The cost of CEO duality: Evidence from French leadership compensation

European Management Journal xxx (2017) 1e15

Contents lists available at ScienceDirect

European Management Journal journal homepage: www.elsevier.com/locate/emj

The cost of CEO duality: Evidence from French leadership compensation
raldine Broye a, *, Abel François b, Yves Moulin c Ge a

EM Strasbourg Business School, Strasbourg University (LaRGE), France Lille 1 University, Sciences and Technologies (LEM), France c Lorraine University (CEREFIGE), France b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 February 2016 Received in revised form 6 December 2016 Accepted 23 January 2017 Available online xxx

This paper aims to provide a detailed analysis of the relationship between board leadership structures and executive compensation. According to agency theory, the combined position of CEO and Chairperson of the Board (COB) entails greater compensation for the CEO in order to reduce conflicts of interest. In the literature, combined board structure is generally considered to generate additional costs for companies. However, the choice of two separate structures implies the payment of incentive compensation for the COB in addition to that defined for the CEO. This paper investigates the financial cost of duality when compensation packages are set for both leaders. Our results suggest that although combined board structure is associated with higher incentive compensation for the CEO, the overall compensation cost to the company is no higher when the chairperson's compensation is considered. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Corporate governance Executive compensation Duality Chairperson of the board France

1. Introduction Separating the positions of Chairperson of the Board (COB) and Chief Executive Officer (CEO) is seen as an essential key to ensuring good corporate governance. Agency theory argues that combining the functions of decision and control reduces the board's ability to effectively monitor the decisions and actions of the CEO, thus giving the latter more latitude to satisfy his own interests rather than those of shareholders (Fama & Jensen, 1983; Jensen, 1993). According to managerial power theory, CEOs holding the chairperson title may use their increased power for rent extraction, thus obtaining higher compensation at the expense of shareholders (Bebchuk & Fried, 2004). Alternatively, optimal contracting theory argues that the incentive compensation of a CEO holding the position of chairperson is likely to increase in order to mitigate the reduced effectiveness of direct monitoring (Jensen & Murphy, 1990; Murphy, 1999). Overall, although a combined chair position may have benefits such as improving the decision-making process and the ability to exchange information (Adams & Ferreira, 2007; Finkelstein & D'Aveni, 1994), this leadership structure may entail greater bonding costs for the firms, mainly due to the higher

* Corresponding author. E-mail address: [email protected] (G. Broye).

compensation that can be accorded to CEO-Chairpersons. The aim of this paper is to shed new light on the relationship between board structure and executive compensation by considering the role of the chairperson, and specifically his/her own compensation contract. Separate leadership requires a compensation contract for the chairperson, in addition to that defined for the CEO. Specifically, the separation of ownership and control in firms can generate conflicts of interest between the chairperson and the shareholders, which can result in incentive compensation packages being set for the chairperson. It is therefore crucial to consider compensation paid to the chairperson when analyzing the impact of the choice of board structure on executive compensation. According to Brickley, Coles, and Jarrell (1997), the decision to separate the titles should be analyzed in terms of costs and benefits. In this light, additional compensation awarded to the chairperson in a separate structure is likely to result in additional costs for firms. Our study argues that although the cost of CEO compensation may be lower in firms with separate titles, the cumulative cost of leadership in these firms is equivalent to the cost associated with CEO compensation in dual firms, if not higher. The main objective of this paper is to analyze how far the choice of board structure impacts the total compensation of leadership, whether single or joint. This study uses a sample of French public firms from 2005 to 2011 to address two main empirical questions. Firstly, how does

http://dx.doi.org/10.1016/j.emj.2017.01.007 0263-2373/© 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

2

G. Broye et al. / European Management Journal xxx (2017) 1e15

duality affect French CEO compensation? In line with previous empirical literature, we investigate whether the choice of a combined structure, and the consequent lack of effective monitoring by a chairperson, entails higher CEO compensation levels. We particularly seek to pinpoint how the choice of board structure affects the design of CEO compensation packages and how this choice shapes the relationship between CEO compensation and firm performance. As the CEOs of firms with separate titles might be more effectively monitored by the board, we examine whether incentive pay contracts are used as complements or alternatively as substitutes. Secondly, how does duality affect French leadership compensation? We consider the total compensation costs for CEO and COB leadership, and investigate whether a separate board structure is still an additional cost to the company when COB compensation is included. By addressing this issue, we identify the determinants of COB compensation. This study makes several new contributions to the literature. First, this paper sheds new light on the relationship between duality and executive compensation and on the predictions of optimal contracting and managerial power theories. Previous USbased empirical studies show diverging results on this issue and do not provide a definitive conclusion about the validity of these two theories (e.g. Capezio, Shields, & O'Donnell, 2011; Cordeiro & Veliyath, 2003; Cyert, Kang, & Kumar, 2002; Dey, Engel, & Liu, 2011). The recent meta-analysis from Van Essen, Otten, and Carberry (2015) suggests that managerial power does have a significant influence over the pay-setting process, but that optimal contracting arrangements may also exist. The present paper assesses the validity of these theories within the French context. Second, this paper is a valuable addition to recent cross-national governance research. Several studies have highlighted the “undercontextualized” nature of agency theory, and the necessity to explain the diversity of corporate governance arrangements across different institutional contexts (Aguilera, Filatotchev, Gospel, & Jackson, 2008; Aguilera & Jackson, 2003; Desender, Aguilera, Crespi, & Garcia-Cestona, 2013; Schiehll & Martins, 2016; Van Essen, Engelen, & Carney, 2013). These studies suggest that the effectiveness of governance prescriptions may be contingent on a variety of legal and institutional factors. Van Essen et al. (2013) specifically document that the positive effects of CEO duality depend on individual country factors such as the rule of law and investor rights (La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 2000). Besides, the governance mechanisms interact at both the firm-level and the country-level, and corporate governance can be viewed as the result of interrelated country- and firm-level factors (Aguilera et al., 2008; Schiehll & Martins, 2016). The present study builds on these arguments to examine complementarities and substitutions among firm-level governance mechanisms such as board composition and incentive compensation, within the French context. The majority of literature on CEO compensation and board structure to date has been conducted in the Anglo-American context. This environment is characterized by dispersed ownership where board composition and contractual incentives are key governance mechanisms (Aguilera & Jackson, 2003; Fernandes, Ferreira, Matos, & Murphy, 2013). This study provides the opportunity to examine an environment where listed firms typically have large shareholders and a high proportion of family-based ownership (Belot, Ginglinger, Slovina and Sushka, 2014). Since large shareholders are likely to exercise more direct monitoring of managers, a concentrated ownership environment should entail less necessity for incentive compensation and a reduced monitoring role of the board. CEO compensation in France is indeed significantly lower and less equity-based than in the US. Nevertheless, compensation levels in France have increased considerably in recent years, as well as the relative proportion of

