
US - Different industries
Examines the relationship between VEDQ and firm value.
Five different industries, both environmental issue sensitive and non-sensitive. Economic Theory Positive disclosure quality is positively and significantly related with CFP
Rhou Singal Koh 2016 CSR and financial performance: The role of CSR awareness in the restaurant industry. International Journal of Hospitality Management Restaurant Industry Investigates the moderating role of CSR awareness, measured by CSR media coverage, on the relationship between CSR and corporate financial performance (CFP) in the restaurant context. 53 firms with 369 firm-year observations Stakeholder theory CSR awareness affects how CSR initiatives can result in improved CFP. Brooks Oikonomou 2018 The effects of environmental, social and governance disclosures and performance on firm value: A review of the literature in accounting and finance
Not specified Examines the relationship between ESG disclosures and CFP. 45 years of empirical research in accounting and finance Many different theories (literature review)
ESG disclosures are linked to better CSP and better CFP. Fatemi Glaum Kaiser 2018 ESG performance and firm value: The moderating role of disclosure
US Examines whether the association between a firm's ESG activities and its valuation is moderated by its ESG-related disclosure. 1640 firm-year observations for publicly traded U.S. firms for the years 2006 to 2011 Neoclassical theory, Stakeholder theory, Voluntary Disclosure theory ESG strengths (weaknesses) increase (decrease) firm value. ESG disclosure decreases firm value.
Hassan 2018 The impact of voluntary environmental disclosure on firm value: Does organizational visibility play a mediation role? International coverage Examines whether organizational visibility plays a mediation
role in the observed relationship between CED and firm value. Sample of S&P Global 1200 companies from 2010 to 2015 with 2,365 firm‐year observations 815 companies from 29 countries Voluntary Disclosure theory, Legitimacy theory, Discretionary Disclosure theory By enhancing a sustainability report a corporate management could improve organizational visibility and attract more financial analysts. Li Gong Zhang Koh 2018 The impact of environmental, social, and governance disclosure on firm value: The role of CEO power
UK Whether or not ESG information disclosure prompts value creation. And, if it does, what are the drivers? 2415 firm-year observations representing 367 individual firms between 2004 and 2013. Stakeholder Theory, Agency Cost Theory, Positive Synergies Theory, Affordability Theory
The level of ESG disclosure is positively related to FV. Also, the interaction between ESG disclosure and CEO power is positively associated to FV. Rezaee
Tuo
2019 Are the Quantity and Quality of Sustainability Disclosures Associated with the Innate and Discretionary Earnings Quality?
US Examines the
quantity and quality of sustainability disclosures and their relationship with earnings quality. 35,110 firm-year observations between 1999 and 2015 Stakeholder Theory, Legitimacy Theory, Voluntary Disclosure Theory Sustainability disclosure quantity and quality is positively associated with earnings quality that is innate. A negative correlation was produced between sustainability reports quantity and quality and discretionary earnings quality.
Hypothesis
Although the results of previous studies are somewhat mixed it seems that most of the recent studies find that both CSP and voluntary sustainability disclosures are significantly positively correlated to CFP, at least in the long-term. Positive correlation between good CSR practices and stakeholder perception of a company leads to improved financial performance, hence it should be rational to assume that opposite is true as well. This implies that there may be some incentive for a firm to illuminate only positive practices and not mention irresponsibility (Boiral, 2013). However, this is not what this paper sets out to investigate as GRI is the ones responsible for authenticating if adherence level has been followed and is correctly illustrating said companies actual CSP. With this said, the contradiction of Voluntary Disclosure Theory and Legitimacy Theory cannot be answered with this study. Furthermore, with the support of the theories, that are highly connected to the research question of this study (besides this contradiction), and previous studies made, the following relationship is hypothesised as:
H0: The adherence level on a company’s GRI Report will have no effect on a company’s firm value.
H1: The adherence level on a company’s GRI Report will have a negative effect on a company’s firm value.
H2: The adherence level on a company’s GRI Report will have a positive effect on a company’s firm value.
Firm Value
Research Methodology & Method
There is a clear distinction between the research methodology and method in this paper. Methodology referring to the theory about how the study should be executed, whereas method stands for the techniques used to collect and dissect data (Saunders, Lewis & Thornhill, 2016). This chapter firstly elaborates on the research methodology used by presenting the underlying conceptual framework and the philosophical framework. This is followed by the research method parts; data collection, variables used in the study, analysis method and finally the criticism and quality criteria of the chosen methodologies and methods.
