The Impact of Employee Board Representation on the

Firm’s CSR Engagement : evidence from the French

Context
Amal Boukadhaba

To cite this version:

Amal Boukadhaba. The Impact of Employee Board Representation on the Firm’s CSR Engagement :
evidence from the French Context. Business administration. Le Mans Université, 2020. English.
�NNT : 2020LEMA2001�. �tel-03084772�

HAL Id: tel-03084772

[Link]
Submitted on 21 Dec 2020

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 THESE DE DOCTORAT DE

LE MANS UNIVERSITE
COMUE UNIVERSITE BRETAGNE LOIRE

ECOLE DOCTORALE N° 597

Sciences Economiques et sciences De Gestion
Spécialité : Sciences de gestion

Par

Amal BOUKADHABA
The Impact of Employee Board Representation on the Firm’s CSR Engagement:
Evidence from the French Context
Impact de la représentation des salariés au sein du conseil d’administration sur l’engagement RSE de
l’entreprise : Etude du contexte français

Thèse présentée et soutenue à Le Mans, le 09 Juin 2020

Unité de recherche : GAINS-ARGUMANS Recherche Gestion (N° EA 2167 CNRS)
Thèse N° : 2020LEMA2001

Rapporteurs avant soutenance :

Nicolas Aubert Professeur des universités, Aix-Marseille Université

Xavier Hollandts Professeur associé, Kedge Business School

Composition du Jury :

Nicolas Aubert Professeur des universités, Aix-Marseille Université

Xavier Hollandts Professeur associé, Kedge Business School
Gilles Paché Professeur des universités, Aix-Marseille Université
Salma Mefteh Wali Professeur, ESSCA School of Management

Directeur de thèse : Mehdi Nekhili Professeur des universités, Le Mans Université

 À mes parents,

À mon mari,

À tous ceux que j’aime,

Et à tous ceux qui m’aiment…

i
ii
Remerciements

Au terme de ce travail, je tiens à exprimer ma profonde gratitude à mon cher directeur de

thèse Monsieur le Professeur Mehdi Nekhili pour son encadrement, son expertise, sa

disponibilité, ses précieux conseils et son énorme soutien tout au long de la période de thèse.

Ce travail lui doit beaucoup.

Je remercie Monsieur le Professeur Nicolas Aubert et Monsieur le Professeur Xavier

Hollandts d’avoir accepté d’être rapporteurs de ce travail de recherche. Je tiens à remercier

également Monsieur le Professeur Gilles Paché et Madame la Professeure Salma Mefteh Wali

d’avoir accepté d’être membre de jury de cette thèse

J’adresse aussi mes vifs remerciements aux membres de mon comité de suivi de thèse

Monsieur le Professeur Wael Louhichi de l’ESSCA et Madame la Professeure Christine

Gonzalez de Le Mans Université pour les discussions enrichissantes ainsi que leurs conseils.

Je remercie également Monsieur Haithem Nagati, enseignant-chercheur à l’emlyon, pour ses

conseils et commentaires précieux.

Par l’occasion, je remercie les membres du laboratoire de recherche GAINS, professeurs et

doctorants pour leurs commentaires et conseils lors des séminaires.

Je ne manquer pas à remercier mon cher mari Moez pour sa grande patience, son soutien

morale et financier, sa relecture et ses précieux conseils.

Je remercie mes parents Lotfi et Moufida, mes sœurs Amani et Eya et mon frère Mohamed

Aziz, ma belle famille et mes ami(e)s Safa, Amira, Islem, Soumaya et Fahim ainsi que toute

personne qui a contribué de près ou de loin à l’achèvement de ce travail.

1
2
 The Impact of Employee Board Representation on the Firm’s
CSR Engagement: Evidence from the French Context

Abstract

This thesis proposes to examine the relationship between employee board representation and

the CSR engagement of the firm. Particularly, we study how investors perceive CSR reporting

and ESG performance when employees are represented on the board.

Using a sample of French firms belonging to the SBF 120 index over two different periods

from 2001 to 2011 and from 2007 to 2017, our results show that the presence of employee

directors on board moderates positively the perception of investors towards CSR reporting

and negatively their perception towards ESG performance. More fine-grained examination

shows that directors elected by employees by their right of employment enhance exclusively

the social performance and moderate negatively the way investors perceive information on

ESG performance. In contrast, directors elected by employee shareholders enhance the overall

ESG performance as well as the environmental and the governance performances and

moderate positively the way investors perceive the ESG performance of the firm. Our

findings highlight a conflict of interest between shareholders and employee directors,

particularly with labour representatives.

This thesis covers several disciplines namely accounting, corporate governance and human

resource management. As a result, it is with great interest for regulators, investors, managers

and shareholders.

Key words: Employee board representation, CSR reporting, ESG performance, Firm

market value.

3
4
 Impact de la représentation des salariés au sein du conseil
d’administration sur l’engagement RSE de l’entreprise : Etude du
contexte français

Résumé

Cette thèse propose d'examiner la relation entre la représentation des salariés dans le conseil

d’administration et l’engagement RSE de l’entreprise. En particulier, nous étudions comment

les investisseurs perçoivent-ils le reporting RSE et la performance ESG lorsque les salariés

sont représentés au sein du conseil d’administration. En utilisant un échantillon de sociétés

françaises appartenant à l'indice SBF 120 sur deux périodes différentes de 2001 à 2011 et de

2007 à 2017, nos résultats montrent que la présence des administrateurs salariés modère

positivement la perception du reporting RSE et négativement la perception de la performance

ESG par les investisseurs. Un examen plus détaillé montre que les administrateurs élus par les

salariés par leur droit de travail améliorent exclusivement la performance sociale et modèrent

négativement la perception des investisseurs vis-à-vis de la performance ESG. Au contraire,

les administrateurs élus par les salariés actionnaires renforcent la performance ESG globale

ainsi que les performances environnementale et de gouvernance et modèrent positivement la

perception de la performance ESG par les investisseurs. Nos résultats mettent en évidence un

conflit d'intérêts entre les actionnaires et les administrateurs salariés, notamment avec les

représentants des travailleurs.

Cette thèse couvre plusieurs disciplines à savoir la comptabilité, la gouvernance d'entreprise

et la gestion des ressources humaines. En conséquence, elle intéresse vivement les

législateurs, les investisseurs, les managers et les actionnaires.

Mots clés : Administrateurs salariés, reporting RSE, performance ESG, performance

boursière.

5
6
 Plan de thèse

Plan de thèse .......................................................................................................................................... 7

Liste des Tableaux ............................................................................................................................... 11
Liste des Annexes ................................................................................................................................ 13
Liste des Acronymes............................................................................................................................ 15
General Introduction ....................................................................................................................... 17
1. What are the Differences between Directors Elected by Employees by Right of Employment
and Directors Elected by Employee-Shareholders? ......................................................................... 20
2. Theories Explaining the Employee Board Representation-CSR Relationship ...................... 22
3. Motivations .................................................................................................................................. 25
3.1. CSR criteria, an important feature for investors’ decision making ...................................... 26
3.2. CSR: a board control task ..................................................................................................... 27
3.3. Employee board representation and CSR ............................................................................. 28
3.4. Employee board representation and the market perception towards the firm’s CSR
engagement........................................................................................................................................ 28
4. Methodology................................................................................................................................. 30
5. Overview of the Thesis’ Structure ............................................................................................. 32
References ............................................................................................................................................ 35
Chapter I: Does Employee Board Representation Matter for the Relevance of CSR
Reporting? .......................................................................................................................................... 43
1. Introduction ................................................................................................................................. 43
2. Conceptual Framework and Hypotheses Development ........................................................... 46
2.1. CSR Reporting : Motives and related Value Relevance ........................................................ 46
2.2. Employee board representation and the value relevance of CSR reporting ......................... 49
3. Methodology................................................................................................................................. 52
3.1. Sample and data .................................................................................................................... 52
3.2. Dependent variable: Tobin’s q ................................................................................................... 53
3.3. Endogenous variable: CSR-related reporting ....................................................................... 53
3.4. Moderating variable: employee directorship ........................................................................ 54
3.5. Control variables................................................................................................................... 54
3.6. Estimation method ................................................................................................................. 55
4. Results .......................................................................................................................................... 57

7
 4.1. Descriptive Statistics .................................................................................................................. 57
4.2. Univariate Analysis .................................................................................................................... 58
4.3. Multivariate Analysis ............................................................................................................ 60
4.4. Supplementary Analysis ........................................................................................................ 65
5. Conclusion .................................................................................................................................... 69
References ............................................................................................................................................ 71
Chapter II: ESG Performance and Market Value: The Moderating Role of Employee
Board Representation ...................................................................................................................... 77
1. Introduction ................................................................................................................................. 77
2. Conceptual Framework and Hypotheses Development ........................................................... 80
2.1. Employee directorship and firm value .................................................................................. 80
2.2. The moderating role of employee directorship ..................................................................... 83
3. Methodology................................................................................................................................. 86
3.1. Sample and data .................................................................................................................... 86
3.2. Dependent variable: Tobin’s q .............................................................................................. 87
3.3. Endogenous variable: ESG Performance ............................................................................. 87
3.4. Moderating variable: employee directorship ........................................................................ 88
3.5. Control variables................................................................................................................... 88
3.6. Estimation method ................................................................................................................. 89
4. Results .......................................................................................................................................... 91
4.1. Descriptive statistics.............................................................................................................. 91
4.2. Multivariate analysis ............................................................................................................. 92
5. Conclusion .................................................................................................................................. 103
References .......................................................................................................................................... 106
Chapter III: The ESG-Financial Performance Relationship: Does the Type of Employee
Board Representation Matter?.................................................................................................... 111
1. Introduction ............................................................................................................................... 111
2. Employee Board Representation and ESG Performance ...................................................... 114
2.1. Employee Board Representation and Social Performance ................................................. 115
2.2. Employee Board Representation and Environmental Performance.................................... 117
2.3. Employee Board Representation and Corporate Governance Performance ...................... 119
3. The Moderating Role of Employee Directors in the ESG-Financial Performance
Relationship ....................................................................................................................................... 121
4. Methodology............................................................................................................................... 122
4.1. Sample and Data ................................................................................................................. 122

8
 4.2. Dependent variable ............................................................................................................. 123
4.3. Endogenous variable ........................................................................................................... 124
4.4. Control variables................................................................................................................. 125
4.5. Estimation Model ................................................................................................................ 126
5. Results and Discussion .............................................................................................................. 128
5.1. Descriptive Statistics ........................................................................................................... 128
5.2. Pairwise Correlation ........................................................................................................... 130
5.3. Propensity Score Matching ................................................................................................. 133
5.4. Test of H1a to H3b .............................................................................................................. 137
5.5. Test of H4a and H4b ........................................................................................................... 143
5.6. Supplementary Analysis ...................................................................................................... 147
6. Conclusion .................................................................................................................................. 155
References .......................................................................................................................................... 158
General Conclusion ........................................................................................................................ 163
References .......................................................................................................................................... 169
Résumé de la thèse .......................................................................................................................... 171
Références .......................................................................................................................................... 206

9
10
 Liste des Tableaux

Table I. 1: Variables and Their Measurement .......................................................................... 56

Table I. 2: Descriptive Statistics .............................................................................................. 58
Table I. 3: Mean Difference Test ............................................................................................. 59
Table I. 4:Pairwise Correlation ................................................................................................ 61
Table I. 5: System GMM Regression of Tobin's q on CSR Reporting and Employee Board
Representation .......................................................................................................................... 63
Table I. 6: System GMM Regression of Tobin's q on Higher CSR Reporting and Employee
Board Representation ............................................................................................................... 65
Table I. 7: System GMM Regression of Tobin's q on Higher Social, Environmental and
Sustainability Reporting and Employee Board Representation ............................................... 67
Table I. 8: System GMM Regression of Tobin's q on the Interaction between Higher Social,
Environmental and Sustainability Reporting and Employee Board Representation................ 69

Table II. 1: Variables and Their Measurement ........................................................................ 90

Table II. 2: Descriptive Statistics ............................................................................................. 92
Table II. 3 : System GMM regression of Tobin’s q on ESG performance and employee board
representation. .......................................................................................................................... 95
Table II. 4: System GMM Regression of Tobin's q on ESG Performance in Firm-Years with
and without Employee Board Representation .......................................................................... 96
Table II. 5: System GMM Regression of Tobin's q on Social, Environmental and Corporate
Governance Performance and Employee Board Representation ............................................. 98
Table II. 6: System GMM Regression of Tobin's q on the Interaction between Social,
Environmental and Corporate Governance Performance and Employee Board Representation
................................................................................................................................................ 100
Table II. 7: System GMM Regression of Tobin's q on Social Performance in Firm-Years with
and without Employee Board Representation ........................................................................ 101
Table II. 8: System GMM Regression of Tobin's q on Environmental Performance in Firm-
Years with and without Employee Board Representation ..................................................... 102
Table II. 9: System GMM Regression of Tobin's q on Corporate Governance Performance in
Firm-Years with and without Employee Board Representation ............................................ 103

Table III. 1: Variables and Their Measurement ..................................................................... 127

Table III. 2: Descriptive Statistics .......................................................................................... 129
Table III. 3: Pairwise Correlation ........................................................................................... 131
Table III. 4: Mean Difference Test between Firm-Years with and without Employee Board
Representation for Entire and Matched Samples ................................................................... 134
Table III. 5: Mean Difference Test between Firm-Years with and without Labour Board
Representation for Entire and Matched Samples ................................................................... 135
Table III. 6: Mean Difference Test between Firm-Years with and without Employee-
Shareholders Representation for Entire and Matched Samples ............................................. 136
Table III. 7: System GMM Regression of ESG Performance on Employee Board
Representation ........................................................................................................................ 139
Table III. 8: System GMM Regression of ESG Performance on Labour Board Representation
................................................................................................................................................ 141
Table III. 9: System GMM Regression of ESG Performance on Employee-Shareholders
Board Representation ............................................................................................................. 142

11
Table III. 10: System GMM Regression of Tobin's q on ESG Performance and Employee
Board Representation ............................................................................................................. 144
Table III. 11: System GMM Regression of Tobin's q on the Interaction between ESG
Performance and Employee Board Representation ................................................................ 146
Table III. 12: System GMM Regression of Tobin's q on Social Performance and Employee
Board Representation ............................................................................................................. 148
Table III. 13: System GMM Regression of Tobin's q on Environmental Performance and
Employee Board Representation ............................................................................................ 149
Table III. 14: System GMM Regression of Tobin's q on Corporate Governance Performance
and Employee Board Representation ..................................................................................... 150
Table III. 15: System GMM Regression of Tobin's Q on the Interaction between Social
Performance and Employee Board Representation ................................................................ 152
Table III. 16: System GMM Regression of Tobin's q on the Interaction between
Environmental Performance and Employee Board Representation ....................................... 153
Table III. 17: System GMM Regression of Tobin's q on the Interaction between Corporate
Governance Performance and Employee Board Representation ........................................... 154

12
 Liste des Annexes

Appendix 1: The Evolution of Employee Board Representation Laws over Time ................. 39
Appendix 2: Comparison between the Different Types of Employee Directors ..................... 41

Appendix I. 1: Items of Grenelle II Act ................................................................................... 75

13
14
 Liste des Acronymes

AGEFI : Agence Economique et Financière

CEO : Chief Executive Officer

CSR : Corporate Social Responsibility

ESG : Environment, Social and Governance

ESO : Employee Share Ownership

F.P : Financial Performance

GMM : General Method of Moments

GRI : Global Reporting Initiative

KLD : Kinder, Lydenberg and Domini

KPI : Key Performance Indicator

NER : Nouvelles Régulations Economiques

PACTE : Plan d’Action pour la Croissance et la Transformation des Entreprises

R&D : Research and Development

RSE : Responsabilité Sociale de l’Entreprise

SBF : Société des Bourses Françaises

VIF : Variance Inflation Factor

15
16
 General Introduction

The starting point of this thesis is that efficient corporate governance, “the system by which

companies are directed and controlled and as a set of relationships between a company’s

management, its board, its shareholders and its other stakeholders” (European Commission,

2011a: 2), is crucial to maintain the confidence of stakeholders in the firm’s outcomes and

that inefficient governance raises the likelihood of financial crises and stakeholders’ doubt.

Indeed, the wave of scandals in Us (WorldCom, Tyco, Arthur Anderson, Enron), Europe

(FlowTex, Parmalat and more recently Thomas Cook) and worldwide (HIH Insurance,

Carrian Group, etc) have questioned both the corporate social responsibility (hereafter CSR)

of organizations and the efficiency of the board of directors as a mechanism of the corporate

governance. These scams highlighted that, on the one hand, shareholders’ supremacy

orientation of firms is no longer appropriate in this new context (Gordon & Roe, 2004) and

that a stakeholder orientation would, in such case, be a better alternative from the corporate

social responsibility perspective. In other words, the outcomes of financial crisis such as

employees’ lay off, unpaid suppliers, investors’ doubt, etc., have shown that while

maximizing the interests of one stakeholder, namely shareholders, other stakeholders’

interests, crucial to the firm’s continuity, have been neglected. On the other hand, they

highlighted the need to enhance the efficiency of the board of directors as the responsible for

the firm’s strategies and policies and also for managerial monitoring. In this respect, the

composition of the board and more particularly, the diversity of its members may be

considered as a solution to enhance its efficiency. Indeed, a diverse board would have

different competencies, knowledge, information and views, offering more capacities to

perform its monitoring duties (Bagdi, 2015; Huse, Nielson, & Hagen, 2009). As for gender

17
diversity and the independence of the board members, the representation of employees on the

board has attracted increasing attention from regulators and researchers.

The co-determination, a translation for the German word “Mitbestimmung”, is the

practice by which workers are represented on the board of directors with voting rights. This

representation provides employees with the opportunity to participate in the decision-making

process, to express their needs and to defend their interests. Initially, this practice started after

the World War II in Germany, particularly in the steel, coal and iron industries, and saw a

widespread interest 60 years later in other European countries (Conchon, 2011; Preuss,

Haunschild, & Matten, 2009). The renewed interest for employees’ participation in the

decision-making process can be justified by their long-term perspective towards the firm

(Kleincknecht, 2015). Referring to the stakeholders’ perspective, the representation of

employees on the board of directors, mainly a European phenomenon1, aims to balance forces

(human and financial capitals) on the board. Different laws and directives were launched to

support the presence of employees on the corporate boards at the country and European level.

At the European level, the launch of the Union Statute of European Companies (EC

2157/2001) and the associated Council Directive (2001/86/ EC) have largely promoted for

workers’ participation at the board level. At the country level, 19 from among 31 European

Economic Area (EEA) countries have legislations with regard to employees’ participation at

the board level. France show an intermediate level of employees’ representation on the board

compared to Germany, which mandates a third (half) of the supervisory board for firms with

more than 500 (2000) employees, respectively and United Kingdom, which required no

employee board level representation. The first French law stipulating labour representation (or

the institutional representation of employees) on the board focused on state-controlled firms

(law of 1983). Eleven years later, regulators start to pay attention to privatized companies by

1
For more details see Conchon, Kluge and Stollt (2015), Worker representation in the 31 European Economic
Area Countries.

18
giving them the possibility to reserve seats for employees and employee shareholders subject

to the acceptance of shareholders at the general assembly (law of 1994). Started in 2013,

French public listed firms are required to reserve seats for directors elected by employees by

right of employment (New laws were launched in 2015 and 2019). Besides the representation

of labour on the board of directors, French firms show another specification by representing

employee-shareholders on their boards. The representation of employee-shareholders became

mandatory for publicly listed French firms in 20062 conditional to holding at least 3% of total

shares. More details on the different French laws regarding employee board level

representation are presented in Appendix 1.

Referring to AGEFI, France is the most advanced country in terms of employee share-

ownership with 3.2 million of employee-shareholders representing 43% of total employee-

shareholders in Europe. These numbers are expected to increase in the next years for two

main reasons. First, referring to BFM Bourse,3 81.4% of firms belonging to the SBF 120

index are expected to make a capital increase in favour of employees. Second, Bruno Le

Maire, the French ministry of economy, has launched an objective for French firms to reach a

level of 10% of employee ownership by 2030. In this context of employee share-ownership

encouragement, the “PACTE law of 2019 proposes measures in favour of employee share

ownership4.

Previous studies have mostly focused on the impact of employee directors on the

corporate governance of the firm (Bøhren & Strøm, 2010; Huse et al., 2009; Hollandts &

Aubert, 2019; Faleye, Mehrotra, & Morck, 2006; Fauver & Fuerst, 2006; etc.). In relation to

the corporate social responsibility of the firm, a number of authors examined the impact of

CSR initiatives on employees’ attitudes (Brammer, Millington, & Rayton, 2007 ; Kim, Lee,

2
The application decree of the law 2002 had not been published which makes the law inapplicable.
3
[Link]
[Link]
4
For more information, refer to the web site [Link]

19
Lee, & Kim, 2010) while others examined the perception of employees towards CSR

initiatives (El Akermi, Gond, Swaen, & Igalens, 2018; Gao & Young, 2016; Jones, Willness,

& Madey, 2014; Triana, Jayasinghe, Pieper, Delgado, & Li, 2018; etc). So far, however, there

has been little discussion about the relationship between the presence of employees on the

board of directors and CSR (Hillman, Keim, & Luce, 2001; Huse et al., 2009; Preuss et al.,

2009). The aim of this thesis is to shine new light on the debates on employee board

representation by focusing on the impact of employee directors on the firm’s CSR

engagement. More particularly, we aim to examine the extent to which the presence of

employee directors on the board may impact the perception of the market towards CSR

reporting and CSR performance (hereafter environmental, social and governance (ESG)

performance).

1. What are the Differences between Directors Elected by Employees by Right of

Employment and Directors Elected by Employee-Shareholders?

Employee participation in the decision-making process may be either a long-term, formal and

consultative participation (i.e. participation on the supervisory boards or boards of directors

with voting rights) or a short-term and informal participation (as in the case of participation in

working councils). However, recent legislations have largely encouraged formal participation

of employees on the board (Law 2006, law 2013, law 2015, law 2019).

France represents an interesting context for studying employee board representation.

Indeed, in French boards, we can find two types of employee directors: directors elected by

employees by their right of employment and directors elected by employee-shareholders.

Referring to Johnson, Daily and Ellestrand (1996), directors on the board have three main

roles : (1) an agency role consisting on assuring that shareholders’ interests are taken into

consideration and that CEO is acting in the same orientation, (2) a resource-dependence role

consisting on facilitating access to valuable resources for the firm and (3) a strategic role

20
consisting on bringing valuable information and knowledge to help managers in their

decision-making. Besides these responsibilities, employee directors are asked to represent

either employees or employee-shareholders’ preferences and interests on the board.

The human capital is as important as the financial capital for the firm’s survival.

However, the representation of traditional employees (without share-ownership) on the board

show mixed results. For their advocators, giving the opportunity for labour to be represented

on the board is beneficial for the firm. More particularly, labour representatives enable

valuable information to be shared with the other directors and contribute to the board’s

efficiency (Huse et al., 2009), enhance the financial performance of the firm by assuming

their control and advice responsibilities on the board (Hollandts, Guedri & Aubert, 2009),

reduce negative shocks in time of crisis (Kleinknecht, 2015) and it is important for firms with

higher needs of coordination with workers (Fauver & Fuerst, 2006). However, for their

opponents the institutional representation of employees on the board reduces the firm’s value

(Bøhren & Strøm, 2010) and the efficiency of self-managed firms (Jensen & Meckling, 1979).

In addition, labour representatives on the board are likely to increase conflict of interests with

other shareholders by reducing payout ratios (Ginglinger, Megginson, & Waxin, 2011) and by

voting for decisions that maximize their own interests (such as decisions which guaranty their

jobs and increase their fixed and residual claims) at the expense of shareholders’ value

maximization. Whereas the presence of workers on the board does not serve shareholders

value maximization, some authors recommend a prudent level (between one-third and one-

half of the board’s seats) of employee board representation to enhance board’s monitoring and

reduce agency costs (Fauver & Fuerst, 2006).

Traditional employees may become shareholders through several ways: they can either

buy the firm’s stocks individually on the market or obtain them through employee share

ownership (hereafter ESO) plans. However, it is noteworthy to mention that ESO plans are

21
the most used tool for employees to become shareholders. Indeed, firms tend to encourage

employee share-ownership plans in order to motivate their employees and reduce their debt-

dependency to creditors. French government and legislators, also, show support for the

employee share ownership plans (the PACTE law of 2019).

Employee shareholders’ representation on the board seems to be legitimated by their

double investments: human and financial. In addition, employee shareholders are more

attached to the survival of the firm (Kleinknecht, 2015). Indeed, unlike others shareholders,

employees have lower opportunity to leave the firm. Referring to Ginglinger et al. (2011),

directors elected by employee-shareholders enhance the French firms’ profitability. However,

according to Hollandts and Aubert (2011), these directors have a hierarchical-dependency

relationship with their managers. Consequently, their representation on board would

encourage managerial entrenchment (Hollandts, Aubert, Abdelhamid, & Prieur, 2018).

More details on the personal characteristics and the process of nomination on the board

of directors are presented in Appendix 2.

2. Theories Explaining the Employee Board Representation-CSR Relationship

Examining the literature, we find that four theories may explain the relationship between

employee board representation and CSR, namely the agency theory, the stakeholder theory,

the neo-institutional theory and the social identity theory.

The agency theory

Employees and shareholders have different interests. While employees seek for safe jobs and

long-term survival of the firm (Kleincknecht, 2015), shareholders usually look after

profitability. Therefore, representing employees on the board of directors with voting rights

may increase conflicts of interests with other directors representing the shareholders. A large

number of studies advance that employee directors primarily maximize their interests on the

22
board (Bøhren & Strøm, 2010) and deviate the firm from shareholders value maximization

(Faleye et al., 2006). Another argument supporting the rise of conflicts between employee

directors and shareholders is provided by Pagano and Volpin (2005). Indeed, the authors

argue that the close relationship between employees and managers, as insiders within the

firm, may result in a powerful coalition against shareholders’ decisions.

Whereas employee directors increase the board’s efficiency in terms of CSR and

strategy controls (Huse et al., 2009), some authors argue that CSR increases the conflict of

interests between inside and outside shareholders (Barnea & Rubin, 2010). In such a case,

increasing CSR engagement when employees are on the board may increase conflicts of

interests with shareholders.

The stakeholder theory

The stakeholder theory aims to tell a new story about the business by advancing that making

money for shareholders is no longer the main purpose of the firm. Instead, firms have to

create value for all stakeholders. This theory is considered as a mix between business and

ethics. More particularly, it advances that firms would create value by considering the needs

of its different stakeholders and by meeting their expectations. Confirming the stakeholder

view of business, previous findings confirm that meeting stakeholders’ needs and expectation

creates value for the firm, particularly by focusing on primary stakeholders, namely

employees, shareholders, customers (Hillman & Keim, 2001; Van der Laan, Van Ees, & Van

Witteloostuijn, 2008).

From a stakeholder perspective, the firm needs to pay more attention to its different

stakeholders rather than focusing only on shareholders. Therefore, the inclusion of directors

elected by employees (either by right of employment or by employee-shareholders) on the

board may sign for better attention to stakeholders’ needs, particularly the workers’ issues.

23
 From a CSR perspective, a stakeholder-oriented board is likely to reach higher levels of

social and environmental performances (Shaukat, Qiu, & Trojanowski, 2016). Along similar

lines, the representation of employees on the board is documented to increase the board’s

engagement in CSR issues (Huse et al., 2009).

The neo-institutional theory

Companies working in the same institutional environment may face different institutional

pressures, namely coercive, mimetic and normative pressures (Meyer & Rowan, 1977;

DiMaggio & Powell, 1983). To gain legitimacy of its societal context, crucial for their

survival, organizations have to respond to these pressures. For Preuss et al. (2009), CSR is an

important institutional pressure in the twenty-first century. CSR stipulates that firms have to

act in a responsible way towards all its stakeholders. In that sense, firms may reserve seats for

their employees on the corporate board to show their compliance with the concept of CSR and

to gain the legitimacy of their workforce and potential job seekers. A change in the corporate

structures or strategies could result from corporate cultural or political-legal processes.

Therefore, firms may appoint employees to their boardroom in response to the different

French laws (Law of 2006, Law of 2013, Law of 2015 and more recently the law of 2019) or

because the culture of the firm encourages the democracy in the decision-making process.

The social identity theory

The social identity theory (hereafter SIT), proposed by Turner (1985) and Tajfel and Turner

(1985), argue that the social identity of individuals is likely to impact their attitudes and

behaviours towards their in-group (group holding similar values) and out-group (group

holding different values). The social identity of individuals is thus generated by their feeling

of membership and their comparison with other group members. Based on the SIT, a large

body of research examining the impact of CSR initiatives on the behaviour of employees has

24
shown that a company that acts responsibly is likely to have more engaged employees

(Brammer et al., 2007, Turker, 2009). However, the irresponsible behaviour of firms

generates negative effects on employee behaviour (Triana et al., 2018). In this sense,

employee directors, concerned about the socially responsible behaviour of their companies,

are likely to strengthen the board's interest for CSR, which will, in turn, intensify the social

identity of employees.

3. Motivations

Representing the employees on the board of directors with voting rights serves to implement a

project of justice and democracy in the decision-making process (Conchon, 2011). Previous

studies have examined the impact of employee directors on the board’s efficiency (Huse et al.,

2009), on corporate governance (Bøhren & Strøm, 2010; Faleye et al., 2006; Fauver & Fuerst,

2006; Hollandts & Aubert, 2019), on company’s resilience (Kleincknecht, 2015), on firm’s

performance (Hollandts et al., 2009), on financial policies (Ginglinger et al., 2011), on value

creation (Poulain-Rehm & Lepers, 2013), on CEO entrenchment (Hollandts et al., 2018), and

CSR (Huse et al., 2009; Preuss et al., 2009). Summarizing previous findings, we claim that

the representation of labour on the corporate board is beneficial for firms which demand a

high level of coordination with employees (Fauver & Fuerst, 2006) and in time of crisis

(Bagdi, 2015; Kleinknecht, 2015). Moreover, Employee directors defund workers’ interests

(Bøhren & Strøm, 2010), increase the board’s involvement in CSR (Huse et al., 2009),

enhance the board’s monitoring and reduce private block holders’ privileges (Fauver &

Fuerst, 2006) and finally reduce the likelihood of strikes by providing credible information

from the board to workers and unions (Fauver & Fuerst, 2006). Nevertheless, employee

directors are likely to deviate the board from shareholder value-maximization (Faleye et al.,

2006) which may raise conflict of interests with other board members. Besides, employee

directors reduce the boards’ efficiency (Jensen & Meckling, 1979) and facilitate CEO’s

25
entrenchment (Hollandts et al., 2018). From a stakeholder standpoint, these findings suggest

that employee directors succeed in defunding the workers’ interests and balancing forces

between the human capital and the financial capital. However, from an agency standpoint,

employee directors may oppose to shareholders’ decisions if they are not aligned with their

interests, resulting in increased shareholders’ mistrust in the board’s outcomes. These

arguments motivated us to examine whether the presence of employees on the board may

moderate the perception of the market participants towards the firm’s CSR engagement.

3.1. CSR criteria, an important feature for investors’ decision making

In the new global economy, CSR has become a central issue for the firm and its stakeholders

and an interesting field of study for researchers. One frequently used definition of CSR is, as

proposed by the Commission of the European Communities (2001), “a concept whereby

companies integrate social and environmental concerns in their business operations and in

their interactions with their stakeholders on a voluntary base” (Dahlsrud, 2008).

French legislation has experienced a “green revolution” during the last decade.

Particularly, three laws were launched to encourage the social and environmental engagement

of firms, namely NRE law (2001), Grenelle I law (2009) and Grenelle II law (2010). Contrary

to the two first laws and inspired by the GRI guidelines, Grenelle II has identified the

different extra-financial information that firms should disclose in their annual reports.

Stakeholders and particularly the market participants care about the CSR engagement of

the firm and use the extra-financial criteria to make their perception and decision (Lydenberg,

2013; Nekhili, Boukadhaba, Nagati, & Chtioui, 2019). Stakeholders may assess the CSR

engagement of the firm through two criteria; the extra-financial reporting and the extra-

financial performance. The extra-financial reporting (CSR reporting) encloses a set of

information related to social, environmental and sustainability investments of the firm. These

information aims primarily to increase the awareness of stakeholders about the CSR activities

26
of the firm. Accordingly, firms may use CSR reporting as a channel to help investors

understanding their ESG performance (Fuente, García-Sanchez, & Lozano, 2017). The extra-

financial performance (ESG performance) is a rating provided by independent extra-financial

agencies aiming to assess the relationship between the firm and its stakeholders. The main

ESG rating and index providers are Thomson Reuters Asset4, MSCI, S&P Dow Jones and

RobecoSAM and FTSE Russell (Pagano, Sinclair, & Yang, 2018).

Investors and more particularly responsible investors pay attention to ESG performance

when making their sell-buy decisions (Lydenberg, 2013). Indeed, the ESG performance helps

investors to analyse the firm’s financial and extra-financial prospects (creation of value for

both shareholders and the different stakeholders), identify firms with values that match with

their owns and finally, identify the firm’s potential risks and rewards through assessing its

relationship with the different stakeholders.

3.2. CSR: a board control task

From a CSR perspective, the board of directors is responsible for setting the CSR strategies

and orientations of the firm, verify that CSR decisions made by the CEO and its management

team meet CSR requirements and assure that CSR expenses are likely to create value for

shareholders and stakeholders. An efficient board, from a CSR point of view, is expected to

enhance the trust of stakeholders by providing higher quality of CSR reporting and by

meeting the stakeholders’ expectations and needs.

The board composition is argued to impact the board’s CSR control task (Huse et al.,

2009), the quality of CSR reporting (Cormier, Aerts, Ledoux, & Magnan, 2009) and the ESG

performance of the firm (Boulouta, 2013; Coffey & Wang, 1998; Dunn & Sainty, 2009; Hafsi

& Turgut, 2013). In accordance, Huse et al. (2009) advance that board diversity and directors’

competencies are the most important qualities that may impact the CSR engagement of the

27
firm. In that sense, the board composition may be a good filter for the market participant to

assess the engagement of the firm towards CSR issues.

3.3. Employee board representation and CSR

The perception of employees towards CSR initiatives has been widely discussed by previous

studies. Related literature shows that a positive perception of CSR initiatives may result in

satisfied and proud workers (El Akermi et al., 2018). On the contrary, a negative perception

of the firm’s actions towards its employees is likely to reduce the psychological health of

workers and increase negative job attitudes (Triana et al., 2018). El Akermi et al. (2018) find

that employees care about all stakeholders and make higher perception over CSR activities

related to employees, shareholders, customers, natural environment, local community and

suppliers.

Considering the representation of workers’ interests as the primary role of employee

directors (Hammer, Curall & Stern, 1991), employee board representation is likely to increase

the board’s involvement in CSR issues (Huse et al., 2009). Studying the Norwegian context,

Huse et al. (2009) find that employee directors enhance the quality of discussion on the board

by providing valuable and different information. Besides, focusing on the American context,

Hillman et al. (2001), show that employee directors enhance the environmental performance

of the firm. This finding suggests that employee directors may provide the board with special

skills and knowledge related to the production process and the working conditions, helping to

reach better ESG performance.

3.4. Employee board representation and the market perception towards the firm’s CSR
engagement

The “business case” for CSR engagement is not yet well understood. Indeed, CSR initiatives

are costly for the firm and CSR related-benefits are expected to manifest in the long-term

perspective (Kanter, 2011). Moreover, the CSR outcomes may result in tangible and

28
intangible advantages (Lydenberg, 2013) and may bring more benefits for inside shareholders

(large shareholders, managers and employees) than for outside shareholders (Barnea & Rubin,

2010).

Previous findings examining the perception of the market participants towards CSR

reporting (Cahan, De Villiers, Jeter, Naiker & Van Staden, 2016; Nekhili, Nagati, Chtioui, &

Nekhili, 2017) and ESG performance (Friede, Busch, & Bassen, 2015; Nekhili et al., 2019;

Orlitzky, Schmidt, & Rynes, 2003) document mixed findings. This inconclusiveness of

findings could be explained by the unclear managerial incentives behind voluntary CSR

reporting and CSR expenses since managers have full discretion over CSR decisions. In such

a case, the scepticism of stakeholders and particularly investors towards the real CSR

engagement of the firm would be more significant.

Referring to the CSR studies, providing a better quality of CSR reporting and meeting

stakeholders’ expectation is likely to be influenced by the board’s composition (Cormier et

al., 2009; Hafsi & Turgut, 2013; Nekhili et al., 2017). Therefore, the appointment of

employees on the boardroom could be used as a filter by investors to assess the CSR

engagement of the firm. Indeed, on the one hand, employee representatives have different

knowledge, expertise, realistic view and valuable information to bring to the board.

Accordingly, the presence of employees on the boardroom may enhance the board’s CSR task

(Huse et al., 2009) and enhance CSR decisions (Preuss et al., 2009). However, on the other

hand, employees have different interests than those of shareholders and especially regarding

CSR initiatives. In that sense, employee directors may be more willing to ally with managers

in order to increase the workers-related initiatives.

Employees and shareholders have not the same interest neither the same benefits from

CSR initiatives (Barnea & Rubin, 2010). In other words, whereas shareholders may benefit

from the enhanced reputation and the firm’s performance resulting from CSR activities,

29
employees have more advantages comparing to shareholders who provide the financial

support to the firm. Employees may profit from enhanced working conditions, equal treatment

and opportunities, more trainings, higher salaries, etc., without supporting any costs.

Together, the arguments discussed above motivated us to dig deeper on the value relevance of

CSR reporting and ESG performance of the firm. In this study, we are going to examine the

moderating effect of employee directors on the CSR reporting-financial performance

relationship and the ESG-financial performance relationship.

4. Methodology

A considerable number of studies on CSR and employee board representation highlight the

need to control for the endogeneity problem (Ginglinger et al., 2011; Hollandts et al., 2018;

Nekhili et al., 2017). Endogeneity problem is a frequent and important issue that may impact

the quality of our results. Three main sources were documented in the literature: unobserved

heterogeneity, simultaneity and dynamic endogeneity.

Unobserved heterogeneity: This empirical concern may result from unobserved

variables such as institutional pressures which may impact employee board representation and

ESG (financial) performance. For instance, the corporate culture or the political context may

play an important role in increasing the representation of employees on the board and in the

CSR engagement resulting in higher CSR reporting and ESG performance.

Simultaneity: Simultaneity refers to the reverse causality which is the situation where

the independent variable can influence the dependent variable and similarly, the dependent

variable can influence the independent variable. In our study, it is likely that a reverse

causality exisits between the financial performance, the extra-financial performance and the

employee board representation. On the one hand, we may expect a reverse causality between

the financial performance and the extra-financial performance. Indeed, companies that

achieve a good extra-financial performance can generate more financial benefits. At the same

30
time, companies that are profitable may increase their social and environmental investments

and disclose information about these investments in order to gain competitive advantages

such as creating a better image, having a good reputation and attracting the interest of

responsible investors and potential jobseekers. On the other hand, we can also expect a

reverse causality between the representation of employees on the board of directors and the

financial performance of the company. In fact, companies with employee directors on the

board are found to be more profitable (Hollandts et al., 2009, Ginglinger et al., 2011). At the

same time, profitable businesses can be a source of attraction for employees who want to

invest their money, which may result in employee-shareholders’ board representation.

Dynamic endogeneity: This problem may occur due to a lagged reverse causality

between the lagged value of the dependent variable and the independent variables; in our case,

between the lagged value of the extra-financial (financial) performance and the appointment

of employees on the boardroom.

These problems if not appropriately treated could generate biased results. A commonly

used solution for endogeneity problem with small samples is to apply the GMM system

(Blundell & Bond, 1998, Roodman, 2009). Examining the literature, this estimator was used

to address the problem of endogeneity in studies examining the extra-financial disclosure

(Nekhili et al., 2017), the extra-financial performance (Boulouta, 2013, Nekhili et al., 2019)

and the representation of employees on the board of directors (Hollandts et al., 2018, Nekhili

et al., 2019). Following these researchers, we use the GMM system estimator in our

regression analyses. Moreover, to ensure the validity of our GMM system estimator, we use

two empirical tests, namely the Arellano and Bond of the second order autocorrelation and the

Sargan/Hansen test for over-identifying restrictions. We also control for the dynamic

behaviour of our dependent variable by adding its lagged variable to the empirical model.

31
5. Overview of the Thesis’ Structure

The aim of this thesis is to answer the following research question: How investors perceive

the firm’s CSR engagement when employees are appointed on the board of directors? ESG

performance could be a proxy for the CSR engagement by translating the ability of the firm to

manage good relationship with its stakeholders. However, to understand this ESG

performance, stakeholders need credible quantitative and qualitative extra-financial

information. CSR reporting is a key channel for firms to help stakeholders understand their

extra-financial performance (Fuente et al., 2017). Consequently, our first chapter focuses on

CSR reporting and particularly aims to answer the following research question: To what

extent the value relevance of CSR reporting is affected by the employee board representation?

In the first section, we present a conceptual framework regarding the motives behind CSR

reporting and its related value relevance in order to understand the perception of stakeholders

towards this extra-financial disclosure. Then, we highlight the role played by employee

directors in moderating the market perceived relevance of CSR information. In the second

section, we describe the chosen sample and the methodology used. The sample used for the

statistical processing consists of 790 observations of French companies belonging the SBF

120 index for the period from 2001 to 2011. It is noteworthy here to mention that in this first

chapter, we use a different period of study, comparing to the two other chapters, in order to

capture the voluntary criteria of CSR reporting. Indeed, after the Grenelle II law (2010) came

into force in 2012, French firms are mandated to disclose particular social, environmental and

sustainability information. In the third section, the empirical results obtained from the system

GMM regression analyses are discussed, highlighting the way in which the capital market

evaluates higher levels of CSR reporting in the presence of employee directors.

Supplementary analyses add to the latter findings by examining our research question for each

category of CSR information disclosed, namely social, environmental and sustainability. A

32
concluding section presents the main results and emphasizes future perspectives that can be

considered.

In the second chapter, we raise the question of how investors perceive ESG

performance depending on whether employees are appointed or not on the boardroom. To

answer this question, first, we review the literature on the impact of employee board

representation on the firm’s value and then, highlight the moderating role of employee

directors on the value relevance of ESG performance. Second, we use a recent study period

(2007-2017) to test our hypothesis and after we present the system GMM regression results.

Supplementary analyses were used to examine our research question for each pillar of ESG

performance namely, social, environmental and corporate governance and then compare the

value relevance of ESG performance and its related pillars for firms with and without

employee board representation. Last section is dedicated to the conclusion and the future

perspectives.

Extending the scope of our second research question, we distinguish, in our last chapter,

between the two types of employee directors appointed on French boards, namely employee

directors elected by right of employment (labour representatives) and directors elected by

employee shareholders (employee-shareholders representatives). First, we examine the impact

of the two types of employee directors on the social, environmental and corporate governance

pillars of ESG performance. Then, we examine how the perception of investors towards the

ESG performance is affected by the type of employee director appointed on the board.

Therefore, the last chapter aims to answer two research questions: (1) How each type of

employee directors impacts the ESG pillars namely, social, environment and corporate

governance? And (2) To what extent the value relevance of ESG performance and its related

pillars is affected by the type of employee director on the board? This chapter is organized

into different sections. First section is split into four subsections. The first subsection

33
highlights the impact of employee directors on the social pillar of the ESG performance while

the second and third subsections highlight their impact on the environmental and corporate

governance pillars, respectively. The final subsection investigates the extent to which each

type of employee directors, namely employee directors elected by right of employment and

directors elected by employee shareholders, may impact the perception of the market towards

the ESG performance. The second section tests our hypothesis using propensity score

matching and system GMM estimation. Supplementary analyses were also performed to

provide new insights on the moderating impact of each type of employee directors on the

value relevance of specific pillars of the ESG performance. Last section concludes and

highlights new areas for future research.

In the general conclusion, we discuss the main results of this thesis, identify the major

contributions and limitations and emphasize future research perspectives.