incentive compensation (Goyer & Jung, 2011). Additionally, as large shareholders have both the power and incentives to discipline CEOs, they may encourage performance-based compensation. While ownership concentration is likely to limit agency conflicts between managers and shareholders, it is also a potential source of conflicts between large and small shareholders, and blockholders can use their discretion to extract private benefit (Schleifer and Vishny, 1997). In a context of weak minority shareholder protection, as observed in the French context (La Porta et al., 2000), the board is likely to play an important monitoring role to protect minority shareholders. In this case the choice of board structure remains a key issue. The French institutional setting provides an interesting framework for the study of duality, as companies can freely choose to unify or split CEO and COB positions within a unitary board (as in the US model) or a two-tier board (as in the German model). This leadership structure model provides a stimulating context in which the relationship between CEO duality and compensation can be examined. Lastly, this study provides greater insight into the ongoing debate surrounding the costs and benefits of implementing good corporate governance mechanisms in different countries (Aguilera et al., 2008; Doidge, Karolyi, & Stulz, 2007), and specifically the costs and benefits of separating roles (Brickley et al., 1997). The paper adds to the literature by providing a close examination of how CEO duality impacts leadership compensation within a novel framework that addresses COB compensation. This contribution seems significant given the scarcity of literature on the role - and more particularly the compensation - of the COB (Roberts, 2002). Nevertheless, the leadership provided by the joint CEO e COB positions is a key issue for companies that have chosen a separate board structure. As the leader of the board, the chairperson plays a pivotal role in the company (Karabadse & Karabadse, 2007; Roberts, 2002; Waelchli & Zeller, 2013). A COB can make an effective and positive contribution to the strategic direction and corporate success of companies. He/she has a central role in determining the effectiveness of a board and its contribution to the quality of executive decision making. The COB receives compensation to hold this key position. Our study provides insights into the financial cost of separating titles when COB compensation is taken into account. 2. Board structure and leadership compensation: The French context 2.1. Board structure in France French companies have the opportunity to choose freely between three board structures. The first type, introduced by French law in 1966, is a two-tier board structure. This two-tier system consists of a supervisory board of non-executive directors and a separate management board of executive managers. The CEO runs the management board and the COB chairs the supervisory board; their respective roles and responsibilities are clearly separated. The management board conducts the day-to-day management and corporate activities of the firm, while the supervisory board supervises the management board and the monitoring of managerial activities. The chairperson sets the agenda for board meetings, leads discussions among non-executive directors and coordinates their activities. He/she plays a key role in ensuring that managerial decisions are in line with the overall strategic guidelines of the company, and in encouraging effective corporate governance practices. The second type of structure is a traditional single-tier board structure, where the board of directors is typically chaired by one

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

G. Broye et al. / European Management Journal xxx (2017) 1e15

person holding the titles of both COB and CEO. Within this structure, the CEO has a combined operational and supervisory role, with full power to act in the firm's name. While chairing the board, the CEO is expected to take on the role and responsibilities of a nonexecutive chairperson. A third type of structure was introduced by French law in 2001, and allows the separation of the positions of COB and CEO within a unitary board of directors. In this case the CEO sits on the board but this “dissociated board” is chaired by a separate, non-executive COB. This structure makes it possible to separate the functions of decision and control within a single board in the same way as in the US context, and the non-executive COB is expected to play an identical role to that of the chairperson in a two-tier structure. This analysis defines dual firms as firms in which CEO and COB positions are combined. Conversely, the term ‘separate structure’ is used for any case where these positions are split, within unitary or two-tier board structures. The term “leadership” refers to the CEO in combined structures and the CEO and the COB in separate structures. Changing between a single-tier and two-tier board requires the involvement of both the board and shareholders (Belot, Ginglinger, Slovin, & Sushka, 2014). The board has the initial responsibility of proposing a change in board structure. It submits a charter amendment proposal to an Extraordinary General Meeting of shareholders. Conversely, a unitary board can freely decide to separate or unify CEO and COB positions without the approval of the shareholders. Since the enactment of the 2001 act, “dissociated boards” have multiplied in France, enhancing a trend to separate the positions of chairperson and CEO. Nevertheless, combined structures remain the most common type, and are used by 57% of French listed companies (AMF, 2015). This percentage is very close to that found recently in the United States by Yang and Zhao (2014), who provide evidence of a significant trend in recent years towards separating the positions of CEO and COB. They report that over 80% of large U.S. companies chose to combine the positions in the early 1990s, and that this figure fell to 54% in 2010. The choice of structure in France is different to that observed in other European countries, where a majority of companies separate CEO and COB positions. This is particularly true in the United Kingdom, Germany, Austria, Denmark and Finland. French companies are amongst those with the highest proportion of combined structure boards in Europe (Heidrick & Struggles, 2011 ). 2.2. CEO and COB compensation in French firms Executive compensation in French firms usually consists of a fixed salary, bonuses and equity-based compensation. According to the Expert Corporate Governance Service survey (ECGS, 2012), 37% of total CEO compensation is equity based, whilst bonuses make up 35% of the sum. France is ranked fifth in Europe in terms of average compensation, and total executive compensation is significantly lower than that of counterparts in the UK, Germany or Switzerland. In the UK, equity-based compensation alone represents 66% of total compensation. In France, the board is responsible for setting CEO compensation packages1. The board defines salary and bonuses. Although the initial approval of the shareholders is a prerequisite to the issuance of any stock option plan or performance shares plan, the board sets the conditions for the stock option allocation plan (i.e. the

1 France recently introduced the principle of advisory say on pay votes by shareholders. While say on pay is not a legal requirement, it has been included in 2013 in the French code of governance AFEP-MEDEF as a soft law (AFEP/MEDEF, 2013).