Research Methodology
According to Collis and Hussey (2009, p.67) “A methodology is an approach to the process of the research, encompassing a body of methods”. This study implements a quantitative and deductive research approach to be able to investigate the purpose and research question, test the hypotheses and fulfil the aim of the research. This research is making use of an experimental study methodology since it investigates mainly the relationship between GRI adherence level and FV (Saunders et al., 2016). The hypotheses of this study investigate how the GRI adherence level affects the dependent variable (Tobin’s Q) thus elaborating on the possible connection between CSP (if it is assumed that the adherence level of a sustainability report represents CSP of a firm) and CFP. This can be seen in the study’s conceptual framework below.
3.1.1 Conceptual Framework
Although the research methodology in this study is deductive rather than inductive or abductive (Collis & Hussey, 2009) a conceptual framework is added. The reason for this is that it may be easier to organise the underlying idea behind the steps taken to achieve the purpose of this research by implementing this framework (Saunders et al., 2016).
Figure 1: The Conceptual Framework
The conceptual framework above is determined by the explanatory variable, GRI adherence level since it is the variable this study is focused on. This same focus can also
Sustainability
Disclosures Adherence GRI
Level Stakeholder Theory Legitimacy Theory Voluntary Disclosure Theory Previous Studies Corporate Social Performance Corporate Financial Performance
be seen in the study’s hypotheses. However, the conceptual framework does not include the two control variables (size of the firm and growth) that are used in this study. These variables are included to be able to oversee the effect of other explanatory variables on the dependent variables. Summarized, this study will look at CSP via sustainability disclosures by considering their adherence level, simultaneously carrying lenses that consider previous studies and relevant theories, to be able to reflect on the possible impact they have on FV, and on a bigger scale, on the CFP.
3.1.2 Philosophical Framework
According to Collis and Hussey (2009) something that steers how research, more specifically scientific research, ought to be carried out is called a research paradigm which is a philosophical framework. A positivist research paradigm or philosophical framework is used in this study. Positivism is a paradigm that assumes an objective and singular truth to science and that the truth or findings do not change just because they are being measured or investigated. Furthermore, this paradigm is mostly used in quantitative research where every reasonable phenomenon can be explained scientifically by logical proof with involving a deductive process into the study (Collis & Hussey, 2009). Thus, this paradigm or philosophy suits this research better than an interpretivist one.
Research Method
According to Collis and Hussey (2009, p.67) “A method is a technique for collecting and/or analysing data”. Additionally, it is vital to make sure that the methods chosen for this study support the philosophical and conceptual assumptions (Collis & Hussey, 2009). Hence, the data collection, the variables and analysis method have been chosen into this study by keeping the underlying assumptions in mind.
3.2.1 Data Collection
This research makes use of secondary data from the GRI database, Statista, and the sample firms’ financial statements/annual reports. Secondary data is defined as data that is not of one’s own experiments or collections but rather data that can be collected from an existing source like databases, internet etc. (Collis & Hussey, 2009). The secondary data used in this paper is compiled of many different sources making it into multiple-source secondary data (Saunders et al., 2016).
3.2.1.1 Sample selection and Delimitations
This research looks at a specific phenomenon, in a specific geographical area, in a specific industry. That is why a non-probability sampling method, purposive sampling, is used to retrieve the sample (Saunders et al., 2016). This study examines 98 European large enterprises and Multinational Enterprises (MNE’s). The reason for this is that large companies, global ones especially, are usually the driving forces for sustainability reporting trends, and these are adopted by others later subsequently (KPMG, 2020). GRI’s definitions for Large Enterprises and MNEs follows the EUs definitions which are:
Large Enterprise: Headcount is larger or equal to 250. Turnover is equal to or larger
than € 50 million or the balance sheet total is larger or equal to € 43 million.
MNE: Same as for large enterprise but multinational (Global Reporting Initiative, n.d.).
Hence, the sample criteria and delimitation for the firms considered in this study is the following: The Large Enterprise or MNE is:
✓ Listed
• Listed firms due to financials being available to the public.
✓ Has a GRI G4 or GRI Standards report published for the years 2017-2019 which are reported ‘in accordance’ (core or comprehensive) or ‘not in accordance’ (undeclared, referenced or none) with GRI Standards.