34
References

Bagdi, K. (2015). The future of employees’ board-level representation in the European Union.
Procedia Economics and Finance, 23, 1394–1400.
Barnea, A., & Rubin, A. (2010). Corporate Social Responsibility as a conflict between
shareholders. Journal of Business Ethics, 97, 71–86.
Bøhren, O., & Strøm, R. O. (2010). Governance and politics: Regulating independence and
diversity in the board room. Journal of Business Finance and Accounting, 37, 1281–1308.
Boulouta, I. (2013). Hidden connections: The link between board gender diversity and
corporate social performance. Journal of Business Ethics, 113(2), 185–197.
Blundell, R., & Bond, S. (2000). GMM estimation with persistent panel data: an application
to production functions. Econometric Reviews, 19(3), 321–340.
Brammer, S., Millington, A., & Rayton, B. (2007). The contribution of corporate social
responsibility to organizational commitment. The International Journal of Human
Resource Management, 18(10), 1701–1719.
Cahan, S. F., De Villiers, C., Jeter, D. C., Naiker, V., & Van Staden, C. J. (2016). Are CSR
disclosures value relevant? Cross-country evidence. The European Accounting Review, 25,
579–611.
Coffey, B. S., & Wang, J. (1998). Board diversity and managerial control as predictors of
corporate social performance. Journal of Business Ethics, 17(14), 1595–1603.
Conchon, A. (2011). Board-level employee representation rights in Europe: Facts and trends.
Brussels: Etui.
Cormier, D., Aerts, W., Ledoux, M. J., & Magnan, M. (2009). Attributes of social and human
capital disclosure and information asymmetry between managers and investors. Canadian
Journal of Administrative Sciences, 26, 71–88.
Dahlsrud, A. (2008). How corporate social responsibility is defined: an analysis of 37
definitions. Corporate Social Responsibility and Environmental Management, 15(1), 1–13.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism
and collective rationality in organizational fields. American Sociological Review, 147–160.
Dunn, P., & Sainty, B. (2009). The relationship among board of director characteristics,
corporate social performance and corporate financial performance. International Journal of
Managerial Finance, 5(4), 407–423.
El Akremi, A., Gond, J. P., Swaen, V., De Roeck, K., & Igalens, J. (2018). How do
employees perceive corporate responsibility? Development and validation of a

35
 multidimensional corporate stakeholder responsibility scale. Journal of Management,
44(2), 619–657.
Faleye, O., Mehrotra, V., & Morck, R. (2006). When labor has a voice in corporate
governance. Journal of Financial and Quantitative Analysis, 41, 489–510.
Fauver, L., & Fuerst, M. E. (2006). Does good corporate governance include employee
representation? Evidence from German corporate boards. Journal of Financial Economics,
82(3), 673–710.
Friede, G., Busch, T., & Bassen, A. (2015). ESG and financial performance: Aggregated
evidence from more than 2000 empirical studies. Journal of Sustainable Finance &
Investment, 5, 210–233.
Fuente, J. A., García-Sanchez, I. M., & Lozano, M. B. (2017). The role of the board of
directors in the adoption of GRI guidelines for the disclosure of CSR information. Journal
of Cleaner Production, 141, 737–750.
Gao, Y., & Yang, H. (2016). Do employees support corporate philanthropy? Evidence from
Chinese listed companies. Management and Organization Review, 12, 747–768.
Ginglinger, E., Megginson, W., & Waxin, T. (2011). Employee ownership, board
representation, and corporate financial policies. Journal of Corporate Finance, 17(4), 868–
887.
Gordon, J. N., & Roe, M. J. (2004). Convergence and persistence in corporate governance.
Cambridge University Press.
Hafsi, T., & Turgut, G. (2013). Boardroom diversity and its effect on social performance:
Conceptualization and empirical evidence. Journal of Business Ethics, 112(3), 463–479.
Hammer, T. H., Currall, S. C., & Stern, R. N. (1991). Worker representation on boards of
directors: A study of competing roles. ILR Review, 44(4), 661–680.
Hillman, A. J., & Keim, G. D. (2001). Shareholder value, stakeholder management, and social
issues: what's the bottom line?. Strategic Management Journal, 22, 125–139.
Hillman, A. J., Keim, G. D., & Luce, R. A. (2001). Board composition and stakeholder
performance: Do stakeholder directors make a difference? Business & Society, 40(3), 295–
314.
Hollandts, X., Guedri, Z., & Aubert, N. (2009). Représentation du travail au conseil
d'administration et performance de l'entreprise: une étude empirique sur le SBF 250 (2000-
2005).
Hollandts, X., & Aubert, N. (2011). La représentation obligatoire des actionnaires salariés au
conseil d'administration : un état des lieux. Gestion 2000, 28(6), 15–26.

36
Hollandts, X., Aubert, N., Abdelhamid, A., & Prieur, V. (2018). Beyond dichotomy: The
curvilinear impact of employee ownership on CEO entrenchment. Management
International, 22, 112–164.
Hollandts, X., & Aubert, N. (2019). La gouvernance salariale: contribution de la
représentation des salariés à la gouvernance d’entreprise. Finance Contrôle Stratégie, A
paraître. <hal-01989060>.
Huse, M., Nielsen, S. T., & Hagen, I. M. (2009). Women and employee-elected board
members, and their contributions to board control tasks. Journal of Business Ethics, 89(4),
581–597.
Jensen, M. C., & Meckling, W. H. (1979). Rights and production functions: An application to
labor– managed firms and codetermination. Journal of Business, 52, 469–506.
Johnson, J. L., Daily, C. M., & Ellstrand, A. E. (1996). Boards of directors: A review and
research agenda. Journal of management, 22(3), 409–438.
Jones, D. A., Willness, C. R., & Madey, S. (2014). Why are job seekers attracted by corporate
social performance? Experimental and field tests of three signal-based mechanisms.
Academy of Management Journal, 57(2), 383–404.
Kanter, R. M. (2011). How great companies think differently.
Kim, H. R., Lee, M., Lee, H. T., & Kim, N. M. (2010). Corporate social responsibility and
employee–company identification. Journal of Business Ethics, 95(4), 557–569.
Kleinknecht, R. H. (2015). Employee participation in corporate governance: Implications for
company resilience. European Journal of Industrial Relations, 21, 57–72.
Lydenberg, S. (2013). Responsible investors: Who they are, what they want. Journal of
Applied Corporate Finance, 25, 44–49.
Meyer, J. W., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth
and ceremony. American Journal of Sociology, 83(2), 340–363.
Nekhili, M., Nagati, H., Chtioui, T., & Nekhili, A. (2017). Gender-diverse board and the
relevance of voluntary CSR reporting. International Review of Financial Analysis, 50, 81–
100.
Nekhili, M., Boukadhaba, A., Nagati, H., & Chtioui, T. (2019). ESG performance and market
value: the moderating role of employee board representation. The International Journal of
Human Resource Management, 1–27.
Orlitzky, M., Schmidt, F. L., & Rynes, S. L. (2003). Corporate social and financial
performance: A meta-analysis. Organization Studies, 24(3), 403–441.

37
Pagano, M., & Volpin, P. F. (2005). Managers, workers, and corporate control. The Journal of
Finance, 60, 841–868.
Pagano, M. S., Sinclair, G., & Yang, T. (2018). Understanding ESG ratings and ESG
indexes. Research Handbook of Finance and ESG, 339.
Poulain-Rehm, T., & Lepers, X. (2013). Does employee ownership benefit value creation?
The case of France (2001–2005). Journal of Business Ethics, 112(2), 325–340.
Preuss, L., Haunschild, A., & Matten, D. (2009). The rise of CSR: Implications for HRM and
employee representation. The International Journal of Human Resource Management,
20(4), 953–973.
Roodman, D. (2009). How to do xtabond2: An introduction to difference and system GMM in
Stata. The Stata Journal, 9(1), 86–136.
Shaukat, A., Qiu, Y., & Trojanowski, G. (2016). Board attributes, corporate social
responsibility strategy, and corporate environmental and social performance. Journal of
Business Ethics, 135(3), 569–585.
Tajfel, H. and Turner, J.C. (1985), The Social Identity Theory of Intergroup Behaviour. In:
Worchel, S. and Austin, W.G., Eds., Psychology of Intergroup Relations, 2nd Edition,
Nelson Hall, Chicago, 7–24.
Triana, M. D. C., Jayasinghe, M., Pieper, J. R., Delgado, D. M., & Li, M. (2019). Perceived
workplace gender discrimination and employee consequences: a meta-analysis and
complementary studies considering country context. Journal of Management, 45(6), 2419–
2447.
Turker, D. (2009). How corporate social responsibility influences organizational
commitment. Journal of Business Ethics, 89(2), 189–204.
Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of
group behavior. In E.E. Lawler (Ed.). Advances in Group Processes, 77–122 Greenwich,
CT: JAI.
Van der Laan, G., Van Ees, H., & Van Witteloostuijn, A. (2008). Corporate social and
financial performance: An extended stakeholder theory, and empirical test with accounting
measures. Journal of Business Ethics, 79(3), 299–310.

38
Appendix 1: The Evolution of Employee Board Representation Laws over Time
Representation
Representation
Date of the of employees by
Text of the Law of employee
French law right
shareholders
of employment
July 26, 1983 State controlled companies (more than 50% of shares are hold by the state) allowed
✓
labour representation on board
October 21, Privately owned firms are allowed to change their statutes in order to have an elected
1986 representatives of employees on board ✓
(Ordonnance)
July 25, 1994 Privatized companies have to change their statutes before sale to reserve seats for
employee representatives on board
(1) 2 seats are reserved for labour representatives and one seat for employee shareholders
representatives if the board’s members is lesser than 15
(2) 3 seats are reserved for labour representatives and one seat for employee shareholders
representatives if they are more than 15 ✓ ✓
N.B: Once the firm was privatized, shareholders have the right to change once again the
statutes, cancelling by that the seats reserved for employee representatives.
However, employees are obliged to submit to a vote for the general meeting a resolution
to ask for seat (s) for employee shareholders once they own at least 5% of total shares.
Though, shareholders have the right to agree or disagree with this resolution.
February 19, The threshold of 5% is reduced to 3%
✓
2001
January 17, Mandated that employee shareholders have to be represented on boards by a nominated
2002 director when their ownership exceeds 3%. However, the application decrees had not
✓
(Modernisation been published.
sociale)
December 30, Employee shareholders of publicly listed firms have to be represented on board when
✓
2006 they hold at least 3% of the capital
June 14, 2013 Employees have to be represented on board by 1 director if the board’s members is less
(Sécurisation de than 12 members and 2 directors if the board is composed from more than 12 members ✓
l’emploi) for firms with at least 5000 employees in France or 10 000 employees in France and

39
 around the world
The threshold of eligibility of employee representation is reduced to 1000 employees in
August 17, 2015 France or 5000 employees in France and around the world.
✓
(loi Rebsamen) Moreover, this law allowed a universal representation even for very small firms (less than
11 employees)
Employees have to be represented on the board by 1 director if the board members are
Mai 22, 2019 less than 8 members and 2 directors if the board is composed from more than 8 members
✓
(loi Pacte) for firms with at least 1000 employees in France or 5000 employees in France and around
the world.

40
Appendix 2: Comparison between the Different Types of Employee Directors
Employee directors elected by right of employment Directors elected by employee shareholders
(Referring to the study of Carley (2005)) (referring to the study of Hollandts & Aubert, 2011)

14% are women 100% are men (except for Thales)

55% are aged between 35-50 years On average, they are over 50 years old
45% are aged between 50-65 years

60% are manual or clerical workers These directors have, on most of cases, executive functions
24% are middle management level (e.g., financial director, marketing manager, branch general
10% are technical and supervisory level manager, management control, etc.)
Personal 5% are senior managers
characteristics
Among the employee directors elected by right of employment Almost all directors elected by employee shareholders in
who respondent to this study, 59% of them are elected by their study are nominated by the internal association of
workers of the firm, 34% are nominated by work councils or employee shareholders (except Renault)
trade unions.
Over three-quarter have between 1-10 years’ experience in this On average, they have 3 years of experience in this function
function

• Elected by the workforce • Elected by employee shareholders

Process of • Nominated by work council • Nominated, mostly, by the internal association

nomination • Nominated by trade union of employee shareholders
• Nominated by the European work council

41
42
 Chapter I: Does Employee Board Representation
Matter for the Relevance of CSR Reporting?

1. Introduction

For the last two decades, corporate social responsibility (hereafter CSR) reporting has been

used as a channel to inform the firm’s stakeholders (shareholders, employees and investors,

etc.) about social and environmental activities. Reporting extra-financial information is

largely considered as a good strategy to gain legitimacy in the eyes of stakeholders, who have

different interests concerning CSR (Sahed-Granger & Boncori, 2014). Nevertheless, the use

of CSR reports may change from stakeholder to another. For example, while investors may

use extra-financial information to assess the competitiveness of the firm and its future

perspectives, shareholders may use it to understand how their funds have been used.

Employees are also interested in such information since they are directly affected by social

investments (work conditions, training, etc.). Indeed, André, Husser, Barbat and Lespinet-

Najib (2011) document that CSR reports of French companies tend to strongly favour the

issue of employees (retirement, training, mobility, etc.). The other stakeholders, such as

governments, institutions, creditors, etc., may use CSR reports to assess the firm’s social and

environmental impact and its reputation as a socially responsible or irresponsible firm.

Whereas increasing the awareness of stakeholders about the firm’s CSR activities is

value relevant and it has grown constantly over time (Allix-Desfautaux and Makany, 2015),

the motivation of managers behind such disclosure is not clear. High level of CSR disclosure

may be used to improve stakeholders’ perception of the firm’s social and environmental

commitment or, conversely, to divert their attention from unfavourable performance or

actions. This indeterminacy is reinforced by the inconclusiveness of findings on the value

43
relevance of CSR reporting. Previous empirical studies have examined the value relevance of

CSR disclosure and report sparse and ambiguous results, thus highlighting the complexity of

the relationship between CSR reporting and firm value (Nekhili, Nagati, Chtioui, &

Rebolledo, 2017a; Nekhili, Nagati, Chtioui, & Nekhili, 2017b). This inconclusiveness

regarding the value relevance of CSR reporting might be explained by the difference in

stakeholders’ perception, as Bebbington, Larrinaga and Moneva (2008) have argued. More

acutely, the credibility of CSR reporting may be called into question. The KPMG

international survey on CSR reporting in 2008 revealed that, for French firms, enhancing

market share is among the top key drivers for CSR reporting, in front of ethical considerations

and innovation.

In this chapter, we argue that a diverse board that includes employee directors may, in

such circumstances, reflect higher credibility of CSR reporting. Indeed, employee directors

have a special concern for social and environmental activities of the firm (Huse, Nielson &

Hagen, 2009). Moreover, they care about the firm’s reputation for being socially responsible

or irresponsible, since they are directly impacted by it. Previous works argue that Employee

representatives can play an important role in CSR decisions (Preuss, Haunschild & Matten,

2009), in the board’s effectiveness concerning CSR tasks (Huse et al., 2009). Barnea and

Rubin (2010), however, document conflict of interest between shareholders with respect to

CSR. More precisely, they argue that inside shareholders (large shareholders, managers and

employees) gain more advantages from CSR than institutions and small investors. Along

similar lines, Guedri and Hollandts (2008) document that allowing non-executive employees

to be represented on the board increases the likelihood of their self-interest maximization at

the expense of shareholders value maximization. In this respect, it is possible to expect an

alliance between managers and employees counterbalancing shareholders supremacy on

board. We, therefore, raise the question of how investors would perceive a high level of CSR

44
reporting when employees are represented on the board. Would more CSR reporting, when

there are employee directors, be conducive to greater credibility for investors or be seen as

over-investment at the expense of shareholders? To our knowledge, the present study is the

first to examine the relationship between voluntary CSR reporting and firm performance

using the moderating role of employee directors.

To do so, first, we use Tobin’s q as a proxy to detect the value relevance of CSR

reporting as perceived by the market. Second, we test whether employee board representation

and CSR reporting jointly affect firm performance as measured with Tobin’s Q using the joint

test technique. We control for endogeneity between the independent variables (CSR reporting

and employee directors) and the dependent variable (firm performance) using the system

GMM approach (Roodman, 2009).

Using a sample of French firms listed on the SBF 120 for the period from 2001 to 2011,

we provide new evidence regarding the impact of employee directors on the value relevance

of CSR reporting. Primarily, we find that firms with employee directors on the board report

more on their social activities than their counterparts. Our regression analyses show that the

capital market positively assesses high CSR reporting when employees are represented on the

board. However, the marginal effect of employee directors on the relationship between CSR

reporting and market value differs among CSR categories. Meaningfully, we find that, when

employees are represented on the board, reporting on environmental and sustainability

information is more relevant than reporting on social activities. This finding supports the

previous evidence on the conflict of interest between different shareholders and employees. A

high level of social reporting may reflect an alliance between managers and employee

directors that counterbalances shareholders’ power on the board. Moreover, a high level of

social reporting may also reflect more advantages for employees at the expense of

shareholders’ wealth.

45
 Our study contributes to the research mainstreams examining the effects of employee

board representation and the voluntary CSR reporting in several ways. First, we show that the

market participants pay attention to the board’s composition and use filters to assess the

credibility of voluntary CSR disclosure. Second, we argue that using moderating variables is

likely to provide more deep insights. Third, we highlight conflict of interest between

employees and shareholders regarding CSR reporting categories namely, social,

environmental and sustainability.

The chapter is organized as follows. We start by examining the motives of the firm

behind engaging in CSR and the value relevance of CSR reporting. We then question the role

of employee directors in moderating the value relevance of CSR reporting and we suggest

hypotheses. Next, after describing the methodology used, we present and discuss the

empirical results. Finally, we present our conclusion and suggest new research perspectives.

2. Conceptual Framework and Hypotheses Development

2.1. CSR Reporting : Motives and related Value Relevance

Referring to the meta-analysis work of Aguinis and Glavas (2012), the incentives of the firm

behind engaging in CSR can be split into three levels: institutional, organizational and

individual. The institutional level suggests that firms engage in CSR activities as a response

for institutional forces, namely regulation, standards and certification (Christmann & Taylor,

2006), or as a response for stakeholders’ pressure and needs (Johnson and Greening 1999).

However, it is noteworthy to mention that under high pressure, firms may engage in symbolic

rather than substantive activities to avoid the negative perception of stakeholders or bad

reputation. From an organizational standpoint, firms may engage in social and environmental

actions to search for legitimacy and competitiveness (Bansal & Roth, 2000) or gain extra-

financial benefits such as enhancing its intangible resources (Surroca, Tribó & Waddock,

2010). Moreover, firms with unexpected CSR reporting may reach higher market value as

46
measured by Tobin’s Q (Cahan, De Villiers, Jeter, Naiker & Van Staden, 2016). Lastly, from

an individual standpoint, values and personal motives of CEOs and employees may influence

the firm’s CSR activities (Bansal, 2003; Markey, McIvor & Wright, 2016; Waldman, Siegel

& Javidan, 2006).

From a stakeholder theory standpoint, firms can be viewed as a set of relations with

different parties (primary and secondary stakeholders). In that sense, meeting the different

needs and expectation of their stakeholders through CSR reporting is a key success for firms.

CSR reporting encloses “any information that a firm makes public, typically within or

alongside its annual accounts or in a stand-alone report, that relates to its performance,

standards or activities under the corporate social responsibility umbrella” (Brooks &

Oikonomou, 2018: p. 2). These social, environmental and sustainability information disclosed

are useful for the different stakeholders and particularly for shareholders allowing them to

assess the potential risks and future profitability of the firm as well as its CSR engagement.

By reporting on their CSR activities, firms are likely to gain some competitive advantages

such as an increased transparency in the eyes of stakeholders (Fuente, García-Sanchez, &

Lozano, 2017), good reputation (Bear, Rahman, & Post, 2010), positive image (Adams,

2002), better treatment from regulators (Aerts & Cormier, 2009), and reduced asymmetry of

information between managers and investors (Cormier et al., 2009).

Whereas CSR reporting is with benefits for the firm, the incentives of managers behind

voluntary disclosure are not yet well understood thus raising the scepticism of stakeholders

towards the informativeness of the extra-financial information. First, managers may report on

their social and environmental activities to cover the poor performance or hostile actions

(such as lay off plans, exceed of CO2 accepted emission rates, gender discrimination at work,

etc). Second, managers could maintain some value-relevant information, namely proprietary

information to protect their position in the market. Last but not least, managers may report on

47
their CSR activities as a result of great pressure from stakeholders. In such a case, the

probability of adopting symbolic CSR policies would be higher, thus affecting the quality of

CSR information disclosed. Michelon, Pilonato and Ricceri (2015) report that CSR reporting

practices (CSR verification by a third independent party, reporting in standalone reports and

compliance with GRI guidelines) expected to enhance the credibility of CSR reports are not

associated with better CSR reporting quality.

Research on the value relevance of social and environmental information for

stakeholders and particularly shareholders provide mixed results. Cahan et al. (2016)

investigate the value relevance of CSR information disclosed by 676 firms from 21 countries.

They find that CSR disclosure, and particularly unexpected information is positively related to

higher firm value (Tobin’s q). Similarly, focusing on a sample of British firms from 2005-

2009, Qiu, Shaukat and Tharyan (2016) report a positive and significant association between

CSR disclosure and the firm’s stock price. Contrary to this positive evidence, Nekhili et al.

(2017b) show that the level of voluntary CSR disclosure of French firms is negatively and

significantly related to the market performance as measured by Tobin’s q. Nevertheless, in

their investigation of the value relevance of CSR disclosure for a sample of US manufacturing

companies during 1997 and 2010, Cho, Michelon, Patten and Roberts (2015) document that

CSR information is not associated with firm value. Similar results were obtained in the study

of Verbeeten, Gamerschlag and Möller (2016). Using data from 130 German companies over

the period 2005-2008, they find that a higher level of CSR information is unrelated to share

price. To better understand the perception of the market participants towards CSR reporting,

previous research investigate the moderating role of board diversity as measured with female

directorship (Nekhili et al. 2017b) and share ownership structure as measured with family

ownership (Nekhili et al. 2017a) on the value relevance of social, environmental and

sustainability information. Using a sample of French firms over the period 2001-2011,

48
Nekhili et al. (2017b) find that female directorship moderates positively the relationship

between CSR reporting and firm value as measured with Tobin’s q suggesting that the market

place greater value on CSR information when female directors are on board. Besides, Nekhili

et al. (2017a) document that CSR reporting is more relevant for the market when disclosed by

family firms, while it is less relevant when disclosed by non-family firms. The previous

pieces of evidence suggest that the market participant use filters to assess the value relevance

of CSR-related information.

The board of directors, argued to be the apex of the decision making, is responsible for

the firm’s strategies and policies, including CSR engagement (Huse et al., 2009). Efficient

boards are argued to enhance the quality of CSR reporting by disclosing extensive and

transparent information (Cormier et al., 2009). More precisely, the board composition

variables such as female directors, independent directors and board CSR committee, are

documented to enhance the CSR reporting (Cormier et al., 2009; Fuente et al., 2017; Nekhili

et al., 2017b). Considering the discussion above, the board composition may be a relevant

filter for the market participant to consider the relevance of social, environmental and

sustainability information.

2.2. Employee board representation and the value relevance of CSR reporting

Besides female and independent directors, the presence of employee representatives on board

is another way to examine the board composition. Providing the opportunity for employees to

be represented on board is expected to increase the firm’s involvement towards CSR (1) by

playing an active role in CSR decisions (Preuss et al., 2009) and (2) by contributing to CSR

controls (Huse et al., 2009).

CSR reporting is an effective tool to keep up stakeholders with the firms’ CSR

activities. However, from all stakeholders, employees may be the most interested in CSR

actions and policies. In fact, employees, an insider stakeholder, care about, assess, react, and

49
contribute to the firm’s CSR activities (El Akermi, Gond, Swaen, & Igalens, 2018; Huse et al.

2009; Jones, Willness, & Madey, 2014). In order to understand how employees assess and

react to the CSR activities of the firm, El Akermi et al. (2018) develop a new measure to

detect the employee CSR perception. Using this new measure, they document that the positive

perception of employees towards CSR activities of their organizations makes them feel pride,

committed and satisfied. Focusing on the firm’s attractiveness to job seekers, Jones et al.

(2014) report that firms with higher community commitment are more able to recruit talented

employees. Employees also pay attention to the social justice within their organizations and

react to irresponsible actions regarding this issue. Using a sample of published and

unpublished studies on gender discrimination effects on employees, Triana, Jayasinghe,

Pieper, Delgado and Li (2018) argue that firms which discriminate between its workers based

on their gender are more likely to deal with negative job attitudes and poorer quality of

psychological health of their workers. Concerning employee contribution to CSR, previous

studies document that workers awareness play an important role in adopting and suggesting

socially responsible actions (Markey et al., 2016; Preuss et al., 2009). Studying the Australian

context, Markey et al. (2016) argue that employee participation in the decision making

process is positively associated with CO2 reduction decision in the workplace. Similarly,

Bansal (2003) finds that the individual values and discretion of employees (and senior

managers) plays a considerable role in the firm environmental engagement, particularly firm’s

response to natural environmental issues. For example, she shows that employees may

propose some CSR initiatives that they esteem relevant in their working place such as energy

management and waste recycling.

The participation of employees in the decision-making process is likely to enhance the

quality of social and environmental information disclosed to stakeholders. First, employees

have more valuable, real and precise information about the firm’s workplace and the workers’

50
needs. Consequently, they would enhance the efficiency of social investments related to

employees. In such case, stakeholders may treat with confidence CSR information and

particularly human capital information disclosed. Second, employees are dependent on the

viability of their organization. Therefore, they may encourage extensive CSR reporting in

order to increase stakeholders’ awareness, essential for their firm’s viability. Third, employee

directors enhance the board’s efficiency (Huse et al., 2009), which is responsible for better

quantitative CSR reporting (Cormier et al., 2009). Last but not least, employees as insiders

witness of the social and environmental actions and policies which makes them different from

other stakeholders. Taken together, we suggest that employee directors may impact positively

the way the market perceives CSR reporting.

H1a: CSR reporting is more relevant for firms with employee directors than for firms

without employee directors.

Nevertheless, previous studies argue an opportunistic behaviour of employees on board.

Meaningfully, employee directors may use their power on board to maximize their own

interests rather than shareholders’ interests (Bøhren & Strøm, 2010; Faleye, Mehrotra, &

Morck, 2006; Guedri & Hollandts, 2008). Employee representatives on board are likely to use

their voting rights to increase the interests of the workers at the expense of those of

shareholders (Faleye et al., 2006). More particularly, they may increase social investments

with benefits to employees such as enhancing the working conditions, more training, higher

salaries, etc. In that sense, extensive CSR reporting may signal a powerful position of

employees on board that threat the shareholders supremacy and/or may indicate a possible

managers-employees coalition. The possible coalition between managers and employees may

result from the close relationship between the two parties within the firm (Pagano & Volpin,

2005). Certainly, the CSR investments enhance the intangibles assets of the firm (Surroca et

al., 2010). However, they would not necessarily increase shareholders’ value (Barnea &

51
Rubin, 2010), neither reflect a real social responsibility. In fact, Barnea and Rubin (2010)

document that high CSR expenditure brings personal benefits for managers, employees and

large shareholders at the expense of minority shareholders. Therefore, the perception of the

market participants towards CSR reporting may be negatively influenced by employee board

representation. Taken together, we suggest that:

H1b: CSR reporting is less relevant for firms with employee directors than for firms

without employee directors.

3. Methodology

3.1. Sample and data

In this study, we focus on French listed firms belongings to the SBF 120 index for a period of

eleven years starting in 2001 to 2011. After excluding financial, assurance and real estate

firms, we end up with a sample composed from 91 firms. The French context shows an

increasing interest in CSR and in CSR reporting specifically. Indeed, our study period (2001-

2011) extends from enactment of the NRE law in 2001 to the year before Grenelle II came

into force in 2012 and encloses the implementation of the Grenelle I in 2009. NRE, Grenelle I

and Grenelle II are French laws encouraging the engagement of firms in social activities.

During our study period, reporting on CSR was made on a wholly voluntary basis. Our data

concerning corporate governance, ownership and CSR variables were manually compiled

from public annual reports and standalone reports. As to financial and accounting data, they

were collected from the ThomsonOne database.

52
3.2. Dependent variable: Tobin’s q

Based on previous value relevance studies, we rely on Tobin’s q to measure the firm’s market

value. Largely used in literature (for example, Cahan et al. 2016; Ginglinger et al. 2011;

Nekhili et al., 2017b), this measure provides a good proxy to studying the market perception

towards CSR reporting (Cahan et al. 2016). Not being influenced by accounting conventions

is another benefit for the use of Tobin’s q.

3.3. Endogenous variable: CSR-related reporting

We use the un-weighted disclosure index method proposed by Botosan (1997) to measure the

CSR reporting level. The first step requires item identification. For that, we use the Grenelle

II grid because of its accuracy, simplicity and its compliance with GRI guidelines. The

Grenelle II grid lists all CSR information required. We identify three main categories (social,

environment and sustainability). For each category there are several required pieces of

information, deemed relevant for disclosure. For social reporting, we find 19 items dealing

with employment, organization of work, training and labour relations, equal treatment, health

and safety, and compliance with the clauses of basic agreements of the International Labour

Organization (ILO). For environmental reporting, 14 items are required, variously related to

the firm’s environment policy, pollution and waste management, the sustainable use of

resources, response to climate change, and the protection of biodiversity. Information required

in relation to sustainability (9 items) includes the territorial, social and economic impact of the

firm’s activities, its relationship with stakeholders, subcontractors and suppliers, honesty in

practices, and the measures taken on behalf of human rights. All items are listed in Appendix

I.1. For each item, we assign the value 1 if it is clearly disclosed in annual reports and/or

standalone reports, and 0 otherwise. We then calculate the level of CSR disclosure index as

the ratio of the assigned total score to the maximum score (42 items).

53
3.4. Moderating variable: employee directorship

In this study, we focus solely on the presence of employee directors on the board, since

shareholders are opposed to both types of employee directors (Bøhren & Strøm, 2010; Faleye

et al., 2006). We measure the moderating role of employee directors using a dummy variable

that takes the value 1 if there is at least one employee director on the board and 0 otherwise.

3.5. Control variables

In this study, we control for several variables that may affect our dependent variable, Tobin’s

q, such as board characteristics, ownership structure and others variables. First, some

corporate governance characteristics are considered as a tool to ensure that managers act to

defend not only shareholders’ interests but also those of interest owners in the broader sense

(Haniffa & Cooke, 2005). Peters and Romi (2014) consider CSR committee and CSR

assurance as two important sustainability-oriented corporate governance mechanisms in order

to manage and monitor sustainability concerns. Firms that verify their CSR reports through a

third party are likely to produce better quality of CSR information (Gillet, 2012). The

presence of CSR committee may signal higher engagement of the board to encouraging

sustainability and having a greater propensity to report their CSR practices/increasing

transparency through the disclosure of more information (Fuente, García-Sánchez, & Lozano,

2017). Independent directors pay attention to all stakeholders (Haniffa & Cooke, 2005);

consequently, they may encourage the firm’s CSR disclosure to increase the awareness of

stakeholders towards the firm’s CSR activities. Board meetings may translate a better

engagement of board members to discuss important issues such as those related to CSR

(Nekhili et al., 2017a, 2017b). CEO power (CEO tenure and CEO/chair duality) on board may

impact the voluntary disclosure of CSR information (Galbreath, 2010; Lewis, Walls &

Dowell, 2014). The second category of control variables concerns ownership structure that

has an influence on the disclosure of social and environmental information (Prado-Lozano et

54
al, 2009). Family owners are likely to report less on CSR duties (Nekhili et al., 2017a). In

contrast, institutional owners may exhibit higher pressure on the firm’s managers and require

greater disclosure of CSR-related information. Employee ownership is considered to be an

important counterbalance to other shareholders and tends to be willing to support CSR

activities and disclosure. The third category of control variables concerns firm characteristics

that are commonly investigated in the literature. Accordingly, we control for leverage, foreign

assets, firm systematic risk (measured by beta), R&D intensity and company size. Finally, to

comply with the French regulatory context, we control the introduction of the low known as

Grenelle I adopted on July 23, 2009.

3.6. Estimation method

CSR reporting may impact the firm’s market value; however, profitable firms may disclose

more information on their CSR activities than less profitable firms. Moreover, employee

board representation may be driven by board size and board size may impact CSR reporting

(Giannarakis, 2014). In such case, if we do not control for board size our results would be

biased. Here, two serious problems of endogeneity, namely reverse causality and omitted

variables, may arise. Consequently, we opt for the use the two-step system GMM (known as

system GMM) estimator, argued to be the better solution for the endogeneity problem in small

samples (Blundell & Bond, 2000). In addition, we control for the dynamic structure of

Tobin’s q by adding its lagged value to our model.

We use the following empirical model to examine the impact of employee board

representation on the value relevance of CSR reporting:

Tobin’s Qit = β0 + β1LagTobin’s Qit + β2CSR_REPit + β3EMP_REPit +

β4CSR_REP*EMPL_BOARD β5CSR_ASSit + β6CSR_COMit + β7BOARD_SIZEit +
β8BOARD_INDit + β9BOARD_MEETit + β10DUALit + β11TENUREit + β12FAM_OWNit +
β13INST_OWNit + β14EMPL_OWNit + β15LEVit + β16BETAit + β17FOR_ASSit + β18R&Dit +
β19SIZEit + β20GRE1it + β21INDUSTRYit + ɛit

55
 All variables are as described in Table I.1.

Examining the consistency of system GMM estimator is crucial to avoid biased results.

Referring to previous works (for example, Nekhili et al., 2017b), we use two statistical tests :

the second-order autocorrelation test and the Sargan/Hansen over-identifying test.

Table I. 1: Variables and Their Measurement

Variable Description Measure5
Dependent variable
Tobin’s Q Tobin’s q Stock market capitalization plus book value of liabilities as a
ratio of total assets
Endogenous variables:
CSR_REP CSR reporting CSR reporting index as the ratio of the assigned total
score to the maximum CSR score (42 items, see Appendix I.1).
SOCIAL_REP Social reporting Social reporting index as the ratio of the assigned total
score to the maximum social score (19 items, see Appendix I.1)
ENVIR_REP Environmental reporting CSR reporting index as the ratio of the assigned total
score to the maximum environmental score (14 items, see
Appendix I.1)
SUST_REP Sustainability reporting CSR reporting index as the ratio of the assigned total
score to the maximum sustainability score (9 items, see Appendix
I.1)
Moderating variable
EMPL_BOARD Employee board Binary variable that takes the value 1 if the firm have at least one
representation employee on board and 0 otherwise
Governance variables
CSR_ASS CSR assurance Binary variable that takes the value 1 if CSR assurance is adopted
and 0 otherwise
CSR_COM CSR committee Binary variable that takes the value 1 if the company has a CSR
committee and 0 otherwise
BOARD_SIZE Board size Natural logarithm of the number of directors on the board
BOARD_IND Board independence Ratio of number of non-executive independent directors to total
number of board directors
BOARD_MEET Board meetings Natural logarithm of the number of annual board meetings
DUAL CEO duality Dummy variable coded 1 if the CEO serves as board chair; 0
otherwise.
TENURE CEO tenure Number of years within the company as CEO
Ownership variables
FAM_OWN Family ownership Percentage of capital held by family
INST_OWN Institutional ownership Percentage of capital held by institutional investors
EMPL_OWN Employee ownership Percentage of capital held by employees
Other control variables
LEV Leverage Ratio of total financial debt to total assets
FOR_ASS Foreign assets Ratio of foreign assets to total assets
BETA Beta Equity beta
R&D R&D intensity Ratio of Research and Development to total sales
SIZE Firm size Natural logarithm of the total assets
Industry Industry Binary variable that takes the value 1 if the company belongs to
the sector in question and 0 otherwise

5
Note: Variables from ThomsonOne are winsorized at the 1% and 99% levels.

56
4. Results

4.1. Descriptive Statistics

Table I.2 presents the descriptive statistics for the sample observation. With regard to market

value, our sample firm-years have an average Tobin’s q of 1.135. As reported in Table I.2,

only 27.85% of firm-years have employee representatives on their boards. Our sample firm-

years disclose on average 44.27% of the total items included in the Grenelle II Act grid. As

regards dimensions of CSR information disclosed by firms in our sample, the average social,

environmental and sustainability disclosure score is 45.31%, 39.71% and 56.62%,

respectively. Only 19.84% of firms provide assurance of their CSR information and 27.84%

of firms have a CSR committee. Boards are, on average, composed from 11 directors, of

whom 42.74% are independent. On average, board meetings are around seven per year.

54.10% of our firm-years present a duality in function of the CEO and the chairman of the

board. Approximately, CEO’s tenure is around nine years. With regard to ownership

structure, family, institutions and employees own 26%, 15% and 2%, respectively of total

shares. Leverage is around 26%, the average proportion of foreign assets is 38%, and the

average beta is 0.885. As reported in Table I.2, average R&D intensity is 1.92%. The

companies’ average size is 16.717 billion euros.

57
Table I. 2: Descriptive Statistics
Mean Median Standard Minimum Maximum
Deviation
Tobin’s Q 1.135 0.888 0.827 0.256 4.557
EMPL_BOARD 27.85% 0 44.85% 0 1
CSR_REP 44.27% 47.62% 25.10% 0 90.48%
SOCIAL_REP 45.31% 52.63% 29.05% 0 100%
ENVIR_REP 39.71% 35.71% 27.88% 0 92.86%
SUST_REP 52.62% 50% 31.29% 0 100%
CSR_ASS 19.84% 0 39.90% 0 1
CSR_COM 27.84% 0 44.84% 0 1
BOARD_SIZE (number of directors) 11.613 12 3.962 3 26
BOARD_IND 42.74% 42.86% 23.46% 0 100%
BOARD_MEET (number of meetings) 7.219 7 3.547 0 30
DUAL 54.10% 1 49.86% 0 1
TENURE (number of years) 9.089 7.14 6.987 0 43
FAM_OWN 26.64% 22.91% 26.20% 0 99.37%
INST_OWN 15.45% 5% 22.65% 0 90%
EMPL_OWN 2.50% 0.99% 4.73% 0 32.75%
LEV 26.21% 25.28% 13.63% 0.93% 60.07%
FOR_ASS 38.81% 37.99% 29.20% 0 97.36%
BETA 0.885 0.899 0.273 0.203 1.575
R&D 1.92% 0 4.46% 0 24.17%
SIZE (in billions of euros) 16.717 5.185 29.784 4 240.559
Note: All variables are as defined in Table I.1.

4.2. Univariate Analysis

Table I.3 presents differences in variables between firms with and without employee

representation. Consistently with Ginglinger et al. (2011), firms without employee

representation achieve better market-based performance, as measured by Tobin’s Q, than

firms with employee representation (1.211 and 0.897, respectively). Our study shows that the

level of CSR disclosure and its components is greater for firms with employee representation

than without employee representation (54.26% and 40.63%, respectively). Provision of

assurance services and the presence of a CSR committee are also more prevalent in firms with

employee representation than in firms without employee representation. Firms with and

without employee representation differ also significantly in terms of board characteristics.

Similarly to Ginglinger et al. (2011), we find that firms with employee representation tend to

have a larger board size (14.25 versus 10.59) and more annual meetings (8.3 versus 6.7) than

firms without employee representation. The separation of chairman and CEO functions is

58
greater than in firms with employee representation (51.41% versus 60.98%). The average

length of tenure for CEOs working in firms with employee representation is higher than in

firms without employee representation (9.8 versus 8.8 years). No significant difference is

observed for board independence between the two panels. Regarding ownership structure,

Table I.3 shows that family shareholders hold more capital in firms without employee

representation (31.37%) than in firms with employee representation (14.39%). Conversely,

employee shareholders hold less of the capital of firms without employee representation

(1.06%) than in firms with employee directors (6.13%). For the other control variables, we

observed significant statistical differences regarding foreign assets, beta, R&D intensity, and

firm size. There are no significant differences in terms of leverage between firms with and

without employee representation.

Table I. 3: Mean Difference Test

Variables Firms with at least one Firms without t-test
employee board employee board
member member
Tobin’s Q 0.897 1.211 5.079***
CSR_REP 54.26% 40.63% 7.629***
SOCIAL_REP 52.85% 42.42% 4.981***
ENVIR_REP 52.27% 34.84% 8.871***
SUST_REP 66.01% 48.46% 8.017***
CSR_ASS 33.33% 15.69% 6.129***
CSR_COM 40.15% 24.27% 4.902***
BOARD_SIZE (number of directors) 14.257 10.594 12.420***a
BOARD_IND 44.02% 42.43% 0.939
BOARD_MEET (number of meetings) 8.367 6.770 7.027***a
DUAL 60.98% 51.41% 2.655***
TENURE (number of years) 9.802 8.812 2.325**a
FAM_OWN 14.39% 31.37% 9.340***
INST_OWN 15.95% 15.26% 0.420
EMPL_OWN 6.13% 1.09% 16.681***
LEV 25.19% 26.44% 1.271
FOR_ASS 32.18% 41.61% 4.483***
BETA 0.928 0.875 2.692***
R&D 2.44% 1.69% 2.294**
SIZE (in millions of euros) 33,652 10,659 9.723***a
Number of observations 264 684
Note:*,**, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
a
t-tests are based on natural logarithm transformed values.
All variables are as defined in Table I.1.

59
4.3. Multivariate Analysis

Before carrying out regression analyses, the pairwise correlation matrix and the VIFs are used

to assess the correlations between the independent variables. Table I.4 reveals no strong

multicollinearity between independent variables, and no correlation exceeds 0.5. This

condition is crucial for the validity of our instruments. Table I.4 also shows an excellent

variance inflation factor (VIF<3).

60
Table I. 4: Pairwise Correlation
1 2 3 4 5 6 7 8 9 10 11 12 VIF

1. Tobin’s Q 1.000
2. CSR_REP –0.157* 1.000 1.82
3. SOCIAL_REP –0.137* 0.932* 1.000 1.55
4. ENVIR_REP –0.109* 0.858* 0.677* 1.000 1.67
5. SUST_REP –0.193* 0.775* 0.631* 0.538* 1.000 1.65
6. EMPL_BOARD –0.167* 0.240* 0.159* 0.277* 0.252* 1.000 1.84
7. CSR_ASS –0.079 0.442* 0.423* 0.346* 0.376* 0.195* 1.000 1.50
8. CSR_COM –0.067 0.389* 0.349* 0.361* 0.287* 0.157* 0.141* 1.000 1.31
9. BOARD_SIZE –0.253* 0.376* 0.297* 0.389* 0.318* 0.375* 0.248* 0.238* 1.000 1.91
10. BOARD_IND –0.179* 0.215* 0.204* 0.168* 0.198* 0.030 0.280* 0.086* 0.096* 1.000 1.45
11. BOARD_MEET –0.059 0.157* 0.144* 0.115* 0.160* 0.223* 0.130* 0.157* 0.045 –0.011 1.000 1.18
12. DUAL –0.085 0.071 0.009 0.128* 0.100* 0.086* –0.044 –0.094* 0.038 –0.175* –0.025 1.000 1.17
13. TENURE 0.082 0.265* 0.205* 0.228* 0.292* 0.075 0.223* 0.136* 0.163* 0.028 –0.074 0.172* 1.24
14. FAM_OWN 0.282* –0.083 –0.037 –0.108* –0.110* –0.290* –0.193* –0.069 –0.148* –0.278* –0.093* –0.020 1.54
15. INST_OWN –0.192* 0.178* 0.156* 0.159* 0.143* 0.013 0.183* 0.049 0.060 0.301* –0.055 0.046 1.41
16. EMPL_OWN –0.236* 0.127* 0.076 0.140* 0.164* 0.476* 0.021 0.031 0.179* –0.017 0.034 0.188* 1.24
17. LEV –0.206* 0.009 –0.001 0.034 –0.011 –0.041 0.017 –0.056 0.023 –0.006 0.042 0.045 1.13
18. BETA 0.022 0.074 0.103* –0.034 0.134* 0.087* 0.176* 0.042 0.001 0.102* 0.254* –0.038 1.24
19. FOR_ASS –0.062 0.007 –0.028 –0.035 0.125* –0.144* 0.018 –0.022 0.059 0.247* 0.078 –0.117* 1.19
20. R&D 0.233* 0.110* 0.109* 0.068 0.110* 0.074 0.127* –0.006 –0.033 0.086* 0.013 –0.067 1.15
21. SIZE –0.281* 0.465* 0.370* 0.427* 0.474* 0.303* 0.421* 0.273* 0.659* 0.308* 0.140* –0.079 2.54
Table I.4: Continued
13 14 15 16 17 18 19 20 21
13. TENURE 1.000
14. FAM_OWN 0.017 1.000
15. INST_OWN –0.061 –0.426* 1.000
16. EMPL_OWN 0.110* –0.173* 0.083 1.000
17. LEV –0.079 –0.046 0.049 –0.093* 1.000
18. BETA 0.046 –0.223* –0.016 –0.069 –0.032 1.000
19. FOR_ASS 0.107* –0.106* 0.063 –0.168* –0.052 0.093* 1.000
20. R&D 0.168* 0.037 –0.067 –0.101* –0.192* 0.085* 0.000 1.000
21. SIZE 0.146* –0.265* 0.091* 0.100* 0.104* 0.200* 0.127* –0.012 1.000
Note:* Represents significance at the 0.01 level. All variables are as defined in Table I.1.

61
 Table I.5 presents the system GMM regression results of Tobin’s q on CSR reporting

and employee representation. Model 1 of Table I.5 shows a positive (0.735) and significant

impact (z = 6.60) from CSR reporting on market-based performance as measured by Tobin’s

q. In line with Cahan et al. (2016) and Qiu et al. (2016), shareholders are more likely to assign

value relevance to CSR information disclosed. Nevertheless, a negative and significant

relationship is found between employee board representation and Tobin’s q suggesting that

the market participants do not appreciate the presence of employees on board. One

explanation is that investors may consider employees’ presence on board as a threat for their

supremacy on board. In other words, employees may use their voting rights to maximise their

own interests at the expense of shareholders (Guedri & Hollandts, 2008).

The positive and significant relationship between CSR reporting and market-based

performance remains unchanged when we introduce employee board representation into the

model. No significant impact is observed for CSR assurance, suggesting that providing

assurance services is not rewarded by a higher valuation in the financial markets. We also

note a negative and significant correlation between the existence of a CSR committee and

market value. These results imply that sustainability-oriented corporate governance

mechanisms may be perceived by market participants as costly and simply as symbolic

actions with regard to sustainability issues.

As regards corporate governance structures, board characteristics are most often

negatively related to firm performance. Table I.5 shows negative and significant coefficients

between board size, board independence, board meetings and CEO duality of function,

suggesting that smaller boards with fewer board outsiders, fewer board meetings and high

separation between CEO and chairman functions are more valued by the market. Concerning

ownership variables, our findings show that family ownership is positively and significantly

62
linked with market performance as measured by the Tobin’s q, whereas employee ownership

is negatively linked with Tobin’s q, suggesting that employee ownership reduces firm value.