3

beneficiaries, number of shares allocated and exercise price). In dual firms, the CEO is awarded a global compensation package without explicitly separating the compensation received for his/her position as chairperson. The French Code of Governance (Code AFEP/MEDEF, 2013) provides several guidelines on compensation policy, and underlines the necessity to publish clear and transparent rules in annual reports with regard to any performance criteria used to define executive bonuses and the granting of options. However, the relative lack of academic research on CEO compensation in France means that the relationship between pay and firm performance remains unclear (Broye & Moulin, 2010). The decision to have a separate board structure entails the provision of a compensation contract for a non-executive chairperson, in addition to that defined for the CEO. Like for CEOs, it is the role of the board to set the level and the structure of COB compensation (Article L.225-47 of the Commercial Code). Part of the compensation is paid in the form of meeting fees. The annual general meeting of shareholders determines an annual amount for meeting fees, to be shared amongst all directors. The board is responsible for deciding how to allocate this sum, and how much of it will be awarded to the COB. Apart from these fees, chairpersons are likely to receive compensation from other sources. The board of directors can decide to grant chairpersons stock options, and award them bonuses for special assignments or duties. Although these payments should normally be enacted by regulated agreement, they are, in practice, rarely under shareholder control. The market authority therefore considers the determination of the COB's compensation to be too opaque, and specifies that the chairperson should not receive a bonus or equity-based compensation unless it is justified by specific circumstances (AMF, 2014, 2015). Overall, the compensation of non-executive French chairpersons is particularly high in comparison to that of other European counterparts (Heidrick & Struggles, 2011). 3. Theoretical framework and hypothesis development The objective of this paper is to examine the costs entailed by setting a compensation contract for the chairperson, then relate this analysis to the costs incurred when a firm separates the positions of CEO and COB. 3.1. CEO compensation and duality Two theories address the relationship between the leadership structure of a firm and the design of CEO compensation contracts, namely optimal contracting theory and managerial power theory. Optimal contracting theory suggests that compensation contracts are set up to solve agency problems and to provide managers with an incentive to maximize shareholder value (Core, Guay, & Larcker, 2003; Jensen & Murphy, 1990; Murphy, 1999). Under this theory, when CEO and COB positions are combined, boards should provide CEOs with efficient incentives to compensate for the reduced effectiveness of direct monitoring. Consequently, CEOs are expected to receive a higher proportion of performance-based compensation, and a compensation package that is closely tied to firm performance (Dey et al., 2011). This emphasis on contingent compensation results in increased compensation risk, and riskaverse CEOs are likely to require higher levels of fixed salary and total compensation. Accordingly, optimal contracting theory predicts greater CEO compensation in dual firms, where levels of both fixed and incentive compensation are higher and the incentive compensation is more strongly related to performance. Alternatively, managerial power theory argues that CEOs holding the title of COB are likely to use their position within the board to exert pressure on directors and achieve their own objectives

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

4

G. Broye et al. / European Management Journal xxx (2017) 1e15

(Bebchuk & Fried, 2003, 2004; Weisbach, 2007). This theory predicts that duality will lead to greater structural power through the authority of the CEO's preeminent formal position (Finkelstein, 1992). The term ‘structural power’ refers to the influence of CEOs over the board. CEOs who also hold the COB position have increased power over the director selection process, and can influence the nomination of friendly board members (Hermalin & Weisbach, 1998; Westphal & Zajac, 1995). Their responsibility for organizing boards meetings and setting the agendas also gives them the opportunity to control the information provided to the board, hence weakening the board's ability to monitor (Adams & Ferreira, 2007; Bebchuk & Fried, 2004; Tuggle, Sirmon, Reutzel, & Bierman, 2010). As CEOs combining both roles have more power over board members, they are able to negotiate for compensation arrangements that serve their own interests, including a generous compensation package (Bebchuk & Fried, 2004; Boyd, 1994; Grabke-Rundell & Gomez-Mejia, 2002). In dual firms, powerful CEOs are therefore likely to request high levels of pay. Conversely, the chairperson and the board in firms with separate structures have more freedom to exercise a more effective monitoring role, and can limit the level of executive compensation. CEO compensation is therefore expected to be higher in combined structures. Managerial power theory also suggests that the design of CEO compensation contracts reflects managerial preference rather than optimal incentive contract (Van Essen et al., 2015). Risk-averse CEOs are likely to prefer a higher salary to riskier, performance-based compensation (Mehran, 1995). Accordingly, powerful CEOs are able to dissociate their compensation from firm performance, making the relationship between compensation and performance weak or non-existent. Van Essen et al. (2015) argue that optimal contracting and managerial power theories are not competing perspectives, and should rather be viewed as complementary theories to explain compensation arrangement. Overall, both theories predict higher levels of salary and total compensation in dual firms, and this implies an increased cost for firms that wish to combine the two positions. However, the two theories differ on the expected level of incentive compensation within the package and the expected relationship between compensation and performance. Empirical studies (largely conducted in the Anglo-American context) have reached diverging conclusions on these alternative theoretical expectations. Core, Holthausen, and Larcker (1999) and Cyert et al. (2002) find that combining the positions positively impacts CEO compensation, and suggest that CEOs use their power and entrenchment within the board to extract rents at the expense of shareholders. Other studies find no relationship between duality and CEO compensation (Brickley et al., 1997; Capezio et al., 2011; Conyon & Peck, 1998; Cordeiro & Veliyath, 2003). Fernandes et al. (2013) describe a positive impact of duality on compensation level for US firms, but a negative effect for non-US firms. Dey et al. (2011) show that executive pay-performance sensitivity was significantly lower when the company opted for a separation of positions, and significantly higher after a switch to a combined structure. Their findings support the optimal contracting hypothesis. Conversely, Firth, Fung, and Rui (2007) find that the relationship between pay and performance is weaker for Chinese firms that have chosen a combined structure. Van Essen et al. (2015) provided a recent meta-analysis of 219 US studies and found that duality is positively related to the level of total compensation, but has no significant effect on the performance-pay relationship. The present study allows analyzing how duality influences CEO compensation arrangements in a context characterized by concentrated ownership and weaker managerial incentives. The presence of controlling shareholders reduces the need to rely on the board for monitoring management, as large shareholders have