• These reports are the ones considered due to both being categorized with similar adherence levels. Also, these reports are the most recent GRI Standards that are used.
✓ Is European.
• When said that a firm is European in this study it is referred to the location of the firm’s headquarters. In the case of subsidiaries European means that the company reports to this region (Global Reporting Initiative, n.d.). ✓ Is in the industry of manufacturing in the sectors automotive, chemicals,
equipment, food and beverage products, or metals products.
• These specific sectors are considered due to them being the largest ones in the European manufacturing industry in value added and employment relative to the other sectors (European Union, 2020).
The afore-mentioned sample selection boiled down to 187 sample companies but after data collection of the variables the sample size shrunk to 98 companies due to some data that could not be found.
3.2.2 Independent Variables
The independent variables of this study are the GRI adherence levels for the publication years 2017, 2018 and 2019. More specifically, the adherence levels stand for the sample companies sustainability reports for the years 2016 to 2018 but these reports are always published the following year whereas the adherence levels become known the following year. This information is retrieved from the GRI database where every sustainability report receives a category depending on its level following the GRI Standards. To be able to measure the adherence levels statistically they are divided into 0 ‘not in accordance’ and 1 ‘in accordance’ with GRI Standards. Not in accordance with GRI Standards refers to reports with the adherence levels: none, undeclared or referenced. Contrarily, in accordance with GRI Standards has the adherence levels of core or comprehensive.
0- Adherence Level: Not in accordance: None, Undeclared, or Referenced. 1- Adherence Level: In accordance: Core or Comprehensive.
3.2.3 Dependent Variables
The main economic performance or CFP measures are either market or accounting based as could be seen in the previous research that consider the relationship between CSP and CFP. After finalizing the literature review it was decided on Tobin’s Q as a market-based measure of FV (Singh, Tabassum, Darwish, & Batsakis, 2018). The Tobin’s Q is this study’s dependent variable of the regression.
3.2.3.1 Tobin’s Q
Tobin’s Q is a market-based long-term performance measure which is commonly used in previous studies to measure FV (Rhou et al., 2016; Fatemi et al., 2018; Li et al., 2018). This performance measure is the ratio between assets’ market value and their replacement value (Tobin, 1969). Thus, in this study the Tobin’s Q is calculated by dividing market capitalization by the book value of equity, which is calculated by multiplying the book value per share by shares outstanding, of the sample firms. This means that if the Tobin’s Q value is higher than 1, the companies are suggested to be efficient in their current asset’s usage. These financials were retrieved from both sample company financial statements and Statista. This market-based measure will be the only dependent variable this study investigated due to the interest in FV.
𝑇𝑜𝑏𝑖𝑛′𝑠 𝑄 = 𝑀𝑎𝑟𝑘𝑒𝑡 𝐶𝑎𝑝𝑖𝑡𝑎𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛
𝐵𝑜𝑜𝑘 𝑉𝑎𝑙𝑢𝑒 𝑝𝑒𝑟 𝑆ℎ𝑎𝑟𝑒 × 𝑆ℎ𝑎𝑟𝑒𝑠 𝑂𝑢𝑡𝑠𝑡𝑎𝑛𝑑𝑖𝑛𝑔 3.2.4 Control Variables
According to Saunders et al. (2016, p.668) control variables are “unwanted but
measurable variables that need to be kept constant to avoid them influencing the effect of the independent variable on the dependent variable”. These variables are used to be able to verify the causality or correlation between the independent and dependent variables (Collis & Hussey, 2009).
3.2.4.1 Size of the sample company
The size of the sample companies is measured by looking at the natural logarithm of the headcount of the sample firms. The number of employees is retrieved from the sample companies’ annual reports or from Statista if the information is available in the database. The size is chosen as a control variable in this study since it is an extensively used variable in previous similar studies (van Beurden & Gössling, 2008; Lopez et al., 2007; Rhou et al., 2016; Hassan, 2018; Fatemi et al., 2018).
3.2.4.2 Growth of the sample company
The growth of the sample companies is widely used as a control variable in previous studies (van Beurden & Gössling, 2008; Plumlee et al., 2015; Fatemi et al., 2018). In this study the growth is measured by the usage of the year-over-year growth rate which stands for the percentage change in revenue for one year for a sample company (Statista, 2021). This measure was retrieved from the Statista database for all the sample companies. However, to check the reliability of this metric in the database the annual reports were considered also.