No significance is found for institutional ownership. For the other control variables, firm

leverage and foreign assets are negatively and significantly related to Tobin’s q. R&D

intensity and beta are found to be positively related to Tobin’s q. Finally, no significance is

observed for firm size.

Table I. 5: System GMM Regression of Tobin's q on CSR Reporting and Employee

Board Representation
Variables Model 1 Model 2

Coef. t–test Coef. t–test

Lag Tobin’s Q 0.089*** 16.46 0.080*** 9.32
CSR_REP 0.735*** 6.60 0.690*** 5.46
EMPL_BOARD –0.302*** –3.49
CSR_ASS –0.043 –1.26 –0.022 –0.55
CSR_COM –0.143*** –4.83 –0.128*** –4.29
BOARD_SIZE –0.265*** –5.56 –0.184*** –3.27
BOARD_IND –0.180** –2.01 –0.189** –2.15
BOARD_MEET –0.089*** –2.92 –0.048 –1.47
DUAL –0.250*** –7.35 –0.209*** –5.81
TENURE 0.001 0.03 0.015 0.46
FAM_OWN 0.461*** 4.12 0.356*** 3.26
INST_OWN –0.051 –0.73 –0.133 –1.60
EMPL_OWN –2.392*** –6.12 –1.237** –2.28
LEV –0.596*** –6.50 –0.590*** –6.07
BETA 0.290*** 6.48 0.305*** 5.38
FOR_ASS –0.273*** –4.15 –0.340*** –5.47
R&D 1.263*** 4.30 1.590*** 3.74
SIZE –0.029* –1.76 –0.021 –1.30
Intercept 2.153*** 6.49 1.832*** 5.85
Year Yes Yes
Industry Yes Yes
Number of observations 790 790
Fisher (Prob. > F) 8812.63 (p = 0.000) 4865.23 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.94 (p = 0.003) –2.90 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): 0.21 (p = 0.837) 0.14 (p = 0.887)
Sargan test (Chi–square, p–value): 734.31 (p = 0.000) 733.83 (p = 0.000)
Hansen test (Chi–square, p–value): 77.83 (p = 0.218) 77.10 (p = 0.211)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table I.1.

In Table I.6, we determine whether CSR reporting and the presence of employee on the

board jointly affect firm market performance. We test this proposition using the joint test

technique. We first derive a dummy variable to represent high CSR reporting, by coding 1 for

63
firm-year observations with CSR reporting greater than the median (47.62%), and 0

otherwise. We then measure the marginal effect of higher CSR reporting on market

performance in firms with emplouee directors by testing the effect of the sum of the

coefficients (HCSR Reporting + HCSR Reporting*EMPL_BOARD) on market value. Model

1 of Table I.6 shows a positive impact of high CSR reporting on Tobin’s q. These results

confirm that shareholders are concerned about CSR information contained in annual or

standalone sustainability reports. Qui et al. (2016) and Cormier, Aerts, Ledoux and Magnan

(2009) document a positive and significant effect of CSR reporting on firm value.

In Model 2 of Table I.6, we determine how high CSR reporting and employee board

representation affect firm performance in combination by including the interaction between

high CSR reporting and employee board representation (HCSR Reporting* EMPL_BOARD).

We then conduct a joint test of the coefficient for HCSR Reporting and HCSR Reporting*

EMPL_BOARD. In line with Huse et al. (2009), who document that employee board

representation may specifically contribute to CSR, Model 2 in Table I.6 shows a strongly

positive and highly significant impact of the interaction term (HCSR_REP *EMPL_BOARD)

on Tobin’s q. These results suggest that a higher level of CSR reporting, when provided by

firms with at least one employee on the board, is rewarded by a higher valuation in the

financial markets. The marginal effect of employee directors on the value relevance of a

higher level of CSR reporting is assessed by the joint test of the sum of the coefficient

(HCSR_REP + HCSR_REP *EMPL_BOARD) on Tobin’s q. For Tobin’s q, the joint

coefficient is positive (1.393) and significant at the 1% level (z = 5.43). In accordance with

hypothesis H1a, we find that market participants value a higher level of CSR reporting

provided by firms with at least one employee on the board.

64
Table I. 6: System GMM Regression of Tobin's q on Higher CSR Reporting and
Employee Board Representation
Variables Model 1 Model 2

Coef. t–test Coef. t–test

Lag Tobin’s Q 0.128*** 16.73 0.107*** 10.12
HCSR_REP 0.141*** 4.15 –0.213*** –3.03
EMPL_BOARD –0.085 –0.81 –1.188*** –3.61
HCSR_REP *EMPL_BOARD 1.606*** 5.64
CSR_VERIF –0.006 –0.15 –0.047 –1.08
CSR_COM –0.048* –1.66 –0.089** –2.86
BOARD_SIZE –0.213*** –4.87 –0.185** –2.37
BOARD_IND –0.153* –1.96 –0.211** –2.31
BOARD_MEET –0.061* –1.40 –0.068 –1.56
DUAL –0.201*** –5.67 –0.150*** –3.62
TENURE 0.025 0.76 –0.029 –0.75
FAM_OWN 0.496*** 4.20 0.453*** 4.01
INST_OWN –0.134 –1.49 –0.225 –1.50
EMPL_OWN –1.84*** –3.07 –1.375 –1.47
LEV –0.521*** –5.50 –1.034*** –8.62
BETA 0.237*** 5.30 0.121* 1.90
FOR_ASS –0.281*** –4.00 –0.086 –1.05
R&D 1.113*** 3.50 0.487 1.08
SIZE 0.001 0.05 0.014 0.62
Intercept 1.291*** 5.04 1.622*** 5.92
Year Yes Yes
Industry Yes Yes
Number of observations 790 790
Fisher (Prob. > F) 12774 (p = 0.000) 5170.61 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.88 (p = 0.004) –2.92 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): 0.34 (p = 0.737) 0.39 (p = 0.698)
Sargan test (Chi–square, p–value): 736.84 (p = 0.000) 669.35 (p = 0.000)
Hansen test (Chi–square, p–value): 75.10 (p = 0.259) 67.11 (p = 0.473)
Joint test: HCSR_REP + (HCSR_REP * EMPL_BOARD) 1.393*** 5.43
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table I.1.

4.4. Supplementary Analysis

The preceding results are obtained for firms based on their combined score of CSR reporting.

We now analyse whether different CSR disclosure score categories (environmental, social and

sustainability) lead to different results. Indeed, the value relevance of CSR information

depends on the specific type of information provided (Qiu et al., 2016; Verbeeten et al.,

2016). Accordingly, we examine the effect of employee board representation on the value-

relevance of each dimension of CSR reporting. We test this proposition using the joint test

technique. We first derive a dummy variable to represent each specific type of high CSR

65
information provided. Table I.7 shows the value relevance of a high level for each category of

CSR reporting. Results suggest that shareholders have different sensitivity levels with regard

to CSR components.

Model 1 in Table I.7 reports a positive and significant impact of high social disclosure

on Tobin’s q, suggesting that the more social information the firm discloses, the greater its

commitment to its employees and other stakeholders, which in turn could lead the market to

expect increased future cash flows for the firm (Qiu et al., 2016). Another explanation is that

social disclosure provides information on the firm’s human capital, which consequently

impacts its future performance (Flammer, 2015). Furthermore, extensive social disclosure

helps create a competitive advantage that can enhance firm value. Extensive social reporting

may also reflect good labour relations and thus reduce potential strikes.

Model 2 in Table I.7 examines the value relevance of high environmental disclosure and

shows a positive impact of high environmental disclosure on Tobin’s Q. Environmental

disclosure provides information on the firm’s interaction with the environment and how it

contributes, for example, to climate change or waste management. This result suggests that

firms with high environmental disclosure are viewed favourably by investors (Radhouane,

Nekhili, Nagati, & Paché, 2018). Indeed, reporting on environmental activities reflects a

positive commitment to the environment and the avoidance of future costly disasters that

could reduce firm value.

However, in contrast to social and environmental disclosure, shareholders assign a

negative value to a higher level of sustainability information disclosed by firms. One

explanation could be that sustainability-related information as described in Appendix I.1 is

ambiguous, thus leading investors to misinterpret the information provided (Nekhili et al.,

2017b).

66
Table I. 7: System GMM Regression of Tobin's q on Higher Social, Environmental and
Sustainability Reporting and Employee Board Representation
Variables Model 1 Model 2 Model 3
Coef. t–test Coef. t–test Coef. t–test
Lag Tobin’s Q 0.056*** 5.65 0.124*** 15.72 0.218*** 18.14
HSOCIAL_REP 0.094** 2.15
HENVIR_REP 0.188*** 5.00
HSUST_REP –0.205*** –3.53
EMPL_BOARD –0.775*** –6.22 –0.137* –1.83 0.032 0.36
CSR_VERIF 0.062 1.33 0.001 0.04 0.011 0.45
CSR_COM –0.024 –0.68 –0.057** –2.29 0.021 0.85
BOARD_SIZE –0.048 –0.72 –0.195*** –4.60 –0.256*** –7.58
BOARD_IND –0.225*** –2.64 –0.149*** –2.64 0.010 0.19
BOARD_MEET –0.015 –0.42 –0.051* –1.85 –0.098*** –3.16
DUAL –0.132*** –3.00 –0.195*** –5.06 –0.115*** –3.82
TENURE 0.085*** 2.92 0.026 0.92 0.008 0.29
FAM_OWN 0.211** 2.31 0.493*** 4.77 0.525*** 6.27
INST_OWN –0.271*** –2.88 –0.074 –0.97 0.000 0.00
EMPL_OWN 0.352 0.43 –1.616*** –3.33 –1.442*** –2.75
LEV –0.573*** –4.50 –0.484*** –5.80 –0.399*** –5.11
BETA 0.327*** 5.31 0.271*** 5.28 0.218*** 4.44
FOR_ASS –0.379*** –4.87 –0.265*** –4.49 –0.167*** –3.53
R&D 1.708*** 4.23 1.108*** 2.76 1.003*** 3.06
SIZE 0.033** 2.01 –0.001 –0.01 0.020 1.30
Intercept 0.443* 1.66 1.128*** 4.64 1.199*** 5.45
Year Yes Yes Yes
Industry Yes Yes Yes
Number of observations 790 790 790
Fisher (Prob. > F) 1436.21 (p = 0.000) 13873.03 (p = 0.000) 12355.82 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.86 (p = 0.004) –2.87 (p = 0.004) –2.91 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): 0.10 (p = 0.923) 0.36 (p = 0. 0.719) 0.81 (p = 0. 0.418)
Sargan test (Chi–square, p–value): 696.56 (p = 0.000) 740.44 (p = 0.000) 726.31 (p = 0.000)
Hansen test (Chi–square, p–value): 70.83 (p = 0.290) 75.00 (p = 0.186) 76.78 (p = 0.218)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table I.1.

In Table I.8, we measure the marginal effect of employee board representation on the

value relevance of a higher level of each component of CSR reporting (social reporting,

environmental reporting and sustainability reporting). For social reporting, our empirical

results of Model 1 in Table I.8 show that both the interaction term

(HSOCIAL_REP*EMPL_BOARD) and the joint coefficient (HSOCIAL_REP +

HSOCIAL_REP*EMPL_BOARD) are not significant, suggesting that high social reporting,

as compared to environmental or to sustainability reporting, reflects more advantages for

employees at the expense of shareholders (Bøhren & Strøm, 2010) and a higher level of social

67
reporting when employees are represented on the board will be financially penalized in terms

of the market value of the firm.

Models 2 and 3 in Table I.8 show a positive and significant link between the interaction

variables (HENVI_REP*EMPL_BOARD) and (HSUST_REP*EMPL_BOARD) and Tobin’s

q, suggesting that the market positively assesses high environmental and sustainability

reporting when employees are included on the board. A possible explanation is that, since

employees are more attached to CSR activities than other directors on the board, investors

may interpret the environmental and sustainability information disclosed as more credible.

The joint coefficients in Model 2 (HENVI_REP + HENVI_REP*EMPL_BOARD) and Model

3 (HSUST_REP + HSUST_REP*EMPL_BOARD) are positive and significant on the market-

based performance, indicating that a higher level of environmental and sustainability

reporting, when provided by firms with at least one employee on the board, are financially

rewarded by market participants.

68
Table I. 8: System GMM Regression of Tobin's q on the Interaction between Higher
Social, Environmental and Sustainability Reporting and Employee Board
Representation
Variables Model 1 Model 2 Model 3
Coef. t–test Coef. t–test Coef. t–test
Lag Tobin’s Q 0.060*** 5.78 0.097*** 9.88 0.148*** 10.65
HSOCIAL_REP 0.151*** 2.56
HENVIR_REP –0.017 –0.30
HSUST_REP –0.670*** –8.45
EMPL_BOARD –0.622*** –3.74 –0.709*** –3.54 –2.836*** –7.29
HSOCIAL_REP *EMPL_BOARD –0.358 –1.39
HENVIR_REP *EMPL_BOARD 0.725*** 3.81
HSUST_REP *EMPL_BOARD 3.277*** 7.52
CSR_VERIF 0.087 1.77 –0.004 –0.10 –0.128*** –3.45
CSR_COM –0.041 –1.08 –0.104*** –3.74 –0.027 –0.68
BOARD_SIZE –0.025 –0.36 –0.138** –2.53 –0.205*** –2.83
BOARD_IND –0.214** –2.47 –0.210*** –3.03 –0.334*** –3.07
BOARD_MEET 0.002 0.06 –0.059** –2.07 –0.164*** –5.15
DUAL –0.130*** –2.91 –0.190*** –4.95 –0.031 –0.73
TENURE 0.089*** 3.02 0.036 1.44 0.025 0.82
FAM_OWN 0.175* 1.81 0.463*** 4.76 0.524*** 4.38
INST_OWN –0.285*** –2.98 –0.077 –1.00 0.007 0.05
EMPL_OWN 0.668 0.79 –1.328** –2.37 –2.685** –2.42
LEV –0.524*** –4.05 –0.681*** –7.63 –1.059*** –13.08
BETA 0.359*** 5.59 0.319*** 5.90 0.115 1.59
FOR_ASS –0.449*** –4.89 –0.220*** –2.88 –0.181** –2.02
R&D 2.001*** 4.26 1.568*** 3.71 0.811* 1.96
SIZE 0.040** 2.23 0.003 0.15 0.063*** 3.60
Intercept 0.179 0.54 1.624*** 5.02 1.823*** 5.28
Year Yes Yes Yes
Industry Yes Yes Yes
Number of observations 790 790 790
Fisher (Prob. > F) 32162.52 (p = 0.000) 3323.06 (p = 0.000) 13127.25 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.84 (p = 0.004) –2.88 (p = 0.004) –2.92 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): 0.12 (p = 0.906) 0.35 (p = 0. 0.725) 0.78 (p = 0. 0.434)
Sargan test (Chi–square, p–value): 694.82 (p = 0.000) 740.11 (p = 0.000) 564.881 (p = 0.000)
Hansen test (Chi–square, p–value): 69.59 (p = 0.295) 73.99 (p = 0.184) 74.06 (p = 0.259)
Joint test: HSOCIAL_REP –0.207 –0.93
+ (HSOCIAL_REP * EMPL_BOARD)
Joint test: HENVIR_REP 0.708*** 4.37
+ (HENVIR_REP * EMPL_BOARD)
Joint test: HSUST_REP 2.607*** 5.50
+ (HSUST_REP * EMPL_BOARD)
Note: *, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table I.1.

5. Conclusion

Setting out from the idea that not all stakeholders on the board have the same attachment to

CSR and that employee directors may use their power on boards to defend their interests at

the expense of shareholders, the presence of employee directors on the board may impact

stakeholders’ perception of CSR reporting. The present study is designed to examine the

69
extent to which the value relevance of voluntary CSR reporting is impacted by the

appointment of employees to boards in the French context.

Using a sample of French firms listed in the SBF120 from 2001 to 2011, we found that

firms with employee directors report more on their CSR activities than their counterparts. We

find that the value relevance of high CSR reporting is enhanced by the presence of employee

directors on the board. However, the value relevance of CSR information depends on the type

of information provided. Accordingly, we carried out supplementary analyses to assess the

impact of employee directors on the value relevance of each category of CSR disclosure. Our

results indicate that employee directors enhance the value relevance of both environmental

and sustainability reporting. Nonetheless, we find that investors do not value a high level of

social reporting in firms with employee directors on the board. A possible explanation is that

higher social reporting by such firms could indicate either over-investment in social duties to

the advantage of employees and a possible alliance between employees and managers to

counterbalance shareholder supremacy on the board.

This study has thrown up many questions in need of further investigation. First, we

believe that it would be interesting to carry out a cross-country study to assess the impact of

the representation of employee directors on the value relevance of CSR reporting. More

information on the impact of employee directors on the value relevance of CSR before and

after the adoption of the Grenelle II Act would also help to establish a greater degree of

accuracy on this subject. Finally, it will be with importance to investigate the type and nature

of decisions that employee directors may impact the most. Indeed, Huse et al. (2009)

document that employee directors influence the strategic control task of the board which is

often of a qualitative nature, but they do not impact the budget control task which is often of a

quantitative nature.

70
References

Aguinis, H., & Glavas, A. (2012). What we know and don’t know about corporate social
responsibility: A review and research agenda. Journal of management, 38(4), 932–968.
Allix-Desfautaux, E., & Makany, L. (2015), Développement durable et gestion d’une
entreprise : croisements fertiles, Management & Avenir, 81, 15–36.
Andre, J.M., Husser, J., Barbat, G., & Lespinet-Najib, V. (2011), Le rapport de
développement durable des entreprises françaises: quelles perspectives pour les parties
prenantes?, Management & Avenir, 48, 37–56.
Bagdi, K. (2015). The future of employees’ board-level representation in the European Union.
Procedia Economics and Finance, 23, 1394–1400.
Barnea, A., & Rubin, A. (2010). Corporate Social Responsibility as a conflict between
shareholders. Journal of Business Ethics, 97, 71–86.
Bebbington, J., Larrinaga, C., & Moneva, J. M. (2008). Corporate social reporting and
reputation risk management. Accounting, Auditing & Accountability Journal, 21, 337–361.
Bhattacharya, C. B., Sen, S., & Korschun, D. (2008). Corporate social responsibility as an
internal marketing strategy. MIT Sloan Management Review, 49, 37–44.
Bøhren, O., & Strøm, R. O. (2010). Governance and politics: Regulating independence and
diversity in the board room. Journal of Business Finance and Accounting, 37, 1281–1308.
Botosan, C. A. (1997). Disclosure level and the cost of equity capital. The Accounting Review,
72, 323–350.
Cahan, S. F., De Villiers, C., Jeter, D. C., Naiker, V., & Van Staden, C. J. (2016). Are CSR
disclosures value relevant? Cross-country evidence. The European Accounting Review, 25,
579–611.
Christmann, P., & Taylor, G. (2006). Firm self-regulation through international certifiable
standards: Determinants of symbolic versus substantive implementation. Journal of
International Business Studies, 37(6), 863–878.
Cormier, D., Aerts, W., Ledoux, M. J., & Magnan, M. (2009). Attributes of social and human
capital disclosure and information asymmetry between managers and investors. Canadian
Journal of Administrative Sciences, 26, 71–88.
Ehnert, I., Parsa, S., Roper, I., Wagner, M., & Muller-Camen, M. (2016). Reporting on
sustainability and HRM: A comparative study of sustainability reporting practices by the
world's largest companies. The International Journal of Human Resource Management,
27, 88–108.

71
Faleye, O., Mehrotra, V., & Morck, R. (2006). When labor has a voice in corporate
governance. Journal of Financial and Quantitative Analysis, 41, 489–510.
Flammer, C. (2015). Does corporate social responsibility lead to superior financial
performance? A regression discontinuity approach. Management Science, 61, 2549–2568.
Fuente, J. A., García-Sánchez, I. M., Lozano, M. B. (2017). The role of the board of directors
in the adoption of GRI guidelines for the disclosure of CSR information. Journal of
Cleaner Production, 141, 737–750.
Galbreath, J. (2010). Corporate governance practices that address climate change: An
exploratory study. Business Strategy & the Environment, 19, 335–350.
Gillet, C. (2012). A study of sustainability verification practices: the French case. Journal of
Accounting & Organizational Change, 8, 62–84.
Ginglinger, E., Megginson, W., & Waxin, T. (2011). Employee ownership, board
representation, and corporate financial policies. Journal of Corporate Finance, 17, 868–
887.
Guedri, Z., & Hollandts, X. (2008). Beyond dichotomy: The curvilinear impact of employee
ownership on firm performance. Corporate governance: An international review, 16, 460–
474.
Haddock-Millar, J., Sanyal, C., & Müller-Camen, M. (2016). Green human resource
management: A comparative qualitative case study of a United States multinational
corporation. The International Journal of Human Resource Management, 27, 192–211.
Haniffa, R., & Cooke, T. (2005). The impact of culture and governance on corporate social
reporting. Journal of Accounting & Public Policy, 24, 391–430.
Hur, W. M., Moon, T. W., & Ko, S. H. (2016). How employees’ perceptions of CSR increase
employee creativity: Mediating mechanisms of compassion at work intrinsic motivation.
Journal of Business Ethics, 138, 1–16.
Huse, M., Nielsen, S. T., & Hagen, I. M. (2009). Women and employee-elected board
members, and their contributions to board control tasks. Journal of Business Ethics, 89,
581–97.
Jensen, M. C., & Meckling, W. H. (1979). Rights and production functions: An application to
labor-managed firms and codetermination. Journal of Business, 52, 469–506.
Lewis, B. W., Walls, J. L., & Dowell, G. W. (2014). Difference in degrees: CEO
characteristics and firm environmental disclosure. Strategic Management Journal, 35(5),
712–722.

72
Lins, K. V., Servaes, H., & Tamayo, A. (2017). Social capital, trust, and firm performance:
The value of corporate social responsibility during the financial crisis. The Journal of
Finance, 72, 1785–1824.
Kleinknecht, R. H. (2015). Employee participation in corporate governance: Implications for
company resilience. European Journal of Industrial Relations, 21, 57–72.
KPMG (2008). KPMG international survey of corporate sustainability reporting.
[Link]
Kramar, R. (2014). Beyond strategic human resource management: Is sustainable human
resource management the next approach?. The International Journal of Human Resource
Management, 25, 1069–1089.
Nekhili, M., Nagati, H., Chtioui, T., & Rebolledo, C. (2017a). Corporate social responsibility
disclosure and market value: Family versus nonfamily firms. Journal of Business
Research, 77, 41–52.
Nekhili, M., Nagati, H., Chtioui, T., & Nekhili, A. (2017b). Gender-diverse board and the
relevance of voluntary CSR reporting. International Review of Financial Analysis, 50, 81–
100.
O'Donohue, W., & Torugsa, N. (2016). The moderating effect of ‘Green’ HRM on the
association between proactive environmental management and financial performance in
small firms. The International Journal of Human Resource Management, 27, 239–261.
Poulain-Rehm, T., & Lepers, X. (2013). Does employee ownership benefit value creation?
The case of France (2001–2005). Journal of Business Ethics, 112, 325–340.
Preuss, L., Haunschild, A., & Matten, D. (2009). The rise of CSR: implications for HRM and
employee representation. The International Journal of Human Resource Management, 20,
953–973.
Qiu, Y., Shaukat, A., & Tharyan, R. (2016). Environmental and social disclosures: Link with
corporate financial performance. The British Accounting Review, 48, 102–116.
Radhouane, I., Nekhili, M., Nagati, H., & Paché, G. (2018). The impact of corporate
environmental reporting on customer-related performance and market value. Management
Decision, 56, 1630–1659.
Richard, O. C., & Johnson, N. B. (2001). Strategic human resource management effectiveness
and firm performance. International Journal of Human Resource Management, 12, 299–
310.
Roodman, D. (2009). How to do xtabond2: An introduction to difference and system GMM in
stata. The Stata Journal, 9, 86–136.

73
Sahed-Grangery, Y., & Boncori, A.L. (2014), La traduction des attentes des parties prenantes
en RSE en indicateurs de la Global Reporting Initiative : vers un mode consultatif de la
gestion des parties prenantes, Management & Avenir, 68, 73–91.
Verbeeten, F. H. M., Gamerschlag, R., & Möller, K. (2016). Are CSR disclosures relevant for
investors? Empirical evidence from Germany. Management Decision, 54, 1359–1382.
Zhu, Y., Xie, Y., Warner, M., & Guo, Y. (2015). Employee participation and the influence on
job satisfaction of the ‘new generation’ of Chinese employees. The International Journal
of Human Resource Management, 26, 2395–2411.

74
 Appendix I. 1: Items of Grenelle II Act
Components Description
1 Social Reporting (19 items)
1.1 Employment 1.1.1 Number of employees and how they are subdivided according to age, gender and
geographic distribution (based on numbered data and diagram)
1.1.2 Hiring and firing
1.1.3 Remuneration and it evolution
1.2 Organization of work 1.2.1 Organization of working time (flexibility of working hours, weekly working hours,
etc.)
1.2.2 Absenteeism
1.3 Labour relations 1.3.1 Social dialogue (information procedures, consultation of staff and
negotiating with employers)
1.3.2 Outcome of collective agreements
1.4 Occupational Health and 1.4.1 Health and safety conditions at work
safety 1.4.2 Outcome of the collective agreements signed with trade unions and the staff
representatives in terms of occupational health and safety
1.4.3 Frequency and seriousness of accidents
1.5 Training 1.5.1 Policies implemented with regard to training
1.5.2 Total number of training hours
1.6 Equal treatment 1.6.1 Measures promoting equality between women and men
1.6.2 Measures promoting employment and integration of people with disabilities
1.6.3 Policy against discrimination
1.7 Respect of the clauses of 1.7.1 Respect for the right to organize and collective bargaining
fundamental conventions of 1.7.2 Abolition of discrimination in employment and occupation
the International Labour 1.7.3 Abolition of forced or compulsory labour
Organization (ILO) 1.7.4 Abolition of child labour
2 Environmental Reporting (14 items)
2.1 Environmental policy 2.1.1 Organization of the company to take into account environmental concerns,
and, if applicable, environmental evaluation and verification approaches
2.1.2 Training and information towards employees on environmental protection
2.1.3 Budget devoted to environmental protection and environmental risk mitigation
2.1.4 Financial provisions for environmental risks
2.2 Pollution and 2.2.1 Prevention, reduction and fixing of air/water/soil emissions
Waste Management 2.2.2 Prevention, recycling and cutting waste
2.2.3 Noise pollution and other type of pollution
2.3 Sustainable use 2.3.1 Water consumption and supply considering local resources
of resources 2.3.2 Consumption of raw materials and measures taken to improve the efficiency of raw
materials use
2.3.3 Energy consumption and measures to improve energy efficiency and the use of
renewable energy
2.3.4 Land use
2.4 Climate change 2.4.1 Greenhouse gas emissions
2.4.2 Measures to adapt to climate change
2.5 Protection of biodiversity 2.5.1 Measures taken to save and develop biodiversity
3 Sustainability reporting (9 items)
3.1 Territorial, economic and 3.1.1 Measures in favour of environment, employment and regional development
social impact of the activity 3.1.2 Measures taken in relation to people living in the area around the business
3.2 Relationships with 3.2.1 Conditions for dialogue with stakeholders
stakeholders 3.2.2 Measures promoting partnership or sponsorship
3.3 Subcontracting 3.3.1 Amount of subcontracting
and suppliers 3.3.2 Taking into account social and environmental responsibility with suppliers and
subcontractors
3.4 Honesty in practices 3.4.1 Measures to prevent corruption
3.4.2 Measures in favour of consumers’ health and safety
3.5 Measures in favour of 3.5.1 Measures preventing all forms of discrimination and promoting equal treatment
human rights

75
76
 Chapter II: ESG Performance and Market Value:
The Moderating Role of Employee Board
Representation

1. Introduction

In an interview published in the French business magazine L’Usine Nouvelle, Antoine Frérot,

CEO of Veolia, one of the biggest SBF 120 French firms, said, “Corporate social

responsibility (CSR) must be considered at the same level as the creation of economic value”.

This statement acknowledges the need for firms to include CSR in their objectives and to

view them as equally important as financial objectives (Lydenberg, 2013). Moreover,

achieving good financial and extra-financial performance is essential if firms are to gain

competitive advantage and attract investors, particularly socially responsible investors

(Lydenberg, 2013).1 Environmental, social and governance (ESG) ratings help investors make

effective responsible investments and anticipate long-term performance advantages (Crifo &

Mottis, 2013). In this context, ESG ratings seem to attract the attention of French

shareholders, as ESG criteria figure among the major points to be discussed during the 2019

general assembly of SBF 120 French firms.2

ESG performance is an important aspect of corporate strategy, which reflects corporate

ability to maintain good relations with its various stakeholders (Friede, Bush & Bassen,

2015). Starting in 1970, a broad range of research has aimed to examine the connection

between ESG performance and financial performance (hereafter ESG-FP). Though previous

1
The concept of CSR has made great progress, affecting both academics' work and corporate behaviour. At this
stage, Moura-Leite and Padgett (2011) argue that whereas CSR was coupled with strategy literature during
the 1990s, CSR became definitively an important strategic issue in the 2000s.
2
[Link]
2019_638610

77
researchers claim that results on the ESG-FP relationship are ambiguous, inconclusive and

contradictory (Tang, Hull, & Rothenberg, 2012), Friede et al. (2015) find that recent

aggregated evidence from 2200 empirical studies suggest that there is a positive relationship.

Moreover, in a more recent literature review, Brooks and Oikonomou (2018) suggest overall

that high ESG performers are more likely to achieve better financial performance than low

ESG performers.

For better ESG performance, some authors recommend the adoption of stakeholder-

oriented boards (Shaukat, Qiu & Trojanowski, 2016), in that the involvement of stakeholders

in the decision-making process adds value to their contribution toward organizational

goals and changes the market perception of ESG performance. A major change in corporate

governance in recent years has been the growing awareness of the importance of the presence

of employee representatives on the board (Huse, Nielson & Hagen, 2009; Ginglinger,

Megginson & Waxin, 2011). However, from the standpoint of shareholders, high ESG

performance when employees are represented on the board may lead to two different

outcomes. On the one hand, employee directors may have a positive impact on ESG

performance, since they may play an important role in CSR decisions (Preuss, Haunschild &

Matten, 2009) and because they are more concerned about socially responsible activities than

other board members (Huse et al., 2009). On the other hand, good ESG performance may

reflect the high impact of employees on the board, since they are the primary beneficiaries of

ESG investment (Barnea & Rubin, 2010). Bøhren and Strøm (2010) argue that employees on

the board may over-use their power and lobby for the maximization of their own interests.

Furthermore, good ESG performance when employees are on the board might stem from a

managerial entrenchment strategy. Indeed, Pagano and Volpin (2005) point out that managers

and employees are natural allies in the firm. Consequently, ESG performance may be used by

78
managers to obtain personal benefits at the expense of shareholders, who bear the underlying

costs (Barnea & Rubin, 2010; Berrone & Gomez-Mejia, 2009; Surroca & Tribó, 2008).

In this chapter, we raise the question of how investors perceive ESG performance and

their supporting pillars (social, environmental and governance), depending on whether or not

employees are represented on the board of directors. We capture the market perception of

ESG performance through the relationship between ESG ratings and firms’ market value

measured by Tobin’s q. We then identify the moderating impact of employee directors by

comparing the value relevance of ESG performance for firms with and without employee

representation on the board. We control for endogeneity of the relationship between ESG

performance and firm performance by using an appropriate econometric specification. To our

knowledge, the present study is the first to examine the ESG-FP relationship using the

moderating role of employee directorship.

Using a sample of French firms listed on the SBF 120 for the period from 2007 to 2017,

we provide new evidence regarding the impact of employee board representation on the value

relevance of ESG performance. Our primary finding is that investors react negatively to the

representation of employees on board. Importantly, our regression analyses show a negative

relationship between market value and ESG performance when employees are represented on

the board. The marginal effect of employee board representation on the value relevance of

ESG pillars (i.e., social, environmental and governance) is negative. In other words, we find

that, when employees are represented on the board, neither social nor environmental and

governance performance are relevant for investors. These findings support previous evidence

on the conflict of interest between shareholders and employees and suggest that higher ESG

performance may reflect an alliance between managers and employee directors that

counterbalances shareholders’ power on the board. Moreover, high ESG performance may

79
result in more advantages for employees at the expense of shareholders and may lead to

greater entrenchment on the part of managers.

Our study makes a number of contributions to the literature on both employee board

representation and the value relevance of ESG performance, thus helping to make good

existing perceived shortcomings. First, we show that the use of moderating variables, such as

employee board representation, enhances understanding of the ESG-FP relationship. Second,

we present new evidence on the conflict of interest between shareholders and employees with

respect to CSR, and highlight the fact that shareholders’ perception with respect to ESG

pillars differs according to whether or not employees are represented on the board.

The chapter is organized as follows. We start by examining the impact of employee

board representation on the firm’s market value. We then question the role of employee

directors in moderating the value relevance of ESG performance. Next, after describing the

sample and the methodology used, we present and discuss the empirical findings. Finally, we

present our conclusion and suggest new research perspectives.

2. Conceptual Framework and Hypotheses Development

2.1. Employee directorship and firm value

Employee representation on boards varies widely around the world and is more likely to be a

European phenomenon (OECD, 2017).3 While a high proportion is mandatory in Germany,

some countries’ regulations, such as those of United Kingdom and United States of America,

do not require there to be employee representation on boards. In France, an intermediate level

3
The 2017 edition of the OECD report on Employment Outlook states that in some countries such as Austria,
Denmark, Finland, France, Germany, Hungary, Luxembourg, Netherlands, Norway, Poland, the Slovak
Republic, Slovenia, and Sweden, companies above a certain size are required to have employee representatives
on the board. This is the case for firms with more than 500 employees in Germany, more than 300 employees in
Austria, more than 35 employees in Denmark, more than 30 employees in Norway and more than 25 employees
in Sweden. In other countries, such as Chile, Greece, Ireland, Poland, Portugal and Spain, employees may be
represented on the boards only of state-owned enterprises.

80
of employee directors is mandatory. Inspired by the German codetermination system,4 the

French law mandates that employees in large publicly listed firms can elect two types of

directors to represent them on the board: the first are elected by employees by virtue of their

right of employment and the second are elected by employee-shareholders (Ginglinger et al.,

2011).5

The appointment of employees to boards has several benefits for the firm. Employee

participation on the board enhances its efficiency by allowing diverse and valuable

information to be shared with the board and by having a variety of backgrounds, which

improves the quality of decision-making (Huse et al., 2009). Zhu, Xie, Warner, and Guo

(2015) consider that when employees participate in the decision-making process, their

satisfaction level increases and they make greater discretionary efforts. Employee

representation on the board may be also viewed as a way of maintaining harmonious relations

with trade unions and increasing workers’ protection, while at the same time acting as

motivation for workers. Bagdi (2015) confirms that representing employees on the board is

beneficial for firms in times of crisis, and shows that, during the last international financial

crisis, firms which strengthened participatory rights of employees on the board experienced

fewer negative effects from the crisis and even managed to remain profitable. Kleinknecht

(2015) documents that employee participation on the board increases firms’ resilience when

they suffer negative shocks. In their study of the French context, Ginglinger et al. (2011)

4
It is, however, worth noting that Germany has a dual-board system with two strictly separated administrative
bodies: the supervisory board (non-executive directors only) and the management board (executive directors
only). Employees in Germany thus sit on the supervisory board but not on the board of directors. In France,
firms may adopt either a board of directors or a dual system with an executive board and a supervisory board.
5
Compared to other European countries, France has shown great interest in employee board representation. The
first French legislation stipulating labour representation on boards focused on state controlled firms (the law of
1983). Eleven years later, regulators start to pay attention to private companies, allowing them to reserve board
seats for employees and employee shareholders subject to acceptance by shareholders at the general assembly
(the law of 1994). Starting in 2006, representation of employee shareholders became mandatory for publicly
listed French firms when employees hold at least 3% of the total shares. As for directors elected by employees
by their right of employment, French boards have been obliged to reserve seats for employee directors since
2013.

81
argue that the presence of directors elected by employee-shareholders enhances firms’ market

value.

There is also counter evidence suggesting that labour participation in the corporate

governance process may be perceived negatively by shareholders. In an early study, Jensen

and Meckling (1979) argue that employees and shareholders have different interests and that

worker representation on the board may be a source of inefficiency in self-managed firms. In

their study of Norwegian listed firms over the period 1989-2002, Bøhren and Strøm (2010)

find that employee directors may have the power to divert the board from shareholder value

maximization to their own advantage. Employee directors may exert pressure on the board to

defend their interests and obtain higher salaries at the expense of shareholders, who bear the

associated cost (Jensen & Meckling, 1979). Similarly, Ginglinger et al. (2011) show that the

institutional participation of employees on boards significantly reduces the dividend payout

ratio. Faleye, Mehrotra and Morck (2006) argue that firms with directors elected by

employee-shareholders invest less in long-term assets, create fewer jobs, take fewer risks,

employ fewer workers and grow more slowly. Last but not least, Guedri and Hollandts (2008)

confirm that employees’ representation on boards facilitates their lobbying for the

maximization of their own interests to the detriment of shareholders’ value maximization.

From the standpoint of shareholders, other arguments against employee directorship can

be found in the literature. Aubert, Garnotel, Lapied and Rousseau (2014) argue that managers

use employee share ownership (ESO) plans as an entrenchment tool. More particularly, they

argue that in order to retain their jobs, low-performing managers tend to offer stock

ownership to workers. Accordingly, Hollandts, Aubert, Abdelhamid and Prieur (2018) have

studied the moderating impact of employee shareholders’ representation on the board on the

ESO-CEO relationship for French firms. They argue that the presence of directors elected by

employee shareholders increases the impact of ESO on CEO entrenchment. Similarly,

82
previous researchers have shown that employee representation on the board is likely to reduce

the board’s monitoring effectiveness (Jensen & Meckling, 1979). Due to the hierarchical

dependency relationship between employees and managers, employee directors may facilitate

amicable monitoring of CEOs (Guedri & Hollandts, 2008). Along similar lines, Bøhren and

Strøm (2010) confirm that having non-executive employees on the board facilitates their

lobbying for the maximization of their own interest. In such cases, a close alliance between

managers and employees against shareholders can be identified (Pagano & Volpin, 2005).

2.2. The moderating role of employee directorship

Unlike the traditional view of the firm which views shareholders as the only stakeholders that

matter, stakeholder theory suggests that all stakeholders’ needs should be taken into account

when initiating action (Freeman, 1984). The reason for this is that stakeholders, as “persons or

groups that have, or claim, ownership, rights, or interest in a corporation and its activities,

past, present, or future” (Clarkson, 1995 p. 106), are crucial for the continuity of the firm.

From this perspective, stakeholder theory is arguably the most appropriate theory for ESG

studies (Weber & Gladstone, 2014, Wood & Jones, 1995). Dunn and Sainty (2009) advance

the idea that the essence of corporate social performance is recognition that a firm has

multiple duties towards various stakeholders over the long term. Hence, ESG performance

may reflect the capacity of the firm to manage relations with its stakeholders. Accordingly,

the better the firm’s ESG performance, the more it is able to maintain good relations with its

stakeholders and the more likely it is to be profitable, in particular when targeting its primary

stakeholders, namely employees, consumers and investors (Hillman & Keim, 2001). Based on

a stakeholder framework, Clarkson (1995) defines the firm as “a system of primary

stakeholder groups, a complex set of relationships between and among interest groups, with

different rights, objectives, expectations, and responsibilities” (p. 107). A more fine-grained

examination of the ESG-FP relationship through the moderating role of stakeholder

83
management provides interesting results (Hillman & Keim, 2001; Van der Laan, Van Ees &

Van Witteloostuijn, 2008; Weber & Gladstone, 2014). Consistently with Clarkson (1995),

Van der Laan et al. (2008) shows that ESG performance targeting primary stakeholders is

more closely related to financial performance than ESG performance targeting secondary

stakeholders (i.e., community, diversity and human rights). Employees, as key stakeholders,

are of great importance with regard to the ESG-FP relationship. For instance, Weber and

Gladstone (2014) document that targeting in terms of ESG apparently improves financial

performance. Similarly, Berman, Wicks, Kotha and Jones (1999) show that maintaining good

relations with employees has a significant impact on firms’ financial performance. Qiu,

Shaukat and Tharyan (2016) find that investors pay more attention to social performance than

environmental performance because they expect economic returns from creating good

relations with employees and other stakeholders (such as the community and consumers). In

that respect, social investments may create competitive advantages for the firm by creating

intangible assets such as human capital and corporate reputation (Marsat & Williams, 2014).

Moreover, firms that pay greater attention to their workers’ needs are more likely to attract

talented job seekers (Jones, Willness & Madey, 2014). Employee satisfaction and productivity

is also likely to be influenced by the level of firms’ ESG performance. Gao and Yang (2016),

for example, examine the impact of corporate philanthropy on employees’ level of

productivity in Chinese firms, and find that, as a form of the corporate social responsibility, it

enhances employees’ productivity.

While earlier studies concentrated on the impact of strategic human resource

management on firms’ performance (Richard & Johnson, 2001), the positive role played by

socially and environmentally aware employees has been studied more recently (Ehnert, Parsa,

Roper, Wagner & Muller-Camen, 2016; Haddock-Millar, Sanyal & Muller-Camen, 2016;

Kramar, 2014; O’Donohue & Torugsa, 2015). Hur, Moon and Ko (2016) show that, when

84
employees have a positive perception of their firm’s CSR engagement, they become more

creative. Further, CSR initiatives could be an excellent “internal marketing program” for

motivating employees. Accordingly, employee directors may make specific contributions to

CSR tasks, in that they are more concerned about health, the environment and social issues

than other directors on the board and have the power to increase the firm’s CSR engagement

(Huse et al., 2009). Through three case studies from Belgium, Germany and France, Preuss et

al. (2009) analyze the implications of the rise of CSR on employee representation on the

board and on human resource management. Their qualitative study reports that employee

directors may play an active role in CSR. Moreover, employees are more dependent on the

long-term survival of the firm. Consequently, employee directors are likely to increase

investment in CSR activities, since the benefits from such investment are manifested in the

long term. Markey, McIvor and Wright (2016) show that employee directors may have the

power to reduce carbon emissions in their work places by developing and implementing

measures to control them.

Nevertheless, ESG performance may be also used by managers to obtain private

benefits at the expense of shareholders, who bear the associated costs (Barnea & Rubin, 2010;

Berrone & Gomez-Mejia, 2009; Surroca & Tribó, 2008). More specifically, ESG performance

may increase managerial entrenchment (Surroca & Tribó, 2008), CEO compensation (Berrone

& Gomez-Mejia, 2009) and conflicts of interest between shareholders and managers (Barnea

& Rubin, 2010). In order to protect their jobs and strengthen their position in the firm,

managers may opt for a collusion strategy with stakeholders other than shareholders (Surroca

& Tribó, 2008). For example, managers may offer high salaries to workers in order to gain

their protection against internal disciplining mechanisms (Surroca & Tribó, 2008). This

concession would be to the advantage of managers at the expense of shareholders. Moreover,

managers may choose to maintain good relations with the community in order to enhance

85
their personal reputation. Coffey and Wang (1998) find that managerial control on board

increases expenditure on corporate philanthropy. In such cases, high ESG performance could

be detrimental to the creation of value for shareholders.

Barnea and Rubin (2010) identify a conflict of interest between shareholders with

respect to CSR. They argue that affiliated shareholders (those influenced by the activity and

the reputation of the firm, such as managers, employees, large shareholders and directors) and

non-affiliated shareholders (such as institutions and small investors seeking profitable

investments in the short term) have different interests in relation to CSR. This conflict of

interest would be greater if employees are represented on the board, since employees are seen

as natural allies for managers (Pagano & Volpin, 2005).

While employee directors may be expected to have a positive impact on ESG

performance, leading in turn to better financial performance, the fact that employees are

considered to be natural allies for managers and may influence their decisions, employee

board representation may negatively moderate the relationship between ESG performance and

market value. In line with these arguments, we consider the following alternative hypotheses:

H1a: Employee board representation positively moderates the relationship between

ESG performance and market value.

H1b: Employee board representation negatively moderates the relationship between

ESG performance and market value.

3. Methodology

3.1. Sample and data

Our aim is to examine to what extent the value relevance of ESG performance is affected by

employee representation on the board. To this end, we use a sample of 91 French firms listed

in the SBF 120 (excluding financial, insurance and real estate companies) over a period of

eleven years (2007 to 2017). For ESG performance, we use the Thomson Reuters/S-Network

86
ESG Best Practice Ratings, which provide information on European firms from December

2007. Our period starts one year after employee shareholders’ representation on board became

mandatory in 2006 for publicly listed firms in cases where employees hold at least 3% of the

capital. Before 2006, information on employee board representation was scarce. Corporate

governance and ownership variables were collected from firms’ annual reports and standalone

reports, while financial and accounting data were obtained from the ThomsonOne database.

3.2. Dependent variable: Tobin’s q

Following Cahan, De Villiers, Jeter, Naiker and Van Staden (2016), Ginglinger et al. (2011)

and Nekhili, Nagati, Chtioui and Nekhili (2017), we use Tobin’s q as a measure of firm

performance. Tobin’s q reflects the market’s assessment of the long-term expected value of a

firm, as explained by Cahan et al. (2016). Accordingly, the benefits of ESG performance on

firm value are more likely to appear in the long term and be captured by Tobin’s q. For this

reason, we use Tobin’s q to reflect market participants’ perception of ESG performance. This

measure has a number of advantages. First, it is a favoured measure that in practice can be

used by investors and other market participants in making relevant decisions. Tobin’s q is also

a market-based instrument, and can therefore detect the perception of the market (by

investors) (Cahan et al., 2016) and is unaffected by accounting conventions.