both the incentives and abilities to actively monitor CEOs (Desender et al., 2013). The greater capacity of large shareholders to exercise direct control on CEOs may also make boards feel less obliged to complete their monitoring role through more incentive compensation (Aguilera & Jackson, 2003). Fernandes et al. (2013) suggest that higher institutional ownership and more independent boards in US firms entail higher CEO compensation and increased use of equity-based pay than in non-US firms. The links between board structure and compensation in the French context may therefore be less evident than those expected in the AngloAmerican context. However, high concentration ownership and weak legal protection for minority shareholders (La Porta et al., 2000) may result in differing interests between blockholders and minority shareholders, leading to agency conflicts between large and small investors (Shleifer & Vishny, 1997). In this context, the board is expected to strengthen its monitoring role to prevent the expropriation of small investors. Specifically, separate board structures are expected to play a role in minimizing agency problems between large and small shareholders (Fraile & Fradejas, 2014; Goyer & Jung, 2011), and directors may be willing to set optimal incentive compensation for managers to protect shareholder value. The context of ownership concentration may also have consequences on duality effects under the managerial power perspective. Desender et al. (2013) argue that duality obstructs the board's monitoring role when its members represent dispersed shareholders. Conversely, the presence of controlling shareholders on the board may overcome the dominance of powerful CEOs holding the COB position. The representation of active large shareholders on the board can also encourage the setting of optimal compensation contracts to motivate CEOs. In a context of concentrated ownership, there may therefore be a decrease in CEO/COB power over the board. This study tests the following hypotheses about the impact of duality on the level and design of French CEO compensation: Hypothesis 1. CEO total compensation is higher in dual firms than in firms with separate board structures. Hypothesis 2a. (optimal contract). Dual firms set a CEO compensation package made up of a higher fixed salary, a higher level of incentive compensation, and incentive compensation that is more strongly linked to firm performance. Hypothesis 2b. (managerial power). Dual firms set a CEO compensation package with a higher fixed salary, lower incentive compensation, and incentive compensation that is unrelated to firm performance. 3.2. Leadership compensation and duality Several publications suggest that the choice of leadership structure requires a trade-off between expected costs and benefits (Brickley et al., 1997; Dey et al., 2011; Linck, Netter, & Yang, 2008). The main expected benefit of splitting the roles lies in reducing agency costs and therefore increasing the firm performance through more effective monitoring (Baliga, Moyer, & Rao, 1996; Krause & Semadeni, 2013). Adams and Ferreira (2007) claim that separated structures allow the cleanest separation of the board's advisory and monitoring roles. Duality may also have strategic leadership benefits, such as improving the decision-making process, creating unity in the command structure of a firm (Finkelstein & D'Aveni, 1994), and fostering information exchange (Adams & Ferreira, 2007). The literature also pinpoints some of the potential costs of separating the positions, such as the costly and potentially

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

G. Broye et al. / European Management Journal xxx (2017) 1e15

incomplete transfer of critical information between the CEO and the COB (Brickley et al., 1997), the potential for rivalry between the CEO and the COB (Anderson & Anthony, 1986), and costs related to confusion about authority (Finkelstein & D'Aveni, 1994). Brickley et al. (1997) also mention the agency costs incurred by the separation of positions. Although the appointment of a chairperson may reduce number of agency conflicts through increased monitoring of CEO behavior, it creates other agency costs inherent to the monitoring of the COB. Indeed, chairpersons are likely to take advantage of their position on the board to favor decisions that will satisfy their own interests at the expense of shareholders. Brickley et al. (1997) mention that curiously, the discussion about separate leadership has completely ignored the critical issue of the non-executive chairperson's incentives. However, the literature has highlighted the value of incentive compensation granted to directors in order to reduce conflicts of interest arising from separation of ownership and control and consequently increase the performance of the company (Cordeiro, Veliyath, & Romal, 2007; Deutsch, 2007; Yermack, 2004). In this line, we expect firms with separate titles to set a specific compensation contract for the chairperson who runs the board of directors, leading to additional costs. We propose to further analyze the cost of separating these two positions by taking the compensation of the COB into account. Section 3.1 documents how the optimal contracting and managerial power arguments both predict higher total compensation for CEOs in combined board structures. This increase in compensation is considered to incur additional cost for firms with combined titles. However, the examination of CEO compensation without factoring the chairperson compensation cannot provide a reliable comparison of how much each board structure costs in terms of compensation. It seems more appropriate to compare the compensation of CEOs in dual firms with the cumulative compensation of both the CEO and the COB in firms with separate positions. It is possible that paying compensation to the chairperson increases the cost to firms with separate positions, and that the compensation of a CEO holding the COB title may be equivalent to the compensation of two leaders who share the same role and responsibility in a firm with separate structure. When examining CEO compensation in dual firms and in firms with separate structures, it is important to bear in mind that we are comparing the compensation of CEOs playing different roles; a CEO who is also the chairperson can be expected to have a more complex job than a counterpart in a firm that separates the two positions, and may thus require higher compensation for holding both positions (Core et al., 1999). This alternative argument may also explain higher CEO compensation levels in dual firms. Examining the aggregate compensation for the CEO and the COB in firms with separate structures makes it possible to compare the compensation awarded for the same job, whether it is carried out by one or two heads. We propose to test the hypothesis that leadership compensation remains similar whether the leadership is single or joint, and that overall, the choice to combine these positions does not incur higher costs in terms of compensation. Hypothesis 3. The compensation of the CEO combining both positions in dual firms is equivalent to the compensation of the CEO added to that of the COB in firms separating these positions. Addressing leadership compensation permits further analysis of the optimal contracting and managerial power arguments. If CEO compensation in dual firms is similar to, or even higher than the leadership compensation in firms with separate structures, the result would be consistent with a greater rent extraction or

5

an optimal incentive contract for the CEO holding the chairperson title. A more thorough analysis is obtained through the examination of how duality impacts the leadership compensation structure. In the case of combined positions, the CEO is likely to assume full responsibility for the firm’s performance, and compensation should therefore include incentive compensation and should be closely linked to performance. Conversely, in the case of separate positions, the CEO and the COB share this responsibility. The role of the chairperson is potentially critical to the performance of the company, and high interaction is necessary between the two leaders to ensure corporate success (Florou, 2005; Waelchli and Zeller, 2013). This sharing of responsibility and the implementation of incentive compensation for the chairperson probably entail a lower incentive for a CEO in a separate structure than for a CEO who also holds the title of chairperson. Yet the compensation of leadership and the compensation of one CEO combining COB title should both be associated with overall firm performance. Alternatively, if the optimal contracting view dominates, CEOs in dual firms should have higher incentives and their compensation should be more strongly dependent on performance compared to the compensation packages for two separate leaders in firms with separate structures. If the managerial power view dominates, CEOs in dual firms should receive less incentive compensation and their remuneration should be less related to performance than that of two separate leaders in firms with separate structures. This study tests the link between firm performance and compensation of the CEO and the chairperson team, versus the link between firm performance and compensation of a CEO who combines both positions. Hypothesis 4a. (neutral). The compensation packages for the CEO and the COB in firms with separate structures are equivalent to the compensation package of the CEO holding the COB position in dual firms. Hypothesis 4b. (optimal contract). The compensation includes more incentive compensation and has a stronger link to firm performance for the CEO in dual firms than for both the CEO and the COB in firms with separate structures. Hypothesis 4c. (managerial power). The compensation includes less incentive compensation and is less connected to firm performance for the CEO in dual firms than for both the CEO and the COB in firms with separate structures. 3.3. Chairperson compensation The skills and behavior of the chairperson have a direct impact on the effectiveness of the board (Leblanc, 2005; Roberts, 2002). Given that the chairperson can play a significant role in the quality of the decision-making process and the strategic direction of the firm, he/she can contribute to its successful financial performance. However, agency problems are likely to emerge from a divergence of interest between the chairperson and shareholders (Brickley et al., 1997). If we assume that the COB is a self-serving agent who does not necessarily act in the interest of shareholders, then the question arises of which factors determine his/her compensation, and more specifically whether this compensation is linked to performance. Previous studies have investigated the determinants of board director compensation (e.g. Boyd, 1996; Cordeiro et al., 2007; Ryan & Wiggins, 2004). This literature provides evidence that outside director compensation is one of the principle incentive mechanisms motivating directors to behave in the interest of shareholders (Yermack, 2004). In the US, Cordeiro et al. (2007) mention that director compensation has considerably shifted towards a greater