3.2.5 Analysis Method 3.2.5.1 Descriptive Statistics
Since this research is set out to investigate a relationship between variables it is crucial to look at the descriptive statistics (Ho & Yu, 2015). Similarly to previous studies (Rhou et al., 2016; Chen et al., 2015; Rezaee & Tuo, 2019) the mean and standard deviation has been evaluated for the data for all the years considered. The mean stands for the average
and standard deviation represents the spread around the mean of the individual observations (Anderson, Sweeney, Williams, Freeman & Shoesmith, 2017).
3.2.5.2 Correlation Analysis – Pearson Correlation
In a correlation analysis the goal is to measure the linear association degree between two variables. A Pearson correlation is used in this study’s research analysis and the value of the correlation coefficient should fall between -1 and +1 and the closer the value is to 1 the stronger the correlation is said to be (Gujarati & Porter, 2009).
3.2.5.3 Panel Data Regression Model
A panel data regression method refers to an empirical analysis method that considers two aspects simultaneously, time and space, namely. This method has become increasingly common within research (Gujarati & Porter, 2009) and thus is used also in this study where one dimension is time (2017, 2018 and 2019) and the other is the European manufacturing companies (98 sample firms). Multiplying the three years with the sample size results in 294, which is the number of observations of this study. According to Gujarati and Porter (2009) this technique gives more variability and more informative data whilst giving less collinearity and more degrees of freedom when compared to time series or cross-sectional methods. Pooled OLS, Fixed Effects Model (FEM) and Random Effects Model (REM) are the three main panel data regression models. Hausman test and the Breusch and Pagan Lagrange Multiplier (BP-LM) test are implemented to decide on the most appropriate model for the data (Gujarati & Porter, 2009).
3.2.5.3.1 Hausman Test
Hausman Test is used to decide between the REM and FEM model. Its null hypothesis is “A REM is suitable” and its alternative hypothesis is “A REM is not suitable. If the null would be rejected the FEM would be a suitable model. If the null cannot be rejected both REM and FEM are suitable (Gujarati & Porter, 2009).
3.2.5.3.2 Breusch and Pagan Lagrange Multiplier Test
The BP LM Test’s null hypothesis states that the variance of the random effect is zero, whereas if accepted a Pooled OLS model would be suitable. The alternative hypothesis states the variance of the random effect is larger than zero, whereas if accepted, REM could be used (Gujarati & Porter, 2009).
3.2.5.4 Fixed Effects Model
The panel data regression model that is used in this study is the FEM. A FEM is a panel data model where the parameters, or the values of independent variables, are fixed (Gujarati & Porter, 2009).
The functional form of the fixed effect model used is the following: Y it = β0 + β1iX1it+ β2iX2it+ β3iX3it+ uit
Where:
Y = Dependent variable
X = Explanatory variable (An independent variable or a control variable) β = The coefficient
0 = the intercept
1, 2 and 3 = Number of explanatory variables u = The error term
i = The company t = The year
The functional form simplified:
Tobin’s Q it= β0i + β1iGRI adherence level it+ β2iFirm Size it + β3iFirm Growth it + u it
3.2.5.5 Data Assumption Tests
To be able to ensure the reliability of the regression results some tests must be carried out to see that the data is normally distributed and has no autocorrelation. The statistical tests that will be performed are the Jarque-Bera Test where normal distribution of residuals is tested and Durbin Watson Test that is used to track possible autocorrelation or serial correlation.
3.2.5.6 R² and Adjusted R²
R² or coefficient of determination explains how well the sample regression line fits the data and is commonly used as a metric for the goodness of fit of a regression line. The higher (between 0 and 1) the value of R² is the better the chosen model will explain the dependent variable variations. The problem of the coefficient of determination is that its value becomes higher the more variables are added. Hence, the adjusted R² is in this case more suitable to use since, as the name says, it is adjusted for the degrees of freedom that are lost when more variables are included into the model (Gujarati & Porter, 2009). The adjusted R² is used in this study as a determinant to verify that the regression is suitable for declaring the variations of the chosen dependent variables.