3.3. Endogenous variable: ESG Performance

To measure ESG performance, we use the Thomson Reuters/S-Network ESG Best Practice

Ratings. This measure provides ratings regarding composite ESG performance and its three

pillars (environmental, social and governance) of over 5000 firms worldwide. Unlike

ASSET4 ratings, which use an equally weighted key performance indicators (KPIs), Thomson

Reuters/S-Network ESG Best Practice Ratings assign a specific weight for each KPI used to

measure the ESG performance. Assigning these specific weights enhances the quality of the

ESG performance measure, making it more consistent and objective.

87
3.4. Moderating variable: employee directorship

Employee representation varies from country to country. While a high proportion is the norm

in Germany, United Kingdom regulations do not require there to be any employee

representation on boards. In France an intermediate level of employee directors is mandatory.

Following Ginglinger et al. (2011) and Guedri and Hollandts (2008), we measure employee

board representation by the number directors elected by employees as a proportion of the total

number of directors on the board.

3.5. Control variables

Following previous empirical studies on ESG, we included various features of firms that may

influence ESG performance and firm performance: governance variables, ownership variables

and other control variables. Verification of CSR reports by independent third parties reflects

firms’ desire to enhance the credibility of their CSR information in the eyes of stakeholders

(Gillet, 2012). The existence of a board committee for CSR issues demonstrates the

commitment of top management towards the firm’s stakeholders, which increases its ESG

performance (Velte, 2016). Board independence provides better alignment with the social

views of external groups (Haniffa & Cooke, 2005), and is likely to influence the firm’s extra-

financial performance (Dunn & Sainty, 2009). For Nekhili, Nagati, Chtioui and Rebolledo

(2017a) and Nekhili, Nagati, Chtioui and Nekhili (2017b), the number of board meetings is a

proxy for diligence and an indicator of directors’ concerns, such as CSR duties. Similarly,

long-tenured CEOs are better able resist pressures and change, and are less likely than newly

appointed CEOs to acquiesce to stakeholder pressure. CEO/chair role duality gives CEOs

greater power, and may lead to CEOs disregarding further involvement in social and

environmental activities (Galbreath, 2010). Family owners are likely to have a positive impact

on ESG initiatives (Bingham, Dyer, Smith & Adams, 2010). Similarly, institutional holders

exert strong control over managerial activities and increase ESG the firm’s performance (Ben

88
Lahouel, Perretti & Autissier, 2014). We control for employee ownership because since it is

considered as an important tool for managers to serve their entrenchment strategy (Aubert et

al., 2014). We also follow Nekhili et al. (2017a, b) by controlling for company size, leverage

and R&D intensity. Previous studies dealing with ESG performance widely control for

industry. We use the Industry Classification Benchmark (ICB) developed by Dow Jones and

FTSE and used by Euronext. Lastly, we control for the labour law reforms of 2013, which

make it mandatory for employees to be represented on the board by one director if there are

fewer than 12 board members, and by two directors if the board has more than 12 members in

the case of firms with at least 5,000 employees in France or 10,000 employees in France and

internationally. This variable equals 1 after the adoption of the labour law reforms in 2013

and 0 otherwise.

3.6. Estimation method

Whereas ESG performance and employee board representation may affect firm performance,

their potential impact may be driven by firms’ characteristics, which at the same time affect

performance. This situation is known as the archetypical endogeneity problem, and it may

arise due to reverse causalities and omitted variables. To check for the endogeneity problem,

we use the two-step General Method of Moments (system GMM).

We use the following equation to estimate the extent to which the value relevance of

ESG performance is affected by employee representation on the board:

TOBINit = β0 + β1 Lag TOBINit + β2 ESG_PERit + β3 EMPL_BOARDit + β4

(ESG_PER*EMPL_BOARD)it + β5 CSR_ASSit + β6 CSR_COMit + β7 BOARD_SIZEit + β8

BOARD_INDit + β9 BOARD_MEETit + β10 DUALit + β11 TENUREit + β12 FAM_OWNit + β13

INST_OWNit + β14 EMPL_OWNit + β15 LEVit + β16 R&Dit + β17 SIZEit + β18 LAW2013t + β19

INDUSTRYi + ɛit

All variables are as described in Table II.1.

89
Table II. 1: Variables and Their Measurement
Variable Description Measure6
Dependent variable
TOBIN Tobin’s q Stock market capitalization plus book value of liabilities
as a ratio of total assets
Endogenous variables
ESG_PER ESG performance Corporate social performance as measured by Thomson
Reuters/S-Network ESG Best Practices Ratings
encompassing environmental, social, and corporate
governance performance.
SOCIAL_PER Social performance Social performance is a pillar of the ESG performance as
measured by Thomson Reuters/S-Network ESG Best
Practices Ratings. This pillar examines employment
quality, health and safety, training and development,
diversity, human rights, community and finally product
responsibility.
ENVIR_PER Environmental Environmental performance is a pillar of the ESG
performance performance as measured by Thomson Reuters/S-
Network ESG Best Practices Ratings. This pillar
examines the resource and emission reduction and
product innovation.
CG_PER Corporate governance Corporate Governance performance is a pillar of the ESG
performance performance as measured by Thomson Reuters/S-
Network ESG Best Practices Ratings. This pillar
examines the board structure, compensation policy, board
functions, shareholders rights and vision and strategy.
Moderating variable
NBEMPL_BOARD Number of employees on Number of employee directors on board
board
EMPL_BOARD Employee board Proportion of employee directors to the total number of
representation board directors
Governance variables
CSR_ASS CSR assurance Binary variable that takes the value 1 if CSR assurance is
adopted and 0 otherwise
CSR_COM CSR committee Binary variable that takes the value 1 if the company has
a CSR committee and 0 otherwise
BOARD_SIZE Board size Natural logarithm of the number of directors on the board
BOARD_IND Board independence Ratio of number of non-executive independent directors
to total number of board directors
BOARD_MEET Board meetings Natural logarithm of the number of annual board
meetings
DUAL CEO duality Dummy variable coded 1 if the CEO serves as board
chair; 0 otherwise.
TENURE CEO tenure Number of years within the company as CEO
Ownership variables
FAM_OWN Family ownership Percentage of capital held by family
INST_OWN Institutional ownership Percentage of capital held by institutional investors
EMPL_OWN Employee ownership Percentage of capital held by employees
Other control variables
LEV Leverage Ratio of total financial debt to total assets
R&D R&D intensity Ratio of Research and Development to total sales
SIZE Firm size Natural logarithm of the total assets
LAW2013 Labour law reform in Binary variable equal to 1 after the adoption of the labour
2013 law reform in 2013 and 0 otherwise
Industry Industry Binary variable that takes the value 1 if the company
belongs to the sector in question and 0 otherwise

6
Note: Variables from ThomsonOne are winsorized at the 1% and 99% levels.

90
 With regard to the consistency of system GMM, two tests are called for. The first is the

second-order autocorrelation test for the error term. This test examines the null hypothesis of

the absence of a second autocorrelation for the error term. The second is the Sargan/Hansen

test of over-identifying restrictions, which examines the overall validity of instruments.

4. Results

4.1. Descriptive statistics

Table II.2 presents the descriptive statistics of variables for the sample observation. With

regard to market value, the firms in our sample have an average Tobin’s q of 1.218. As

reported in Table II.2, employee directors represent 6.48% of total directors on the board. Our

sample firm-years have on average a combined ESG performance of 63.28%. As regards ESG

pillars, the average social, environmental and governance performance is 69.17%, 70.87%

and 53.26%, respectively. A large majority of firms in the SBF 120 provide assurance of their

CSR information (73.93%) and have a CSR committee (62.18%). The board size of our

sampled firms is about 13 directors, of whom 51.96% are independent. The average number

of meetings held each year is seven. The CEO is also the chairman of the board in 58.38% of

cases and the average tenure of the CEO is almost eight years. The average proportion of total

shares owned by family, institutions and employees of the sample firms is 20.82%, 34.55%

and 2.65%, respectively. Leverage is approximately 24.25% and the average R&D intensity is

2.76%. As reported in Table II.2, the average market value of the sample firms is 24.192

billion euros.

91
Table II. 2: Descriptive Statistics
Mean Median Standard Minimum Maximum
Deviation
TOBIN 1.218 0.882 1.106 0.252 7.026
NBEMPL_BOARD 0.664 0 1.345 0 7
EMPL_BOARD 6.48% 0 9.42% 0 41.67%
ESG_PER 63.28% 64.33% 10.02% 24.11% 83.30%
SOCIAL_PER 69.17% 70.60% 11.91% 20.71% 97.09%
ENVIR_PER 70.87% 73.32% 12.41% 29.03% 94.50%
CG_PER 53.26% 51.50% 11.87% 19.25% 91.19%
CSR_ASS 73.93% 1 43.92% 0 1
CSR_COM 62.18% 1 48.52% 0 1
BOARD_SIZE (number of directors) 12.898 13 3.032 4 22
BOARD_IND 51.96% 50% 19.21% 0 1
BOARD_MEET (number of meetings) 7.327 7 3.190 2 24
DUAL 58.38% 1 49.32% 0 1
TENURE (number of years) 7.759 5 7.696 1 56
FAM_OWN 20.82% 9.3% 23.78% 0 91.85%
INST_OWN 34.55% 32.73% 30.58% 0 90.66%
EMPL_OWN 2.65% 1.3% 4.29% 0 29.20%
LEV 24.25% 21.54% 15.63% 0.10% 82.36%
R&D 2.76% 0.53% 5.01% 0 25.69%
SIZE (in billions of euros) 24.192 9.371 38.491 7.169 278.941
Note: All variables are as defined in Table II.1

4.2. Multivariate analysis

Table II.3 presents the system GMM regression results of Tobin’s q on ESG performance and

employee representation on the board. Model 1 of Table II.3 shows a positive and significant

impact of ESG performance on market-based performance as measured by Tobin’s q (β2 =

0.185, t = 5.33). In line with Friede et al., (2015) and Brooks and Oikonomou (2018), the

market is likely to assign value relevance to ESG performance. Nevertheless, a negative and

significant relationship is found between employee board representation and Tobin’s q. This

result highlights the conflict of interests between different directors on the board (Jensen &

Meckling, 1979). The negative coefficient is in accordance with previous studies (Bøhren &

Strøm, 2010; Kleinknecht, 2015), suggesting that, from a shareholder standpoint, employee

directors might use their power on board to deviate from maximization of shareholder value

(Guedri & Hollandts, 2008; Faleye et al., 2006), and oppose decisions that do not serve their

interests (Bøhren & Strøm, 2010). Another explanation is that since managers and employees

are natural allies (Pagano & Volpin, 2005), employee directors may facilitate amicable

92
monitoring of managers (Guedri & Hollandts, 2008) and may act as an additional

entrenchment tool (Hollandts et al., 2018).

The impact of ESG performance on market-based performance is still stable and

positively and significantly related to Tobin’s q when employee directors on the board are

introduced into the model. A positive and significant impact is observed for CSR assurance,

suggesting that providing assurance services is rewarded by a higher valuation in the financial

markets. We also note a positive and significant correlation between the existence of a CSR

committee and market value. These results imply that sustainability-oriented corporate

governance mechanisms, such as CSR assurance and having a CSR committee, are positively

valued by the market.

As regards corporate governance structures, board size has a positive and significant

impact on Tobin’s q. However, it loses its significance when we introduce employee directors

into the model. In contrast, board independence gains significance when we introduce

employee directors into the empirical model. Indeed, we note a positive and significant

impact of board independence on Tobin’s q in Model 2. Board duality and CEO tenure show

a negative and significant impact on Tobin’s q, suggesting that greater separation between the

CEO and chairman functions and CEOs with shorter tenure are more valued by the market.

As regards ownership variables, our findings show that family ownership is positively and

significantly linked to market performance as measured by the Tobin’s q, whereas neither

institutional nor employee ownership has a significant impact. For the other control variables,

firm leverage shows a positive and significant coefficient on Tobin’s q, while R&D intensity

and firm size have negative and significant impacts on market performance. Finally, the 2013

labour law reform shows a positive and significant impact on Tobin’s q.

In Model 3 of Table II.3 we measure the marginal effect of ESG performance on market

performance in firms with employee directors, using a joint-test technique. The key test is the

93
joint test of the sum of the coefficients on ESG performance (ESG_PER) and the interaction

term (ESG_PER*EMPL_BOARD). In accordance with our hypothesis H1b and in contrast to

our hypothesis H1a, employee board representation negatively moderates the relationship

between ESG performance and market value. Model 3 shows that the joint coefficient is

strongly negative (β2 + β4 = –7.814) and highly significant (t = –6.71). This finding suggests

that ESG performance, when provided by firms with employee directors on the board, is more

likely to be destructive of value. Investors may perceive such extra-financial performance as a

proxy for an entrenchment strategy to the benefit of employees and managers at the expense

of shareholders (Surroca & Tribó, 2008). Another explanation is that employee directors may

have the power to divert the board from shareholder value maximization and may exert

pressure on the board to defend their interests at the expense of shareholders (Bøhren &

Strøm, 2010; Jensen & Meckling, 1979).

94
Table II. 3: System GMM regression of Tobin’s q on ESG performance and employee
board representation
Variables Model 1 Model 2 Model 3:
ESG_PER*EMPL_BOARD
Coef. t–test Coef. t–test Coef. t–test
Lag TOBIN 0.833*** 156.32 0.813*** 179.73 0.807*** 125.99
ESG_PER 0.185*** 5.33 0.220*** 5.39 0.752*** 10.75
EMPL_BOARD –1.852*** –11.35 8.864*** 9.19
ESG_PER*EMPL_BOARD –8.566*** –7.38
CSR_ASS 0.065*** 5.10 0.051*** 2.71 –0.033* –1.72
CSR_COM 0.045*** 5.32 0.044*** 2.61 0.047** 2.16
BOARD_SIZE 0.107*** 3.59 0.004 0.07 0.040 0.65
BOARD_IND –0.026 –0.92 0.353*** 7.79 0.381*** 9.70
BOARD_MEET 0.010 0.79 0.010 0.40 –0.038 –1.43
DUAL –0.119*** –8.73 –0.135*** –7.62 –0.140*** –7.05
TENURE –0.032*** –4.49 –0.020** –2.13 –0.026** –2.41
FAM_OWN 0.220*** 5.03 0.508*** 9.32 0.478*** 7.21
INST_OWN –0.003 –0.10 –0.038 –1.11 –0.081** –2.52
EMPL_OWN 0.340 1.38 0.403 1.12 0.093 0.29
LEV 0.299*** 6.99 0.358*** 5.82 0.374*** 5.99
R&D –0.241* –1.75 0.001 0.01 0.493*** 2.98
SIZE –0.031*** –6.27 –0.064*** –4.98 –0.067*** –6.13
LAW2013 0.067*** 8.92 0.004 0.38 0.028*** 3.10
Intercept –0.181*** –2.58 0.017 0.14 –0.293** –1.99
Industry Yes Yes Yes
Number of observations 741 741 741
Fisher (Prob. > F) 1965.33 (p = 0.000) 6145.54 (p = 0.000) 7333.52 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.75 (p = 0.003) –2.75 (p = 0.003) –2.90 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.43 (p = 0.670) –0.69 (p = 0.887) 0.14 (p = 0.887)
Sargan test (Chi–square, p–value): 594.41 (p = 0.000) 591.31 (p = 0.000) 733.83 (p = 0.000)
Hansen test (Chi–square, p–value): 69.84 (p = 0.382) 67.33 (p = 0.431) 77.10 (p = 0.211)
Joint test: ESG_PER + (ESG_PER*EMPL_BOARD) –7.814*** –6.71
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table II.1.

For further examination of the value relevance of ESG performance when employees

are represented on the board, we divided our sample into two sub-samples: firm-years with

employee directors (314 firm-year observations) and firm-years without employee directors

(427 firm-year observations). We then regressed market value (Tobin’s q) on ESG

performance for each sub-sample. Results of Table II.4 show that ESG performance provided

by firms with employee directors has a negative and significant impact at the 1% level on

Tobin’s q (β2 = –0.479, t = –3.43), while ESG performance provided by firms with no

employee directors has a positive and highly significant impact on Tobin’s q (β2 = 0.332, t =

4.01). These findings suggest that the market cares about firms’ extra-financial performance.

However, the representation of employees on the board apparently changes their perception

95
towards good ESG performance. When employees are on the board, investors do not perceive

ESG performance as a greater commitment of the firm towards its stakeholders. Instead, they

are more likely to view it as a strategy used by managers to get close to employees and other

non-shareholder stakeholders so as to neutralize pressure from shareholders (Surroca & Tribó,

2008).

Table II. 4: System GMM Regression of Tobin's q on ESG Performance in Firm-Years

with and without Employee Board Representation
Variables Model 1: Model 2:
Firm-years with Firm-years without
employee board employee board
representation representation
Coef. t–test Coef. t–test
Lag TOBIN 0.848*** 71.47 0.862*** 78.35
ESG_PER –0.479*** –3.43 0.332*** 4.01
CSR_ASS 0.073 1.65 0.064*** 2.80
CSR_COM 0.018 0.31 0.050** 2.28
BOARD_SIZE 0.251* 1.73 0.154 1.27
BOARD_IND 0.008 0.05 0.178 1.28
BOARD_MEET 0.035 0.77 0.021 0.57
DUAL –0.097*** –3.53 –0.121*** –3.94
TENURE –0.068*** –3.94 –0.050** –2.39
FAM_OWN –0.119 –0.66 0.355** 2.40
INST_OWN –0.058 –0.82 –0.139* –1.68
EMPL_OWN –0.115 –0.05 3.996 0.62
LEV 0.253*** 2.69 0.307* 1.76
R&D 1.065 1.54 0.679 1.46
SIZE –0.022 –0.70 –0.048*** –2.56
LAW2013 0.134*** 5.87 0.010 0.63
Intercept 0.267 1.00 –0.513 –0.58
Industry Yes Yes
Number of observations 314 427
Fisher (Prob. > F) 5908.04 (p = 0.000) 5920.80 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.91 (p = 0.004) –2.68 (p = 0.005)
Arellano–Bond test AR(2) (z, p–value): –1.24 (p = 0.737) 0.23 (p = 0.821)
Sargan test (Chi–square, p–value): 219.28 (p = 0.000) 380.39 (p = 0.000)
Hansen test (Chi–square, p–value): 30.54 (p = 0.538) 46.21 (p = 0.587)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table II.1.

4.3 Supplementary analysis

The preceding results were obtained for firms based on their combined ESG performance. We

now analyse whether each ESG pillar (i.e., environmental, social and governance) leads to

different results. Indeed, the value relevance of ESG pillars depends on the specific type of

performance provided (Friede et al., 2015; Marsat & Williams, 2014; Qiu et al., 2016).

96
Accordingly, we examine the effect of employee board representation on the value relevance

of each pillar of ESG performance. We test this proposition using the joint test technique.

Table II.5 reports results from assessing the value relevance of each of the ESG pillars. For

the three models in Table II.5, we note a constant, negative and significant impact of

employee board representation on Tobin’s q. However, results suggest that shareholders have

differing sensitivity levels with regard to ESG pillars.

Model 1 in Table II.5 reports a highly negative (β2 = –1.568) and strongly significant (t

= –16.06) impact of social performance on Tobin’s q, suggesting that social performance is

value destructive. This finding is in contrast with previous studies (e.g., Marsat & Williams,

2014; Qiu et al., 2016), that argue that investing in the social pillar creates intangible assets

and may result in future cash flows for the firm. One argument in favour of our finding is that

the market may perceive social performance as an entrenchment strategy by managers to get

close to non-shareholder stakeholders such as employees and the community (Surroca &

Tribó, 2008), in which case, social performance could be detrimental to the creation of

shareholder value.

In Model 2 of Table II.5, we examine the value relevance of environmental

performance. Environmental performance provides information on the firm’s interaction with

the environment (e.g., waste management), and the firm’s willingness to address climate

change. We note a positive (β2 = 0.781) and significant (t = 10.65) coefficient for

environmental performance on Tobin’s q. This result suggests that firms with good

environmental performance are financially rewarded by market participants (Kassinis &

Soteriou, 2003). Indeed, good environmental performance reflects a positive commitment to

the environment and the avoidance of future costly disasters that could reduce firm value.

In Model 3 in Table II.5, we investigate the value relevance of the governance pillar of

ESG performance, which expresses the ability of the firm to manage good relationships with

97
shareholders and shows how its board of directors is governed (board function, board

structure, vision and strategy and compensation policy). Model 3 reports a positive and

significant impact of corporate governance performance on Tobin’s q (β2 = 0.344, t = 7.24).

This finding suggests that corporate governance performance is value relevant to investors

since well governed firms are more able to create value (Huse et al., 2009).

Table II. 5: System GMM Regression of Tobin's q on Social, Environmental and

Corporate Governance Performance and Employee Board Representation
Variables Model 1: Model 2: Model 3:
SOCIAL_PER ENVIR_PER CG_PER
Coef. t–test Coef. t–test Coef. t–test
Lag TOBIN 0.825*** 112.26 0.851*** 146.64 0.843*** 181.25
SOCIAL_PER –1.568*** –17.06
ENVIR_PER 0.781*** 10.65
CG_PER 0.344*** 7.24
EMPL_BOARD –1.220*** –5.42 –1.410*** –10.05 –0.427*** –3.33
CSR_ASS 0.247*** 12.01 0.037** 1.98 0.060*** 4.09
CSR_COM 0.135*** 6.27 0.025** 2.07 0.018 1.33
BOARD_SIZE 0.292*** 6.08 0.111** 2.36 0.115*** 3.57
BOARD_IND –0.093 –1.16 –0.399*** –7.68 –0.155*** –2.62
BOARD_MEET 0.068** 2.35 –0.007 –0.36 0.014 0.96
DUAL –0.140*** –5.37 –0.133*** –7.69 –0.079*** –5.22
TENURE –0.042** –2.08 –0.029*** –3.10 –0.029*** –3.26
FAM_OWN 0.008 0.10 0.089 1.59 0.220*** 3.45
INST_OWN 0.028 0.55 0.056 1.42 0.032 0.90
EMPL_OWN 0.936** 2.40 0.513* 1.65 0.513* 1.79
LEV 0.105 1.21 0.381*** 7.02 0.283*** 5.95
R&D –0.532 –1.49 –0.778*** –3.23 –0.352* –1.68
SIZE 0.038*** 3.45 –0.019** –2.38 –0.023*** –3.98
LAW2013 0.042*** 3.25 0.115*** 11.85 0.130*** 12.90
Intercept 0.167 0.98 –0.560*** –5.46 –0.266*** –3.50
Industry Yes Yes Yes
Number of observations 741 741 741
Fisher (Prob. > F) 2253.56 (p = 0.000) 11194.02 (p = 0.000) 9467.88 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.90 (p = 0.003) –2.74 (p = 0.004) –2.82 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.09 (p = 0.926) –0.19 (p = 0.849) 0.20 (p = 0.982)
Sargan test (Chi–square, p–value): 170.43 (p = 0.000) 118.95 (p = 0.000) 189.25 (p = 0.000)
Hansen test (Chi–square, p–value): 59.25 (p = 0.150) 62.22 (p = 0.264) 63.54 (p = 0.179)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table II.1.

In Table II.6, we measure the marginal effect of employee board representation on the

value relevance of each pillar of ESG performance (social, environmental and governance)

using the joint test approach. The results of Models 1, 2 and 3 show a negative impact of the

interaction term of each pillar of ESG performance on Tobin’s q. For the social pillar, our

98
empirical results of Model 1 show that the impact of the interaction term

(SOCIAL_PER*EMPL_BOARD) is strongly negative (β4 = –15.103) and highly significant

(t = –8.58) on Tobin’s q, suggesting that investors perceive social investments, especially

when employees are represented on board, as a collusion strategy between managers and

employees that serves managerial entrenchment and the interests of workers at the expense of

shareholders (Surroca & Tribó, 2008). The marginal effect of employee board representation

on the value relevance of social performance is assessed by the joint test of the sum of the

coefficient (SOCIAL_PER + SOCIAL_PER*EMPL_BOARD) on Tobin’s q. The joint

coefficient in Model 1 is negative (β2 + β4 = –14.258) and highly significant (t = –8.52),

suggesting that a higher level of social performance when provided by firms with employee

directors on the board is financially penalized by the financial market.

Models 2 and 3 of Table II.6 also show a negative and significant relationship between

the sum of the coefficient (ENVIR_PER + ENVIR_PER*EMPL_BOARD) and Tobin’s q, on

the one hand, and between (CG_PER + CG_PER*EMPL_BOARD) and Tobin’s q, on the

other. These results suggest that the market negatively assesses environmental performance

and governance performance when employees are appointed to the board. One possible

explanation is that managers may use a long-term commitment to the environment to reduce

the firm’s attractiveness to raiders (Pagano & Volpin, 2005) or to gain private benefits such as

a reputation for being “good global citizens” (Barnea & Rubin, 2010).

99
Table II. 6: System GMM Regression of Tobin's q on the Interaction between Social, Environmental and Corporate Governance
Performance and Employee Board Representation
Variables Model 1: Model 2: Model 3:
SOCIAL_PER*EMPL_BOARD ENVIR_PER*EMPL_BOARD CG_PER*EMPL_BOARD
Coef. t–test Coef. t–test Coef. t–test
Lag TOBIN 0.819*** 126.87 0.813*** 94.42 0.811*** 101.48
EMPL_BOARD 15.674*** 10.37 8.666*** 3.66 6.851*** 11.92
SOCIAL_PER 0.844*** 6.22
SOCIAL_PER*EMPL_BOARD –15.103*** –8.58
ENVIR_PER –0.258* –1.67
ENVIR_PER*EMPL_BOARD –8.521*** –3.42
CG_PER 0.933*** 12.26
CG_PER*EMPL_BOARD –8.767*** –9.30
CSR_ASS –0.021 –1.00 0.059* 1.83 –0.003 –0.11
CSR_COM 0.050** 2.31 0.091*** 3.07 0.031 1.60
BOARD_SIZE 0.027 0.49 0.019 0.29 0.047 0.76
BOARD_IND 0.445*** 6.85 0.447*** 4.80 0.579*** 6.35
BOARD_MEET –0.008 –0.32 0.003 0.11 –0.109*** –3.07
DUAL –0.125*** –5.43 –0.109*** –4.64 –0.109*** –4.53
TENURE –0.023* –1.75 –0.056*** –3.06 –0.014 –1.11
FAM_OWN 0.447*** 5.96 0.325*** 3.76 0.650*** 6.17
INST_OWN –0.156*** –3.46 –0.037 –0.72 –0.056 –1.39
EMPL_OWN –0.422 –1.41 0.534 1.17 0.278 0.71
LEV 0.252*** 2.55 0.172** 2.40 0.434*** 5.97
R&D 0.385** 2.07 0.808* 1.89 0.631*** 2.90
SIZE –0.077*** –6.73 –0.013 –0.99 –0.072*** –4.93
LAW2013 –0.012 –0.98 0.018 0.90 0.042*** 3.85
Intercept –0.361*** –2.61 0.114 0.63 –0.296** –1.98
Industry Yes Yes Yes
Number of observations 741 741 741
Fisher (Prob. > F) 1348.08 (p = 0.000) 1753.64 (p = 0.000) 1592.60 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.81 (p = 0.004) –2.79 (p = 0.004) –2.87 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.56 (p = 0.906) –0.41 (p = 0.680) –0.20 (p = 0.434)
Sargan test (Chi–square, p–value): 573.49 (p = 0.000) 177.46 (p = 0.000) 550.07 (p = 0.000)
Hansen test (Chi–square, p–value): 62.12 (p = 0.299) 58.90 (p = 0.333) 58.18 (p = 0.413)
Joint test: SOCIAL_PER + (SOCIAL_PER*EMPL_BOARD) –14.258*** –8.52
Joint test: ENVIR_PER + (ENVIR_PER*EMPL_BOARD) –8.780*** –3.64
Joint test: CG_PER + (CG_PER*EMPL_BOARD) –7.834*** –8.57
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively. All variables are as defined in Table II.1.

100
 To shed more light on the results of Table II.6, we represent the results of the system

GMM regression of Tobin’s q on social performance in firm-years with employee directors

(Model 1 of Table II.7) and without employee directors on the board (Model 2 of Table II.7).

The results of both Models 1 and 2 show a negative and significant impact of social

performance on market value as measured by Tobin’s q, with a higher impact for the sub-

sample of firm-years with employee directors (β2 = –1.284; t = –7.45) than for the sub-sample

of firm-years without employee directors (β2 = –0.864; t = –5.03). These findings suggest that

the presence of employees on the board strengthens the negative relationship between social

performance and market value.

Table II. 7: System GMM Regression of Tobin's q on Social Performance in Firm-Years

with and without Employee Board Representation
Variables Model 1: Model 2:
Firm-years with Firm-years without
employee board employee board
representation representation
Coef. t–test Coef. t–test
Lag TOBIN 0.812*** 19.33 0.846*** 87.85
SOCIAL_PER –1.284*** –7.45 –0.864*** –5.03
CSR_ASS 0.203*** 3.10 0.226*** 6.69
CSR_COM –0.004 –0.06 0.134*** 4.93
BOARD_SIZE 0.587** 2.51 0.217*** 3.27
BOARD_IND –0.068 –0.55 0.092 0.75
BOARD_MEET 0.022 0.41 –0.016 –0.45
DUAL –0.263*** –3.45 –0.179*** –6.52
TENURE –0.048 –1.45 –0.059*** –3.09
FAM_OWN –0.247 –1.40 0.223*** 2.76
INST_OWN –0.010 –0.11 –0.056 –0.85
EMPL_OWN –1.137 –0.53 4.887*** 2.94
LEV 0.278 1.35 0.187 1.33
R&D –2.724 –1.18 0.014 0.04
SIZE 0.068 1.46 –0.017 –0.89
LAW2013 0.072*** 2.73 0.022 0.95
Intercept –0.451 –0.63 0.184 0.42
Industry Yes Yes
Number of observations 314 427
Fisher (Prob. > F) 30553.57 (p = 0.000) 20512.42 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.09 (p = 0.008) –2.60 (p = 0.006)
Arellano–Bond test AR(2) (z, p–value): –1.50 (p = 0.135) 0.53 (p = 0.595)
Sargan test (Chi–square, p–value): 129.45 (p = 0.000) 90.66 (p = 0.000)
Hansen test (Chi–square, p–value): 40.73 (p = 0.264) 44.87 (p = 0.101)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table II.1.

101
 Note that environmental performance impacts Tobin’s q differently depending on

whether or not employees are represented on the board. The results of Model 1 in Table II.8

show a negative and significant impact of environmental performance on Tobin’s q (β2 = –

1.262; t = –4.08) for the sub-sample of firm-years with employee directors. In contrast, the

results of Model 2 of Table II.8 indicate a positive and significant impact of environmental

performance on Tobin’s q (β2 = 0.458; t = 3.89) for the sub-sample of firm-years without

employee directors. These findings confirm the result from the joint test that employee board

representation negatively moderates the relationship between environmental performance and

Tobin’s q.

Table II. 8: System GMM Regression of Tobin's q on Environmental Performance in

Firm-Years with and without Employee Board Representation
Variables Model 1: Model 2:
Firm-years with Firm-years without
employee board employee board
representation representation
Coef. t–test Coef. t–test
Lag TOBIN 0.846*** 46.29 0.845*** 86.87
ENVIR_PER –1.262*** –4.08 0.458*** 3.89
CSR_ASS 0.092 1.51 0.058** 2.44
CSR_COM 0.107*** 2.72 0.081*** 4.28
BOARD_SIZE 0.045 0.39 0.044 0.40
BOARD_IND 0.212** 2.22 0.1454 1.29
BOARD_MEET 0.085* 1.65 0.017 0.46
DUAL –0.018 –0.38 –0.134*** –5.89
TENURE –0.050* –1.89 –0.033* –1.93
FAM_OWN –0.272** –2.19 0.326*** 4.47
INST_OWN –0.051 –0.70 –0.110 –1.59
EMPL_OWN 0.384 0.77 1.684 0.97
LEV 0.281*** 2.78 0.422** 2.39
R&D 4.559* 1.93 0.208 0.46
SIZE 0.050 1.41 –0.038** –2.31
LAW2013 0.088*** 3.88 0.008 0.42
Intercept 0.470 0.91 –0.380 –0.57
Industry Yes Yes
Number of observations 314 427
Fisher (Prob. > F) 9752.21 (p = 0.000) 11300.03 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.71 (p = 0.005) –2.60 (p = 0.006)
Arellano–Bond test AR(2) (z, p–value): –1.34 (p = 0.180) 0.53 (p = 0.595)
Sargan test (Chi–square, p–value): 220.45 (p = 0.000) 90.66 (p = 0.000)
Hansen test (Chi–square, p–value): 46.89 (p = 0.402) 44.87 (p = 0.141)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table II.1.

102
 Similarly to the case of environmental performance, the results of Models 1 and 2 in

Table II.9 show opposing coefficients for governance performance. While the result of Model

1 shows a negative and significant impact of governance performance on Tobin’s q (β2 = –

0.374; t = –2.38), the result of Model 2 shows a positive and significant impact of governance

performance on Tobin’s q (β2 = –0.469; t = 4.28). These coefficients indicate that governance

performance is less relevant for the financial market when provided by firms with employee

representation on the board.

Table II. 9: System GMM Regression of Tobin's q on Corporate Governance

Performance in Firm-Years with and without Employee Board Representation
Variables Model 1: Model 2:
Firm-years with Firm-years without
employee board employee board
representation representation
Coef. t–test Coef. t–test
Lag TOBIN 0.977*** 70.30 0.862*** 74.32
CG_PER –0.374** –2.38 0.469*** 4.28
CSR_ASS 0.193 1.45 0.054* 1.83
CSR_COM 0.336*** 4.53 0.042 1.40
BOARD_SIZE 0.449*** 3.22 0.115 1.30
BOARD_IND –0.746*** –3.32 0.180 1.56
BOARD_MEET –0.273** –2.02 0.041 1.00
DUAL –0.592*** –3.37 –0.081*** –3.33
TENURE –0.229*** –3.42 –0.022 –1.29
FAM_OWN 0.385 0.91 0.442*** 4.15
INST_OWN –0.522*** –2.63 –0.068 –1.17
EMPL_OWN –0.793 –0.55 –0.142 –0.09
LEV 1.712*** 4.05 0.296** 2.06
R&D 2.508 1.10 0.514 1.43
SIZE –0.360*** –5.78 –0.053*** –3.07
LAW2013 0.227*** 3.82 0.082*** 4.78
Intercept 0.320 0.36 –0.354 –0.60
Industry Yes Yes
Number of observations 314 427
Fisher (Prob. > F) 12250.54 (p = 0.000) 21878.36 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.99 (p = 0.002) –2.80 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –1.42 (p = 0.155) 0.72 (p = 0.475)
Sargan test (Chi–square, p–value): 101.84 (p = 0.000) 102.14 (p = 0.000)
Hansen test (Chi–square, p–value): 35.68 (p = 0.300) 42.85 (p = 0.142)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table II.1.

5. Conclusion

Consistent with the conclusions of previous studies that ESG investments may serve

managerial entrenchment and that employee directors may use their power on board to defend

103
their interests at the expense of shareholders, the presence of employee directors on the board

may impact stakeholders’ perception of ESG performance. The present study is designed to

examine the extent to which the value relevance of ESG performance is moderated by the

appointment of employees to the board.

Using a sample of French firms listed in the SBF 120 from 2007 to 2017, we find that

investors react positively to ESG performance but negatively to the presence of employees on

the board. Nonetheless, we find that ESG performance is less relevant for market participants

when employees are represented on the board compared to when there are no employee

representatives. The value relevance of ESG performance also depends on the type of ESG

pillar (i.e., social, environment and governance). Accordingly, we carried out supplementary

analyses to assess the impact of employee directors on the value relevance of each ESG pillar.

Using the joint test approach, our results show that investors financially penalize social,

environmental and governance performances when provided by firms with employee

directors. A possible explanation is that high investments in ESG pillars by such firms may

point to an alliance between managers and employee directors that counterbalances the power

of shareholders on the board. Moreover, high performance in terms of ESG pillars reflects

higher CSR expenditure that may lead to greater advantages for employees at the expense of

shareholders. To go further, we examined the value relevance of ESG performance as well as

its related pillars for firms with and without employee directors. Our findings provide

additional evidence that the reaction of market participants to ESG performance and its

related pillars varies considerably, according to whether or not employees are appointed to the

board.

Our results question the way employee board representation is currently viewed today

by financial market participants. These findings should prompt companies to make a greater

communication effort to advocate the adoption of stakeholder-oriented boards. Our study also

104
raises a number of questions regarding the role of employee directors in the decision-making

process. We are aware of the limitations of our study, but these in turn offer new directions

for future research. In addition to market-based performance, accounting-based performance

measures (i.e., return on assets, return on equity) should be considered. While Tobin’s q

reflects shareholders' expectations, accounting-based performance measures the efficiency of

assets in producing income and may reflect better employee performance. The moderating

role of employee directors in the ESG-FP relationship may differ according to the firm’s

ownership structure (concentration/dispersion), the profile of shareholders (e.g., family,

institutional), the industry and firm size. Statutory or demographic attributes of employee

directors (e.g., backgrounds, diligence, expertise, age, tenure, etc.) may also influence

strategic decision-making and overlap with the ESG-FP relationship. Finally, it would be

interesting to carry out a cross-country study to assess the relationship between employee

representation on the board, ESG performance and firm performance in different financial

market systems and different jurisdictions.

105
References

Aubert, N., Garnotel, G., Lapied, A., & Rousseau, P. (2014). Employee ownership: A
theoretical and empirical investigation of management entrenchment vs. reward
management. Economic Modelling, 40, 423–434.
Bagdi, K. (2015). The future of employees’ board-level representation in the European Union.
Procedia Economics and Finance, 23, 1394–1400.
Barnea, A., & Rubin, A. (2010). Corporate Social Responsibility as a conflict between
shareholders. Journal of Business Ethics, 97, 71–86.
Ben Lahouel, B., Peretti, J. M., & Autissier, D. (2014). Stakeholder power and corporate
social performance: The ownership effect. Corporate Governance, 14, 363–381.
Berman, S. L., Wicks, A. C., Kotha, S., & Jones, T. M. (1999). Does stakeholder orientation
matter? The relationship between stakeholder management models and firm financial
performance. Academy of Management Journal, 42, 488–506.
Berrone, P., & Gomez-Mejia, L. R. (2009). Environmental performance and executive
compensation: An integrated agency-institutional perspective. Academy of Management
Journal, 52, 103–126.
Bingham, J. B., Dyer, W. G., Smith, I., & Adams, G. L. (2011). A stakeholder identity
orientation approach to corporate social performance in family firms. Journal of Business
Ethics, 99, 565–585.
Bøhren, O., & Strøm, R. O. (2010). Governance and politics: Regulating independence and
diversity in the board room. Journal of Business Finance and Accounting, 37, 1281–1308.
Brooks, C., & Oikonomou, I. (2018). The effects of environmental, social and governance
disclosures and performance on firm value: A review of the literature in accounting and
finance. The British Accounting Review, 50, 1–15.
Cahan, S. F., De Villiers, C., Jeter, D. C., Naiker, V., & Van Staden, C. J. (2016). Are CSR
disclosures value relevant? Cross-country evidence. The European Accounting Review, 25,
579–611.
Clarkson, M. E. (1995). A stakeholder framework for analyzing and evaluating corporate
social performance. Academy of Management Review, 20, 92–117.
Coffey, B. S., & Wang, J. (1998). Board diversity and managerial control as predictors of
corporate social performance. Journal of Business Ethics, 17, 1595–1603.
Crifo, P., & Mottis, N. (2016). Socially responsible investment in France. Business & Society,
55, 576–593.

106
Dunn, P., & Sainty, B. (2009). The relationship among board of director characteristics,
corporate social performance and corporate financial performance. International Journal of
Managerial Finance, 5, 407–423.
Ehnert, I., Parsa, S., Roper, I., Wagner, M., & Muller-Camen, M. (2016). Reporting on
sustainability and HRM: A comparative study of sustainability reporting practices by the
world's largest companies. The International Journal of Human Resource Management,
27, 88–108.
Faleye, O., Mehrotra, V., & Morck, R. (2006). When labor has a voice in corporate
governance. Journal of Financial and Quantitative Analysis, 41, 489–510.
Freeman, R. (1984). Edward: Strategic Management. A Stakeholder Approach, New York.
Friede, G., Busch, T., & Bassen, A. (2015). ESG and financial performance: Aggregated
evidence from more than 2000 empirical studies. Journal of Sustainable Finance &
Investment, 5, 210–233.
Galbreath, J. (2010). Corporate governance practices that address climate change: An
exploratory study. Business Strategy & the Environment, 19, 335–350.
Gao, Y., & Yang, H. (2016). Do employees support corporate philanthropy? Evidence from
Chinese listed companies. Management and Organization Review, 12, 747–768.
Gillet, C. (2012). A study of sustainability verification practices: the French case. Journal of
Accounting & Organizational Change, 8, 62–84.
Ginglinger, E., Megginson, W., & Waxin, T. (2011). Employee ownership, board
representation, and corporate financial policies. Journal of Corporate Finance, 17, 868–
887.
Guedri, Z., & Hollandts, X. (2008). Beyond dichotomy: The curvilinear impact of employee
ownership on firm performance. Corporate Governance: An International Review, 16,
460–474.
Haddock-Millar, J., Sanyal, C., & Müller-Camen, M. (2016). Green human resource
management: A comparative qualitative case study of a United States multinational
corporation. The International Journal of Human Resource Management, 27, 192–211.
Haniffa, R., & Cooke, T. (2005). The impact of culture and governance on corporate social
reporting. Journal of Accounting & Public Policy, 24, 391–430.
Hillman, A. J., & Keim, G. D. (2001). Shareholder value, stakeholder management, and social
issues: what's the bottom line?. Strategic Management Journal, 22, 125–139.

107
Hur, W. M., Moon, T. W., & Ko, S. H. (2016). How employees’ perceptions of CSR increase
employee creativity: Mediating mechanisms of compassion at work intrinsic motivation.
Journal of Business Ethics, 138, 1–16.
Huse, M., Nielsen, S. T., & Hagen, I. M. (2009). Women and employee-elected board
members, and their contributions to board control tasks. Journal of Business Ethics, 89,
581–597.
Hollandts, X., Aubert, N., Abdelhamid, A., & Prieur, V. (2018). Beyond Dichotomy: The
Curvilinear Impact of Employee Ownership on CEO entrenchment. Management
International, 22, 112–164.
Jensen, M. C., & Meckling, W. H. (1979). Rights and production functions: An application to
labor-managed firms and codetermination. Journal of Business, 52, 469–506.
Jones, D. A., Willness, C. R., & Madey, S. (2014). Why are job seekers attracted by corporate
social performance? Experimental and field tests of three signal-based mechanisms.
Academy of Management Journal, 57, 383–404.
Kleinknecht, R. H. (2015). Employee participation in corporate governance: Implications for
company resilience. European Journal of Industrial Relations, 21, 57–72.
Kramar, R. (2014). Beyond strategic human resource management: Is sustainable human
resource management the next approach? The International Journal of Human Resource
Management, 25, 1069–1089.
Lydenberg, S. (2013). Responsible investors: Who they are, what they want. Journal of
Applied Corporate Finance, 25, 44–49.
Markey, R., McIvor, J., & Wright, C. F. (2016). Employee participation and carbon emissions
reduction in Australian workplaces. The International Journal of Human Resource
Management, 27, 173–191.
Marsat, S., & Williams, B. (2014). Does the market value social pillar? European Financial
Management Association (EFMA). Working paper number 0296, Available at
[Link]
Rome/papers/EFMA2014_0296_fullpaper.pdf
Moura-Leite, R. C., & Padgett, R. C. (2011). Historical background of corporate social
responsibility. Social Responsibility Journal, 7, 528–539.
Nekhili, M., Nagati, H., Chtioui, T., & Rebolledo, C. (2017a). Corporate social responsibility
disclosure and market value: Family versus nonfamily firms. Journal of Business
Research, 77, 41–52.

108
Nekhili, M., Nagati, H., Chtioui, T., & Nekhili, A. (2017b). Gender-diverse board and the
relevance of voluntary CSR reporting. International Review of Financial Analysis, 50, 81–
100.
O'Donohue, W., & Torugsa, N. (2016). The moderating effect of ‘Green’ HRM on the
association between proactive environmental management and financial performance in
small firms. The International Journal of Human Resource Management, 27, 239–261.
OECD (2017). Collective bargaining in a changing world of work. Employment Outlook.
Pagano, M., & Volpin, P. F. (2005). Managers, workers, and corporate control. The Journal of
Finance, 60, 841–868.
Preuss, L., Haunschild, A., & Matten, D. (2009). The rise of CSR: implications for HRM and
employee representation. The International Journal of Human Resource Management, 20,
953–973.
Qiu, Y., Shaukat, A., & Tharyan, R. (2016). Environmental and social disclosures: Link with
corporate financial performance. The British Accounting Review, 48, 102–116.
Richard, O. C., & Johnson, N. B. (2001). Strategic human resource management effectiveness
and firm performance. International Journal of Human Resource Management, 12, 299–
310.
Shaukat, A., Qiu, Y., & Trojanowski, G. (2016). Board attributes, corporate social
responsibility strategy, and corporate environmental and social performance. Journal of
Business Ethics 135, 569–585.
Surroca, J., & Tribó, J. A. (2008). Managerial entrenchment and corporate social
performance. Journal of Business Finance & Accounting, 35, 748–789.
Tang, Z., Hull, C. E., & Rothenberg, S. (2012). How corporate social responsibility
engagement strategy moderates the CSR–financial performance relationship. Journal of
Management Studies, 49, 1274–1303.
Van der Laan, G., Van Ees, H., & Van Witteloostuijn, A. (2008). Corporate social and
financial performance: An extended stakeholder theory, and empirical test with accounting
measures. Journal of Business Ethics 79, 299–310.
Velte, P. (2016). Women on management board and ESG performance. Journal of Global
Responsibility, 7, 98–109.
Weber, J., & Gladstone, J. (2014). Rethinking the corporate financial–social performance
relationship: Examining the complex, multi-stakeholder notion of corporate social
performance. Business and Society Review, 119, 297–336.