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

6

G. Broye et al. / European Management Journal xxx (2017) 1e15

use of stock-based incentive compensation since the mid-1990s, as incentive pay is expected to be an effective means of aligning the interests of shareholders and directors. Empirical evidence shows a relationship between incentive compensation to directors and firm performance (Cordeiro et al., 2007; Deutsch, 2007). In France, outside directors cannot be awarded stock options or restricted shares. Although this equity-based compensation can be used as a reward mechanism to remunerate chairpersons of the board, it is a rare occurrence in French firms. We nevertheless expect the COB compensation contract to be closely tied to firm performance, thus providing COBs with stronger incentives to monitor managerial discretion and contribute to corporate performance. Based on these arguments, we hypothesize as follows: Hypothesis 5. In firms with separate structures, the compensation of the COB is positively related to firm performance. 4. Methods The cost of separating positions is assessed using a sample of French companies listed on the SBF 120 (Soci
et
e des Bourses Françaises 120 index). This index lists the 120 largest equities listed in France. In this section the sample is described, then the variables are presented. 4.1. Sample To construct the sample, we start with firms that have been present in the SBF 120 Index for at least one year during the 2005e2011 period. Financial companies are excluded due to the specificity of their accounts. Observations with missing data are also excluded, as well as any companies that changed their CEO and/or chairperson during the year. Information relating to the compensation of CEOs and COBs was collected manually from annual reports. Financial and market data were gathered from Amadeus and Eurofidai databases. The final sample contains 577 observations (corresponding to 113 observed firms). 4.2. Dependent variables CEO total compensation includes salary, bonus and equity-based compensation. The latter includes the value of any stock options and shares granted during the year, which is estimated using the Black and Scholes formula. We also examine CEO fixed salary and CEO incentive compensation (i.e. the sum of bonus and equity-based compensation) as a proxy for performance-based compensation. CEO incentive compensation is also calculated in relative terms to capture the proportion of compensation that is contingent on performance. CEO relative incentive is the ratio of incentive compensation to total compensation. Leadership total compensation corresponds to the compensation of leadership within the firm. This represents the CEO's total compensation in firms with combined structures, given that the CEO receives one global compensation for his/her position as CEO and COB. In firms with separate structures, leadership compensation is the sum of CEO and COB total compensations. We also consider Leadership fixed salary, Leadership incentive compensation, and Leadership relative incentive. In firms with separate structures, COB total compensation is composed of fixed meeting fees and incentive compensation, which includes bonuses and equity-based compensation. The logarithm form is used to analyze compensation variables in the regressions, thus reducing the influence of observations on tail and size distribution.

4.3. Independent variables For the assessment of how duality affects CEO and Leadership compensation, we used the Dual variable which equals 1 if the two positions are combined (duality), and zero when the two positions are separated. In line with agency theory, we predict a close and positive link between CEO compensation and firm performance that should help align the interests of shareholders with those of the CEOs. Firm performance is estimated through two financial measures commonly used in compensation literature (Tosi, Werner, Katz, & Gomez-Mejia, 2000), i.e. an accounting measure of performance, Return on assets (ROA), and a market-based performance measure, Stock return. Both variables are lagged to measure prior performance. We also use two interactive variables, Dual x ROA and Dual x Stock return, to study the impact of duality on the link between pay and performance. We control for a number of other variables that could impact CEO and COB compensations. The literature shows that size is a crucial factor when defining the levels of executive compensation (Jensen & Murphy, 1990; Tosi et al., 2000). Running a larger firm involves a higher level of responsibility with more hierarchical levels to coach, demanding greater experience and skills. A larger company should also require chairpersons to accomplish more extensive and complex work. We therefore expect larger firms to offer higher levels of compensation for CEOs and COBs. Firm size is measured using Log assets, the logarithm of total assets. As CEOs are responsible for developing growth opportunities, they should be rewarded when these opportunities are high (Baber, Janakiraman, & Kang, 1996). These growth opportunities also represent a higher risk factor for CEOs, and this should be reflected by a higher level of compensation. Similarly, the COBs of firms with increased growth opportunities are expected to deal with many diverse issues that demand a more complicated monitoring and evaluation of CEO performances, thus adding to the COB's workload. These growth opportunities are measured through Market to book, i.e. the ratio comparing the company's market capitalization to the book value of its equity, and we expect to see a positive association with CEO and COB compensation level. As ownership structure plays a key role in compensation practices (Croci, Gonenc, & Ozkan, 2012), we control for various ownership proxies. French ownership is characterized by concentrated ownership, with a high proportion of family-based firms. This means that large shareholders, and specifically the large ownership of CEOs or family, are likely to significantly influence compensation policies. Institutional ownership, although much less developed than in Anglo-Saxon businesses, has recently increased in French listed firms and may also be influential (Goyer & Jung, 2011). Previous literature argues that large shareholders, and particularly family or institutional shareholders, are likely to play a more active role in monitoring the management (Jensen & Meckling, 1976), hence limiting agency conflicts not only between shareholders and CEOs but also between shareholders and chairpersons. As these investors have both the power and incentives to discipline CEOs, they are expected to encourage performancebased compensation and ensure that pay levels are not excessive (Croci et al., 2012; Fernandes et al., 2013; Van Essen et al., 2015). When this active monitoring is also applied to COBs, it should encourage incentive compensation for the chairperson and also limit the level of his/her compensation. Large, family and institutional ownerships are therefore expected to be associated with lower total compensation and higher incentive compensation for the CEO and COB. Besides, as large CEO ownership is likely to mitigate agency conflicts, it is expected to limit excessive compensation levels and reduce the need for incentive compensation packages (Core et al., 1999). We predict a negative