3.2.5.7 Type I and Type II Errors
A type I error occurs if a true null hypothesis is rejected, controversially, a type II error occurs if a false null hypothesis is not being rejected. A significance level in statistics stands for the probability of committing a type I error. In an ideal situation the errors (both type I and II) should be minimized, however, it is needed to assume that a type I error can happen thus a significance level (α) must be chosen which is usually as low as 0.05 or 0.1. Simultaneously when a fixed significance level has been chosen, the goal is to minimize the type II error by maximizing the power of the test by minimizing the β (Gujarati & Porter, 2009). In this study a significance level of 0.05 (5%) is used to test the hypothesis.
Criticism and Quality Criteria
The conceptual framework of the study could consist of other theories or other perspectives which would possibly lead to other conclusions than this research will. This is however highly speculative, and the conceptual framework is similar to previous studies, thus arguably relevant.
Positivism, which is the paradigm used in this study, has been criticised for its objective nature. The methodology may cause constraints on the results and ignore findings of relevant nature. More specifically, when explaining a complex phenomenon by only one metric some misleading answers may manifest themselves. Additionally, a researcher is never objective thus objectivism can be somewhat difficult to maintain (Collis & Hussey, 2009). These shortcomings and the study’s quality criteria’s; reliability, validity and generalizability were considered carefully when the data was retrieved, and when the results were analysed and screened.
3.3.1 Reliability
When discussing the consistency of the results of this study, if it was repeated or repeated by someone else, it is referred to reliability of a research. Reliability is one of the aspects of credibility of a scientific research and is important to consider in a positivist study since a study can be counted as credible if its results can be replicated (Collis & Hussey, 2009). Reliability is ensured in this study by examining multiple years (2017, 2018 and 2019) instead of only one to ensure the replicability of the results. In addition, to reach a good level of reliability the calculations used are double checked multiple times and the
statistical relations are realized by the usage of EViews, which is a trustworthy Windows based statistical tool (EViews, 2020).
3.3.2 Validity
The validity of a research describes how well the findings of the research represents the phenomena of interest in the study and it is also the other aspect of credibility (Collis & Hussey, 2009). In a positivist study validity can usually be divided into two categories: construct validity and internal validity. Construct validity considers to what extent the research can measure the actual phenomena of interest. This study fulfils the criteria of construct validity by considering previous studies carefully and using similar methodology and method to reach the results. Internal validity is established in this study also by carefully looking at the relationship between CSP and CFP in previous literature and by concluding that a causal relationship can be found between these variables but not necessarily between GRI adherence level and FV which takes the CSP and CFP relationship on a much smaller detailed scale (Saunders et al., 2016).
3.3.3 Generalizability
According to Collis and Hussey (2009) the generalizability in a research is met if the sample analysed in a study can be used over a whole population. This study considers 98 European large enterprises and MNE’s in the manufacturing industry (specific sub-sectors) who use the GRI Standards to report their CSP. With the same categorization it is possible to find 342 large enterprises from the GRI database which means that this sample size stands for 28.7% of the population. Thus, arguably this study can be generalized over the whole population.
Empirical Results
The results of this study are presented in this chapter. In the beginning The GRI adherence level averages between sample firms will be discussed. This is followed by the descriptive statistics, correlation, and regressions for all the sample firms and all the years. EViews is used as the statistical tool to retrieve the results in this research.
GRI Adherence Level
Graph 1. GRI Adherence Level Average for sample firms (2017 – 2019).
Graph 1 shows the average adherence level that every sample firm has for the period 2017 to 2019. The average has been calculated from the adherence levels in accordance with GRI Standards = 1 (Adherence level core or comprehensive) and not in accordance with the GRI Standards = 0 (Adherence level referenced, undeclared or none). As can be seen in the graph 1 there are only 27 firms that choose to report in accordance with the GRI Standards for all the 3 years (100% adherence average).
0% 20% 40% 60% 80% 100% 120%
GRI Adherence Level Average (2017-2019)
Acerinox AGRANA Group Akzo Nobel NV ALRO
AMAG Austria Metall AmRest Apator APERAM
Arkema Atlas Copco AUDI AG Aurubis
Autogrill S.p.A. Autoneum Barry Callebaut AG BASF SE Bekaert SA Bell Food Group Berentzen-Gruppe BMW Group BONDUELLE SA Bucher Industries AG Bulten AB (publ) Cargotec
Carlsberg Group Ciech Clariant Cloetta
CNH Industrial Coca-Cola HBC AG Continental Corbion
Covestro AG Croda Daimler Diageo
Dialog Semiconductor DSM Ebro ElvalHalcor S.A.