109
Wood, D. J., & Jones, R. E. (1995). Stakeholder mismatching: A theoretical problem in
empirical research on corporate social performance. The International Journal of
Organizational Analysis, 3, 229–267.
Zhu, Y., Xie, Y., Warner, M., & Guo, Y. (2015). Employee participation and the influence on
job satisfaction of the ‘new generation’ of Chinese employees. The International Journal
of Human Resource Management, 26, 2395–2411.

110
 Chapter III: The ESG-Financial Performance
Relationship: Does the Type of Employee Board
Representation Matter?

1. Introduction

While the board of directors is the formal corporate body that determines the orientations of

the company acting in the best interest of its shareholders, board diversity is claimed to be a

determinant for non-financial performance, a corporate strategy that reflects relations between

the firm and its stakeholders (Boulouta, 2013; Coffey & Wang, 1998; Dunn & Sainty, 2009;

Hafsi & Turgut, 2013). Board diversity is defined as the variation among board members

stemming from their personalities, learning styles, knowledge, age, values and expertise

(Coffey & Wang, 1998). From the standpoint of stakeholders, a more diverse board may be

viewed as indicative of sensitivity on the part of higher management to their expectations

(Luoma & Goodstein, 1999). In the same vein, Huse, Nielson, and Hagen (2009) argue that

diversity, including female directors and employee representatives, is an important criterion

when selecting board members.

As for board gender diversity (e.g., Bennouri, Chtioui, Nagati, & Nekhili, 2018), the

representation of employees on the board has attracted increasing attention from regulators

and researchers. The representation of labour may be seen as a policy to balance forces

(labour and capital) on the board and to present the concerns of different stakeholders rather

than focusing solely on shareholders. From an agency theory perspective, employee board

representatives do not necessarily have the same interests as the other directors representing

outside shareholders, and may be more concerned about CSR performance (Barnea & Rubin,

111
2010). Setting out from the supposition that (1) a CSR-oriented board may achieve better

social and environmental performance (Shaukat, Qiu, & Trojanowski, 2016) and (2)

employees are more concerned about socially responsible activities (Huse et al., 2009) as well

as the long-term survival of the firm (Kleinknecht, 2015), employee board representation may

be an explicit signal of a firm’s greater engagement with all its stakeholders (Hillman, Keim,

& Luce, 2001).

Previous studies have focused on the impact of employee board representation on

boards efficiency (Huse et al., 2009), agency costs (Fauver & Fuerst, 2006), CEO

entrenchment (Hollandts, Aubert, Abdelhamid, & Prieur, 2018), payout policies (Ginglinger,

Megginson, & Waxin, 2011), and the implications for human resource management (Preuss,

Haunschild, & Matten, 2009). However, to the best of our knowledge, with the exception of

Nekhili, Boukadhaba, Nagati, and Chtioui (2019) employee board representation has not been

examined with respect to environmental, social and corporate governance (ESG)

performance. ESG performance is an important aspect of corporate strategy and can be used

by equity analysts and market participants as a proxy for management quality (Eccles,

Serafeim, & Krzus, 2011). Nekhili et al. (2019) examine the moderating impact of employee

directors on the value relevance of firms’ extra-financial performance. They find that

employee board representation reduces the relevance of ESG performance in terms of market

value. Extending the scope of that study, we consider two different types of employee

directors, namely labour board representatives and employee-shareholder board

representatives, who may have different interests and different attitudes toward ESG

performance. We then examine the moderating role of each type of employee directors on the

value relevance of overall ESG performance and each of the three pillars of ESG performance

(i.e., environmental, social, and corporate governance). We are thereby able to gain further

insights into the relationship between employee directorship and the firm’s extra-financial

112
performance and to determine to what extent the type of employee directors is important for

the financial market in the assessment of ESG performance.

France provides an interesting and unique context for studying the effect of employee

board representation. Indeed, on French boards, there are two types of employee directors:

labour board representatives and employee-shareholder board representatives. Whereas the

representation of employees on the board is likely to be a European model of corporate

governance (Gordon & Roe, 2004; Gold & Waddington, 2019), French regulators show

greater interest in employee board representation compared to other European countries. 12

Starting on 30 December 2006, the representation of employee-shareholders became

mandatory for publicly listed French firms when they hold at least 3 % of total shares. In this

respect, French firms have experienced a spectacular increase during the past decade in

employee ownership (Guedri & Hollandts, 2008; Poulain-Rehm & Lepers, 2013). As for the

representation of labour on the board, French firms, depending on their board size, have been

required to reserve seats for employee directors elected by right of employment since 2013.13

It is, however, noteworthy that both types of directors have the same voting and other rights

and responsibilities on the board as all the other directors (Ginglinger et al., 2011).

Based on a matched sample of French firms listed on the SBF 120 index for the period

2007-2017, we report that the two types of employee directors have different impacts on the

firm’s extra-financial (ESG) performance. Our results show that labour board representation

tends exclusively to improve social performance, whereas employee-shareholder board

representation is found to enhance the overall ESG performance and especially environmental

12
An overview of the evolution of French regulation on the employee representation on corporate boards is
given in Ginglinger et al. (2011).
13
Employees have to be represented on board by one director if the board has fewer than 12 members and two
directors if the board comprises more than 12 members for firms with at least 5000 employees in France or
10,000 employees in France and around the world. The Rebsamen Law of 17 August 2015 reduced the threshold
of eligibility of employee representation to 1000 employees in France and 5000 in France and around the world.
In order to improve employee representation, the French PACTE (Action Plan for Business Growth and
Transformation) corporate reform law adopted by Parliament on April 2019 reduces the former threshold from
12 to eight board members.

113
performance and corporate governance performance. Our results support the findings of

Bøhren and Strøm (2010) that labour representatives on the board advantage employees at the

expense of shareholders’ wealth. In accordance with our initial argument, we find that while

the presence of employee-shareholders on the board moderates positively market participants’

perception of ESG performance, the presence of labour board representatives moderate

negatively this relationship. Outside shareholders seem to be more reluctant to have labour

board representatives, as these may counterbalance shareholders’ power on the board and be

inclined to support employees' interests. Conversely, outside shareholders are more prone to

view employee-shareholder board representatives as having closer relationships with their

representatives on board.

The chapter is structured as follows. Section 2 consists of a literature review of

employee board representation and ESG performance. Section 3 describes our methodology,

including the sample, variables and empirical model. Section 4 discusses the results. Finally,

section 5 concludes and suggests future research perspectives.

2. Employee Board Representation and ESG Performance

Whereas financial performance has received much attention, the non-financial outcomes of

the stakeholders’ representation on corporate boards have not been thoroughly examined

(Hillman et al., 2001). Nekhili et al. (2019) show recently that the capital market perception

of ESG performance varies markedly in accordance with board-level employee

representation. Nonetheless, the authors do not discuss the direct relationship between

employee board representation and ESG performance. Moreover, they consider employee

directors as a homogenous group without distinguishing the way they are represented on the

board. Distinguishing between the two types of employee directors (i.e., labour board

representatives and employee-shareholder board representatives) and their impact on ESG

performance would provide a deeper understanding on the outcome of the stakeholders’ board

114
representation. Starting from the fact that ESG performance is a multi-stakeholder notion

(Weber & Gladstone, 2014), further examination of ESG performance through its underlying

components would enhance our understanding on the extent to which outside shareholders

take into account employee board representation in their assessment of ESG performance.

2.1. Employee Board Representation and Social Performance

According to social identity theory, ESG performance plays an important role in employee

motivation (e.g., Brammer, Millington, & Rayton, 2007; Kim, Lee, Lee, & Kim, 2010;

Turker, 2009). Social identity theory stipulates that individuals classify themselves and others

into social groups and then compare the in-group (their social group) and the out-group (the

other social group) (Tajfel & Turner, 1985), the comparison being likely to influence their

behaviour. In this respect, the behaviour and attitudes of employees as members of an

organization is thus influenced by their perception of the socially responsible behaviour of

their organization (El Akermi, Gond, Swaen, De Roeck, & Igalens, 2018; Brammer et al.,

2007; Kim et al., 2010; Turker, 2009). In particular, employees’ perception of the firm’s CSR

initiatives is found to be linked with affective commitment (Brammer et al., 2007),

organizational commitment (Turker, 2009), employee-company identification (Kim et al.,

2010), the firm’s attractiveness to employees (Jones, Willness, & Madey, 2014; Turban &

Greening, 1997) and employees’ work attitudes (Triana, Jayasinghe, Pieper, Delgado, & Li,

2019; Zhang, Di Fan, & Zhu, 2014). Taken together, these findings suggest that employees

and potential workers prefer organizations that behave in a responsible manner towards their

stakeholders, because they can identify with such organizations. Likewise, being responsible

towards its stakeholders is likely to increase employees’ satisfaction (El Akermi et al., 2018 ;

Brammer et al., 2007) and to lead to a better quality of and greater number of job applicants

(Turban & Greening, 1997). Socially responsible behaviour is then likely to result in

115
competitive advantages for organizations and in turn to be beneficial for shareholders in terms

of corporate financial performance (Van der Laan, Van Ees, & Van Witteloostuijn, 2008).

Based on the above discussion, it may be presumed that the presence of employees on

the board is clearly linked to good social performance. Surprisingly, however, this issue has

been only marginally addressed in the literature. Hammer, Curall and Stern (1991: p. 665)

suggest that the role behaviour of worker representatives on the board consists of “initiating,

participating, voting and communicating with other constituents”. By providing different

information and engaging to a greater extent in creative discussions with other directors,

employee directors may be more able to fulfil their needs and focus on salient issues through

participating in CSR initiatives (Huse et al., 2009). Hillman et al. (2001) consider two types

of stakeholder directors – employee directors and community directors 14 – likely to have

significant impact on the firm’s stakeholder performance, as representatives of the interests of

the firm’s various stakeholder groups such as suppliers, employees, customers, and

communities. While community directors are found to increase community performance,

Hillman et al. (2001) show that employee directors have no significant impact on employee

relations (union relations, long-term policy of company-wide cash profit sharing, substantial

sense of worker involvement/ownership, sharing of financial information with employees,

employee participation in management decision-making, high employee retirement benefits,

other innovative benefits), on diversity issues or on product quality.

With regard to our focus on the type of employee directors, labour board representatives

may play a strategic role on the board by providing more valuable labour-related information.

In this respect, they may increase the awareness of other directors on the board as to the real

needs and wishes of the workers, particularly the importance of CSR initiatives for their

social identity and the negative impact of irresponsible behaviour on their attitudes in the

14
Hillman et al. (2001, p. 303) define community directors as “directors whose primary occupation is academic,
political, or governmental service, a minister or other religious affiliation, non profit affiliation, or in a few cases,
other non business professionals such as athletes or celebrities”.

116
workplace. The board would therefore be in a better position to develop effective strategies

and policies towards human capital, resulting in improved social performance. Nevertheless,

since they have discretionary power over CSR decisions, managers may increase human

capital expenditure in order to gain the support of employees against shareholders (Surroca &

Tribo, 2008) and to retain their own personal advantages (Coffey & Wang, 1998). In such

cases, higher levels of social expenditure may lead to the destruction of value. Compared to

labour board representatives, employee-shareholder board representatives are expected to act

like other directors representing outside shareholders and to increase market value, as

confirmed by Ginglinger et al. (2011). There is reason to be skeptical, however, regarding

outcomes in terms of social initiatives. Consistently with their role of representing employee-

shareholders, they may not deviate from maximizing shareholder value. Thus Ginglinger et al.

(2011) find that wages paid to employees and officers are higher in firms with labour board

representatives than in those with directors representing employee-shareholders.

H1a: The proportion of labour board representatives is positively associated with the

firm’s social performance.

H1b: The proportion of employee-shareholder board representatives is negatively

associated with the firm’s social performance.

2.2. Employee Board Representation and Environmental Performance

Employees are both directly and indirectly impacted by environmental concerns and practices

(Markey, McIvor, & Wright, 2016). They are directly impacted by the quality of their work

environment, which is crucial for their effectiveness in areas such as CO2 emissions reduction,

noise reduction, green buildings, reduced toxic chemicals, etc. They are also indirectly

impacted via broader environmental issues such as the ozone layer depletion, greenhouse gas

emissions, climate change, and so forth. Using a sample of 53 firms from United Kingdom

and Japan, Bansal and Roth (2000) argue that individual concerns, issue salience and field

117
cohesion are factor drivers for the organization’s environmental responsiveness. Accordingly,

employees may suggest environmental practices or adopt environmental measures to improve

safety at their workplace (Markey et al., 2016).

Board composition in general and board diversity in particular play an important role in

the firm’s environmental performance (Huse et al. 2009; Walls, Berrone, & Phan, 2012).

CSR-oriented boards (e.g., those with employee directors) develop a more proactive and

comprehensive board CSR strategy, which in turn leads to better social and environmental

performance (Shaukat et al., 2016). Clearly, employee directors are likely to be concerned

about environmental performance, and may encourage the board to comply with legislation so

as to avoid costly fines and penalties and acquiring a bad reputation, resulting in turn in

decreased shareholder value. Employee directors are more inclined to provide valuable

information about the production process, thereby helping the board to adopt effective

strategies and resource allocation for product innovation. They are also more likely to

increase the board’s engagement with environmental issues that may enhance the quality of

the environment in the workplace. To our knowledge, very little literature deals with the

effect of employee board representation of environmental performance. One exception is the

study by Hillman et al. (2001) who find, in the US context, a positive and significant

relationship between employee directorship and the firm’s environmental performance.

Given that environmental performance is the responsibility of all stakeholders and is

critical to the firm as a whole, we propose that employee directors, whether representing

labour or employee-shareholders, have the same sensitivity to environmental obligation and

may have a positive impact on the firm’s environmental performance.

H2a: The proportion of labour board representatives is positively associated with the
firm’s environmental performance.
H2b: The proportion of employee-shareholder board representatives is positively
associated with the firm’s environmental performance.

118
2.3. Employee Board Representation and Corporate Governance Performance

The corporate governance pillar of ESG performance comprises information on shareholders’

rights, board structure, board functions, compensation policy, vision and strategy. In

alignment with the stakeholder perspective, employee board representation is seen as formal

recognition of the firm’s stakeholders, particularly its human capital. Nevertheless, the

outcomes of employee board representation in terms of board efficiency are still ambiguous

(Fauver & Fuerst, 2006; Huse et al., 2009; Jensen & Meckling, 1979). Judicious

representation of labour on the board is argued to enhance the board’s monitoring, increase

the firm’s efficiency and market value, and reduce agency costs (Fauver & Fuerst, 2006). In

contrast, labour board representation is found to reduce managerial control (Huse et al.,

2009), to decrease the dividends payout ratio (Ginglinger et al., 2011), to hamper the board’s

effectiveness (Jensen & Meckling, 1979) and to reduce firm performance (Bøhren & Strøm,

2010). A recent study by Nekhili et al. (2019) shows that ESG performance creates value for

shareholders only when employees are not represented on the board, suggesting the existence

of a potential conflict of interests between shareholders and employees regarding CSR

strategy.

The main role of labour representatives on the board is to protect the workers’ interests

(Hammer et al., 1991). Accordingly, labour board representatives may act in opposition way

to what is intended by shareholders (Bøhren & Strøm, 2010), such as promoting higher wages

(Jensen & Meckling, 1979), reducing dividend distribution (Ginglinger et al., 2011) and

constraining the board’s effectiveness (Huse et al., 2009). To protect their jobs, labour board

representatives may adopt a risk-averse strategy, resulting in lower value creation in the short

term (Kleinknecht, 2015). Huse et al. (2009) investigate the impact of employee directors on

the board’s control tasks, namely CSR, strategy, behaviour and budget. These authors find

that, although labour representatives foster board engagement toward CSR issues, they may

119
ultimately mitigate the effectiveness of the board with regard to behaviour control tasks,

including the evaluation of the CEO and management team’s performance and discussion of

the CEO’s and the top-management team’s compensation system. All this suggests that labour

representatives may hinder effective managerial control, since employees and managers may

have close relationships, such as alliances against takeover threats (Pagano & Volpin, 2005).

Doubts can also be raised about the competence of employee-elected directors in terms of

their contribution to financial and budget control issues (Bøhren & Strøm, 2005).

It has been argued that employee shareholding leads to closer relationships between

directors elected by employee-shareholders and directors representing outside shareholders

(Ginglinger et al., 2011). For Huse et al. (2009), esteem and cohesiveness of board members

are two features assumed to be critical for effective board task performance. Using survey

data on board performance in Norway, Huse et al. (2009) find that employee-elected directors

perceive themselves to be less esteemed and viewed as ‘second class’ members of the board.

Accordingly, we expect that directors elected by employee-shareholders to have higher

esteem and provide better board cohesiveness than labour board representatives, for at least

three reasons. First, compared to labour board representatives, employee-shareholder

directors, by sharing the same interests, may have close relationships with directors

representing outside shareholders, thus reducing managerial discretion (Ginglinger et al.,

2011). Second, apart from financial considerations in terms of the residual equity-based

claim, employee-shareholder board representatives often hold management positions

(engineers, heads of mission, corporate lawyers, managers, etc.) and possess greater

knowledge and skills (Hollandts & Aubert, 2011). Likewise, the application of knowledge

and skills is promoted by the esteem and cohesiveness of directors and is viewed as one of the

major challenges of board leadership (Huse et al., 2009). Third, as representatives of

120
employee-shareholders on the board, they may influence the vote of all employee-

shareholders in general meetings, thereby providing additional control devices over managers.

H3a: The proportion of labour board representatives is negatively associated with the

firm’s corporate governance performance.

H3b: The proportion of directors representing employee-shareholders is positively

associated with the firm’s corporate governance performance.

3. The Moderating Role of Employee Directors in the ESG-Financial Performance

Relationship

Nekhili et al. (2019) recently investigated the relationship between ESG performance and

market performance, focusing on the moderating role of employee board representation in the

French context. The authors document a negative market perception of overall ESG

performance and each of the key ESG pillars (social, environment and corporate governance)

when employees are represented on the board. Their results highlight the potential conflict of

interests between employee directors and shareholders and suggest that, by achieving a better

ESG performance, both employees and managers may attempt to counterbalance

shareholders’ supremacy on the board. Without questioning the originality of their findings,

the authors did not, however, distinguish between the two specific types of employee

directors, namely labour board representatives and employee-shareholder board

representatives. On the basis of our discussion above, we expect that these two types of

employee directors may affect the market perception of ESG performance differently.

Starting from the fact that ESG performance may result in more advantages for

employees and managers at the expense of shareholders (Barnea & Rubin, 2010; Berrone &

Gomez-Mejia, 2009; Surroca & Tribó, 2008), enhancing ESG performance with directors

elected by and drawn from employees may be perceived by shareholders as a managerial

entrenchment strategy running counter to the shareholders’ interests. It is also noteworthy

121
that, within this strategy, both managers and employees may share the same interest. While

more entrenched managers obtain personal benefits at the expense of shareholders, employees

are also found to be better paid by entrenched managers (Cronqvist, Heyman, Nilsson,

Svaleryd, & Vlachos, 2009). Comparatively, directors elected by employee-shareholders

would have, in addition to their knowledge and skills, closer relationships with directors

representing outside shareholders (Ginglinger et al., 2011), probably leading to greater esteem

and allowing better board cohesiveness (Huse et al., 2009). Combined with the fact that

shareholders are rather sensitive to the firm’s ESG performance (Nekhili et al., 2019),

employee-shareholder representation on the board will moderate in a positive direction the

relationship between ESG performance and corporate financial performance.

In line with the above arguments, we formulate the following hypotheses.

H4a: The proportion of labour board representatives negatively moderates the

relationship between ESG performance and market value.

H4b: The proportion of employee-shareholder board representatives positively

moderates the relationship between ESG performance and market value.

4. Methodology

4.1. Sample and Data

In this study, we use a sample of SBF 120 French firms covering a period of eleven years

(2007-2017). From this starting sample, we eliminate financial, insurance and real estate firms

and foreign companies not subject to French regulation. Our study period starts in 2007

because ESG information as provided by the Thomson Reuters/S-Network ESG Best

Practices Ratings is available for European firms since 2007 (Thomson Reuters’ press release

2014). Not all firms are rated, resulting in a final sample of 817 firm-year observations.

Corporate governance and ownership variables were collected from annual reports available

on the official websites of individual companies and the French financial markets regulator

122
(AMF) website, whereas financial and accounting data were obtained from the ThomsonOne

database.

4.2. Dependent variable

The aim of our study is two-fold: (1) to examine the impact of differing types of employee

board representation (i.e., labour board representation and employee-shareholder board

representation) on ESG performance and (2) to investigate whether the perception of ESG

performance by market participants is influenced by the way employees are represented on

board. Two dependent variables are then considered in the analysis: ESG performance and

market-based value as measured by Tobin’s q.

ESG performance is a complex notion that requires more precision when being

measured than financial performance (Weber & Gladstone, 2014). According to Pagano,

Sinclair, and Yang (2018), the four major ESG rating and index providers are Thomson

Reuters Asset4, MSCI, S&P Dow Jones and RobecoSAM and FTSE Russell. Previous studies

have used different measures for corporate social performance, such as Asset4 ESG ratings

(Shaukat et al., 2016; Velte, 2016) and Kinder, Lydenberg and Domini (KLD) ratings

(Boulouta, 2013; Hafsi & Turgut, 2013; Van der Laan et al., 2008; Weber & Gladstone,

2014). In their comparison of ESG data from Asset4 and KLD, Bloomberg, Halbritter and

Dorfleitner (2015) show that none of the ESG ratings, from the three different (lesquels ?)

providers, is able to detect a significant return difference between high and low ESG

portfolios. In this study, we use the Thomson Reuters/S-Network ESG Best Practices Ratings

that measure the environmental, social, governance and composite ESG performance of over

5,000 companies around the world. These ESG ratings are considered to be an improvement

on the Asset4 ESG ratings. Indeed, whereas Asset4 uses equally weighted key performance

indicators (KPI), the Thomson Reuters/ S-Network attributes a specific weight to each KPI

based on specific considerations and their relative importance. ESG ratings provided by the

123
Thomson Reuters Asset4 have been widely used in the literature as a measure for corporate

social performance (Shaukat et al., 2016).

We use Tobin’s q as a measure to reflect the financial market response to both the

firm’s ESG performance (e.g., Nekhili et al., 2019) and employee board representation (e.g.,

Ginglinger et al., 2011; Nekhili et al., 2019). Tobin’s q is a market-based measure that

translates the way the market perceives future earnings. As compared to accounting

performance measures, Tobin’s q is less influenced by accounting conventions and by the

recognized manipulations of earnings (Dechow, Sloan, & Sweeney, 1996). In our study,

Tobin’s q is measured as stock market capitalization plus book value of liabilities as a ratio of

total assets.

4.3. Endogenous variable

In our study, the two types of employee board representation, namely labour board

representation and employee-shareholder board representation, are endogenously determined.

Endogeneity may result from the unobserved heterogeneity, reverse causality and dynamic

behaviour of the firm (Bennouri et al., 2018). First, unobserved variables (political context,

culture, managerial ability, etc.) may impact both employee board representation and firm

ESG performance. Second, reverse causality may be manifested in that firms with higher ESG

performance may be more willing to represent their stakeholders on their board as a strategy

to satisfy their needs. Conversely, firms with employee board representation would pay more

attention to their different stakeholders, resulting in better ESG performance. Lastly, dynamic

endogeneity, a recurrent problem in corporate governance studies, refers to the lagged reverse

causality between ESG performance and employee board representation.

Similarly to previous studies (Ginglinger et al., 2011; Guedri & Hollandts, 2008;

Nekhili et al., 2019), we measure employee board representation as the number of employee

directors (elected by labour or by employee-shareholders) divided by the total number of

124
board directors. The presence of labour board representatives is measured by the proportion of

employee directors elected by right of employment in the total number of directors on the

board. The presence of employee-shareholder board representatives is measured by the

proportion of directors elected by employee-shareholders in the total number of board

directors.

4.4. Control variables

In our study, we control for stand-alone reports and CSR assurance, because firms that make

higher CSR investments are more likely to disclose them in a separate report and verify them

through an external auditor. We believe that controlling for CSR committees is important in

ESG performance studies. Indeed, the role of CSR committees is to meet the firm’s social and

environmental commitments. The existence of these committees could reflect a greater

commitment towards environmental and social issues, thus resulting in higher levels of ESG

performance (e.g., Nekhili et al., 2019; Velte, 2016). Following previous studies, we control

for board size (e.g., Hafsi & Turgut, 2013), board independence (e.g., Dunn & Sainty, 2009;

Johnson & Greening, 1999; Shaukat et al., 2016), board gender diversity (e.g., Boulouta,

2013; Hafsi & Turgut, 2013, Macaulay, Richard, Peng, & Hasenhuttl, 2018; Velte, 2016) and

board meetings (e.g., Nekhili et al., 2019). Like Hafsi and Turgut (2013) and Shaukat et al.

(2016), we control for CEO duality, which can be expected to influence the board’s

commitment towards ESG duties. We control for CEO tenure in relation to ESG performance,

because CEOs who have held this position for a long time may have more influence on the

board (Nekhili et al., 2019). As well as board characteristics, we control for ownership

structure. Family ownership (e.g., Rees & Rodionova, 2015; Bingham, Dyer, Smith, &

Adams, 2011; Nekhili et al., 2019) and institutional ownership (e.g., Graves & Waddock,

1994; Johnson & Greening, 1999; Nekhili et al., 2019) are expected to affect ESG

performance. Employee ownership gives labour both contractual and residual claims,

125
impacting corporate governance and the firm’s strategic orientations (e.g., Faleye, Mehrotra,

& Morck, 2006; Poulain-Rehm & Lepers, 2013). Following an extensive research stream

(e.g., Bingham et al., 2011; Weber & Gladstone, 2014; Nekhili et al., 2019), we control for

debt and ROA. McWilliams and Siegel (2000) and Boulouta (2013) highlight the importance

of R&D as a control variable in ESG-related studies. Following previous studies, we also

control for firm size and the industry concerned (e.g., Boulouta, 2013; Graves & Waddock,

1994; Shaukat et al., 2016). Finally, in line with Nekhili et al. (2019), we control for the law

of 14 June 2013 that mandates the representation of labour on the boards of French firms. All

variables are as described in Table III.1.

4.5. Estimation Model

To deal with the endogeneity problems discussed above, we use the two-step system GMM

estimation developed by Blundell and Bond (2000). Our model is the following:

ESG performancei,t = β0 + β1Lag ESG performancei,t + β2 Employee directorshipi,t + β3

Standalonei,t + β4 CSR assurancei,t + β5 CSR committeei,t + β6 Board sizei,t + β7 Board

independencei,t + β8 Board gender diversityi,t + β9 Board meetingi,t + β10 Dualityi,t + β11 CEO

tenurei,t + β12 Family ownershipi,t + β13 Institutional ownershipi,t + β14 Leveragei,t + β15 ROAi,t

+ β16 R&Di,t + β17 Firm size i,t + β18 Law 2013_FE + β19 Industry_FE + εi,t

Where i indexes firms and t indexes time. ESG performance is broken down into its

three components (i.e., social, environmental and corporate governance). Employee

directorship is decomposed into labour board representation and employee-shareholders

representation. ε is the error term. All variables are as defined in Table III.1.

126
Table III. 1: Variables and their Measurement
Variable Measure15
Dependent variables
Tobin’s q Stock market capitalization plus book value of liabilities as a ratio of total assets
Moderating variables
ESG performance Corporate ESG (ESG) performance as measured by Thomson Reuters/S–Network
ESG Best Practices Ratings encompassing environmental, social, and corporate
governance performance.
Social performance Social performance is a pillar of the ESG performance. This pillar examines
employment quality, health and safety, training and development, diversity,
human rights, community and finally product responsibility.
Environmental Environmental performance is a pillar of the ESG performance. This pillar
performance examines the resource and emission reduction and product innovation.
Corporate Governance Corporate Governance performance is a pillar of the ESG performance. This pillar
performance examines the board structure, compensation policy, board functions, shareholders
rights and vision and strategy.
Endogenous variables
Employee directorship Proportion of employee directors to the total number of board directors
Labour representation Proportion of employee directors elected by right of employment to the total
number of board directors
Employee-shareholders Proportion of employee-shareholders director to the total number of board
representation directors
Control variables
Standalone CSR report Dummy variable taking the value one if the company issue a standalone CSR
report and zero otherwise.
CSR assurance Binary variable that takes the value one if CSR assurance is provided by external
auditor and zero otherwise
CSR committee Dummy variable taking the value one if the company has a CSR committee and
zero otherwise.
Board size The log of the total number of board directors.
Board independence Ratio of the number of non-executive independent directors to the total number of
board directors.
Board gender diversity Proportion of female directors on the board.
Board meetings Log of the number of annual board meetings.
CEO duality Dummy variable coded one if the CEO is the chair of the board; zero otherwise.
CEO tenure Number of years at a company after being appointed to a CEO position.
Family ownership Percentage of capital held by family members.
Institutional ownership Percentage of capital held by institutional investors.
Employee ownership Percentage of capital held by employees.
Leverage Total financial debt reported to total assets.
ROA Ratio of EBITDA and total assets.
R&D Ratio of R&D expenditure to total sales.
Firm size Log of total assets.
Industry Binary variable that takes the value 1 if the company belongs to the sector in
question and 0 otherwise
Law 2013 Binary variable equal to one after the adoption of the labour law reform in 2013
and zero otherwise.

15 We winsorize all variables from ThomsonOne at 0.01 and 0.99 tails.

127
5. Results and Discussion

5.1. Descriptive Statistics

Table III.2 presents the descriptive statistics of our sample. We find that on average firms in

our sample have good ESG, social, environmental and corporate governance performance

(63.28 %, 69.17 %, 70.87 % and 53.26 % respectively). For our firm-years, employee

directors represent 6.48 % of total directors on the board. More precisely, labour board

representatives represent 4.39 % and employee-shareholder board representatives represent

2.09 %. On average, 47.37 % of firms in our sample have stand-alone CSR reports. Regarding

corporate governance structure, we report that 73.93 % of firms have CSR assurance carried

out by an external auditor and 62.18 % have CSR committees. Boards are, on average,

composed of 13 members, half of them outsiders and 23.25 % female. The average number of

board meetings is seven. In 58 % of cases, the CEO holds the chairman position and the

average tenure of the CEO is about seven and a half years. With regard to ownership

variables, the average proportion of total shares owned by family, institutions and employees

of the sampled firms is 20.82 %, 34.55 % and 2.65 %, respectively. Finally, the sampled firms

have an average size of 24.192 billion euros, an average debt ratio of 24.25 % and an average

R&D intensity of 2.76 %.

128
Table III. 2: Descriptive Statistics
Mean Standard Minimum Maximum 25th 50th 75th
Deviation percentile percentile percentile
Tobin’s q 1.218 1.106 0.252 7.026 0.609 0.882 1.348
ESG performance (%) 63.28 10.02 24.11 83.3 57.97 64.33 70.34
Social performance (%) 69.17 11.91 20.71 97.09 62.73 70.6 77.39
Environmental performance (%) 70.87 12.41 29.03 94.50 62.48 73.31 79.82
Corporate governance performance (%) 53.26 11.87 19.25 91.19 44.94 51.5 61.54
Employee directorship (%) 6.48 9.42 0 41.67 0 0 11.11
Labour representation (%) 4.39 8.40 0 41.67 0 0 7.69
Employee-shareholders representation (%) 2.09 4.31 0 21.43 0 0 0
Standalone (%) 47.37 49.96 0 1 0 0 1
CSR assurance (%) 73.93 43.93 0 1 0 1 1
CSR committee (%) 62.18 48.52 0 1 0 1 1
Board size (Number of directors) 12.90 3.032 4 22 11 13 15
Board independence (%) 51.96 19.21 0 1 40 50 64.28
Board gender diversity (%) 23.25 14.78 0 66.67 9.76 23.08 35.29
Board meetings (Number of meetings) 7.327 3.190 2 24 5 7 9
Duality (%) 58.38 49.32 0 1 0 1 1
CEO tenure (Number of years) 7.759 7.695 1 56 3 5 10
Family ownership (%) 20.82 23.78 0 91.85 0 9.3 40.85
Institutional ownership (%) 34.55 30.58 0 90.66 1.63 32.73 61.05
Employee ownership (%) 2.65 4.29 0 29.2 0.26 1.3 2.96
Leverage (%) 24.25 15.63 0.10 82.36 13.15 21.54 32.93
ROA (%) 4.66 4.63 –11.99 21.7 2.21 4.14 6.67
R&D (%) 2.76 5.01 0 25.69 0 0.53 3.48
Firm size (in billions of euros) 24.192 38.491 0.007 278.941 4.066 9.371 28.114
All variables are as defined in Table III.1.

129
5.2. Pairwise Correlation

Table III.3 presents bivariate correlation coefficients between all the variables considered in

our study. This statistical test is used to detect the presence of multicollinearity between

variables. In this regard, Table III.3 does not show any strong multicollinearity, except for the

coefficient of correlation between employee ownership and employee-shareholder directors,

which is greater than 0.7. Consequently, we eliminate employee ownership from the

following statistical tests.

130
Table III. 3: Pairwise Correlation
1 2 3 4 5 6 7 8 9 10 11
1. Tobin’s q 1.000
2. ESG performance –0.187* 1.000
3. Social performance –0.180* 0.766* 1.000
4. Environmental performance –0.211* 0.826* 0.571* 1.000
5. Corporate governance performance –0.132* 0.502* 0.398* 0.325* 1.000
6. Employee directorship –0.076 0.209* 0.182* 0.262* 0.038 1.000
7. Labour representation –0.046 0.145* 0.160* 0.205* –0.033 0.889* 1.000
8. Employee-shareholders representation –0.078 0.181* 0.092* 0.178* 0.153* 0.452* –0.004 1.000
9. Standalone –0.152* 0.355* 0.362* 0.352* 0.187* 0.189* 0.178* 0.066 1.000
10. CSR assurance –0.048 0.475* 0.367* 0.500* 0.181* 0.274* 0.226* 0.160* 0.356* 1.000
11. CSR committee 0.058 0.264* 0.235* 0.267* 0.157* 0.072 0.116* –0.065 0.123* 0.192* 1.000
12. Board size –0.223* 0.295* 0.332* 0.338* 0.118* 0.404* 0.353* 0.197* 0.327* 0.243* 0.160*
13. Board independence –0.095* 0.310* 0.102* 0.203* 0.361* –0.243* –0.287* 0.031 0.013 0.171* 0.066
14. Board gender diversity 0.084 0.267* 0.126* 0.308* –0.004 0.221* 0.206* 0.081 0.136* 0.515* 0.141*
15. Board meeting 0.010 0.062 0.015 0.039 0.022 0.076 0.129* –0.088 0.007 0.050 0.102*
16. Duality –0.193* 0.010 0.053 0.157* –0.159* 0.146* 0.078 0.170* 0.089 0.115* 0.063
17. CEO tenure –0.004 –0.024 0.022 0.013 –0.113* –0.119* –0.103* –0.059 0.062 0.093* 0.028
18. Family ownership 0.335* –0.353* –0.168* –0.256* –0.368* –0.292* –0.148* –0.351* –0.135* –0.196* 0.002
19. Institutional ownership –0.229* 0.297* 0.164* 0.273* 0.276* 0.228* 0.140* 0.226* 0.106* 0.203* 0.075
20. Employee ownership –0.168* 0.166* 0.148* 0.176* 0.098* 0.323* 0.004 0.701* 0.027 0.112* –0.001
21. Leverage 0.134* –0.091* –0.102* –0.146* –0.034 0.017 0.004 0.030 –0.090 –0.004 –0.019
22. ROA 0.569* –0.101* –0.043 –0.132* –0.069 –0.123* –0.092* –0.088 –0.079 –0.076 0.036
23. R&D 0.059 –0.109* –0.128* –0.049 –0.085 –0.020 0.018 –0.079 –0.151* –0.037 –0.018
24. Firm size –0.374* 0.527* 0.421* 0.552* 0.343* 0.442* 0.393* 0.202* 0.335* 0.310* 0.151*
25. Law 2013 0.067 0.243* 0.093* 0.356* –0.127* 0.166* 0.165* 0.044 0.099* 0.539* 0.187*
* Represents significance at 0.01 level. All variables are as defined in Table III.1.

131
Table III.3: Continued.
12 13 14 15 16 17 18 19 20 21 22 23 24
12. Board size 1.000
13. Board independence –0.173* 1.000
14. Board gender diversity 0.039 0.129* 1.000
15. Board meetings –0.028 –0.010 0.080 1.000
16. Duality 0.203* –0.174* 0.024 –0.052 1.000
17. CEO tenure –0.002 0.049 0.171* –0.036 0.045 1.000
18. Family ownership –0.143* –0.372* –0.091* –0.173* –0.065 0.157* 1.000
19. Institutional ownership 0.182* 0.311* 0.046 0.056 0.017 0.043 –0.484* 1.000
20. Employee ownership 0.203* –0.082 0.053 –0.164* 0.217* –0.026 –0.290* 0.100* 1.000
21. Leverage –0.064 0.041 0.022 0.120* –0.050 –0.101* –0.150* –0.031 0.118* 1.000
22. ROA –0.161* –0.067 –0.059 –0.046 –0.109* –0.025 0.285* –0.144* –0.102* –0.043 1.000
23. R&D –0.213* –0.028 –0.005 0.180* –0.020 0.085 0.057 –0.056 –0.126* –0.193* –0.093* 1.000
24. Firm size 0.528* 0.070 0.078 0.039 0.116* –0.091* –0.390* 0.303* 0.275* –0.010 –0.290* –0.135* 1.000
25. Law 2013 0.063 0.148* 0.722* 0.092* 0.052 0.172* –0.039 0.025 –0.018 –0.028 –0.076 0.017 0.030
* Represents significance at 0.01 level. All variables are as defined in Table III.1.

132
5.3. Propensity Score Matching

Following Rosenbaum and Rubin (1983), we use the propensity score matching (PSM)

technique to control for firm-level characteristics. We apply the PSM procedure for employee

board representation and each type of employee directors (i.e., labour board representation

and employee-shareholder board representation). Accordingly, we derive a dummy variable to

represent firm-year observations with and without employee board representation. Tables

III.4, III.5 and III.6 report statistics on pre- and post-match pairwise differences of all

variables considered in our study with regard to each matched variable. When we compare

firm-year observations with employee board representation (treatment) to similar firm-year

observations without employee board representation (control), we find that treatment and

control groups present non-significant differences for the post-matched sample as compared

to the pre-matched sample. These results indicate that our matching is effective for reducing

the overlaps between the presence of employee board representation and the firms’ other

characteristics.

133
Table III. 4: Mean Difference Test between Firm-Years with and without Employee Board Representation for Entire and Matched
Samples
Variables Total sample Matched sample
Firm-years with Firm-years without t–value Treated Control t–value
employee employee (n = 242) (n = 242)
directorship directorship
(n = 334) (n = 473)
Tobin’s q 1.143 1.269 1.60 1.268 1.122 1.38
ESG performance (%) 65.87 61.49 6.24*** 65.30 64.78 0.64
Social performance (%) 71.26 67.81 4.08*** 70.20 70.40 0.20
Environmental performance (%) 74.76 68.16 7.68*** 73.38 71.66 1.61
Corporate governance performance (%) 54.10 52.62 1.73* 55.19 55.96 0.72
Standalone (%) 54.49 41.65 3.62*** 51.65 50.01 0.36
CSR assurance (%) 88.02 64.06 7.92*** 81.47 78.03 1.04
CSR committee (%) 66.77 58.56 2.37** 65.70 61.98 0.85
Board size (Number of directors) 14.234 11.994 11.05*** 13.574 13.095 1.55
Board independence (%) 49.35 53.42 2.99*** 53.48 54.07 0.35
Board gender diversity (%) 28.03 19.77 8.14*** 26.35 24.31 1.53
Board meetings (Number of meetings) 7.554 7.230 1.42 7.182 7.376 0.68
Duality (%) 65.27 53.07 3.48*** 59.92 57.44 0.55
CEO tenure (Number of years) 7.614 7.844 0.42 8.405 8.203 0.26
Family ownership (%) 12.91 26.62 8.41*** 16.64 18.25 0.86
Institutional ownership (%) 43.24 28.37 6.98*** 40.47 37.62 1.00
Leverage (%) 25.49 23.48 1.80* 25.45 25.94 0.33
ROA (%) 3.98 5.12 3.46*** 4.26 4.45 0.43
R&D (%) 2.02 3.26 3.50*** 2.25 1.93 0.99
Firm size (in billions of euros) 38.168 14.439 9.03*** 24.852 21.235 1.39
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

134
Table III. 5: Mean Difference Test between Firm-Years with and without Labour Board Representation for Entire and Matched
Samples
Variables Total sample Matched sample
Firm-years with Firm-years without t-value Treated Control t-value
labour representation labour representation (n = 166) (n = 166)
(n = 229) (n = 588)
Tobin’s q 1.279 1.192 1.01 1.369 1.103 1.95*
ESG performance (%) 65.54 62.42 4.02*** 65.23 65.11 0.13
Social performance (%) 71.21 68.45 2.97*** 70.51 71.19 0.56
Environmental performance (%) 75.25 69.16 6.41*** 74.38 73.31 0.84
Corporate governance performance (%) 52.01 53.72 1.83* 53.35 53.75 0.32
Standalone (%) 57.64 42.73 3.86*** 57.23 50.60 1.21
CSR assurance (%) 90.39 67.47 6.87*** 86.75 84.34 0.62
CSR committee (%) 71.62 58.13 3.58*** 69.28 65.06 0.82
Board size (Number of directors) 14.555 12.273 10.20*** 14.069 13.631 1.49
Board independence (%) 44.59 54.57 6.85*** 47.77 47.86 0.04
Board gender diversity (%) 31.02 20.08 10.06*** 28.98 29.75 0.51
Board meetings (Number of meetings) 7.812 7.187 2.53** 7.157 7.121 0.10
Duality (%) 62.45 56.40 1.57 61.45 58.43 0.56
CEO tenure (Number of years) 7.913 7.683 0.38 8.002 9.145 1.21
Family ownership (%) 17.62 22.27 2.51** 20.17 22.74 1.04
Institutional ownership (%) 42.72 31.27 4.84*** 38.93 34.94 1.60
Leverage (%) 25.24 23.94 1.07 24.84 25.73 0.47
ROA (%) 4.23 4.81 1.59 4.40 4.34 0.12
R&D (%) 2.30 2.92 1.58 2.64 1.92 1.63
Firm size (in billions of euros) 43.407 16.674 9.35*** 30.198 27.764 1.62
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively. All variables are as defined in Table III.1.

135
Table III. 6: Mean Difference Test between Firm-Years with and without Employee-Shareholders Representation for Entire and
Matched Samples
Variables Total sample Matched sample
Firm-years with Firm-years without t-value Treated Control t-value
employee- employee- (n = 154) (n = 154)
shareholders shareholders
representation representation
(n = 194) (n = 623)
Tobin’s q 0.949 1.296 3.74*** 1.016 0.927 0.93
ESG performance (%) 67.78 61.98 7.10*** 67.94 66.49 1.72*
Social performance (%) 72.17 68.37 3.82*** 72.62 73.48 0.88
Environmental performance (%) 76.58 69.21 7.28*** 76.31 73.39 2.62***
Corporate governance performance (%) 57.54 51.96 5.68*** 57.81 55.69 1.74*
Standalone (%) 54.89 44.62 2.46** 55.84 60.39 0.81
CSR assurance (%) 88.04 69.82 5.02*** 87.66 85.07 0.66
CSR committee (%) 63.04 61.64 0.34 66.88 68.83 0.36
Board size (Number of directors) 14.381 12.492 7.67*** 14.045 14.071 0.08
Board independence (%) 53.92 51.09 1.76* 55.55 57.23 0.70
Board gender diversity (%) 25.70 22.44 2.64*** 25.75 25.90 0.09
Board meetings (Number of meetings) 7.134 7.432 1.11 7.578 7.695 0.34
Duality (%) 73.91 53.45 5.01*** 70.13 68.18 0.37
CEO tenure (Number of years) 6.668 8.067 2.16** 7.208 6.266 1.31
Family ownership (%) 4.38 25.84 11.61*** 5.24 3.95 1.29
Institutional ownership (%) 47.75 30.62 6.84*** 46.93 42.78 1.18
Leverage (%) 25.66 23.91 1.34 26.41 26.32 0.05
ROA (%) 3.66 4.94 3.32*** 3.89 3.52 0.84
R&D (%) 1.89 2.99 2.65*** 1.76 1.68 0.26
Firm size (in billions of euros) 37.518 20.344 5.40*** 35.127 35.397 0.05
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

136
5.4. Test of H1a to H3b

Table III.7 presents the system GMM regression results of ESG performance and its related

components (i.e., environmental, social and corporate governance) on employee board

representation. Findings show that the impact of employee directors is negative and

significant on ESG performance and its three components, as documented in Models 1, 2, 3

and 4. These results suggest that board-level employee representation reduces the extra-

financial performance of the firm. Unexpectedly, results of Model 2 report a negative impact

of employee directors on social performance, suggesting that increasing the board-level

employee representation would result in less engagement in social issues. Our finding from

Model 3 concerning the environmental pillar is contrary to Hillman et al. (2001), who report a

significant and positive impact of employee directorship on firms’ environmental

performance. One explanation is that the competence and skills of employee directors may

not be very relevant for improving firms’ environmental performance. Results of Model 4

show a negative and significant relationship between employee directors and corporate

governance performance, suggesting that board-level employee representation adversely

affects the efficiency of the corporate board. This finding could be explained by the

divergence in interests between employee directors and other directors representing outside

shareholders (Ginglinger et al., 2011).