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

G. Broye et al. / European Management Journal xxx (2017) 1e15

association between CEO ownership and total and incentive compensation for the CEO and COB. The following variables are used to control for the impact of ownership structure on the compensation of CEOs and COBs: Largest shareholder, which is the percentage of shares held by the largest shareholder, Institutional main shareholder, which is equal to 1 if the main owner is an institutional investor, CEO main shareholder, which is equal to 1 if the main owner is the CEO2, and Family firm, which is equal to 1 if the family ultimately holds more than 20% of the shares (following Faccio & Lang, 2002). We also control for additional board characteristics. The disciplinary role played by the board is likely to be more effective when it includes a higher proportion of independent outside directors (Jensen, 1993). Independent directors are expected to be more resistant to management pressure and also ensure that CEOs do not expropriate wealth from shareholders in the form of excess pay. A higher proportion of independent directors is therefore expected to lower CEO compensation and favor incentive compensation linked to performance (Chhaochharia & Grinstein, 2009; Croci et al., 2012). We expect to note a negative (positive) association with total (incentive) compensation for the CEO. Inversely, we can expect independent directors to favor a higher level of pay for the COB as an incentive to monitor the management more effectively (Ryan & Wiggins, 2004). A higher proportion of independent directors is expected to be positively associated with the level of COB compensation. The independent directors variable is the percentage of independent directors on the board. We also control for Board size. Larger boards are likely to be associated with less effective monitoring, as they entail coordination and communication problems, as well as difficulties in decision making (Jensen, 1993). Smaller boards are expected to supervise compensation practices more efficiently, meaning lower total compensation and higher incentive compensation. We control for CEO characteristics. CEO age and CEO tenure variables reflect accumulated experience, the acquisition of expertise in this role, and specific knowledge of the business and its environment, all of which should be compensated by higher pay. Alternatively, these variables may reflect CEO power, in which case the CEO is likely to exert more pressure on the board and influence compensation policy for his own advantage (Hill & Phan, 1991). Moreover, we can expect a negative relationship between CEO tenure or age and the level of COB compensation. Indeed, an entrenched CEO can use his power to limit the effective control of the board by reducing the incentive compensation of directors (Ryan & Wiggins, 2004), and thereby, that of the chairperson. To explain COB compensation, we additionally control for COB tenure and COB age. It is expected that the longer tenure and age of individuals indicates greater experience, and thus a higher level of compensation for the chairperson. Having held a previous position as CEO of the firm or being the founder of the company also gives the COB significant experience and a thorough knowledge of the company. This managerial experience should be associated with a higher level of compensation (Broye & Moulin, 2014). We use the dummy variable COB previous CEO, which is equal to 1 if the COB was previously the CEO of the company, and COB founder, which is equal to 1 if the COB is the founder. We also control for COB ownership where a large ownership is likely to limit the need for incentive compensation. The dummy variable COB main shareholder, which is equal to 1 if the main owner is the COB, is expected to be negatively associated with total and incentive compensation

2 In France, CEO main shareholders are often the founders of the company. We alternatively used in our models a dummy variable measuring whether the CEO is founder, and we obtained similar results.

7

for the COB. We also control for COB in committee, which is equal to 1 if the COB sits on the compensation committee3. We suggest that the COB's involvement in this committee is likely to provide him more power to negotiate a higher compensation. Finally, year and industry fixed effects are controlled in each regression. We use Industry dummy variables according to ICB classification4. 5. Results 5.1. Descriptive statistics Table 1 presents descriptive statistics for our sample. For the full sample (577 observations), the mean CEO compensation is V2,295,000 and total compensation varies strongly, ranging from V109,000 to V25,449,000. Equity-based compensation is the largest element of this compensation, with an average of V918,000 (40% of total compensation). While this component of compensation can reach up to V18,200,000, firms did not grant equity-based compensation to the CEO in 45% of the observations. This result indicates that French CEOs receive a much lower fraction of their compensation in the form of stock options than US CEOs (Fernandes et al., 2013). Fixed salary and bonus represent an average compensation of V702,000 and V683,000, respectively. In firms with separate positions (233 observations), COBs receive an annual average compensation of V370,000. This compensation also varies widely, as it ranges from V0 to V3,587,000, with a median of V196,000. On average, this compensation is made up of 9% fees, 85% bonus and 6% stock options and shares. Bonuses are therefore the main component of total COB compensation, while the awarding of stock options to chairpersons remains a relatively exceptional event in French companies. Only 11 of our observations illustrate a firm offering equity-based compensation to their COB. Regarding board leadership structures, the two positions had been split in 40% of cases. The French context is therefore similar to the US context, in which 46% of US companies adopted separate board structures (Yang & Zhao, 2014). The average proportion of independent directors is 46.5%. The largest shareholder holds an average 33.4% of shares, which illustrates the high ownership concentration in French firms. Moreover, 44% of our observations are family firms. Table 1 also provides firm and CEO characteristics for dual firms and firms with separate structures. Firm size is revealed to be significantly higher in dual firms, whereas ROA is significantly lower. Dual firms also have a significantly smaller proportion of independent directors on their board, and have larger boards. This result suggests that board structure and board independence are complementary governance mechanisms used to enhance board effectiveness, rather than substitutes designed to constrain agency conflicts (Faleye, 2007). This is also consistent with the managerial power argument, in which CEOs holding the chairperson title use their power to favor internal directors. Our results also show that dual firm CEOs are older, with longer tenure and more ownership. This suggests that the greater experience of older and long-tenure CEOs makes them more likely to hold the position as chairperson. It may also indicate that CEOs are more entrenched and use their

3 We did not control for the CEO sitting on the compensation committee, as this situation is very rare in France, and has been illegal since 2008. In our sample, there are only 15 observations of the CEO sitting on the compensation committee, and the variable is not significant. 4 The Industry Benchmark Categorization (ICB) distinguishes ten sectors: “oil and gas”, “basic materials”, “industrials”, “consumer goods”, “health care”, “consumer services”, “telecommunications”, “utilities”, “financials” and “technology”.

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

8

G. Broye et al. / European Management Journal xxx (2017) 1e15

Table 1 Descriptive statistics. Dual firms (N ¼ 344)

Full sample (N ¼ 577)

Firms with separate structures (N ¼ 233)

Mean

Median

Std. Dev.

Mean

Median

Std. Dev.

Mean

Median

Std. Dev.