Emmi Group Evonik Industries AG Fagerhult Group FAMUR S.A.
FCA Ferrari Fiskars Fleury Michon
GEA Gestamp Givaudan International SA Gruppo Campari
Gurit HEXPOL HKScan Corporation Johnson Matthey
Kapsch TrafficCom AG Kemira Krones Landis+Gyr
LANXESS AG Legrand Lerøy Seafood Group Metro Group
Michelin Mikron Nestlé Norsk Hydro
OCI Orior OutoKumpu Piaggio Group
Raisio Rémy Cointreau Renault SAS Rheinmetall
SABAF SAF-HOLLAND SE Sandvik Sartorius
SBM Offshore Amsterdam B.V. Schaeffler Gruppe Schindler Semperit Group
SKF Group SSAB Symrise AG Synthomer plc
Systemair Tikkurila Umicore Valeo
voestalpine Volkswagen Aktiengesellschaft Volvo Group Wacker Chemie AG Wacker Neuson Group Zwack Unicum Nyrt
Descriptive Statistics (2017 – 2019)
4.2.1 2017
Table 2: Descriptive Statistics 2017
Tobin's Q Adherence
Level Growth Size
Mean 3.146 0.571 7.04% 9.383
Standard Deviation 2.643 0.497 12.76% 1.565
Observations 98 98 98 98
The GRI adherence is a binary variable, which means that the variable can take on only a value of 1 or 0, and thus the mean and standard deviation for the variable is reasonable in Table 2. Furthermore, the mean of the variable tells us that the average reporting rate, when 100% is “in accordance” and 0% means “not in accordance”, is 57.1%. Simply put, in the publication year 2017 the collective adherence level of European manufacturing companies is 57.1%. The mean of Tobin’s Q is quite remarkable (3.146) in 2017 since already a Tobin’s Q value over 1 describes that the companies are efficient in their current asset’s usage.
4.2.2 2018
Table 3: Descriptive Statistics 2018
Tobin's Q Adherence
Level Growth Size
Mean 2.398 0.582 6.36% 9.413
Standard Deviation 1.971 0.496 13.67% 1.555
Observations 98 98 98 98
In 2018 the mean and standard deviation of Tobin’s Q is lower than in 2017 as can be seen in Table 3. However, the mean of the GRI adherence level is higher (58.2%) this year compared to the previous year.
4.2.3 2019
Table 4: Descriptive Statistics 2019
Tobin's Q Adherence
Level Growth Size
Mean 2.819 0.490 2.44% 9.437
Standard Deviation 2.855 0.502 11.67% 1.541
Observations 98 98 98 98
The lowest GRI adherence level mean can be found in 2019 which corresponds to 49% (Table 4). Contrarily the lowest Tobin’s Q mean could be found in 2018 and not in 2019. However, the standard deviation of Tobin’s Q is quite remarkable this year (2.855).
Pearson Correlation (2017 - 2019)
4.3.1 2017
Table 5: Pearson Correlation 2017
Tobin's Q GRI Adherence Growth Size
Tobin's Q 1 -0.164 0.113 -0.06
GRI Adherence -0.164 1 -0.145 0.324
Growth 0.113 -0.145 1 -0.025
Size -0.06 0.324 -0.025 1
As can be seen in the above Pearson correlation table there is no significant correlation between the GRI adherence level and Tobin’s Q in 2017.
4.3.2 2018
Table 6: Pearson Correlation 2018
Tobin's Q GRI Adherence Growth Size
Tobin's Q 1 -0.021 0.018 -0.127
GRI Adherence -0.021 1 0.017 0.170
Growth 0.018 0.017 1 -0.092
Size -0.127 0.170 -0.092 1
As could be seen for 2017 the same results hold for 2018, there is no significant correlation between GRI adherence level and Tobin’s Q.
4.3.3 2019
Table 7: Pearson Correlation 2019
Variable Tobin's Q GRI Adherence Growth Size
Tobin's Q 1 0.11 0.121 -0.054
GRI Adherence 0.110 1 -0.276 0.282
Growth 0.121 -0.276 1 -0.148
Size -0.054 0.282 -0.148 1
The correlation of GRI adherence and Tobin’s Q in 2019 holds the same result as for the two previous years, no significant correlation is found.
Panel Data Regression Results
4.4.1 Hausman TestTable 8: Hausman Test
Chi-square Stat.
Chi-Sq. d.f. Prob