With regard to the control variables, results of Table III.7 show that firms adopting a

CSR orientation – having a standalone report, CSR assurance and a CSR committee – are

more likely to have higher levels of social, environmental and corporate governance

performance. Similarly to previous studies (Boulouta, 2013; Hafsi & Turgut, 2013; Johnson

& Greening, 1999), we find that board characteristics significantly impact the ESG

performance of the firm. Indeed, firms with large boards exhibit lower ESG and corporate

governance performance. In line with Boulouta (2013) and Dunn and Sainty (2009), we find

137
that independent and female directors enhance overall ESG performance. In particular,

independent directors enhance environmental and corporate governance performance, but

reduce social performance. Female directors are shown to enhance only the overall ESG

performance. Surprisingly, Table III.7 shows that boards that meet frequently have lower

social and environmental performance. Duality is found to enhance the overall ESG

performance, as well as environmental and social performance, but reduces corporate

governance performance. CEO tenure is found to decrease the ESG performance in general

and corporate governance performance in particular, suggesting that CEOs with long tenure

are less acquiescent to stakeholder pressures. Our findings show that family ownership

influences the extra-financial performance of the firm. In line with Rees and Rodionova

(2015), we find that family firms have lower overall ESG, social and corporate governance

performance, but that their environmental performance is unaffected. As regards institutional

ownership, we confirm the finding of Barnea and Rubin (2010) by reporting no significant

impact on corporate social performance. Firms with higher level of debt are more likely to

have lower ESG performance in general and social performance in particular. Financial

performance as measured by ROA has a negative and significant impact on the overall ESG

and the social performance component. We also show that firms with higher R&D intensity

have higher environmental and corporate governance performance but lower social

performance. Finally, large firms are more likely to have higher ESG performance, considered

as a whole or in terms of the underlying pillars (i.e., social, environmental and corporate

governance).

138
Table III. 7: System GMM Regression of ESG Performance on Employee Board Representation
Model 1: Model 2: Model 3: Model 4:
ESG Social Environmental Corporate governance
Variables
performance performance performance performance
Coef. t-test Coef. t-test Coef. t–test Coef. t-test
Lag ESG performance 0.552*** 47.58
Lag Social performance 0.681*** 46.56
Lag Environmental performance 0.747*** 87.29
Lag Corporate governance performance 0.483*** 39.08
Employee directorship –0.088*** –5.50 –0.046** –2.18 –0.095*** –5.15 –0.174*** –7.70
Standalone 0.005* 1.95 0.006** 2.53 0.006*** 3.26 0.012*** 3.89
CSR assurance 0.014*** 3.25 –0.001 –0.16 0.013*** 4.38 0.040*** 7.85
CSR committee 0.011*** 4.06 0.012*** 5.41 0.011*** 3.47 0.013*** 5.05
Board size –0.018*** –2.87 –0.004 –0.73 –0.009 –1.46 –0.031*** –2.95
Board independence 0.045*** 5.40 0.008 0.96 0.040*** 6.46 0.036*** 3.06
Board gender diversity 0.036*** 2.98 –0.001 –0.02 –0.014 –1.22 0.005 0.41
Board meetings –0.001 –0.27 –0.008** –2.52 –0.006*** –2.71 –0.005 –1.37
Duality 0.005* 1.87 0.008*** 3.90 0.012*** 5.90 –0.023*** –5.35
CEO tenure –0.003*** –3.30 –0.001 –0.53 –0.002 –1.61 –0.008*** –3.32
Family ownership –0.020*** –2.91 –0.030*** –3.30 0.006 1.28 –0.093*** –6.19
Institutional ownership 0.001 0.16 0.001 0.17 –0.001 –0.35 –0.010 –1.39
Leverage –0.015*** –2.62 0.003 0.32 –0.012 –1.61 –0.018* –1.97
ROA –0.059** –2.13 –0.056** –2.15 –0.034 –1.08 0.067 1.25
R&D 0.044 1.26 –0.142*** –3.33 0.155*** 2.94 0.125** 2.43
Firm size 0.011*** 9.04 0.008*** 5.98 0.005*** 3.75 0.012*** 4.47
Law 2013 –0.003* –1.90 –0.002 –1.12 0.020*** 11.21 –0.045*** –14.23
Intercept 0.130*** 6.77 0.125*** 5.79 0.101*** 4.84 0.195*** 5.43
Industry Yes Yes Yes Yes
Number of obs. 444 444 444 444
F (Prob > F) 1952.51 (p = 0.000) 652.11 (p = 0.000) 7797.37 (p = 0.000) 2829.84 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –5.21 (p = 0.000) –4.80 (p = 0.000) –5.40 (p = 0.000) –4.67 (p = 0.000)
Arellano–Bond test AR(2) (z, p–value): 0.73 (p = 0.361) 0.62 (p = 0.354) 0.81 (p = 0.205) –0.38 (p = 0.420)
Sargan test (Chi–square, p–value): 365.65 (p = 0.000) 380.32 (p = 0.000) 362.33 (p = 0.000) 405.44 (p = 0.000)
Hansen test (Chi–square, p–value): 64.26 (p = 0.105) 58.03 (p = 0.232) 64.17 (p = 0.102) 61.80 (p = 0.143)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively. All variables are as defined in Table III.1.

139
 Tables III.8 and III.9 present the results of the system GMM regressions of ESG

performance and the underlying pillars (i.e., social, environmental and corporate governance)

on labour board representation and employee-shareholder board representation, respectively.

Our results are consistent with the fact that the two types of employee directors do not share

the same interests and do not use their power on the board in the same way. Table III.8 shows

that the impact of labour board representation is negative and significant on ESG performance

as a whole. Going further, when we split ESG performance into its different components, we

find that labour board representation enhances the social component of ESG performance

(Model 2) and negatively impacts environmental performance (Model 3) and corporate

governance performance (Model 4). Consequently, we confirm H1a and H3a and reject H2a.

In contrast to Table III.8, Table III.9 reveals a positive and significant relationship between

the proportion of directors representing employee-shareholders and ESG performance. Table

III.9 also shows that directors elected by employee-shareholders act in the opposite direction

to that of labour board representatives with respect to each component of ESG performance.

Specifically, the ESG orientations of directors representing employee-shareholders are

directed towards environmental and governance performance rather than towards social

performance. The relationships predicted by H1b, H2b, and H3b are valid.

140
Table III. 8: System GMM Regression of ESG Performance on Labour Board Representation
Model 1: Model 2: Model 3: Model 4:
ESG performance Social Environmental Corporate governance
Variables
performance performance performance
Coef. t-test Coef. t-test Coef. t-test Coef. t-test
Lag ESG performance 0.545*** 58.11
Lag Social performance 0.756*** 37.29
Lag Environmental performance 0.665*** 55.14
Lag Corporate governance performance 0.475*** 49.73
Labour representation –0.113*** –8.43 0.067** 2.51 –0.067*** –5.32 –0.196*** –12.79
Standalone 0.013*** 5.25 0.002 0.69 0.013*** 4.37 0.014*** 3.45
CSR assurance 0.020*** 6.80 –0.005 –0.82 0.014*** 5.98 0.009* 1.96
CSR committee 0.005* 1.67 –0.001 –0.12 0.008*** 2.78 –0.001 –0.17
Board size –0.034*** –5.25 –0.011 –0.98 –0.021*** –2.62 –0.042*** –2.94
Board independence 0.064*** 12.41 0.041*** 2.87 0.061*** 6.64 0.023** 2.21
Board gender diversity 0.084*** 8.90 0.009 0.63 0.053*** 7.23 0.029** 2.14
Board meetings –0.006*** –4.62 –0.002 –0.71 –0.007** –2.53 –0.001 –0.26
Duality 0.019*** 10.73 0.009** 2.04 0.022*** 6.24 –0.011** –2.43
CEO tenure –0.009*** –10.09 –0.005** –2.29 –0.007*** –5.54 –0.013*** –6.60
Family ownership 0.002 0.25 –0.019* –1.95 0.012 1.19 –0.050*** –6.46
Institutional ownership 0.028*** 6.81 0.012* 1.69 0.025*** 3.51 0.026*** 3.07
Leverage –0.017*** –3.01 –0.009 –0.81 –0.023*** –3.06 –0.041*** –6.60
ROA 0.022 0.75 –0.075 –1.60 –0.034 –0.76 0.114*** 2.74
R&D –0.025 –1.05 –0.198*** –3.42 0.032 0.59 0.106** 2.05
Firm size 0.012*** 9.23 0.002 0.95 0.010*** 5.64 0.013*** 5.17
Law 2013 –0.020*** –13.14 –0.008*** –2.70 0.011*** 6.95 –0.034*** –16.93
Intercept 0.132*** 5.67 0.173*** 4.84 0.078*** 2.62 0.218*** 7.47
Industry Yes Yes Yes Yes
Number of obs. 313 313 313 313
F (Prob > F) 3769.38 (p = 0.000) 839.24 (p = 0.000) 7950.47 (p = 0.000) 6676.49 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –3.95 (p = 0.000) –3.77 (p = 0.000) –4.80 (p = 0.000) –3.42 (p = 0.001)
Arellano–Bond test AR(2) (z, p–value): 1.04 (p = 0.302) 1.44 (p = 0.150) 1.24 (p = 0.125) –0.94 (p = 0.349)
Sargan test (Chi–square, p–value): 251.65 (p = 0.000) 65.42 (p = 0.001) 279.88 (p = 0.000) 289.78 (p = 0.000)
Hansen test (Chi–square, p–value): 58.82 (p = 0.211) 40.89 (p = 0.228) 60.50 (p = 0.223) 56.87 (p = 0.266)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively. All variables are as defined in Table III.1.

141
Table III. 9: System GMM Regression of ESG Performance on Employee-Shareholders Board Representation
Model 1: Model 2: Model 3: Model 4:
ESG performance Social Environmental Corporate governance
Variables
performance performance performance
Coef. t-test Coef. t-test Coef. t-test Coef. t-test
Lag ESG performance 0.441*** 28.52
Lag Social performance 0.496*** 22.32
Lag Environmental performance 0.721*** 34.47
Lag Corporate governance performance 0.376*** 25.91
Employee-shareholders representation 0.449*** 8.59 –0.606*** –6.39 0.414*** 7.67 1.047*** 10.01
Standalone –0.008** –2.23 0.002 0.48 –0.005 –1.63 0.003 0.69
CSR assurance 0.025*** 5.95 0.003 0.39 0.005 1.03 0.022* 1.70
CSR committee 0.018*** 5.39 –0.001 –0.17 0.018*** 6.34 0.036*** 4.71
Board size –0.022 –1.61 0.021* 1.67 –0.011 –1.37 –0.039* –1.71
Board independence 0.085*** 7.03 0.019 1.04 0.049*** 4.65 0.104*** 5.86
Board gender diversity 0.051*** 4.73 0.074*** 4.49 –0.016 –1.12 –0.045* –1.69
Board meetings 0.021*** 3.96 0.006 0.47 0.007* 1.66 0.011 1.44
Duality –0.004 –0.96 0.012** 2.04 0.013*** 4.11 –0.049*** –9.73
CEO tenure 0.003 1.17 0.014*** 4.91 –0.005** –2.33 0.004 0.72
Family ownership 0.037 1.25 –0.083*** –2.96 0.057** 2.04 0.034 0.72
Institutional ownership –0.022** –2.36 –0.027*** –3.48 –0.009 –1.37 –0.018 –1.34
Leverage –0.009 –0.54 0.027 1.57 –0.017 –1.12 –0.001 –0.06
ROA 0.027 0.62 0.245*** 8.97 0.054 0.94 –0.168*** –3.15
R&D 0.015 0.11 0.110* 1.69 –0.026 –0.47 0.122 0.48
Firm size 0.009*** 4.94 0.007** 2.53 0.006*** 3.98 0.002 0.97
Law 2013 –0.004** –2.19 –0.022*** –6.60 0.032*** 10.56 –0.041*** –8.10
Intercept 0.146*** 3.73 0.167*** 3.04 0.062* 1.84 0.322*** 6.22
Industry Yes Yes Yes Yes
Number of obs. 288 288 288 288
F (Prob > F) 1742.01 (p = 0.000) 2377.08 (p = 0.000) 7455.16 (p = 0.000) 8314.40 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –3.48 (p = 0.000) –3.46 (p = 0.000) –3.25 (p = 0.001) –3.47 (p = 0.001)
Arellano–Bond test AR(2) (z, p–value): –1.07 (p = 0.252) 0.06 (p = 0.956) 1.44 (p = 0.150) –1.83 (p = 0.167)
Sargan test (Chi–square, p–value): 231.65 (p = 0.000) 248.87 (p = 0.000) 249.77 (p = 0.000) 254.85 (p = 0.000)
Hansen test (Chi–square, p–value): 42.88 (p = 0.169) 43.14 (p = 0.162) 41.97 (p = 0.194) 44.62 (p = 0.128)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively. All variables are as defined in Table III.1.

142
5.5. Test of H4a and H4b

To test H4a and H4b, we perform a system GMM regression of Tobin’s q (as a proxy of

market perception) on ESG performance, according to the way employees are represented on

the board. We use the joint test procedure, which simultaneously tests the main effect of ESG

performance and its interaction with employee board representation. Our estimation model is

as follows:

Tobin’s qi,t = β0 + β1 Lag Tobin’s qi,t + β2 ESG performancei,t + β3 Employee

directorshipi,t + β4 (ESG performancei,t * Employee directorshipi,t) + β5 Standalonei,t + β6

CSR assurancei,t + β7 CSR committeei,t + β8 Board sizei,t + β9 Board independencei,t + β10

Board gender diversityi,t + β11 Board meetingi,t + β12 Dualityi,t + β13 CEO tenurei,t + β14 Family

ownershipi,t + β15 Institutional ownershipi,t + β16 Leveragei,t + β17 ROAi,t + β18 R&Di,t + β19

Firm size i,t + β20 Law 2013_FE + β21 Industry_FE + εi,t

where i and t stand for firms and time respectively. Tobin’s q is our dependent variable.

Employee directorship and ESG performance are endogenously determined. Employee

directorship is decomposed into labour board representation and employee-shareholder board

representation, while ESG performance is broken down to illustrate the impact of each of its

components (i.e., social, environmental and corporate governance). ε is the error term. All

variables are as defined in Table III.1.

In accordance with Nekhili et al. (2019), results of Table III.10 show a negative impact

of employee board representative on Tobin’s q, suggesting that market participants are

unlikely to favour employee board representation. Nevertheless, opposing results with regard

to employee directorship are observed in Model 2 and Model 3. Based on the value they

provide to the company, market participants are likely to be more favourable towards

directors elected by employee-shareholders than to labour board representation. In accordance

with Ginglinger et al. (2011), directors elected by employee-shareholders may act in the same

143
way as other directors representing outside shareholders. Results of Table III.10 also show

that the overall ESG performance has a positive impact on Tobin’s q in Model 3, albeit not

significant in Model 1 and Model 2. This finding suggests that outside shareholders interpret

and perceive ESG performance differently, depending on to the way employees are

represented on the board.

Table III. 10: System GMM Regression of Tobin's q on ESG Performance and
Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-
Variables directorship representation shareholders
representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.717*** 71.30 0.884*** 85.16 0.929*** 100.38
ESG performance 0.078 0.30 0.303 1.22 0.376*** 3.89
Employee directorship –1.855*** –5.55
Labour representation –2.575*** –8.17
Employee-shareholders representation 1.659*** 5.09
Standalone 0.043 1.58 0.021 0.61 0.041*** 3.80
CSR assurance 0.157*** 5.04 0.044 0.89 –0.013 –0.52
CSR committee 0.024 0.53 0.003 0.09 0.026** 2.25
Board size 0.110 1.10 0.201* 1.97 0.002 0.06
Board independence –0.229* –1.78 –0.137 –1.04 0.082* 1.87
Board gender diversity –0.008 –0.07 0.054 0.41 0.044 0.79
Board meetings 0.003 0.08 –0.005 –0.11 0.039*** 2.62
Duality –0.089** –2.36 –0.084** –2.08 –0.007 –0.46
CEO tenure –0.022 –1.25 –0.047** –2.28 –0.023*** –3.28
Family ownership 0.117 0.89 0.267** 2.21 0.305*** 5.34
Institutional ownership 0.016 0.18 0.102 1.10 0.006 0.23
Leverage 0.457*** 5.03 0.025 0.26 0.218*** 4.21
ROA 3.030*** 8.14 2.037*** 8.39 1.039*** 5.45
R&D 1.022 1.00 2.240*** 3.85 –0.273 –0.76
Firm size –0.035 –1.39 0.012 0.43 –0.031*** –4.03
Law 2013 0.099*** 3.84 0.042 1.25 0.067*** 4.17
Intercept 0.342 0.94 –0.753 –1.57 –0.025 –0.15
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 2167.21 (p = 0.000) 3128.26 (p = 0.000) 6785.96 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.43 (p = 0.003) –3.06 (p = 0.000) –2.58 (p = 0.003)
Arellano–Bond test AR(2) (z, p–value): –0.74 (p = 0.462) –1.02 (p = 0.308) –1.10 (p = 0.269)
Sargan test (Chi–square, p–value): 131.71 (p = 0.000) 107.63 (p = 0.000) 136.28 (p = 0.000)
Hansen test (Chi–square, p–value): 43.00 (p = 0.193) 46.57 (p = 0.159) 41.48 (p = 0.199)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

In Table III.11, we measure the marginal effect of employee directorship, labour board

representation and employee-shareholder board representation on the ESG-financial

performance relationship. For the overall employee board representation (Model 1), the sum

144
of coefficient (ESG performance + (ESG performance * Employee directorship)) on Tobin’s

q is negative (β2 + β4 = –2.739) and significant (t = –3.54), suggesting that outside

shareholders are reluctant to favour an increase in the level of ESG performance of firms

when employee directors are represented on the board. Models 2 and 3 show that the results

of the marginal effect of employee directorship on the value relevance of ESG performance

varies depending on the way employees are represented on board. Meaningfully, the joint

coefficient (ESG performance + (ESG performance * Labour representation)) in Model 2 is

strongly negative (β2 + β4 = –16.652) and significant at the 1 % level (t = –4.06). Our results

clearly show that this negative effect comes mainly from the adversely perceived labour board

representation and not from the firm’s orientation in terms of ESG performance. In contrast,

with regard to the marginal effect of directors representing employee-shareholders, results of

Model 3 indicate that the impact of the sum of coefficients (ESG performance + (ESG

performance * employee-shareholders representation)) on Tobin’s q is highly positive (β2 + β4

= 20.964) and significant (t = 4.65). In accordance with H4a and H4b, these findings

emphasize that outside shareholders are highly sensitive to the type of employee directors

when assessing the relevance of the firm’s ESG performance.

145
Table III. 11: System GMM Regression of Tobin's q on the Interaction between ESG
Performance and Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.497*** 96.13 0.830*** 51.34 0.956*** 101.15
ESG performance 0.521*** 4.77 1.885*** 3.87 –0.763*** –4.25
Employee directorship 2.121*** 3.88
ESG performance * Employee –3.261*** –3.87
directorship
Labour representation 11.859*** 3.97
ESG performance * Labour –18.537*** –4.16
representation
Employee-shareholders representation –13.538*** –4.08
ESG performance * Employee- 21.727*** 4.70
shareholders representation
Standalone 0.010 0.56 –0.029 –0.84 0.063*** 5.47
CSR assurance 0.081*** 3.76 –0.109** –2.29 0.011 0.44
CSR committee 0.077*** 3.61 –0.015 –0.47 0.026 1.52
Board size –0.060 –1.04 0.161* 1.88 0.004 0.11
Board independence –0.029 –0.68 0.005 0.05 0.127** 2.45
Board gender diversity –0.031 –0.57 –0.026 –0.21 0.027 0.58
Board meetings 0.031 1.57 –0.056 –1.59 0.012 0.57
Duality –0.116*** –5.32 –0.097** –2.52 0.042*** 2.77
CEO tenure –0.022* –1.64 –0.026 –1.58 –0.018** –2.22
Family ownership 0.283*** 3.85 0.335*** 3.44 0.191** 2.23
Institutional ownership –0.115*** –3.44 –0.062 –0.92 0.038* 1.70
Leverage 0.802*** 16.27 0.118 1.28 0.145** 2.32
ROA 5.651*** 34.40 2.675*** 9.95 1.087*** 5.62
R&D 2.391*** 7.14 2.251*** 3.78 0.576** 2.14
Firm size –0.118*** –10.09 –0.028 –1.16 –0.025*** –3.31
Law 2013 0.073*** 6.00 0.091*** 3.79 0.058*** 4.50
Intercept 1.743*** 8.48 –0.943** –2.05 0.599*** 3.26
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 42872.70 (p = 0.000) 7356.76 (p = 0.000) 6051.65 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.37 (p = 0.006) –3.09 (p = 0.001) –2.64 (p = 0.003)
Arellano–Bond test AR(2) (z, p–value): –1.09 (p = 0.279) –0.93 (p = 0.272) –0.76 (p = 0.446)
Sargan test (Chi–square, p–value): 367.43 (p = 0.000) 222.57 (p = 0.000) 227.75 (p = 0.000)
Hansen test (Chi–square, p–value): 50.72 (p = 0.367) 35.74 (p = 0.297) 37.50 (p = 0.231)
Joint test: ESG performance + (ESG –2.739*** –3.54
performance * Employee directorship)

Joint test: ESG performance + (ESG –16.652*** –4.06

performance * Labour representation)
Joint test: ESG performance + (ESG 20.964*** 4.65
performance * Employee-shareholders
representation)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

146
5.6. Supplementary Analysis

Tables III.12, III.13 and III.14 present the results for the effect of each pillar of ESG

performance, namely social, environmental and corporate governance on Tobin’s q,

respectively. Table III.12 shows that social performance is negatively perceived by the

financial market for firm-years with employee directorships (Model 1) and particularly for

firm-years with labour board representation (Model 2). Non-significance is, however,

observed for firms with employee-shareholder board representation (Model 3). One

explanation is that investors may consider high levels of social performance as an excessive

and costly investment or the result of an entrenchment strategy (Surroca & Tribó, 2008;

Nekhili et al., 2019), especially when employees are appointed to the board in compliance

with employment rights. The results of Models 1 and 2 in Table III.13 show that

environmental performance and the presence of employees on the board are negatively

perceived by market participants. Corporate governance performance is financially rewarded

by the market, as reported in Models 1, 2 and 3 of Table III.14. This positive relationship

between corporate governance performance and Tobin’s q indicates that outside shareholders

find information on such ESG performance pillar relevant and that well-governed firms are

likely to attract market participants (Fauver & Fuerst, 2006). Once again, we find that the

negative relationship between employee directorship and Tobin’s q is mainly due to the

presence of labour board representatives (Model 2) and not to the presence of directors

representing employee-shareholders (Model 3).

147
Table III. 12: System GMM Regression of Tobin's q on Social Performance and
Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-
Variables directorship representation shareholders
representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.724*** 43.44 0.879*** 66.61 0.935*** 86.68
Social representation –1.677*** –4.01 –1.975*** –9.81 –0.103 –0.82
Employee directorship –4.165*** –7.34
Labour representation –2.565*** –6.92
Employee-shareholders representation 2.274*** 6.21
Standalone 0.109** 2.37 0.083** 2.18 0.058*** 4.32
CSR assurance 0.321*** 6.04 0.103** 2.17 –0.043 –1.51
CSR committee 0.060 1.02 0.070* 1.64 0.062*** 3.21
Board size 0.330** 2.06 0.094 0.89 –0.001 –0.05
Board independence –0.610*** –3.04 0.110 0.76 0.143*** 2.65
Board gender diversity 0.135 0.62 0.104 0.72 0.107* 1.65
Board meetings 0.010 0.17 0.031 0.77 0.036* 1.70
Duality –0.067 –0.96 –0.024 –0.46 –0.005 –0.28
CEO tenure –0.027 –0.97 –0.078*** –3.04 –0.012 –1.29
Family ownership –0.195 –1.04 0.254* 1.92 0.361*** 3.47
Institutional ownership 0.104 0.73 0.127* 1.67 –0.023 –0.89
Leverage 0.704*** 3.83 –0.221** –2.32 0.197*** 2.58
ROA 2.647*** 4.69 2.044*** 6.67 1.003*** 4.25
R&D 0.395 0.27 1.103* 1.66 0.050 0.13
Firm size 0.034 0.73 0.078** 2.39 –0.026** –2.40
Law 2013 0.052 1.37 –0.028 –0.69 0.054*** 3.69
Intercept –0.018 –0.03 –0.148 –0.26 0.157 0.89
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 411.06 (p = 0.000) 6189.08 (p = 0.000) 7164.21 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.48 (p = 0.005) –3.08 (p = 0.000) –2.60 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.60 (p = 0.549) –0.95 (p = 0.342) –0.94 (p = 0.349)
Sargan test (Chi–square, p–value): 108.85 (p = 0.000) 101.30 (p = 0.000) 235.84 (p = 0.000)
Hansen test (Chi–square, p–value): 41.93 (p = 0.112) 46.65 (p = 0.158) 37.08 (p = 0.286)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

148
Table III. 13: System GMM Regression of Tobin's q on Environmental Performance
and Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-
Variables directorship representation shareholders
representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.702*** 49.54 0.868*** 177.65 0.935*** 123.86
Environmental performance –0.676*** –2.84 –0.103 –1.41 0.195** 2.17
Employee directorship –2.061*** –5.30
Labour representation –0.905*** –7.58
Employee-shareholders representation 1.472*** 5.42
Standalone 0.053 1.44 0.019** 2.34 0.037** 2.54
CSR assurance 0.181*** 4.50 –0.004 –0.30 –0.002 –0.08
CSR committee 0.063 1.62 0.042*** 2.76 0.034*** 2.93
Board size 0.098 0.86 0.013 0.53 –0.010 –0.26
Board independence –0.118 –0.93 0.082* 1.75 0.071 1.60
Board gender diversity –0.075 –0.49 –0.086* –1.72 0.070 1.11
Board meetings –0.054 –1.19 0.005 0.44 0.046*** 3.11
Duality –0.020 –0.43 –0.043*** –2.86 –0.010 –0.74
CEO tenure –0.020 –1.04 –0.042*** –5.97 –0.020*** –2.57
Family ownership 0.165 1.22 0.202*** 5.96 0.288*** 4.38
Institutional ownership 0.066 0.66 0.062** 2.21 0.009 0.39
Leverage 0.516*** 5.89 –0.048* –1.79 0.217*** 4.94
ROA 3.312*** 8.01 2.456*** 35.76 0.989*** 5.66
R&D 1.548 1.22 1.229*** 7.04 –0.309 –1.06
Firm size –0.013 –0.45 –0.009 –0.97 –0.032*** –4.19
Law 2013 0.131*** 4.45 0.062*** 5.26 0.059*** 3.85
Intercept 0.466 0.97 0.215* 1.75 0.106 0.71
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 1435.92 (p = 0.000) 5393.59 (p = 0.000) 12785.57 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.45 (p = 0.005) –3.06 (p = 0.002) –2.58 (p = 0.005)
Arellano–Bond test AR(2) (z, p–value): –0.71 (p = 0.478) –1.07 (p = 0.284) –1.03 (p = 0.303)
Sargan test (Chi–square, p–value): 125.68 (p = 0.000) 230.74 (p = 0.000) 133.73 (p = 0.000)
Hansen test (Chi–square, p–value): 41.08 (p = 0.130) 46.22 (p = 0.546) 40.25 (p = 0.124)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

149
Table III. 14: System GMM Regression of Tobin's q on Corporate Governance
Performance and Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.705*** 73.29 0.858*** 92.25 0.944*** 149.87
Corporate governance performance 1.207*** 5.85 0.404*** 7.09 0.188*** 3.33
Employee directorship –2.547*** –7.36
Labour representation –0.675** –2.46
Employee-shareholders representation 1.417*** 4.31
Standalone 0.070** 2.05 0.009 0.42 0.047*** 5.01
CSR assurance 0.123*** 2.54 –0.007 –0.28 –0.043** –2.23
CSR committee 0.025 0.47 –0.001 –0.01 0.023** 2.15
Board size 0.191* 1.76 0.057 1.11 0.028 0.82
Board independence –0.504*** –3.38 –0.039 –0.67 0.060** 2.39
Board gender diversity 0.140 0.89 –0.151** –2.20 0.106* 1.87
Board meetings –0.010 –0.21 0.006 0.21 0.036*** 2.56
Duality –0.056 –1.27 –0.031 –1.21 0.011 1.12
CEO tenure –0.007 –0.28 –0.017 –1.42 –0.018*** –2.63
Family ownership 0.255* 1.79 0.235*** 4.82 0.237*** 3.22
Institutional ownership 0.039 0.42 –0.006 –0.10 0.018 0.76
Leverage 0.699*** 7.25 0.034 0.56 0.178*** 3.93
ROA 3.591*** 8.23 2.398*** 13.15 0.997*** 5.16
R&D 1.028 0.80 1.280*** 3.89 –0.063 –0.25
Firm size –0.019 –0.75 –0.021 –1.50 –0.024*** –3.79
Law 2013 0.171*** 6.31 0.084*** 4.42 0.065*** 3.91
Intercept –0.733* –1.86 0.049 0.19 –0.042 –0.32
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 1725.49 (p = 0.000) 4852.70 (p = 0.000) 6874.50 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.46 (p = 0.000) –3.06 (p = 0.004) –2.61 (p = 0.003)
Arellano–Bond test AR(2) (z, p–value): –0.75 (p = 0.455) –1.08 (p = 0.281) –0.95 (p = 0.342)
Sargan test (Chi–square, p–value): 115.33 (p = 0.000) 229.80 (p = 0.000) 240.27 (p = 0.000)
Hansen test (Chi–square, p–value): 38.89 (p = 0.187) 41.92 (p = 0.137) 36.88 (p = 0.294)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

150
 Tables III.15, III.16 and III.17 present the results of the marginal effect of employee

directorship on the value relevance of social, environmental and corporate governance

performance, respectively. The results of Model 1 in Table III.15 are in accordance with those

of Nekhili et al. (2019). Going further, we show that this marginal effect differs according to

the way employees are represented on the board. Results of Model 2 show that the joint

coefficient (Social performance + (Social performance * Labour representation)) is strongly

negative (β2 + β6 = –18.370) and highly significant (t = –7.48). Conversely, the joint test

carried out in Model 3 shows a positive (β2 + β4 = 14.934) and significant (t = 3.69)

relationship between the joint coefficient (Social performance + (Social performance *

Employee-shareholders representation)) and Tobin’s q. These results suggest that, in contrast

to the presence of directors representing employee-shareholders, the presence of labour board

representatives indirectly affects firm valuation by altering outside shareholders’ perception

of a higher level of social performance. With respect to environmental performance, results of

Models 1, 2 and 3 in Table III.16 highlight a negative and significant impact for all the joint

tests considered. The results obtained for corporate governance performance in Table III.17

are similar to those observed for social performance. Once again, these results provide

evidence that market participants react positively to the presence of directors representing

employee-shareholders, as this category of employee directors may operate along the same

lines as other directors representing outside shareholders (Ginglinger et al., 2011).

151
Table III. 15: System GMM Regression of Tobin's Q on the Interaction between Social
Performance and Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.831*** 153.07 0.821*** 56.76 0.934*** 86.12
Social performance –1.718*** –7.56 0.301 1.44 –0.619*** –2.57
Employee directorship 6.106*** 4.10
Social performance * Employee –9.827*** –4.80
directorship
Labour representation 14.790*** 7.56
Social performance * Labour –18.671*** –7.29
representation
Employee-shareholders representation –8.624*** –2.90
Social performance * Employee- 15.553*** 3.68
shareholders representation
Standalone 0.117*** 5.68 0.028 1.02 0.071*** 3.36
CSR assurance 0.150*** 8.49 –0.044 –1.25 0.009 0.38
CSR committee 0.099*** 5.66 0.051 1.53 0.057** 2.19
Board size 0.070 1.03 –0.012 –0.11 0.027 0.43
Board independence –0.067 –0.85 0.258*** 2.85 0.177*** 3.04
Board gender diversity –0.049 –0.43 –0.069 –0.65 0.097 1.49
Board meetings –0.024 –0.85 0.020 0.54 0.027 1.28
Duality –0.068** –2.17 –0.043 –1.09 0.011 0.64
CEO tenure –0.026*** –3.00 –0.059*** –3.15 –0.028*** –3.42
Family ownership –0.124* –1.79 0.137* 1.77 0.215 1.62
Institutional ownership –0.007 –0.15 –0.161*** –2.70 0.007 0.23
Leverage 0.174*** 3.15 –0.108 –1.18 0.131* 1.95
ROA 1.547*** 7.33 2.245*** 8.39 1.208*** 5.02
R&D –0.016 –0.05 0.189 0.40 0.398 0.86
Firm size 0.030** 2.05 –0.043** –2.38 –0.036*** –2.82
Law 2013 0.079*** 3.05 0.067** 2.52 0.060*** 4.02
Intercept 0.572** 2.30 0.552 1.39 0.577*** 2.82
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 9734.58 (p = 0.000) 2576.80 (p = 0.000) 9855.65 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.50 (p = 0.005) –3.06 (p = 0.001) –2.55 (p = 0.005)
Arellano–Bond test AR(2) (z, p–value): –0.29 (p = 0.773) –1.04 (p = 0.297) –0.95 (p = 0.340)
Sargan test (Chi–square, p–value): 125.32 (p = 0.000) 90.19 (p = 0.000) 226.78 (p = 0.000)
Hansen test (Chi–square, p–value): 45.53 (p = 0.453) 35.61 (p = 0.221) 36.54 (p = 0.266)
Joint test: Social performance + (Social –11.545*** –6.12
performance * Employee directorship)
Joint test: Social performance + (Social –18.370*** –7.48
performance * Labour representation)
Joint test: Social performance + (Social 14.934*** 3.69
performance * Employee-shareholders
representation)

*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

152
Table III. 16: System GMM Regression of Tobin's q on the Interaction between
Environmental Performance and Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.798*** 157.17 0.836*** 81.54 0.908*** 111.02
Environmental performance 2.229*** 15.64 0.911*** 2.96 0.638** 2.52
Employee directorship 18.445*** 13.72
Environmental performance * Employee –24.089*** –13.05
directorship
Labour representation 11.600*** 4.43
Environmental performance * Labour –17.274*** –4.82
representation
Employee-shareholders representation 11.833*** 4.94
Environmental performance * –14.131*** –4.49
Employee-shareholders representation
Standalone 0.043** 2.50 0.056** 2.05 0.040 3.22
CSR assurance –0.052* –1.83 –0.103*** –3.04 –0.037 –1.49
CSR committee 0.041** 2.45 0.039 1.07 0.038*** 2.58
Board size 0.068* 1.67 0.037 0.58 0.001 0.01
Board independence –0.082 –1.63 0.083 0.95 0.087** 2.26
Board gender diversity –0.067 –0.75 0.143 1.15 0.304*** 5.89
Board meetings –0.003 –0.14 –0.082** –2.33 0.040*** 2.84
Duality –0.180*** –8.22 –0.113*** –2.97 –0.019 –1.52
CEO tenure –0.018** –2.26 –0.030* –1.95 –0.012 –1.54
Family ownership 0.092* 1.96 0.400*** 4.83 0.256*** 3.43
Institutional ownership –0.056 –1.62 0.152** 2.26 0.025 1.20
Leverage 0.437*** 7.44 0.097 1.27 0.345*** 6.19
ROA 2.363*** 15.49 3.025*** 13.58 0.908*** 5.37
R&D 1.158** 2.42 2.854*** 4.81 0.082 0.26
Firm size –0.079*** –6.39 0.019 1.02 –0.025*** –2.91
Law 2013 0.076*** 4.44 0.072*** 2.88 0.042*** 4.38
Intercept –0.378 –1.39 –1.008*** –3.34 –0.400** –2.33
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 20798.85 (p = 0.000) 5261.69 (p = 0.000) 31680.82 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.46 (p = 0.004) –3.09 (p = 0.001) –2.58 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.45 (p = 0.652) –0.85 (p = 0.396) –1.19 (p = 0.235)
Sargan test (Chi–square, p–value): 337.58 (p = 0.000) 226.36 (p = 0.000) 110.80 (p = 0.000)
Hansen test (Chi–square, p–value): 60.88 (p = 0.119) 42.50 (p = 0.102) 41.21 (p = 0.127)
Joint test: Environmental performance + –21.860*** –12.36
(Environmental performance *
Employee directorship)
Joint test: Environmental performance + –16.363*** –4.90
(Environmental performance * Labour
representation)
Joint test: Environmental performance + –13.493*** –4.63
(Environmental performance *
Employee-shareholder board
representation)

*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1

153
Table III. 17: System GMM Regression of Tobin's q on the Interaction between
Corporate Governance Performance and Employee Board Representation
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.486*** 85.29 0.765*** 50.92 0.919*** 59.15
Corporate governance performance 1.819*** 10.46 1.578*** 8.18 –0.690** –2.40
Employee directorship 11.322*** 6.97
Corporate governance performance * –21.289*** –7.37
Employee directorship
Labour representation 2.118*** 3.81
Corporate governance performance * –3.020*** –4.23
Labour representation
Employee-shareholders representation –14.205*** –4.61
Corporate governance performance * 27.316*** 5.25
Employee-shareholders representation
Standalone 0.015 0.50 0.022 0.69 0.093*** 5.68
CSR assurance 0.121*** 3.62 –0.019 –0.56 –0.034 –1.18
CSR committee 0.064* 1.85 –0.004 –0.09 0.032* 1.79
Board size 0.162** 2.17 0.122 1.23 –0.071 –1.21
Board independence 0.034 0.43 0.107 1.14 0.004 0.07
Board gender diversity –0.043 –0.32 0.023 0.20 0.123** 2.01
Board meetings –0.049* –1.97 0.025 0.82 –0.032 –1.15
Duality –0.147*** –4.54 –0.041 –1.05 0.065*** 3.24
CEO tenure –0.036*** –2.73 –0.006 –0.39 –0.018* –1.90
Family ownership 0.539*** 5.58 0.380*** 4.34 0.178** 2.35
Institutional ownership –0.086* –1.64 –0.173*** –2.68 0.053* 1.64
Leverage 1.076*** 15.10 0.277** 2.49 0.266*** 3.71
ROA 6.216*** 25.84 2.742*** 12.73 1.644*** 7.77
R&D 3.495*** 5.56 1.267* 1.80 1.292** 2.52
Firm size –0.119*** –6.63 –0.076*** –4.72 –0.002 –0.14
Law 2013 0.125*** 7.61 0.154*** 4.50 0.113*** 5.47
Intercept 0.490* 1.80 –0.063 –0.17 0.337* 1.76
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 3732.47 (p = 0.000) 5951.90 (p = 0.000) 8409.00 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.39 (p = 0.006) –3.07 (p = 0.000) –2.43 (p = 0.006)
Arellano–Bond test AR(2) (z, p–value): –1.14 (p = 0.253) –1.16 (p = 0.248) –1.18 (p = 0.238)
Sargan test (Chi–square, p–value): 334.83 (p = 0.000) 97.44 (p = 0.000) 93.84 (p = 0.000)
Hansen test (Chi–square, p–value): 58.00 (p = 0.153) 32.81 (p = 0.285) 35.11 (p = 0.239)
Joint test: Corporate governance –19.469*** –7.03
performance + (Corporate governance
performance * Employee directorship)
Joint test: Corporate governance –1.442*** –2.35
performance + (Corporate governance
performance * Labour representation)
Joint test: Corporate governance 26.625*** 5.38
performance + (Corporate governance
performance * Employee-shareholders
representation)

*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.
All variables are as defined in Table III.1.

154
6. Conclusion

Going further than prior studies (Hillman et al., 2001; Huse et al., 2009; Nekhili et al., 2019),

we hypothesize that the type of employee directors may make a difference to the perception of

ESG issues by market participants. Using a matched sample of French listed firms belonging

to the SBF 120 index, we apply the system GMM estimation to investigate the nature of the

relationship between employee directorship and ESG performance as well as its supporting

pillars (social, environment and corporate governance). We first find that firms with

employee-shareholder board representation achieve higher ESG performance in general and

higher environmental and corporate governance performance in particular. The presence of

labour board representatives is found to be positively and significantly associated with social

performance only. These findings suggest that labour board representatives operate mainly in

the interest of labour, whereas employee-shareholder board representatives are more

concerned about environmental and corporate governance performance than about social

performance. Based on these results, we investigate the moderating role of each type of

employee directors on the value relevance of ESG performance. In accordance with Nekhili et

al. (2019), we find that market participants tend not to favour the presence of employee

representatives as a whole on the board and attribute less value to ESG performance of firms

with employee board representation. Going further than the study by Nekhili et al. (2019), we

find evidence that the reaction of outside shareholders to the level of ESG performance differs

considerably depending on the type of employee directors considered in the model. In

particular, we document that, while market participants are opposed to ESG performance

when employees are represented on board by right of employment, they reward ESG

performance when directors representing employee-shareholders are on the board. These

results may be explained by the fact that directors representing employee-shareholders may

have, through sharing specific information and the same interests, close relationships with

155
other directors representing outside shareholders (Ginglinger et al., 2011). Consequently, the

presence of employee-shareholder board representatives may provide better cohesiveness and

esteem among the board members, which is important for ensuring the effective use of

employees’ knowledge and skills and for improving the effectiveness of the board (Huse et

al., 2009).

This research complements previous studies (Bøhren & Strøm, 2010; Ginglinger et al.,

2011; Nekhili et al., 2019) in the literature by providing new insights into employee board

representation. Labour board representatives and employee-shareholder board representatives

are found to complement each other in terms of their effect on the different pillars of ESG

performance. With regard to assessing the value relevance of ESG performance, our results

indicate that outside shareholders are highly sensitive to the identity of employee directors,

whether they are representing labour or employee-shareholders. A meaningful inference to be

drawn from our findings is that employee ownership and employee representation on the

board as employee-shareholders may bring about basic changes in employees’ attitudes and

behaviour on the board and may, in this respect, be viewed as a credible way of improving

labour relations and creating constructive relationships between employees and shareholders.

Future directions for research can be drawn from this study and may be divided into

three areas. First, it will be interesting, following Huse et al. (2009), to explore the extent to

which esteem and cohesiveness of board members are better achieved by representing

employee-shareholders than by representing labour on the board. Second, competence,

experience and other demographic attributes (i.e., age, gender, educational level, etc.) of

employee directors should also to be taken into consideration as key factors influencing their

attitudes and behaviours on the board. Third, further investigation is needed to establish

whether and how companies provide support to employee-directors’ training, which clearly

156
influences their ability to make a contribution to the board’s decision-making and to align

their aspirations with the needs of the organization.

157
References

Bansal, P., & Roth, K. (2000). Why companies go green: A model of ecological
responsiveness. Academy of Management Journal, 43(4), 717–736.
Barnea, A., & Rubin, A. (2010). Corporate social responsibility as a conflict between
shareholders. Journal of Business Ethics, 97(1), 71–86.
Bennouri, M., Chtioui, T., Nagati, H., & Nekhili, M. (2018). Female board directorship and
firm performance: What really matters? Journal of Banking & Finance, 88, 267–291.
Berrone, P., & Gomez-Mejia, L. R. (2009). Environmental performance and executive
compensation: An integrated agency-institutional perspective. Academy of Management
Journal, 52(1), 103–126.
Bingham, J. B., Dyer, W. G., Smith, I., & Adams, G. L. (2011). A stakeholder identity
orientation approach to corporate social performance in family firms. Journal of Business
Ethics, 99(4), 565–585.
Blundell, R., & Bond, S. (2000). GMM estimation with persistent panel data: an application
to production functions. Econometric Reviews, 19(3), 321–340.
Bøhren, Ø., & Strøm, R. Ø. (2005). The value creating board: Theory and evidence.
Norwegian School of Management. Research Report 8/2005 (BI, Oslo).
Bøhren, Ø., & Strøm, R. Ø. (2010). Governance and politics: Regulating independence and
diversity in the board room. Journal of Business Finance & Accounting, 37(9‐10), 1281–
1308.
Boulouta, I. (2013). Hidden connections: The link between board gender diversity and
corporate social performance. Journal of Business Ethics, 113(2), 185–197.
Brammer, S., Millington, A., & Rayton, B. (2007). The contribution of corporate social
responsibility to organizational commitment. The International Journal of Human
Resource Management, 18(10), 1701–1719.
Coffey, B. S., & Wang, J. (1998). Board diversity and managerial control as predictors of
corporate social performance. Journal of Business Ethics, 17(14), 1595–1603.
Cronqvist, H., Heyman, F., Nilsson, M., Svaleryd, H., & Vlachos, J. (2009). Do entrenched
managers pay their workers more? The Journal of Finance, 64(1), 309–339.
Dechow, P. M., Sloan, R. G., & Sweeney, A. P. (1996). Causes and consequences of earnings
manipulation: An analysis of firms subject to enforcement actions by the SEC.
Contemporary Accounting Research, 13(1), 1–36.