CEO salary CEO bonus CEO equity-based compensation

702.0 683.3 918.1

650.0 560.9 180.6

386.7 628.1 1779.4

705.5 666.9 971.4

647.5 524.3 171.7

406.8 636.1 1975.8

696.9 707.6 839.5

670.0 615.3 220.9

355.8 616.6 1442.1

CEO compensation (1000V)

2295.0

1689.3

2420.1

2334.8

1592.9

2676.6

2236.3

1800.0

1985.8

COB fees COB bonus COB equity-based compensation

34.6 312.7 22.9

24.5 154.8 0

43.9 501.2 119.5

COB compensation (1000V)

370.2

196.5

541.3

2606.5

1998.8

2308.8

8.43 18.35 13,37 2.12 31.8 0.45 0.27 0.1 49.7 11.05 5.5 54.4 4.53 64.4 0.15 0.46 0.26 0.39

6.45 14.83 4,53 1.83 29.3 0 0 0 46.1 11 6 56 3 65 0 0 0 0

8.21 47.33 20,28 1.82 22.6 0.50 0.44 0.3 20 3.16 5.6 6.6 3.95 8.6 0.36 0.5 0.44 0.49

Leadership compensation (1000V) Dual ROA (%) Stock return (%) Total assets (MV) Market to book (%) Largest shareholder (%) Family firm Instit. main shareholder CEO main shareholder Independent directors (%) Board size CEO tenure CEO age COB tenure COB age COB founder COB previous CEO COB main shareholder COB in committee

2444.5

1795.0

2535.9

0.60 7.04 19.80 16,86 2.12 33.4 0.44 0.26 0.22 46.5 11.41 8.5 56.1

1 6.45 16.03 4,65 1.81 28.1 0 0 0 45.5 11 6 56

0.49 8.1 50.6 31,08 1.72 23.4 0.50 0.44 0.42 20.3 3.80 8 6.6

6.09*** 20.78 19,26** 2.12 34.6 0.44 0.26 0.31*** 44.3*** 11.66* 10.6*** 57.2***

6.2 18.41 4,68 1.79 28 0 0 0 44.3 11 8 57

7.9 52.70 36,47 1.66 23.9 0.50 0.44 0.46 20.0 4.16 8.8 6.4

Dual ¼ 1 if CEO and COB positions are combined, 0 otherwise; ROA ¼ lagged return on assets, i.e. ratio of operating profit to total assets; Stock return ¼ lagged stock return; Total assets ¼ firm total assets; Market to book ¼ Ratio of market equity value to book value; Largest shareholder ¼ Percentage of shares held by the largest shareholder; Family firm ¼ 1 if a family holds more than 20% of the shares, 0 otherwise; Instit. main shareholder ¼ 1 if an institutional investor is the main shareholder, 0 otherwise; CEO main shareholder ¼ 1 if the CEO is the main shareholder, 0 otherwise; Independent directors ¼ percentage of outside independent directors; Board size ¼ number of directors in the board; COB founder ¼ 1 if the COB is the founder of the company, 0 otherwise; COB previous CEO ¼ 1 if the COB was previously the CEO of the firm, 0 otherwise; COB main shareholder ¼ 1 if the COB is the main shareholder, 0 otherwise; COB in committee ¼ 1 if the COB is in the compensation committee, 0 otherwise. Means comparisons between dual firms and firms with separate structures (t-test). ***, ** and * mean respectively t-test is significant at 1%, 5% and 10%.

power to favor combined structures and influence the choice of internal directors. Institutional and largest shareholder ownership levels are similar for both types of firm.

5.2. The impact of duality on CEO and leadership compensation The results of the estimations based on OLS regressions using pooled panel are reported in Table 2. We first analyze the impact of the leadership structure on CEO compensation (Table 2, Panel A), then examine its effect on leadership compensation (Table 2, Panel B). Dummy variables are included in each model to control for year and industry fixed effects (the coefficients are not tabulated for convenient reason). For each regression model, we examined variance inflation factors (VIF) to check for multicollinearity. Maximum VIF values were far below 10, the conventional threshold for excessive multicollinearity. The potential endogeneity problem of the dual variable was addressed by performing a DurbineWueHausman test. In a first step, a logit model was run to determine the board structure and obtain a Pearson residual value5. In a second step, this residual

5

We used the following variables to determine the Dual variable: Log assets, Market to book, Log largest shareholder, Institutional main shareholder, Independent directors, Board size, Family firm, CEO tenure, CEO age, CEO main shareholder, Year dummies and Industry dummies. The pseudo-R2 of the estimation is 0.22.

variable was incorporated into the original regression models. For each model, the coefficients of the residual were not significant at 5%, meaning that endogeneity is not a major concern in this empirical work6. Additional robustness checks were conducted. The suppression of fixed effects related to either the industries or the years did not significantly affect the estimation outcomes. In the same vein, the successive exclusions of each firm or each year from the sample led to very similar results. Lastly, we checked that the distributions of the residuals did not denote any latent problem. These robustness tests underline the quality of our empirical results. In Panel A, regression 1a, total CEO compensation is significantly higher in combined structures than in separate structures. This result provides strong support for Hypothesis 1, and is consistent with the findings of Van Essen et al. (2015). In line with both optimal contracting and managerial power theories, this result suggests that the choice of a combined structure entails an additional cost to firms through the provision of higher CEO compensation. More specifically, we find that the higher compensation

6 The coefficients (standard errors) are respectively 0.04 (0.06), 0.02 (0.06), 0.04 (0.03), 0.04 (0.03), 0.12 (0.19), 0.07 (0.18), 0.11 (0.27) and 0.07 (0.26) for regressions 1a, 1b, 2a, 2b, 3a, 3b, 4a and 4b in Panel A; and 0.04 (0.05), 0.02 (0.05), 0.03 (0.03), 0.03 (0.03), 0.12 (0.21), 0.07 (0.20), 0.12 (0.27) and 0.08 (0.26) for regressions 5a, 5b, 6a, 6b, 7a, 7b, 8a and 8b in Panel B.