158
Dunn, P., & Sainty, B. (2009). The relationship among board of director characteristics,
corporate social performance and corporate financial performance. International Journal of
Managerial Finance, 5(4), 407–423.
Eccles, R. G., Serafeim, G., & Krzus, M. P. (2011). Market interest in nonfinancial
information. Journal of Applied Corporate Finance, 23(4), 113–127.
El Akremi, A., Gond, J. P., Swaen, V., De Roeck, K., & Igalens, J. (2018). How do
employees perceive corporate responsibility? Development and validation of a
multidimensional corporate stakeholder responsibility scale. Journal of Management,
44(2), 619–657.
Faleye, O., Mehrotra, V., & Morck, R. (2006). When labor has a voice in corporate
governance. Journal of Financial and Quantitative Analysis, 41, 489–510.
Fauver, L., & Fuerst, M. E. (2006). Does good corporate governance include employee
representation? Evidence from German corporate boards. Journal of Financial Economics,
82(3), 673–710.
Ginglinger, E., Megginson, W., & Waxin, T. (2011). Employee ownership, board
representation, and corporate financial policies. Journal of Corporate Finance, 17(4), 868–
887.
Gold, M., & Waddington, J. (2019), Board-level employee representation in Europe: State of
play, European Journal of Industrial Relations, 25(3), 205–218.
Graves, S. B., & Waddock, S. A. (1994). Institutional owners and corporate social
performance. Academy of Management Journal, 37(4), 1034–1046.
Guedri, Z., & Hollandts, X. (2008). Beyond dichotomy: The curvilinear impact of employee
ownership on firm performance. Corporate Governance: An international Review, 16,
460–474.
Gordon, J. N., & Roe, M. J. (2004). Convergence and persistence in corporate governance.
Cambridge University Press.
Hafsi, T., & Turgut, G. (2013). Boardroom diversity and its effect on social performance:
Conceptualization and empirical evidence. Journal of Business Ethics, 112(3), 463–479.
Halbritter, G., & Dorfleitner, G. (2015). The wages of social responsibility—where are they?
A critical review of ESG investing. Review of Financial Economics, 26, 25–35.
Hammer, T. H., Currall, S. C., & Stern, R. N. (1991). Worker representation on boards of
directors: A study of competing roles. ILR Review, 44(4), 661–680.

159
Hillman, A. J., Keim, G. D., & Luce, R. A. (2001). Board composition and stakeholder
performance: Do stakeholder directors make a difference? Business & Society, 40(3), 295–
314.
Hollandts, X., Aubert, N., Abdelhamid, A., & Prieur, V. (2018). Beyond dichotomy: The
curvilinear impact of employee ownership on CEO entrenchment. Management
International, 22, 112–164.
Hollandts, X., & Aubert, N. (2011). La représentation obligatoire des actionnaires salariés au
conseil d'administration : un état des lieux. Gestion 2000, 28(6), 15–26.
Huse, M., Nielsen, S. T., & Hagen, I. M. (2009). Women and employee-elected board
members, and their contributions to board control tasks. Journal of Business Ethics, 89(4),
581–597.
Jensen, M. C., & Meckling, W. H. (1979). Rights and production functions: An application to
labor– managed firms and codetermination. Journal of Business, 52, 469–506.
Johnson, R. A., & Greening, D. W. (1999). The effects of corporate governance and
institutional ownership types on corporate social performance. Academy of Management
Journal, 42(5), 564–576.
Jones, D. A., Willness, C. R., & Madey, S. (2014). Why are job seekers attracted by corporate
social performance? Experimental and field tests of three signal-based mechanisms.
Academy of Management Journal, 57(2), 383–404.
Kim, H. R., Lee, M., Lee, H. T., & Kim, N. M. (2010). Corporate social responsibility and
employee–company identification. Journal of Business Ethics, 95(4), 557–569.
Kleinknecht, R. H. (2015). Employee participation in corporate governance: Implications for
company resilience. European Journal of Industrial Relations, 21, 57–72.
Luoma, P., & Goodstein, J. (1999). Stakeholders and corporate boards: Institutional
influences on board composition and structure. Academy of Management Journal, 42(5),
553–563.
Macaulay, C. D., Richard, O. C., Peng, M. W., & Hasenhuttl, M. (2018). Alliance network
centrality, board composition, and corporate social performance. Journal of Business
Ethics, 151(4), 997–1008.
Markey, R., McIvor, J., & Wright, C. F. (2016). Employee participation and carbon emissions
reduction in Australian workplaces. The International Journal of Human Resource
Management, 27(2), 173–191.

160
McWilliams, A., & Siegel, D. (2000). Corporate social responsibility and financial
performance: correlation or misspecification? Strategic Management Journal, 21(5), 603–
609.
Nekhili, M., Boukadhaba, A., Nagati, H., & Chtioui, T. (2019). ESG performance and market
value: the moderating role of employee board representation. The International Journal of
Human Resource Management, 1–27.
Pagano, M., & Volpin, P. F. (2005). Managers, workers, and corporate control. The Journal of
Finance, 60, 841–868.
Pagano, M. S., Sinclair, G., & Yang, T. (2018). Understanding ESG ratings and ESG
indexes. Research Handbook of Finance and ESG, 339.
Poulain-Rehm, T., & Lepers, X. (2013). Does employee ownership benefit value creation?
The case of France (2001–2005). Journal of Business Ethics, 112(2), 325–340.
Preuss, L., Haunschild, A., & Matten, D. (2009). The rise of CSR: Implications for HRM and
employee representation. The International Journal of Human Resource Management,
20(4), 953–973.
Rees, W., & Rodionova, T. (2015). The influence of family ownership on corporate social
responsibility: An international analysis of publicly listed companies. Corporate
Governance: An International Review, 23(3), 184–202.
Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the propensity score in
observational studies for causal effects. Biometrika, 70(1), 41–55.
Shaukat, A., Qiu, Y., & Trojanowski, G. (2016). Board attributes, corporate social
responsibility strategy, and corporate environmental and social performance. Journal of
Business Ethics, 135(3), 569–585.
Surroca, J., & Tribó, J. A. (2008). Managerial entrenchment and corporate social
performance. Journal of Business Finance & Accounting, 35(5‐6), 748–789.
Tajfel, H. and Turner, J.C. (1985), The Social Identity Theory of Intergroup Behaviour. In:
Worchel, S. and Austin, W.G., Eds., Psychology of Intergroup Relations, 2nd Edition,
Nelson Hall, Chicago, 7–24.
Triana, M. D. C., Jayasinghe, M., Pieper, J. R., Delgado, D. M., & Li, M. (2019). Perceived
workplace gender discrimination and employee consequences: a meta-analysis and
complementary studies considering country context. Journal of Management, 45(6), 2419–
2447.

161
Turban, D. B., & Greening, D. W. (1997). Corporate social performance and organizational
attractiveness to prospective employees. Academy of Management Journal, 40(3), 658–
672.
Turker, D. (2009). How corporate social responsibility influences organizational
commitment. Journal of Business Ethics, 89(2), 189–204.
Van der Laan, G., Van Ees, H., & Van Witteloostuijn, A. (2008). Corporate social and
financial performance: An extended stakeholder theory, and empirical test with accounting
measures. Journal of Business Ethics, 79(3), 299–310.
Velte, P. (2016). Women on management board and ESG performance. Journal of Global
Responsibility, 7(1), 98–109.
Walls, J. L., Berrone, P., & Phan, P. H. (2012). Corporate governance and environmental
performance: Is there really a link? Strategic Management Journal, 33(8), 885–913.
Weber, J., & Gladstone, J. (2014). Rethinking the corporate financial–social performance
relationship: Examining the complex, multi-stakeholder notion of corporate social
performance. Business and Society Review, 119(3), 297–336.
Zhang, M., Di Fan, D., & Zhu, C. J. (2014). High-performance work systems, corporate social
performance and employee outcomes: Exploring the missing links. Journal of Business
Ethics, 120(3), 423–435.

162
 General Conclusion

Our thesis is based on a current and highly important issue in the world and particularly in the

French context. The French legislation has shown a growing interest in the topics of employee

board representation and employee share ownership (law of 2006, law of 2013, law of 2015

and law of 2019) and for the social responsibility of firms (NRE 2001, Grenelle I 2009,

Grenelle II 2010). Although a significant relationship between employees and CSR initiatives

has been documented, little studies examined the impact of employee directors on CSR.

Believing that CSR is at the heart of our understanding of the impact of employee directors on

the firm and motivated by the gap in the literature, it was interesting to examine how

investors, as potential shareholders, perceive the commitment to CSR when employees are

represented on the board of directors. To reach this objective, four research questions have

been asked in three chapters:

To what extent the value relevance of CSR reporting is affected by the employee board
representation?

Examining the French context for the period 2001-2011, we report two main results. First, our

findings suggest that firms with employee directors disclose more extra-financial information

than their counterparts. Second, our results show that the market participants find

environmental and sustainability information more relevant when employees are appointed on

the boardroom, suggesting that the presence of employee directors may consider a good filter

for the market to assess the quality of CSR reporting. Nevertheless, social information is not

considered as relevant, suggesting that higher social reporting may result from an over-

investment to the benefit of employees and at the expense of shareholders who support the

costs behind. This finding highlights a conflict of interest between shareholders and

employees regarding social initiatives. Considering the answers found for our first research

163
question, we suggest that since market participants make use of employees’ presence on the

board of directors to assess the relevance of CSR information, they may also use it when

making their perception towards the ESG performance resulting from the CSR expenses of

the firm. Therefore, we put forward our second research question.

How do investors perceive ESG performance when employees are appointed on the
boardroom?

Examining a sample of French firms belonging to the SBF 120 index from 2007 to 2017, our

results show that the market participants reward a good ESG performance. Nevertheless, their

perception is moderated negatively when firms appoint employees on their board of directors.

More interestingly, none of the ESG pillars, namely social, environmental and corporate

governance is rewarded when boards include employee representatives. For investors, a

higher ESG performance may result from an alliance between employee directors and

managers. Such an alliance would probably threaten the shareholders’ supremacy on the

board and would benefit both managers and employees. Given that French boards may have

two types of employee directors, labour representatives and employee-shareholders

representatives, we believe that examining the perception of investors towards the ESG

performance by considering the way employee directors are appointed on the board would

provide a deeper understanding with this relationship. Firstly, it is worthy to examine how

each type of employee directors impact the different pillars of ESG performance. Therefore,

we put forward our third research question.

How each type of employee directors impacts the ESG pillars namely, social, environment
and corporate governance?

Our findings show an opposite impact of the two types of employee directors on ESG

performance and its related pillars. Specifically, we find that employee directors elected by

right of employment enhance the social pillar exclusively while reducing significantly the

164
ESG performance and the environmental and corporate governance pillars. Conversely,

directors elected by employee shareholders are shown to increase the ESG performance and

the environmental and corporate governance pillars and to reduce the social pillar. These

findings could be explained on the one hand, by the different interests of the two types of

employee directors and on the other hand, by the different information, knowledge,

competences and skills which they bring to the corporate board. Our last research question is

the following.

To what extent the value relevance of ESG performance and its related pillars is affected by
the type of employee director on the board?

Interesting answers were documented for this research question. Indeed, our findings show

that the market participants reward firms for appointing employee shareholders on their

boards while they penalize them for representing pure labour. Moreover, we find that the

market participants value positively a higher ESG, social and corporate governance

performance when directors elected by employee shareholders are on the board. In contrast,

they value negatively a higher ESG, social and corporate governance when employee

directors elected by right of employment are on the board.

Three main conclusions could be drawn from our thesis. First, findings indicate that the

market participants rely on the diversity of the board in terms of the presence of employee

directors to assess the relevance of CSR reporting. Second, the results of this study show that

investors pay attention to the way employee directors are elected on the board when making

their perception of the firm’s CSR engagement. Finally, we suggest that both types of

employee directors are likely to bring benefits to the firm. In other words, by increasing their

social performance, firms with labour representatives on the board are likely to motivate their

employees and attract potential job seekers. In addition, by increasing their corporate

governance performance and the overall ESG performance, firms with employee-shareholders

165
representatives on the board are likely to attract potential investors and particularly socially

responsible investors.

Major contributions

Our study makes several contributions to the current literature. One of the most significant

contributions of this thesis is that it examines the impact of employee board representation on

a highly important issue in our century namely CSR, a relationship that has been neglected by

the previous studies. This is, to our knowledge, the first attempt to examine the impact of

employee board representation on CSR reporting and ESG performance. Another contribution

is that we did not only examine the presence of employee directors but we also distinguish

between the differing types of employee directors, namely employee directors elected by right

of employment and directors elected by employee shareholders. Considering the effort of

governments and legislators to represent employees on the boardroom (Law 2006, law 2013,

law 2015, law 2019), it is with interest to examine the perception of the market towards such

policy. This is most important because such human capital representation may counterbalance

the supremacy of financial capital providers on the boardroom. Therefore, our findings

provide a new understanding of the impact of the differing types of employee directors on the

firm.

Limitations

Our thesis suffers from two main limitations: a problem of generalized conclusions and small

sample. Focusing on the French context, our findings might not be generalizable to other

countries different from France in terms of legislations (no mandatory representation of

employees on board) and governance. Moreover, we rely on a sample of French firms

belonging to the SBF120 index and we eliminated financial, assurance, real-estate and foreign

companies. Financial, assurance and real-estate firms were excluded because they have

166
special laws comparing to other industries while foreign firms are not subject to the same

laws as other French companies. We choose a sample of firms belonging to the SBF 120

index for several reasons. First, only large firms, with 5000 employees in France and 10 000

employees in France and around the world (law of 2013) and after for firms with 1000

employees in France and 5000 employees in France and around the world (law of 2015), were

mandated to represent their employees on their board level. Another argument is advanced by

Hollandts and Aubert (2011). The authors advance that those firms likely to represent their

employee shareholders on the board level belong most to CAC 40. Consequently, our sample

is relatively small. For our third chapter, we relied on matched samples because we found

significant differences between firms with and without employee board representation due to

the presence of employee directors. Whereas such method enhances the quality of our results,

it further reduces our firm-years observations.

Future area of research

For future research, it would be interesting to examine how employee board representation

may impact CSR reporting and ESG performance by considering the specific attributes of

each type of employee directors (experience, certifications, gender, previous function,

trainings, etc). Indeed, as argued in literature, female directors have positive impacts on ESG

performance and particularly on CSR concerns (Boulouta, 2013; Hafsi & Turgut, 2013);

consequently, female employee directors may make different decisions comparing to male

employee directors. Another example is the experience of employee directors. That is, more

experienced employee directors may make more efficient decisions than less experienced

employee directors. Moreover, a cross country study to compare the impact of specific types

of employee directors on CSR reporting and ESG performance in different contexts may

provide new pieces of evidence on this relationship.

167
 Another area of research would be to carry out a cross-country study to assess the

impact of employee directors on the value relevance of CSR reporting and ESG performance.

More information on the impact of employee directors on the value relevance of CSR before

and after the adoption of the Grenelle II Act would also help to establish a greater degree of

accuracy on this subject.

Future research directions could explore the type and nature of decisions most impacted

by employee directors. Indeed, Huse et al. (2009) document that employee directors influence

the strategic control task of the board which is often of a qualitative nature, but they do not

impact the budget control task which is often of a quantitative nature. Finally, we believe that

examining the impact of employee board representation on the CSR engagement of firms

belonging to controversial industries would provide another piece to the puzzle.

168
References

Boulouta, I. (2013). Hidden connections: The link between board gender diversity and
corporate social performance. Journal of Business Ethics, 113(2), 185–197.
Hafsi, T., & Turgut, G. (2013). Boardroom diversity and its effect on social performance:
Conceptualization and empirical evidence. Journal of Business Ethics, 112(3), 463–479.
Hollandts, X., & Aubert, N. (2011). La représentation obligatoire des actionnaires salariés au
conseil d'administration : un état des lieux. Gestion 2000, 28(6), 15–26.
Huse, M., Nielsen, S. T., & Hagen, I. M. (2009). Women and employee-elected board
members, and their contributions to board control tasks. Journal of Business Ethics, 89(4),
581–597.

169
170
 Impact de la représentation des salariés au sein du
conseil d’administration sur l’engagement RSE de
l’entreprise : Etude du contexte français

Résumé de la thèse

Ce travail doctoral s’appuie sur une actualité dans le contexte Français. En effet, durant les

deux dernières décennies, la législation française a montré un intérêt croissant, d’une part,

pour la représentation des salariés dans le conseil d'administration et pour l'actionnariat salarié

dans les entreprises (loi 2006, loi 2013, loi 2015 et loi 2019) et de l’autre part, pour la

responsabilité sociale de l’entreprise (ci-après RSE) (loi NRE, loi Grenelle I, loi Grenelle II).

Les recherches antérieures ont montré que les salariés sont plus concernés et plus

influencés par les initiatives RSE de leurs entreprises (Brammer, Millington, & Rayton, 2007)

et que la représentation des salariés dans le conseil d’administration avec des droits de vote

améliore significativement l’efficacité du conseil au sujet de l’engagement RSE de

l’entreprise (Huse, Nielson, & Hagen, 2009). L’engagement RSE, qui reflète la prise en

compte des attentes et des besoins des différentes parties prenantes, est devenu un critère

important pour les investisseurs afin de prendre leurs décisions d’investissement ou de

désinvestissement. En revanche, selon la théorie de l’agence, les salariés et les actionnaires ne

présentent pas les mêmes intérêts ni profitent des mêmes avantages issus de l’engagement

RSE de l’entreprise. Dans ce sens, la représentation des salariés dans le conseil

d’administration pourrait influencer la perception que font les investisseurs de son

engagement RSE.

L’objectif de cette thèse est de contribuer à la littérature sur la représentation des

salariés dans le conseil d’administration en répondant à la question de recherche suivante :

171
Comment les investisseurs perçoivent-ils l’engagement RSE de l’entreprise lorsque les

salariés sont nommés dans le conseil d’administration ? Deux aspects de l’engagement RSE

sont étudiés dans ce travail de recherche à savoir ; le reporting extra-financier (i.e., reporting

RSE) et la performance extra-financière (i.e., performance environnementale, sociale et de

gouvernance (ESG)). La performance ESG pourrait être un indicateur de l’engagement RSE

de l’entreprise en traduisant sa capacité à gérer de bonnes relations avec ses différentes parties

prenantes. Cependant, pour mieux comprendre cette performance ESG, les parties prenantes

ont besoin d'informations extra-financières quantitatives et qualitatives crédibles. Les rapports

RSE constituent un moyen essentiel pour aider les parties prenantes, et en particulier les

actionnaires, à comprendre la performance extra-financière des entreprises (Fuente, García-

Sanchez, & Lozano, 2017).

Notre premier chapitre porte sur le reporting RSE et vise en particulier à répondre à la

question de recherche suivante : Dans quelle mesure la présence des salariés dans le conseil

d’administration peut-elle impacter la perception des investisseurs vis-à-vis du reporting RSE

? L’utilisation d’un échantillon d’entreprises françaises cotées appartenant à l’indice SBF 120

pour la période allant de 2001 jusqu’à 2011 montre que les entreprises qui nomment des

administrateurs salariés dans leurs conseil d’administration divulguent plus d’information

extra-financière que d’autres entreprises qui n’ont pas d’administrateurs salariés. Nos résultats

montrent aussi que les investisseurs considèrent que les informations relatives à

l'environnement et au développement durable sont plus pertinentes lorsque les salariés sont

nommés au conseil d'administration. Néanmoins, les informations sociales ne sont pas

considérées comme pertinentes pour les investisseurs. Ces résultats montrent que les

administrateurs salariés jouent un rôle modérateur sur la perception des investisseurs vis-à-vis

de la pertinence des rapports RSE et mettent en évidence un conflit d’intérêts entre les

actionnaires et les salariés vis-à-vis des initiatives RSE.

172
 Pour mieux comprendre l’effet modérateur des administrateurs salariés sur la perception

des investisseurs à l’égard de l’engagement RSE, nous examinons, dans ce deuxième chapitre,

leur effet sur la pertinence de l’information sur la performance ESG. La question de recherche

est de savoir comment les investisseurs perçoivent la performance ESG selon que les salariés

sont nommés ou non au conseil d'administration. En examinant un échantillon d’entreprises

françaises appartenant à l’indice SBF 120 de 2007 jusqu’à 2017, nos résultats montrent que

les investisseurs récompensent une bonne performance ESG. Néanmoins, leur perception est

modérée négativement par la présence des salariés au sein du conseil d'administration. Plus

intéressant encore, aucun des piliers de la performance ESG, à savoir le sociale,

l’environnementale et de gouvernance, n'est récompensé lorsque le conseil d'administration

inclut des représentants des salariés. Pour les investisseurs, une performance ESG plus élevée

peut résulter d'une alliance entre les administrateurs salariés et les dirigeants.

En élargissant le champ de notre deuxième question de recherche, nous avons distingué,

dans notre dernier chapitre, entre les deux types d’administrateurs salariés nommés au sein du

conseil d’administration des entreprises françaises, à savoir les administrateurs élus par les

salariés par le droit de travail (ci-après représentants des travailleurs) et les administrateurs

élus par les salariés actionnaires (ci-après représentants des salariés actionnaires). Nous

examinons dans un premier temps l’impact de chaque type d’administrateur salarié sur chaque

pilier de la performance ESG à savoir le social, l’environnemental et de la gouvernance.

Ensuite, nous examinons comment la perception des investisseurs vis à vis de la performance

ESG est affectée par la nomination des représentants des travailleurs d’une part et des

représentants des salariés actionnaires au conseil d’administration de l’autre part. En utilisant

un échantillon de sociétés françaises appartenant à l’indice SBF 120 de 2007 jusqu’à 2017,

nos résultats montrent que les représentants des travailleurs renforcent exclusivement le pilier

social tout en réduisant de manière significative la performance ESG ainsi que les deux piliers

173
environnement et gouvernance. Inversement, les représentants des salariés actionnaires

améliorent la performance ESG et les piliers environnement et gouvernance et réduisent le

pilier social. Nos résultats montrent aussi que les investisseurs récompensent financièrement

les entreprises ayant une bonne performance ESG et particulièrement une bonne performance

sociale et en matière de gouvernance lorsque les salariés actionnaires sont représentés dans le

conseil d’administration. En revanche, ils les pénalisent lorsque les représentants des

travailleurs font partie du conseil d'administration.

Nous présentons dans ce qui suit un résumé des trois chapitres de notre thèse. Chaque

résumé de chapitre comprend une brève présentation de la littérature suivie de la formulation

des hypothèses. Ensuite, nous présentons la méthodologie utilisée et en particulier

l’échantillon, les sources de données utilisées ainsi que les définitions des variables. Enfin,

nous exposons les principaux résultats.

174
Chapitre I : La représentation des salariés au sein du
conseil d'administration est-elle importante pour la
pertinence des rapports RSE ?

Les rapports RSE sont principalement utilisés pour informer les parties prenantes de

l’entreprise sur ses activités sociales, environnementales et de développement durable. Cette

communication d'information extra-financière est considérée comme une bonne stratégie pour

gagner en légitimité aux yeux des parties prenantes, ayant des intérêts différents en matière de

RSE (Sahed-Granger & Boncori, 2014).

La sensibilisation des parties prenantes aux activités RSE de l'entreprise est importante

et elle n'a cessé de croître au fil du temps (Allix-Desfautaux & Makany, 2015). En revanche,

la motivation des managers derrière cette divulgation n'est pas encore claire. En effet, un

niveau élevé de divulgation RSE peut être utilisé pour améliorer la perception des parties

prenantes vis-à-vis de l'engagement social et environnemental de l'entreprise ou, à l'inverse,

pour détourner leur attention des mauvaises performances ou actions défavorables. Des

recherches antérieures montrent que les investisseurs prennent en considération la

composition et plus particulièrement la présence des femmes au sein du conseil

d’administration pour juger la crédibilité des informations RSE divulguées (Nekhili, Nagati,

Chtioui, & Nekhili, 2017b). De surcroît, la présence des salariés au sein du conseil

d’administration peut servir aussi de filtre pour les investisseurs afin d’apprécier la crédibilité

des rapports RSE.

Dans ce chapitre, nous examinons l’effet modérateur de la présence des administrateurs

salariés sur la relation entre la divulgation des rapports RSE et la valeur du marché de

175
l’entreprise. En utilisant un échantillon d'entreprises françaises cotées appartenant à l’indice

SBF 120 de 2001 jusqu’à 2011, nous montrons que la présence des salariés dans le conseil

d’administration joue un rôle important en matière d’engagement RSE de l’entreprise.

Néanmoins, nous constatons que l'effet marginal de la présence des administrateurs salariés

sur la relation entre le reporting RSE et la valeur du marché de l’entreprise diffère selon la

catégorie RSE examinée. Plus particulièrement, nous constatons que les investisseurs trouvent

les informations environnementales et de développement durable plus pertinentes que les

informations sociales lorsque les salariés sont représentés dans le conseil d’administration. Ce

résultat suggère qu'un niveau élevé de reporting social peut refléter plus d'avantages aux

salariés aux dépens des actionnaires qui en supportent les coûts.

Notre étude contribue à la littérature examinant les effets de la représentation des

salariés dans le conseil d'administration et la pertinence des rapports RSE divulgués

volontairement. Premièrement, nous montrons que les investisseurs font attention à la

composition du conseil d’administration et utilisent des filtres pour évaluer la crédibilité des

rapports RSE volontaires. Deuxièmement, nous soutenons que l'utilisation de variables

modératrices est susceptible de fournir une analyse plus approfondie. Troisièmement, nous

mettons en évidence un conflit d'intérêts entre les salariés et les actionnaires concernant le

reporting RSE et ses différentes catégories à savoir sociale, environnementale et de

développement durable.

I.1 Revue de la littérature et développement des hypothèses

La participation des salariés au processus décisionnel est susceptible d'améliorer la qualité des

informations sociales et environnementales communiquées aux parties prenantes.

Premièrement, les employés disposent d’informations plus précieuses, réelles et précises sur

les conditions de travail de l’entreprise et les besoins des travailleurs. Par conséquent, ils

peuvent améliorer l'efficacité des investissements sociaux liés aux salariés. Dans un tel cas,

176
les parties prenantes peuvent traiter avec confiance les informations RSE et en particulier les

informations divulguées sur le capital humain. Deuxièmement, les salariés dépendent de la

survie et la continuité de leur organisation. Par conséquent, ils vont être plus concernés par la

réputation de leur entreprise et ils peuvent encourager un reporting RSE de meilleure qualité

afin de sensibiliser les parties prenantes, essentielles pour la survie de leur entreprise.

Troisièmement, les administrateurs salariés améliorent l’efficacité du conseil d’administration

(Huse et al., 2009), qui est responsable de la qualité des informations RSE divulguées

(Cormier, Aerts, Ledoux, & Magnan, 2009). Finalement, les salariés en tant qu'internes à

l’entreprise sont témoins des actions et de politiques sociales et environnementales ce qui les

différencient des autres administrateurs. Néanmoins, des études antérieures montrent un

comportement opportuniste des salariés dans le conseil d’administration. En d’autres termes,

les administrateurs salariés peuvent utiliser leur pouvoir pour maximiser leurs propres intérêts

plutôt que ceux des actionnaires (Bøhren & Strøm, 2010; Faleye, Mehrotra & Morck, 2006).

En ce sens, un niveau élevé de reporting RSE peut signaler une position puissante des salariés

dans le conseil d’administration qui menacerait la suprématie des actionnaires et / ou peut

indiquer une éventuelle coalition entre les managers et les salariés. En conséquence nous

proposons deux hypothèses :

H1a : Le reporting RSE est plus pertinent pour les entreprises ayant des administrateurs

salariés que pour les entreprises sans administrateurs salariés.

H1b : Le reporting RSE est moins pertinent pour les entreprises ayant des

administrateurs salariés que pour les entreprises sans administrateurs salariés.

I.2. Méthodologie

Cette section décrit notre échantillon ainsi que les différentes variables utilisées dans la partie

empirique de cette étude.

177
I.2.1 Echantillon

Pour examiner l’effet modérateur des administrateurs salariés sur la relation entre la

divulgation RSE et la valeur de marché de l’entreprise, nous utilisons un échantillon de

91 sociétés françaises appartenant à l’indice SBF 120 (les sociétés financières, d’assurance et

immobilières ayant été exclues), sur une période de 11 ans, allant de 2001 jusqu’à 2011. Nous

avons collecté les données financières et comptables à partir de la base de données ThomsonOne.

Les données concernant la gouvernance, l’actionnariat de l’entreprise et les données sur la RSE

ont été collecté manuellement à partir des rapports annuels et rapports RSE (appelés aussi rapports

de développement durable).

I.2.2 Les variables de régression

I.2.2.1 Variable dépendante : Q de Tobin

Suivant Ginglinger, Megginson, et Waxin (2011), Cahan, De Villiers, Jeter, Naiker et Van

Staden (2016) et Nekhili et al. (2017b), nous utilisons le Q de Tobin (Tobin’s Q) comme mesure

de la valeur de marché des entreprises. Cette mesure reflète l’évaluation par le marché de la valeur

attendue d’une entreprise à long terme, comme l’expliquent Cahan et al. (2016).

I.2.2.2 Variable endogène : Reporting RSE

A la suite de Nekhili, Nagati, Chtioui et Rebolledo (2017a) et Nekhili et al. (2017b), la méthode

d’indice de divulgation non pondéré est utilisée pour mesurer le niveau de reporting RSE. Un

indice d’analyse de contenu, fondé sur la grille de la loi Grenelle II, a été créé compte tenu de la

conformité de ladite loi à la fois aux normes européennes et internationales et aux directives de la

GRI (initiatives de reporting mondial). Cette grille contient 42 éléments subdivisés en trois

catégories, à savoir le reporting social (19 éléments), le reporting environnemental (14 éléments),

et le reporting sur le développement durable (9 éléments).

178
 Le niveau de divulgation (CSR_REP) est la somme des notes obtenues dans les trois

catégories d’informations RSE (reporting social, environnemental et développement durable).

L’indice proprement dit est le rapport entre la note globale attribuée au reporting RSE et la note

maximale possible. Pour chaque item, nous attribuons la valeur 1 si l’item en question est

explicitement divulgué dans les rapports annuels et/ou dans les rapports RSE, et 0 sinon. Le

niveau de l’indice de divulgation de la RSE est alors considéré comme le rapport entre la note

totale attribuée et la note maximale en référence aux 42 items précités.

I.2.2.3 Variables modératrice : Administrateurs salariés

Dans cette étude, on cherche à détecter l’effet modérateur de la présence des salariés dans le

conseil d’administration. Pour ce faire, nous utilisons une variable muette (EMPL_BOARD) qui

prend la valeur 1 s’il y a au moins un salarié au conseil d’administration, et 0 sinon.

I.2.2.4 Variables de contrôle

Suivant les recherches antérieures (par exemple Nekhili et al., 2017a ; 2017b), nous utilisons un

ensemble de variables de contrôles censées influencer le reporting RSE et la valeur de marché de

l’entreprise. Nous contrôlons pour la vérification RSE (CSR_ASS), la présence de comité RSE

(CSR_COM), la taille du conseil d’administration (BOAD_SIZE), l’indépendance du conseil

(BOARD_IND), le nombre de réunion (BOARD_MEET), la dualité des fonctions du CEO

(DUAL), le mandat du CEO (TENURE), l’actionnariat familial (FAM_OWN), l’actionnariat

institutionnel (INST_OWN), l’actionnariat salarié (EMPL_OWN), l’effet de levier (LEV), le

risque systématique (BETA), les actifs étrangers (FOR_ASS), l’intensité des investissements

R&D (R&D), la taille de l’entreprise (SIZE), la loi Grenelle I (GRE1), et finalement l’industrie

(INDUSTRY).

179
I.2.2.5 Modèle empirique

Tobin’s Qit = β0 + β1 Lag Tobin’s Qit + β2 CSR_REPit + β3 EMP_REPit + β4

CSR_REP*EMPL_BOARD + β5 CSR_ASSit + β6 CSR_COMit + β7 BOARD_SIZEit + β8

BAORD_INDit + β9 BOARD_MEETit + β10 DUALit + β11 TENUREit + β12 FAM_OWNit + β13

INST_OWNit + β14 EMPL_OWNit + β15 LEVit + β16 BETAit + β17 FOR_ASSit + β18 R&Dit +

β19 SIZEit + β20 GRE1it + β21 INDUSTRYit + ɛit

I.3. Résultats

L’objectif du tableau I.1 est de déterminer si le reporting RSE et la présence des salariés au sein

du conseil d’administration influent conjointement sur la valeur de marché de l’entreprise (Q de

Tobin). Pour cela, nous utilisons la technique du test conjoint. D’abord, nous créons une variable

muette pour représenter un niveau élevé (H) de reporting RSE, en attribuant la valeur 1 pour les

observations ayant un reporting RSE supérieur à la médiane (47,62 %), et 0 sinon. Nous mesurons

ensuite l’effet marginal de la présence des salariés dans le conseil d’administration sur la

pertinence d’un niveau plus élevé de reporting RSE en testant l’effet de la somme des coefficients

(HCSR_REP + HCSR_REP *EMPL_BOARD) sur le Q de Tobin.

Le Modèle 1 du Tableau I.1 souligne un impact positif et significatif d’un niveau élevé de

reporting RSE sur le Q de Tobin. Le Modèle 2 du Tableau I.1 cherche à comprendre comment un

niveau élevé de reporting RSE et de la présence simultanée des salariés au sein du conseil

d’administration affecte la valeur de marché de l’entreprise en incluant l’interaction entre la

variable modératrice et la variable endogène spécifique (HCSR_REP *EMPL_BOARD). Nous

effectuons un test conjoint du coefficient pour HCSR_REP et HCSR_REP *EMPL_BOARD.

Le Modèle 2 du Tableau I.1 montre un impact positif et significatif du terme d’interaction

(HCSR_REP *EMPL_BOARD) sur le Q de Tobin. L’effet marginal de la présence des salariés

au sein du conseil d’administration sur l’existence d’une valeur plus élevée du niveau de reporting

RSE est évalué par le test conjoint de la somme du coefficient sur le Q de Tobin. Modèle 2 du

180
tableau I.1 montre un coefficient positif (1,393) et significatif (z = 5,43). Conformément à

l’hypothèse de recherche H1a, nous constatons ainsi que les investisseurs réagissent positivement

à un niveau plus élevé de reporting RSE fourni par les entreprises ayant des représentants salariés

au sein de leur conseil d’administration. Nous rejetons par conséquence l’hypothèse de recherche

H1b.

Table I.1: Régression GMM du Q de Tobin sur un niveau élevé de reporting RSE et la
présence des salariés au sein du conseil d’administration
Variables Model 1 Model 2

Coef. t–test Coef. t–test

Lag Tobin’s Q 0.128*** 16.73 0.107*** 10.12
HCSR_REP 0.141*** 4.15 –0.213*** –3.03
EMPL_BOARD –0.085 –0.81 –1.188*** –3.61
HCSR_REP *EMPL_BOARD 1.606*** 5.64
CSR_VERIF –0.006 –0.15 –0.047 –1.08
CSR_COM –0.048* –1.66 –0.089** –2.86
BOARD_SIZE –0.213*** –4.87 –0.185** –2.37
BOARD_IND –0.153* –1.96 –0.211** –2.31
BOARD_MEET –0.061* –1.40 –0.068 –1.56
DUAL –0.201*** –5.67 –0.150*** –3.62
TENURE 0.025 0.76 –0.029 –0.75
FAM_OWN 0.496*** 4.20 0.453*** 4.01
INST_OWN –0.134 –1.49 –0.225 –1.50
EMPL_OWN –1.84*** –3.07 –1.375 –1.47
LEV –0.521*** –5.50 –1.034*** –8.62
BETA 0.237*** 5.30 0.121* 1.90
FOR_ASS –0.281*** –4.00 –0.086 –1.05
R&D 1.113*** 3.50 0.487 1.08
SIZE 0.001 0.05 0.014 0.62
Intercept 1.291*** 5.04 1.622*** 5.92
Year Yes Yes
Industry Yes Yes
Number of observations 790 790
Fisher (Prob. > F) 12774 (p = 0.000) 5170.61 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.88 (p = 0.004) –2.92 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): 0.34 (p = 0.737) 0.39 (p = 0.698)
Sargan test (Chi–square, p–value): 736.84 (p = 0.000) 669.35 (p = 0.000)
Hansen test (Chi–square, p–value): 75.10 (p = 0.259) 67.11 (p = 0.473)
Joint test: HCSR_REP + (HCSR_REP * EMPL_BOARD) 1.393*** 5.43
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.

181
I.4 Analyses Supplémentaires

Nous réalisons des analyses supplémentaires pour examiner l’effet de la présence des

administrateurs salariés sur la pertinence des différentes catégories de reporting RSE (social,

environnemental, et de développement durable). Pour le reporting sociale, le Modèle 1 du

Tableau I.2 montre que le coefficient d’interaction (HSOCIAL_REP * EMPL_BOARD) n'a

pas d'impact significative sur le Q de Tobin, ce qui suggère qu'un reporting social élevé,

reflète plus d'avantages pour les salariés. Les coefficients des Modèles 2 et 3 du tableau I.2

montrent une relation positive et significative entre les variables d'interaction (HENVI_REP *

EMPL_BOARD) et (HSUST_REP * EMPL_BOARD) et le Q de Tobin. Ces résultats

suggèrent que le marché évalue positivement un niveau élevé de reporting environnementaux

et de développement durable lorsque les salariés sont représentés au sein du conseil

d’administration. Les coefficients du test conjoints des Modèles 2 et 3 sont positifs et

significatifs sur la valeur de marché, ce qui indique qu'un niveau plus élevé de reporting

environnementale et de reporting en matière de développement durable, lorsqu'ils sont fournis

par des entreprises avec au moins un salarié nommé au conseil d'administration, sont

financièrement récompensés par les investisseurs.

182
Table I.2: Régression GMM du Q de Tobin sur l’interaction entre un niveau élevé de
reporting social, reporting environnemental et reporting en matière de développement
durable et la présence des salariés au sein du conseil d’administration
Variables Model 1 Model 2 Model 3
Coef. t–test Coef. t–test Coef. t–test
Lag Tobin’s Q 0.060*** 5.78 0.097*** 9.88 0.148*** 10.65
HSOCIAL_REP 0.151*** 2.56
HENVIR_REP –0.017 –0.30
HSUST_REP –0.670*** –8.45
EMPL_BOARD –0.622*** –3.74 –0.709*** –3.54 –2.836*** –7.29
HSOCIAL_REP *EMPL_BOARD –0.358 –1.39
HENVIR_REP *EMPL_BOARD 0.725*** 3.81
HSUST_REP *EMPL_BOARD 3.277*** 7.52
CSR_VERIF 0.087 1.77 –0.004 –0.10 –0.128*** –3.45
CSR_COM –0.041 –1.08 –0.104*** –3.74 –0.027 –0.68
BOARD_SIZE –0.025 –0.36 –0.138** –2.53 –0.205*** –2.83
BOARD_IND –0.214** –2.47 –0.210*** –3.03 –0.334*** –3.07
BOARD_MEET 0.002 0.06 –0.059** –2.07 –0.164*** –5.15
DUAL –0.130*** –2.91 –0.190*** –4.95 –0.031 –0.73
TENURE 0.089*** 3.02 0.036 1.44 0.025 0.82
FAM_OWN 0.175* 1.81 0.463*** 4.76 0.524*** 4.38
INST_OWN –0.285*** –2.98 –0.077 –1.00 0.007 0.05
EMPL_OWN 0.668 0.79 –1.328** –2.37 –2.685** –2.42
LEV –0.524*** –4.05 –0.681*** –7.63 –1.059*** –13.08
BETA 0.359*** 5.59 0.319*** 5.90 0.115 1.59
FOR_ASS –0.449*** –4.89 –0.220*** –2.88 –0.181** –2.02
R&D 2.001*** 4.26 1.568*** 3.71 0.811* 1.96
SIZE 0.040** 2.23 0.003 0.15 0.063*** 3.60
Intercept 0.179 0.54 1.624*** 5.02 1.823*** 5.28
Year Yes Yes Yes
Industry Yes Yes Yes
Number of observations 790 790 790
Fisher (Prob. > F) 32162.52 (p = 0.000) 3323.06 (p = 0.000) 13127.25 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.84 (p = 0.004) –2.88 (p = 0.004) –2.92 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): 0.12 (p = 0.906) 0.35 (p = 0. 0.725) 0.78 (p = 0. 0.434)
Sargan test (Chi–square, p–value): 694.82 (p = 0.000) 740.11 (p = 0.000) 564.881 (p = 0.000)
Hansen test (Chi–square, p–value): 69.59 (p = 0.295) 73.99 (p = 0.184) 74.06 (p = 0.259)
Joint test: HSOCIAL_REP –0.207 –0.93
+ (HSOCIAL_REP * EMPL_BOARD)
Joint test: HENVIR_REP 0.708*** 4.37
+ (HENVIR_REP * EMPL_BOARD)
Joint test: HSUST_REP 2.607*** 5.50
+ (HSUST_REP * EMPL_BOARD)
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.

183
184
 Chapitre II : La performance ESG et la valeur de
marché de l’entreprise : Le rôle modérateur des
administrateurs salariés

La performance ESG est un aspect important de la stratégie d'entreprise, qui reflète la capacité

de cette dernière à maintenir de bonnes relations avec ses différentes parties prenantes

(Friede, Bush & Bassen, 2015).

Pour une meilleure performance ESG, certains auteurs recommandent des conseils

d’administration orientés vers les parties prenantes (Shaukat, Qiu & Trojanowski, 2016), dans

la mesure où la participation des parties prenantes au processus décisionnel modifie la

perception du marché vis à vis des performances ESG. Un changement majeur dans la

gouvernance d'entreprise au cours des dernières années a été la présence des représentants des

salariés au sein du conseil d'administration (Huse et al., 2009 ; Ginglinger et al., 2011).

Dans ce chapitre, nous s’intéressons à la perception des investisseurs envers la

performance ESG et ses trois piliers (sociaux, environnementaux et de gouvernance), selon

que les salariés sont ou non représentés au conseil d'administration. En utilisant un échantillon

d’entreprise française appartenant à l’indice SBF 120 pour la période allant de 2007 jusqu’à

2017, nos résultats montrent que les investisseurs s’intéressent à l’information sur la

performance ESG de l’entreprise. En revanche, ils réagissent négativement à la représentation

des salariés dans le conseil d’administration. Nos analyses montrent aussi que la relation entre

la performance ESG et la valeur de marché de l’entreprise est négativement modérée par la

présence des salariés dans le conseil d’administration. Nous constatons que, lorsque les

185
salariés sont représentés au conseil d’administration, ni la performance sociale, ni la

performance environnementale et de gouvernance ne sont pertinentes pour les investisseurs.

Ce chapitre confirme les résultats du premier chapitre et présente de nouvelles preuves

du conflit d'intérêts entre les actionnaires et les salariés en matière de RSE. Ce chapitre

souligne aussi le fait que la perception des investisseurs à l'égard des piliers de la performance

ESG (environnement, social et de gouvernance) diffère selon que les salariés sont représentés

ou non au conseil d'administration.

II.1 Revue de la littérature et développement des hypothèses

Du point de vue des actionnaires, une bonne performance ESG lorsque les salariés sont

représentés au sein du conseil d'administration peut conduire à deux situations. D'une part, les

administrateurs salariés peuvent avoir un impact positif sur la performance ESG, car ils

peuvent jouer un rôle important dans les décisions RSE (Preuss, Haunschild & Matten, 2009)

et parce qu'ils se préoccupent davantage des activités socialement responsables de leurs

entreprises que les autres membres du conseil d'administration (Huse et al., 2009). De l’autre

côté, une bonne performance ESG peut refléter le pouvoir des salariés sur le conseil

d'administration, car ils sont parmi les principaux bénéficiaires de l'investissement ESG

(Barnea & Rubin, 2010). En conséquence, nous suggérons que :

H1a : La représentation des salariés au sein du conseil d'administration modère

positivement la relation entre la performance ESG et la valeur de marché de l’entreprise.

H1b : La représentation des salariés au sein du conseil d'administration modère

négativement la relation entre la performance ESG et la valeur de marché de l’entreprise.

II.2 Méthodologie

Cette section décrit notre échantillon et les différentes variables utilisées dans notre étude

empirique.

186
II.2.1 Echantillon

Pour examiner l’effet modérateur des administrateurs salariés sur la relation entre la

performance ESG et la valeur de marché de l’entreprise, nous utilisons un échantillon de 91

sociétés françaises cotées au SBF 120 (les sociétés financières, d’assurance et immobilières ayant

été exclues), sur une période de 11 ans, de 2007 à 2017. Les données sur la performance ESG et

ses trois piliers (social, environnement et de gouvernance) sont collectées à partir du site web de

Thomson Reuters/S-Network ESG Best Practice Ratings. Nous avons collecté les données

financières et comptables à partir de la base de données ThomsonOne. Les données concernant la

gouvernance et l’actionnariat de l’entreprise ont été collecté manuellement à partir des rapports

annuels et des rapports RSE.

II.2.2 Les variables de régression

II.2.2.1 Variable dépendante : Q de Tobin

Suivant Ginglinger et al. (2011), Cahan et al. (2016) et Nekhili et al. (2017b), nous utilisons le Q

de Tobin (Tobin’s Q) comme mesure de la valeur de marché des entreprises. Cette mesure reflète

l’évaluation par le marché de la valeur attendue d’une entreprise à long terme, comme

l’expliquent Cahan et al. (2016).

II.2.2.2 Variable endogène : Performance ESG

Pour mesurer la performance ESG, nous utilisons les notations ESG fournies par Thomson

Reuters / S-Network. Cette mesure fournit des évaluations de la performance ESG et ses trois

piliers (environnemental, social et gouvernance) de plus de 5000 entreprises dans le monde.

187
II.2.2.3 Variables modératrice : Administrateurs salariés

Suivant Ginglinger et al., (2011) et Guedri et Hollandts (2008), nous mesurons la représentation

des salariés au sein du conseil d'administration par le nombre d'administrateurs élus par les

salariés en proportion du nombre total d'administrateurs au sein du conseil d’administration.

II.2.2.4 Variables de contrôle

Suivant les recherches antérieures (par exemple Nekhili et al., 2017a ; 2017b), nous utilisons un

ensemble de variables de contrôles censées influencer la performance ESG et la valeur de marché

de l’entreprise. Nous contrôlons pour la vérification RSE (CSR_ASS), la présence de comité RSE

(CSR_COM), la taille du conseil d’administration (BOAD_SIZE), l’indépendance du conseil

(BOARD_IND), le nombre de réunion (BOARD_MEET), la dualité des fonctions du CEO

(DUAL), le mandat du CEO (TENURE), l’actionnariat familial (FAM_OWN), l’actionnariat

institutionnel (INST_OWN), l’actionnariat salarié (EMPL_OWN), l’effet de levier (LEV),

l’intensité des investissements R&D (R&D), la taille de l’entreprise (SIZE), la loi de 2013

(LAW2013), et finalement l’industrie (INDUSTRY).

II.2.2.5 Modèle empirique

TOBINit = β0 + β1 Lag TOBINit + β2 ESG_PERit + β3 EMPL_BOARDit + β4

(ESG_PER*EMPL_BOARD)it + β5 CSR_ASSit + β6 CSR_COMit + β7 BOARD_SIZEit + β8

BOARD_INDit + β9 BOARD_MEETit + β10 DUALit + β11 TENUREit + β12 FAM_OWNit + β13

INST_OWNit + β14 EMPL_OWNit + β15 LEVit + β16 R&Dit + β17 SIZEit + β18 LAW2013t + β19

INDUSTRYi + ɛit

II.3 Résultats

Le Modèle 1 du tableau II.1 montre un impact positif et significatif de la performance ESG

sur la valeur de marché de l’entreprise, mesurée par le Q de Tobin (β2 = 0,185, t = 5,33). Pour

le Modèle 2, l’impact de la performance ESG sur la valeur de marché est toujours stable et

188
positive et significative sur le Q de Tobin lorsque les administrateurs salariés sont introduits

dans le modèle empirique.

Dans le Modèle 3 du Tableau II.1, nous mesurons l'effet marginal de la performance

ESG sur la valeur de marché des entreprises avec des administrateurs salariés, en utilisant la

technique de test conjoint. Le test clé est le test conjoint de la somme des coefficients de

performance ESG (ESG_PER) et du terme d'interaction (ESG_PER * EMPL_BOARD).

Conformément à notre hypothèse H1b et contrairement à notre hypothèse H1a, le coefficient

(ESG_PER + ESG_PER * EMPL_BOARD) montre un impact négatif et significatif sur le Q

de Tobin et suggère que la représentation des salariés au conseil d'administration modère

négativement la relation entre la performance ESG et la valeur de marché de l’entreprise.

Table II.1 : Régression GMM du Q de Tobin sur la performance ESG et la présence des
salariés au sein du conseil d’administration
Variables Model 1 Model 2 Model 3:
ESG_PER*EMPL_BOARD
Coef. t–test Coef. t–test Coef. t–test
Lag TOBIN 0.833*** 156.32 0.813*** 179.73 0.807*** 125.99
ESG_PER 0.185*** 5.33 0.220*** 5.39 0.752*** 10.75
EMPL_BOARD –1.852*** –11.35 8.864*** 9.19
ESG_PER*EMPL_BOARD –8.566*** –7.38
CSR_ASS 0.065*** 5.10 0.051*** 2.71 –0.033* –1.72
CSR_COM 0.045*** 5.32 0.044*** 2.61 0.047** 2.16
BOARD_SIZE 0.107*** 3.59 0.004 0.07 0.040 0.65
BOARD_IND –0.026 –0.92 0.353*** 7.79 0.381*** 9.70
BOARD_MEET 0.010 0.79 0.010 0.40 –0.038 –1.43
DUAL –0.119*** –8.73 –0.135*** –7.62 –0.140*** –7.05
TENURE –0.032*** –4.49 –0.020** –2.13 –0.026** –2.41
FAM_OWN 0.220*** 5.03 0.508*** 9.32 0.478*** 7.21
INST_OWN –0.003 –0.10 –0.038 –1.11 –0.081** –2.52
EMPL_OWN 0.340 1.38 0.403 1.12 0.093 0.29
LEV 0.299*** 6.99 0.358*** 5.82 0.374*** 5.99
R&D –0.241* –1.75 0.001 0.01 0.493*** 2.98
SIZE –0.031*** –6.27 –0.064*** –4.98 –0.067*** –6.13
LAW2013 0.067*** 8.92 0.004 0.38 0.028*** 3.10
Intercept –0.181*** –2.58 0.017 0.14 –0.293** –1.99
Industry Yes Yes Yes
Number of observations 741 741 741
Fisher (Prob. > F) 1965.33 (p = 0.000) 6145.54 (p = 0.000) 7333.52 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.75 (p = 0.003) –2.75 (p = 0.003) –2.90 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.43 (p = 0.670) –0.69 (p = 0.887) 0.14 (p = 0.887)
Sargan test (Chi–square, p–value): 594.41 (p = 0.000) 591.31 (p = 0.000) 733.83 (p = 0.000)
Hansen test (Chi–square, p–value): 69.84 (p = 0.382) 67.33 (p = 0.431) 77.10 (p = 0.211)
Joint test: ESG_PER + (ESG_PER*EMPL_BOARD) –7.814*** –6.71
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively.

189
II.4 Analyses Supplémentaires

Dans le tableau II.2, nous examinons l’effet modérateur de la présence des salariés au sein du

conseil d’administration sur la relation entre les trois piliers de la performance ESG, à savoir

le social, l’environnemental et en matière de gouvernance et la valeur de marché de

l’entreprise, mesurée par le Q de Tobin. Pour le pilier social, les résultats empiriques du

Modèle 1 montrent que l'impact de la variable d'interaction (SOCIAL_PER + SOCIAL_PER

* EMPL_BOARD) est négatif (β4 = –14.258) et fortement significatif (t = –8.52) sur le Q de

Tobin, suggérant que les investisseurs pénalisent financièrement une entreprise ayant une

bonne performance sociale lorsque les salariés sont représentés au sein du conseil

d’administration.

Les Modèles 2 et 3 du tableau II.2 montrent également une relation négative et

significative entre la somme du coefficient (ENVIR_PER + ENVIR_PER * EMPL_BOARD)

et le Q de Tobin, d'une part, et entre (CG_PER + CG_PER * EMPL_BOARD) et le q de

Tobin, d'autre part. Ces résultats suggèrent que le marché évalue négativement les

performances environnementales et les performances de gouvernance lorsque les salariés sont

nommés au sein du conseil d’administration.

190
Table II.2 : Régression GMM du Q de Tobin sur l’interaction entre la performance Sociale, Environnementale et en matière de Gouvernance et la
présence des salariés au sein du conseil d’administration.
Variables Model 1: Model 2: Model 3:
SOCIAL_PER*EMPL_BOARD ENVIR_PER*EMPL_BOARD CG_PER*EMPL_BOARD
Coef. t–test Coef. t–test Coef. t–test
Lag TOBIN 0.819*** 126.87 0.813*** 94.42 0.811*** 101.48
EMPL_BOARD 15.674*** 10.37 8.666*** 3.66 6.851*** 11.92
SOCIAL_PER 0.844*** 6.22
SOCIAL_PER*EMPL_BOARD –15.103*** –8.58
ENVIR_PER –0.258* –1.67
ENVIR_PER*EMPL_BOARD –8.521*** –3.42
CG_PER 0.933*** 12.26
CG_PER*EMPL_BOARD –8.767*** –9.30
CSR_ASS –0.021 –1.00 0.059* 1.83 –0.003 –0.11
CSR_COM 0.050** 2.31 0.091*** 3.07 0.031 1.60
BOARD_SIZE 0.027 0.49 0.019 0.29 0.047 0.76
BOARD_IND 0.445*** 6.85 0.447*** 4.80 0.579*** 6.35
BOARD_MEET –0.008 –0.32 0.003 0.11 –0.109*** –3.07
DUAL –0.125*** –5.43 –0.109*** –4.64 –0.109*** –4.53
TENURE –0.023* –1.75 –0.056*** –3.06 –0.014 –1.11
FAM_OWN 0.447*** 5.96 0.325*** 3.76 0.650*** 6.17
INST_OWN –0.156*** –3.46 –0.037 –0.72 –0.056 –1.39
EMPL_OWN –0.422 –1.41 0.534 1.17 0.278 0.71
LEV 0.252*** 2.55 0.172** 2.40 0.434*** 5.97
R&D 0.385** 2.07 0.808* 1.89 0.631*** 2.90
SIZE –0.077*** –6.73 –0.013 –0.99 –0.072*** –4.93
LAW2013 –0.012 –0.98 0.018 0.90 0.042*** 3.85
Intercept –0.361*** –2.61 0.114 0.63 –0.296** –1.98
Industry Yes Yes Yes
Number of observations 741 741 741
Fisher (Prob. > F) 1348.08 (p = 0.000) 1753.64 (p = 0.000) 1592.60 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.81 (p = 0.004) –2.79 (p = 0.004) –2.87 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.56 (p = 0.906) –0.41 (p = 0.680) –0.20 (p = 0.434)
Sargan test (Chi–square, p–value): 573.49 (p = 0.000) 177.46 (p = 0.000) 550.07 (p = 0.000)
Hansen test (Chi–square, p–value): 62.12 (p = 0.299) 58.90 (p = 0.333) 58.18 (p = 0.413)
Joint test: SOCIAL_PER + (SOCIAL_PER*EMPL_BOARD) –14.258*** –8.52
Joint test: ENVIR_PER + (ENVIR_PER*EMPL_BOARD) –8.780*** –3.64
Joint test: CG_PER + (CG_PER*EMPL_BOARD) –7.834*** –8.57
Note:*, **, *** Represent significance at the 0.10, 0.05 and 0.01 levels, respectively. All variables are as defined in Table II.1.

191
192
 Chapitre III : La relation entre la performance ESG et
la performance financière : le type d’administrateur
salarié est-il important ?

La représentation des salariés au sein du conseil d’administration est une politique ayant pour

objectif d’équilibrer les forces entre le capital humain et le capital financier, et pour mieux prendre

en considération les préoccupations des différentes parties prenantes plutôt que de se concentrer

uniquement sur les actionnaires. En France, les conseils d’administration peuvent inclure deux types

d’administrateurs salariés. Ces deux types d’administrateur ont, en effet, des intérêts différents ce

qui peut impacter leur comportement au sein du conseil d’administration et, plus particulièrement,

au sujet de la performance ESG.

Dans ce chapitre, nous essayons d’abord d’examiner l’impact de chaque type d’administrateur

salarié sur les différents piliers de la performance ESG, et d’examiner par la suite si les

investisseurs sont sensibles au type d’administrateur salarié nommé au sein du conseil

d’administration lorsqu’ils évaluent la performance ESG de l’entreprise. En utilisant un échantillon

apparié d'entreprises françaises cotées appartenant à l'indice SBF 120 pour la période 2007-2017,

nous trouvons que les deux types d’administrateurs salariés ont des impacts différents sur la

performance ESG de l'entreprise. Nos résultats montrent que la représentation des travailleurs au

sein du conseil d'administration améliore exclusivement la performance sociale, alors que la

représentation des salariés actionnaires améliore la performance ESG, la performance

environnementale et la performance en matière de gouvernance d'entreprise. Nous constatons que si

la présence des représentants des salariés actionnaires au conseil d’administration modère

positivement la perception des investisseurs envers la performance ESG, la présence des

représentants des travailleurs modère négativement cette perception.

193
III.1 Développement des hypothèses

En examinant la littérature, nous trouvons que les représentants des travailleurs au sein du conseil

d’administration cherchent à maximiser les intérêts des salariés alors que les représentants des

salariés actionnaires se comportent comme des administrateurs traditionnels (Ginglinger et al.,

2011). En conséquence, nous formulons les hypothèses suivantes :

H1a : La proportion des représentants des travailleurs au sein du conseil d’administration est

positivement associée à la performance sociale de l’entreprise.

H1b : La proportion des représentants des salariés actionnaires au sein du conseil

d’administration est négativement associée à la performance sociale de l’entreprise.

H2a : La proportion des représentants des travailleurs au sein du conseil d’administration est

positivement associée à la performance environnementale de l’entreprise.

H2b : La proportion des représentants des salariés actionnaires au sein du conseil

d’administration est positivement associée à la performance environnementale de l’entreprise.

H3a : La proportion des représentants des travailleurs au sein du conseil d’administration est

négativement associée à la performance de l’entreprise en matière de gouvernance.

H3b : La proportion des représentants des salariés actionnaires au sein du conseil

d’administration est positivement associée à la performance de l’entreprise en matière de

gouvernance.

H4a : La proportion des représentants des travailleurs au sein du conseil d’administration

modère négativement la relation entre la performance ESG et la valeur de marché de l’entreprise.

H4b : La proportion des représentants des salariés actionnaires au sein du conseil

d’administration modère positivement la relation entre la performance ESG et la valeur de marché

de l’entreprise.

194
III.2. Méthodologie

Cette section décrit l’échantillon, les sources de données et les différentes variables utilisées dans

notre étude empirique.

III.2.1 Echantillon

Pour examiner nos hypothèses de recherche, nous utilisons un échantillon de sociétés françaises cotées

au SBF 120 (les sociétés financières, d’assurance et immobilières ayant été exclues), sur une période de

11 ans, de 2007 à 2017. Notre échantillon final se compose de 817 observations. Les données sur la

performance ESG et ses trois piliers (social, environnement et de gouvernance) sont collectées à partir

du site web de Thomson Reuters/S-Network ESG Best Practice Ratings. Nous avons collecté les

données financières et comptables à partir de la base de données ThomsonOne. Les données concernant

la gouvernance et l’actionnariat de l’entreprise ont été collectées manuellement à partir des rapports

annuels et des rapports RSE.

III.2.2 Les variables de régression

III.2.2.1 Variables dépendantes

L'objectif de ce chapitre est double : (1) examiner l'impact des différents types de représentation des

salariés (c'est-à-dire la représentation des travailleurs et la représentation des salariés actionnaires) sur la

performance ESG et (2) déterminer si la perception de la performance ESG par les investisseurs est

influencée par la manière dont les salariés sont représentés au sein du conseil d’administration. Deux

variables dépendantes sont ainsi considérées dans notre étude : la performance ESG (ESG performance)

et la valeur de marché (Tobin’s q), mesurée par le q de Tobin.

III.2.2.2 Variable endogène : Administrateurs salariés

Suivant les études précédentes (Ginglinger et al., 2011; Guedri & Hollandts, 2008), nous mesurons la

représentation des salariés au sein du conseil d’administration (Employee directorship) par le ratio du

nombre d'administrateurs salariés (élus par les travailleurs et élus par les salariés actionnaires) divisé par

195
le nombre total d'administrateurs au sein du conseil d’administration. La présence de représentants

travailleurs (Labour representation) est mesurée par la proportion d'administrateurs élus par les salariés

par leur droit de travail par rapport au nombre total d'administrateurs au sein du conseil. La présence des

représentants des salariés actionnaires au conseil (Employe-shareholders representation) est mesurée par

la proportion d'administrateurs élus par les salariés actionnaires par rapport au nombre total

d'administrateurs.

III.2.2.3 Variables de contrôle

Nous utilisons un ensemble de variables de contrôles censées influencer la performance ESG et la

valeur de marché de l’entreprise. Nous contrôlons pour la présence des rapports RSE (Standalone) la

vérification RSE (CSR assurance), la présence de comité RSE (CSR committee), la taille du conseil

d’administration (Board size), l’indépendance du conseil (Board independence), la présence des femmes

au sein du conseil d’administration (Board gender diversity) le nombre de réunion (Board meeting), la

dualité des fonctions du CEO (Duality), le mandat du CEO (CEO tenure), l’actionnariat familial

(Family ownership), l’actionnariat institutionnel (Institutional ownership), l’effet de levier (Leverage), la

rentabilité des actifs (ROA), l’intensité des investissements R&D (R&D), la taille de l’entreprise (Firm

size), la loi de 2013 (Law 2013), et l’industrie (Industry).

III.2.2.5 Modèle empirique

Dans ce chapitre on utilise deux modèles empiriques.

ESG performancei,t = β0 + β1 Lag ESG performancei,t + β2 Employee directorshipi,t + β3

Standalonei,t + β4 CSR assurancei,t + β5 CSR committeei,t + β6 Board sizei,t + β7 Board

independencei,t + β8 Board gender diversityi,t + β9 Board meetingi,t + β10 Dualityi,t + β11 CEO

tenurei,t + β12 Family ownershipi,t + β13 Institutional ownershipi,t + β14 Leveragei,t + β15 ROAi,t + β16

R&Di,t + β17 Firm size i,t + β18 Law 2013_FE + β19 Industry_FE + εi,t

196
 Tobin’s qi,t = β0 + β1 Lag Tobin’s qi,t + β2 ESG performancei,t + β3 Employee directorshipi,t +

β4 (ESG performancei,t * Employee directorshipi,t) + β5 Standalonei,t + β6 CSR assurancei,t + β7 CSR

committeei,t + β8 Board sizei,t + β9 Board independencei,t + β10 Board gender diversityi,t + β11 Board

meetingi,t + β12 Dualityi,t + β13 CEO tenurei,t + β14 Family ownershipi,t + β15 Institutional ownershipi,t

+ β16 Leveragei,t + β17 ROAi,t + β18 R&Di,t + β19 Firm size i,t + β20 Law 2013_FE + β21 Industry_FE

+ εi,t

III.3 Résultats

Le Tableau III.1 présente les résultats de la régression du système GMM de la performance ESG sur la

représentation des travailleurs au sein du conseil d’administration. Nos résultats montrent que la

représentation des travailleurs réduit significativement la performance ESG de l’entreprise. En

examinant les trois piliers de la performance ESG, nous constatons que la représentation des travailleurs

améliore exclusivement le pilier social (Modèle 2) et a réduit la performance environnementale (Modèle

3) et la performance en matière de gouvernance d'entreprise (Modèle 4). Par conséquent, nous

confirmons nos hypothèses H1a et H3a et rejetons l’hypothèse H2a.

Contrairement au Tableau III.1, le Tableau III.2 révèle une relation positive et significative entre

la présence des représentants des salariés actionnaires et la performance ESG. Le Tableau III.2 montre

également que les représentants des salariés actionnaires améliorent la performance environnementale et

en matière de gouvernance et réduisent la performance sociale. Les relations prédites par les hypothèses

H1b, H2b et H3b sont alors confirmées.

197
Table III.1 : Régression du système GMM de la performance ESG sur la représentation des travailleurs au sein du conseil
d’administration
Model 1: Model 2: Model 3: Model 4:
ESG performance Social Environmental Corporate governance
Variables
performance performance performance
Coef. t-test Coef. t-test Coef. t-test Coef. t-test
Lag ESG performance 0.545*** 58.11
Lag Social performance 0.756*** 37.29
Lag Environmental performance 0.665*** 55.14
Lag Corporate governance performance 0.475*** 49.73
Labour representation –0.113*** –8.43 0.067** 2.51 –0.067*** –5.32 –0.196*** –12.79
Standalone 0.013*** 5.25 0.002 0.69 0.013*** 4.37 0.014*** 3.45
CSR assurance 0.020*** 6.80 –0.005 –0.82 0.014*** 5.98 0.009* 1.96
CSR committee 0.005* 1.67 –0.001 –0.12 0.008*** 2.78 –0.001 –0.17
Board size –0.034*** –5.25 –0.011 –0.98 –0.021*** –2.62 –0.042*** –2.94
Board independence 0.064*** 12.41 0.041*** 2.87 0.061*** 6.64 0.023** 2.21
Board gender diversity 0.084*** 8.90 0.009 0.63 0.053*** 7.23 0.029** 2.14
Board meetings –0.006*** –4.62 –0.002 –0.71 –0.007** –2.53 –0.001 –0.26
Duality 0.019*** 10.73 0.009** 2.04 0.022*** 6.24 –0.011** –2.43
CEO tenure –0.009*** –10.09 –0.005** –2.29 –0.007*** –5.54 –0.013*** –6.60
Family ownership 0.002 0.25 –0.019* –1.95 0.012 1.19 –0.050*** –6.46
Institutional ownership 0.028*** 6.81 0.012* 1.69 0.025*** 3.51 0.026*** 3.07
Leverage –0.017*** –3.01 –0.009 –0.81 –0.023*** –3.06 –0.041*** –6.60
ROA 0.022 0.75 –0.075 –1.60 –0.034 –0.76 0.114*** 2.74
R&D –0.025 –1.05 –0.198*** –3.42 0.032 0.59 0.106** 2.05
Firm size 0.012*** 9.23 0.002 0.95 0.010*** 5.64 0.013*** 5.17
Law 2013 –0.020*** –13.14 –0.008*** –2.70 0.011*** 6.95 –0.034*** –16.93
Intercept 0.132*** 5.67 0.173*** 4.84 0.078*** 2.62 0.218*** 7.47
Industry Yes Yes Yes Yes
Number of obs. 313 313 313 313
F (Prob > F) 3769.38 (p = 0.000) 839.24 (p = 0.000) 7950.47 (p = 0.000) 6676.49 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –3.95 (p = 0.000) –3.77 (p = 0.000) –4.80 (p = 0.000) –3.42 (p = 0.001)
Arellano–Bond test AR(2) (z, p–value): 1.04 (p = 0.302) 1.44 (p = 0.150) 1.24 (p = 0.125) –0.94 (p = 0.349)
Sargan test (Chi–square, p–value): 251.65 (p = 0.000) 65.42 (p = 0.001) 279.88 (p = 0.000) 289.78 (p = 0.000)
Hansen test (Chi–square, p–value): 58.82 (p = 0.211) 40.89 (p = 0.228) 60.50 (p = 0.223) 56.87 (p = 0.266)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.

198
Table III.2 : Régression du système GMM de la performance ESG sur la représentation des salariés actionnaires au sein du conseil
d’administration
Model 1: Model 2: Model 3: Model 4:
ESG performance Social Environmental Corporate governance
Variables
performance performance performance
Coef. t-test Coef. t-test Coef. t-test Coef. t-test
Lag ESG performance 0.441*** 28.52
Lag Social performance 0.496*** 22.32
Lag Environmental performance 0.721*** 34.47
Lag Corporate governance performance 0.376*** 25.91
Employee-shareholders representation 0.449*** 8.59 –0.606*** –6.39 0.414*** 7.67 1.047*** 10.01
Standalone –0.008** –2.23 0.002 0.48 –0.005 –1.63 0.003 0.69
CSR assurance 0.025*** 5.95 0.003 0.39 0.005 1.03 0.022* 1.70
CSR committee 0.018*** 5.39 –0.001 –0.17 0.018*** 6.34 0.036*** 4.71
Board size –0.022 –1.61 0.021* 1.67 –0.011 –1.37 –0.039* –1.71
Board independence 0.085*** 7.03 0.019 1.04 0.049*** 4.65 0.104*** 5.86
Board gender diversity 0.051*** 4.73 0.074*** 4.49 –0.016 –1.12 –0.045* –1.69
Board meetings 0.021*** 3.96 0.006 0.47 0.007* 1.66 0.011 1.44
Duality –0.004 –0.96 0.012** 2.04 0.013*** 4.11 –0.049*** –9.73
CEO tenure 0.003 1.17 0.014*** 4.91 –0.005** –2.33 0.004 0.72
Family ownership 0.037 1.25 –0.083*** –2.96 0.057** 2.04 0.034 0.72
Institutional ownership –0.022** –2.36 –0.027*** –3.48 –0.009 –1.37 –0.018 –1.34
Leverage –0.009 –0.54 0.027 1.57 –0.017 –1.12 –0.001 –0.06
ROA 0.027 0.62 0.245*** 8.97 0.054 0.94 –0.168*** –3.15
R&D 0.015 0.11 0.110* 1.69 –0.026 –0.47 0.122 0.48
Firm size 0.009*** 4.94 0.007** 2.53 0.006*** 3.98 0.002 0.97
Law 2013 –0.004** –2.19 –0.022*** –6.60 0.032*** 10.56 –0.041*** –8.10
Intercept 0.146*** 3.73 0.167*** 3.04 0.062* 1.84 0.322*** 6.22
Industry Yes Yes Yes Yes
Number of obs. 288 288 288 288
F (Prob > F) 1742.01 (p = 0.000) 2377.08 (p = 0.000) 7455.16 (p = 0.000) 8314.40 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –3.48 (p = 0.000) –3.46 (p = 0.000) –3.25 (p = 0.001) –3.47 (p = 0.001)
Arellano–Bond test AR(2) (z, p–value): –1.07 (p = 0.252) 0.06 (p = 0.956) 1.44 (p = 0.150) –1.83 (p = 0.167)
Sargan test (Chi–square, p–value): 231.65 (p = 0.000) 248.87 (p = 0.000) 249.77 (p = 0.000) 254.85 (p = 0.000)
Hansen test (Chi–square, p–value): 42.88 (p = 0.169) 43.14 (p = 0.162) 41.97 (p = 0.194) 44.62 (p = 0.128)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.

199
 Le Tableau III.3 présente les résultats de la régression GMM du Q de Tobin sur la

performance ESG et la présence de chaque type d’administrateur salarié. Conformément au résultat

de notre deuxième chapitre, la présence des administrateurs salariés modère négativement la

relation entre la performance ESG et la valeur de marché de l’entreprise. Les résultats des Modèles

2 et 3 montrent clairement que cet effet négatif provient principalement de la représentation des

travailleurs au sein du conseil d’administration. Nos résultats suggèrent par conséquence que les

investisseurs sont favorables à la représentation des salariés dans le conseil d’administration tandis

qu’ils sont contre la représentation des travailleurs. Les hypothèses H4a and H4b sont, par

conséquence, confirmées.

200
Table III.3 : Régression du système GMM du q de Tobin sur l’interaction entre la
performance ESG et la représentation des salariés au sein du conseil d’administration
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.497*** 96.13 0.830*** 51.34 0.956*** 101.15
ESG performance 0.521*** 4.77 1.885*** 3.87 –0.763*** –4.25
Employee directorship 2.121*** 3.88
ESG performance * Employee directorship –3.261*** –3.87
Labour representation 11.859*** 3.97
ESG performance * Labour representation –18.537*** –4.16
Employee-shareholders representation –13.538*** –4.08
ESG performance * Employee-shareholders 21.727*** 4.70
representation
Standalone 0.010 0.56 –0.029 –0.84 0.063*** 5.47
CSR assurance 0.081*** 3.76 –0.109** –2.29 0.011 0.44
CSR committee 0.077*** 3.61 –0.015 –0.47 0.026 1.52
Board size –0.060 –1.04 0.161* 1.88 0.004 0.11
Board independence –0.029 –0.68 0.005 0.05 0.127** 2.45
Board gender diversity –0.031 –0.57 –0.026 –0.21 0.027 0.58
Board meetings 0.031 1.57 –0.056 –1.59 0.012 0.57
Duality –0.116*** –5.32 –0.097** –2.52 0.042*** 2.77
CEO tenure –0.022* –1.64 –0.026 –1.58 –0.018** –2.22
Family ownership 0.283*** 3.85 0.335*** 3.44 0.191** 2.23
Institutional ownership –0.115*** –3.44 –0.062 –0.92 0.038* 1.70
Leverage 0.802*** 16.27 0.118 1.28 0.145** 2.32
ROA 5.651*** 34.40 2.675*** 9.95 1.087*** 5.62
R&D 2.391*** 7.14 2.251*** 3.78 0.576** 2.14
Firm size –0.118*** –10.09 –0.028 –1.16 –0.025*** –3.31
Law 2013 0.073*** 6.00 0.091*** 3.79 0.058*** 4.50
Intercept 1.743*** 8.48 –0.943** –2.05 0.599*** 3.26
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 42872.70 (p = 0.000) 7356.76 (p = 0.000) 6051.65 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.37 (p = 0.006) –3.09 (p = 0.001) –2.64 (p = 0.003)
Arellano–Bond test AR(2) (z, p–value): –1.09 (p = 0.279) –0.93 (p = 0.272) –0.76 (p = 0.446)
Sargan test (Chi–square, p–value): 367.43 (p = 0.000) 222.57 (p = 0.000) 227.75 (p = 0.000)
Hansen test (Chi–square, p–value): 50.72 (p = 0.367) 35.74 (p = 0.297) 37.50 (p = 0.231)
Joint test: ESG performance + (ESG –2.739*** –3.54
performance * Employee directorship)

Joint test: ESG performance + (ESG –16.652*** –4.06

performance * Labour representation)
Joint test: ESG performance + (ESG 20.964*** 4.65
performance * Employee-shareholders
representation)
*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.

201
III.4 Analyses Supplémentaires

Pour mieux comprendre l’impact de chaque type d’administrateur salarié sur la relation entre la

performance ESG et la valeur de marché de l’entreprise, nous avons procédé à des analyses

supplémentaires. En effet, on a distingué entre les trois piliers de la performance ESG à savoir le

pilier social, le pilier environnemental et le pilier de gouvernance et on a examiné leurs impacts sur

la valeur de marché de l’entreprise en considérant le type d’administrateur salarié nommé au conseil

d’administration. Les résultats des Tableaux III.4 et III.6, montrent que la présence des

représentants des travailleurs au sein du conseil d’administration modère négativement la relation

entre la performance sociale (en matière de gouvernance) et la valeur de marché de l’entreprise.

Inversement, la présence des représentants des salariés actionnaires au sein du conseil

d’administration modère positivement la relation entre la performance sociale (en matière de

gouvernance) et la valeur de marché de l’entreprise. En ce qui concerne le pilier environnemental,

les résultats du tableau III.5 montrent que la performance environnementale n’est pas valorisée par

les investisseurs quelque soit le type d’administrateur salarié nommé au sein du conseil

d’administration.

202
Table III.4 : Régression du système GMM du q de Tobin sur l’interaction entre la
performance sociale et la représentation des salariés au sein du conseil d’administration
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.831*** 153.07 0.821*** 56.76 0.934*** 86.12
Social performance –1.718*** –7.56 0.301 1.44 –0.619*** –2.57
Employee directorship 6.106*** 4.10
Social performance * Employee –9.827*** –4.80
directorship
Labour representation 14.790*** 7.56
Social performance * Labour –18.671*** –7.29
representation
Employee-shareholders representation –8.624*** –2.90
Social performance * Employee- 15.553*** 3.68
shareholders representation
Standalone 0.117*** 5.68 0.028 1.02 0.071*** 3.36
CSR assurance 0.150*** 8.49 –0.044 –1.25 0.009 0.38
CSR committee 0.099*** 5.66 0.051 1.53 0.057** 2.19
Board size 0.070 1.03 –0.012 –0.11 0.027 0.43
Board independence –0.067 –0.85 0.258*** 2.85 0.177*** 3.04
Board gender diversity –0.049 –0.43 –0.069 –0.65 0.097 1.49
Board meetings –0.024 –0.85 0.020 0.54 0.027 1.28
Duality –0.068** –2.17 –0.043 –1.09 0.011 0.64
CEO tenure –0.026*** –3.00 –0.059*** –3.15 –0.028*** –3.42
Family ownership –0.124* –1.79 0.137* 1.77 0.215 1.62
Institutional ownership –0.007 –0.15 –0.161*** –2.70 0.007 0.23
Leverage 0.174*** 3.15 –0.108 –1.18 0.131* 1.95
ROA 1.547*** 7.33 2.245*** 8.39 1.208*** 5.02
R&D –0.016 –0.05 0.189 0.40 0.398 0.86
Firm size 0.030** 2.05 –0.043** –2.38 –0.036*** –2.82
Law 2013 0.079*** 3.05 0.067** 2.52 0.060*** 4.02
Intercept 0.572** 2.30 0.552 1.39 0.577*** 2.82
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 9734.58 (p = 0.000) 2576.80 (p = 0.000) 9855.65 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.50 (p = 0.005) –3.06 (p = 0.001) –2.55 (p = 0.005)
Arellano–Bond test AR(2) (z, p–value): –0.29 (p = 0.773) –1.04 (p = 0.297) –0.95 (p = 0.340)
Sargan test (Chi–square, p–value): 125.32 (p = 0.000) 90.19 (p = 0.000) 226.78 (p = 0.000)
Hansen test (Chi–square, p–value): 45.53 (p = 0.453) 35.61 (p = 0.221) 36.54 (p = 0.266)
Joint test: Social performance + (Social –11.545*** –6.12
performance * Employee directorship)
Joint test: Social performance + (Social –18.370*** –7.48
performance * Labour representation)
Joint test: Social performance + (Social 14.934*** 3.69
performance * Employee-shareholders
representation)

*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.

203
Table III.5 : Régression du système GMM du q de Tobin sur l’interaction entre la
performance environnementale et la représentation des salariés au sein du conseil
d’administration
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.798*** 157.17 0.836*** 81.54 0.908*** 111.02
Environmental performance 2.229*** 15.64 0.911*** 2.96 0.638** 2.52
Employee directorship 18.445*** 13.72
Environmental performance * Employee –24.089*** –13.05
directorship
Labour representation 11.600*** 4.43
Environmental performance * Labour –17.274*** –4.82
representation
Employee-shareholders representation 11.833*** 4.94
Environmental performance * Employee- –14.131*** –4.49
shareholders representation
Standalone 0.043** 2.50 0.056** 2.05 0.040 3.22
CSR assurance –0.052* –1.83 –0.103*** –3.04 –0.037 –1.49
CSR committee 0.041** 2.45 0.039 1.07 0.038*** 2.58
Board size 0.068* 1.67 0.037 0.58 0.001 0.01
Board independence –0.082 –1.63 0.083 0.95 0.087** 2.26
Board gender diversity –0.067 –0.75 0.143 1.15 0.304*** 5.89
Board meetings –0.003 –0.14 –0.082** –2.33 0.040*** 2.84
Duality –0.180*** –8.22 –0.113*** –2.97 –0.019 –1.52
CEO tenure –0.018** –2.26 –0.030* –1.95 –0.012 –1.54
Family ownership 0.092* 1.96 0.400*** 4.83 0.256*** 3.43
Institutional ownership –0.056 –1.62 0.152** 2.26 0.025 1.20
Leverage 0.437*** 7.44 0.097 1.27 0.345*** 6.19
ROA 2.363*** 15.49 3.025*** 13.58 0.908*** 5.37
R&D 1.158** 2.42 2.854*** 4.81 0.082 0.26
Firm size –0.079*** –6.39 0.019 1.02 –0.025*** –2.91
Law 2013 0.076*** 4.44 0.072*** 2.88 0.042*** 4.38
Intercept –0.378 –1.39 –1.008*** –3.34 –0.400** –2.33
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 20798.85 (p = 0.000) 5261.69 (p = 0.000) 31680.82 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.46 (p = 0.004) –3.09 (p = 0.001) –2.58 (p = 0.004)
Arellano–Bond test AR(2) (z, p–value): –0.45 (p = 0.652) –0.85 (p = 0.396) –1.19 (p = 0.235)
Sargan test (Chi–square, p–value): 337.58 (p = 0.000) 226.36 (p = 0.000) 110.80 (p = 0.000)
Hansen test (Chi–square, p–value): 60.88 (p = 0.119) 42.50 (p = 0.102) 41.21 (p = 0.127)
Joint test: Environmental performance + –21.860*** –12.36
(Environmental performance * Employee
directorship)
Joint test: Environmental performance + –16.363*** –4.90
(Environmental performance * Labour
representation)
Joint test: Environmental performance + –13.493*** –4.63
(Environmental performance * Employee-
shareholder board representation)

*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.

204
Table III.6 : Régression du système GMM du q de Tobin sur l’interaction entre la
performance en matière de gouvernance et la représentation des salariés au sein du conseil
d’administration
Model 1: Model 2: Model 3:
Employee Labour Employee-shareholders
Variables
directorship representation representation
Coef. t-test Coef. t-test Coef. t-test
Lag Tobin’s q 0.486*** 85.29 0.765*** 50.92 0.919*** 59.15
Corporate governance performance 1.819*** 10.46 1.578*** 8.18 –0.690** –2.40
Employee directorship 11.322*** 6.97
Corporate governance performance * –21.289*** –7.37
Employee directorship
Labour representation 2.118*** 3.81
Corporate governance performance * Labour –3.020*** –4.23
representation
Employee-shareholders representation –14.205*** –4.61
Corporate governance performance * 27.316*** 5.25
Employee-shareholders representation
Standalone 0.015 0.50 0.022 0.69 0.093*** 5.68
CSR assurance 0.121*** 3.62 –0.019 –0.56 –0.034 –1.18
CSR committee 0.064* 1.85 –0.004 –0.09 0.032* 1.79
Board size 0.162** 2.17 0.122 1.23 –0.071 –1.21
Board independence 0.034 0.43 0.107 1.14 0.004 0.07
Board gender diversity –0.043 –0.32 0.023 0.20 0.123** 2.01
Board meetings –0.049* –1.97 0.025 0.82 –0.032 –1.15
Duality –0.147*** –4.54 –0.041 –1.05 0.065*** 3.24
CEO tenure –0.036*** –2.73 –0.006 –0.39 –0.018* –1.90
Family ownership 0.539*** 5.58 0.380*** 4.34 0.178** 2.35
Institutional ownership –0.086* –1.64 –0.173*** –2.68 0.053* 1.64
Leverage 1.076*** 15.10 0.277** 2.49 0.266*** 3.71
ROA 6.216*** 25.84 2.742*** 12.73 1.644*** 7.77
R&D 3.495*** 5.56 1.267* 1.80 1.292** 2.52
Firm size –0.119*** –6.63 –0.076*** –4.72 –0.002 –0.14
Law 2013 0.125*** 7.61 0.154*** 4.50 0.113*** 5.47
Intercept 0.490* 1.80 –0.063 –0.17 0.337* 1.76
Industry Yes Yes Yes
Number of obs. 444 313 288
F (Prob > F) 3732.47 (p = 0.000) 5951.90 (p = 0.000) 8409.00 (p = 0.000)
Arellano–Bond test AR(1) (z, p–value): –2.39 (p = 0.006) –3.07 (p = 0.000) –2.43 (p = 0.006)
Arellano–Bond test AR(2) (z, p–value): –1.14 (p = 0.253) –1.16 (p = 0.248) –1.18 (p = 0.238)
Sargan test (Chi–square, p–value): 334.83 (p = 0.000) 97.44 (p = 0.000) 93.84 (p = 0.000)
Hansen test (Chi–square, p–value): 58.00 (p = 0.153) 32.81 (p = 0.285) 35.11 (p = 0.239)
Joint test: Corporate governance –19.469*** –7.03
performance + (Corporate governance
performance * Employee directorship)
Joint test: Corporate governance –1.442*** –2.35
performance + (Corporate governance
performance * Labour representation)
Joint test: Corporate governance 26.625*** 5.38
performance + (Corporate governance
performance * Employee-shareholders
representation)

*, **, *** Represent significance at 0.10, 0.05 and 0.01 levels, respectively.

205
Références

Allix-Desfautaux, E., & Makany, L. (2015), Développement durable et gestion d’une entreprise :
croisements fertiles, Management & Avenir, 81, 15–36.
Barnea, A., & Rubin, A. (2010). Corporate Social Responsibility as a conflict between shareholders.
Journal of Business Ethics, 97, 71–86.
Bøhren, O., & Strøm, R. O. (2010). Governance and politics: Regulating independence and
diversity in the board room. Journal of Business Finance and Accounting, 37, 1281–1308.
Brammer, S., Millington, A., & Rayton, B. (2007). The contribution of corporate social
responsibility to organizational commitment. The International Journal of Human Resource
Management, 18(10), 1701–1719.
Cahan, S. F., De Villiers, C., Jeter, D. C., Naiker, V., & Van Staden, C. J. (2016). Are CSR
disclosures value relevant? Cross-country evidence. The European Accounting Review, 25, 579–
611.
Cormier, D., Aerts, W., Ledoux, M. J., & Magnan, M. (2009). Attributes of social and human
capital disclosure and information asymmetry between managers and investors. Canadian
Journal of Administrative Sciences, 26, 71–88.
Faleye, O., Mehrotra, V., & Morck, R. (2006). When labor has a voice in corporate governance.
Journal of Financial and Quantitative Analysis, 41, 489–510.
Friede, G., Busch, T., & Bassen, A. (2015). ESG and financial performance: Aggregated evidence
from more than 2000 empirical studies. Journal of Sustainable Finance & Investment, 5, 210–
233.
Fuente, J. A., García-Sanchez, I. M., & Lozano, M. B. (2017). The role of the board of directors in
the adoption of GRI guidelines for the disclosure of CSR information. Journal of Cleaner
Production, 141, 737–750.
Ginglinger, E., Megginson, W., & Waxin, T. (2011). Employee ownership, board representation,
and corporate financial policies. Journal of Corporate Finance, 17(4), 868–887.
Guedri, Z., & Hollandts, X. (2008). Beyond dichotomy: The curvilinear impact of employee
ownership on firm performance. Corporate governance: An international review, 16, 460–474.
Huse, M., Nielsen, S. T., & Hagen, I. M. (2009). Women and employee-elected board members,
and their contributions to board control tasks. Journal of Business Ethics, 89, 581–597.
Nekhili, M., Nagati, H., Chtioui, T., & Rebolledo, C. (2017a). Corporate social responsibility
disclosure and market value: Family versus nonfamily firms. Journal of Business Research, 77,
41–52.

206
Nekhili, M., Nagati, H., Chtioui, T., & Nekhili, A. (2017b). Gender-diverse board and the relevance
of voluntary CSR reporting. International Review of Financial Analysis, 50, 81–100.
Preuss, L., Haunschild, A., & Matten, D. (2009). The rise of CSR : Implications for HRM and
employee representation. The International Journal of Human Resource Management, 20(4),
953–973.
Sahed-Grangery, Y., & Boncori, A.L. (2014), La traduction des attentes des parties prenantes en
RSE en indicateurs de la Global Reporting Initiative : vers un mode consultatif de la gestion des
parties prenantes, Management & Avenir, 68, 73–91.
Shaukat, A., Qiu, Y., & Trojanowski, G. (2016). Board attributes, corporate social responsibility
strategy, and corporate environmental and social performance. Journal of Business Ethics,
135(3), 569–585.

207
Titre : Impact de la représentation des salariés au sein du conseil d’administration sur

l’engagement RSE de l’entreprise : Etude du contexte français

Mots clés : Administrateurs salariés, Reporting RSE, Performance ESG, Performance boursière

Résumé : Cette thèse propose d’examiner la relation sociale et modèrent négativement la perception des
entre la représentation des salariés dans le conseil investisseurs vis-à-vis de la performance ESG. Au
d’administration et l’engagement RSE de contraire, les administrateurs élus par les salariés
l’entreprise. En particulier, nous étudions comment actionnaires renforcent la performance ESG globale
les investisseurs perçoivent-ils le reporting RSE et la ainsi que les performances environnementale et de
performance ESG lorsque les salariés sont gouvernance et modèrent positivement la
représentés au sein du conseil d’administration. perception de la performance ESG par les
En utilisant un échantillon de sociétés françaises investisseurs. Nos résultats mettent en évidence un
appartenant à l’indice SBF 120 sur deux périodes conflit d’intérêts entre les actionnaires et les
différentes de 2001 à 2011 et de 2007 à 2017, nos administrateurs salariés, notamment avec les
résultats montrent que la présence des représentants des travailleurs.
administrateurs salariés modère positivement la Cette thèse couvre plusieurs disciplines à savoir la
perception du reporting RSE et négativement la comptabilité, la gouvernance d’entreprise et la
perception de la performance ESG par les gestion des ressources humaines. En conséquence,
investisseurs. Un examin plus détaillé montre que les elle intérèsse vivement les législateurs, les
administrateurs élus par les salariés par leur droit de investisseurs, les managers et les actionnaires.
travail améliorent exclusivement la performance

Title : The Impact of Employee Board Representation on the Firm’s CSR Engagement:

Evidence from the French Context

Keywords : Employee board representation, CSR reporting, ESG performance, Firm market value

Abstract : This thesis proposes to examine the exclusively the social performance and moderate
relationship between employee board representation negatively the way investors perceive information
and the CSR engagement of the firm. Particularly, on ESG performance. In contrast, directors elected
we study how investors perceive CSR reporting and by employee shareholders enhance the overall ESG
ESG performance when employees are represented performance as well as the environmental and the
on the board. governance performances and moderate positively
Using a sample of French firms belonging to the SBF the way investors perceive the ESG performance of
120 index over two different periods from 2001 to the firm. Our findings highlight a conflict of interest
2011 and from 2007 to 2017, our results show that between shareholders and employee directors,
the presence of employee directors on board particularly with labour representatives.
moderates positively the perception of investors This thesis covers several disciplines namely
towards CSR reporting and negatively their accounting, corporate governance and human
perception towards ESG performance. More fine- ressource management. As a result, it is with great
grained examination shows that directors elected by interest for regulators, investors, managers and
employees by their right of employment enhance shareholders.