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

G. Broye et al. / European Management Journal xxx (2017) 1e15

9

Table 2 Board leadership structure and leadership compensation. This table reports coefficient estimates using logarithm of CEO compensation in Panel A, and logarithm of leadership compensation in Panel B as dependent variables. Standard errors (in parentheses) are corrected following the method of White (1980). ***, ** and * indicate significance at the 1%, 5% and 10% level, respectively. Panel A e CEO compensation (N ¼ 577) Log Total compensation

Dual ROA Stock return

Log largest shareholder Family firm Instit. main shareholder CEO main shareholder Independent directors Board size CEO tenure CEO age Industry fixed effects Years fixed effects Constant Adj. R2 max VIF

Log Relative incentive

(1a)

(1b)

(2a)

(2b)

(3a)

(3b)

(4a)

(4b)

0.087*** (0.033) 0.002 (0.002) 0.015 (0.028)

0.33*** (0.094) 0.089* (0.047) 0.42*** (0.11) 0.14 (0.21) 0.47*** (0.17) 0.85** (0.34) 1.20** (0.53) 0.21*** (0.037) 0.012 (0.016) 0.007 (0.017) yes yes 4.43** (1.82)

0.15 (0.24) 0.016 (0.012) 0.12 (0.19) 0.014 (0.020) 0.68** (0.26) 0.32*** (0.095) 0.096** (0.047) 0.43*** (0.11) 0.13 (0.21) 0.47*** (0.17) 0.84** (0.34) 1.20** (0.53) 0.21*** (0.037) 0.013 (0.016) 0.006 (0.017) yes yes 4.69** (1.82)

0.070 (0.14) 0.019* (0.010) 0.45*** (0.14)

0.22*** (0.017) 0.019 (0.012) 0.039* (0.022) 0.060 (0.043) 0.047 (0.037) 0.13** (0.061) 0.49*** (0.11) 0.011 (0.007) 0.003 (0.003) 0.006* (0.003) yes yes 9.18*** (0.31)

0.052 (0.039) 0.003 (0.002) 0.062 (0.055) 0.0029 (0.0029) 0.073 (0.060) 0.22*** (0.018) 0.020* (0.012) 0.040* (0.022) 0.057 (0.043) 0.047 (0.037) 0.13** (0.061) 0.49*** (0.11) 0.010 (0.007) 0.003 (0.003) 0.006* (0.003) yes yes 9.23*** (0.32)

0.086 (0.16) 0.023** (0.010) 0.57*** (0.16)

0.30*** (0.030) 0.040** (0.018) 0.16*** (0.042) 0.17** (0.079) 0.098 (0.065) 0.37*** (0.11) 0.74*** (0.19) 0.051*** (0.012) 0.002 (0.005) 0.004 (0.005) yes yes 8.69*** (0.59)

0.08 (0.004) 0.004 (0.005) 0.017 (0.071) 0.002 (0.006) 0.25*** (0.094) 0.30*** (0.030) 0.042** (0.017) 0.16*** (0.041) 0.17** (0.079) 0.097 (0.065) 0.36*** (0.11) 0.74*** (0.19) 0.050*** (0.012) 0.002 (0.005) 0.003 (0.005) yes yes 8.74*** (0.60)

0.060 (0.085) 0.056 (0.041) 0.25*** (0.094) 0.14 (0.18) 0.45*** (0.14) 0.62** (0.31) 0.65 (0.51) 0.18*** (0.033) 0.016 (0.015) 0.0001 (0.016) yes yes 0.89 (1.64)

0.25 (0.22) 0.014 (0.010) 0.11 (0.17) 0.010 (0.019) 0.52** (0.23) 0.055 (0.085) 0.062 (0.041) 0.26*** (0.094) 0.14 (0.17) 0.45*** (0.14) 0.62** (0.31) 0.65 (0.51) 0.18*** (0.033) 0.017 (0.015) 0.001 (0.016) yes yes 0.70 (1.63)

0.57 2.58

0.57 3.43

0.63 2.58

0.63 3.43

0.44 2.58

0.44 3.43

0.34 2.58

0.34 3.43

Dual x Stock return

Market to book

Log Incentive compensation

0.14** (0.061) 0.0049* (0.003) 0.18*** (0.053)

Dual x ROA

Log assets

Log Fixed salary

Panel B e Leadership compensation (N ¼ 577)

Dual ROA Stock return

Log Total compensation

Log Fixed salary

(5a)

(5b)

(6a)

0.010 (0.059) 0.004 (0.003) 0.18*** (0.052)

0.083 (0.072) 0.003 (0.005) 0.0043 (0.065) 0.003 (0.006) 0.27*** (0.088) 0.30*** (0.029) 0.042** (0.016) 0.17*** (0.040) 0.18** (0.076) 0.068 (0.062) 0.39*** (0.11) 0.75*** (0.19) 0.052***

0.024 (0.033) 0.002 (0.002) 0.020 (0.028)

Dual x ROA Dual x Stock return Log assets Market to book Log largest shareholder Family firm Instit. main shareholder CEO main shareholder Independent directors Board size

0.30*** (0.029) 0.040** (0.017) 0.17*** (0.041) 0.18** (0.076) 0.069 (0.062) 0.39*** (0.11) 0.75*** (0.19) 0.052***

0.21*** (0.017) 0.019 (0.012) 0.045** (0.022) 0.068 (0.043) 0.050 (0.036) 0.12** (0.060) 0.48*** (0.10) 0.010

Log Incentive compensation

Log Relative incentive

(6b)

(7a)

(7b)

(8a)

(8b)

0.008 (0.038) 0.003 (0.002) 0.071 (0.054) 0.002 (0.003) 0.079 (0.059) 0.21*** (0.017) 0.020* (0.012) 0.045** (0.022) 0.066 (0.043) 0.051 (0.036) 0.12** (0.060) 0.48*** (0.10) 0.010

0.27* (0.15) 0.017* (0.009) 0.56*** (0.15)

0.58 *** (0.21) 0.004 (0.008) 0.14 (0.15) 0.026 (0.018) 0.65*** (0.23) 0.33*** (0.088) 0.11*** (0.038) 0.37*** (0.10) 0.30 (0.19) 0.30** (0.15) 1.35*** (0.29) 1.43*** (0.51) 0.20***

0.31** (0.13) 0.012 (0.009) 0.45*** (0.13)

0.56 *** (0.19) 0.000 (0.005) 0.16 (0.13) 0.023 (0.017) 0.45** (0.20) 0.038 (0.078) 0.080*** (0.030) 0.18** (0.082) 0.055 (0.16) 0.27** (0.12) 1.23*** (0.26) 0.97** (0.48) 0.17***

0.34*** (0.088) 0.098*** (0.038) 0.37*** (0.10) 0.32* (0.19) 0.30** (0.15) 1.34*** (0.30) 1.42*** (0.51) 0.20***

0.049 (0.079) 0.072** (0.029) 0.18** (0.083) 0.076 (0.16) 0.27** (0.12) 1.22*** (0.27) 0.96** (0.48) 0.17***

(continued on next page)

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

10

G. Broye et al. / European Management Journal xxx (2017) 1e15

Table 2 (continued ) Panel B e Leadership compensation (N ¼ 577) Log Total compensation

CEO tenure CEO age Industry fixed effects Years fixed effects Constant Adj. R2 Max VIF

Log Fixed salary

Log Incentive compensation

Log Relative incentive

(5a)

(5b)

(6a)

(6b)

(7a)

(7b)

(8a)

(8b)

(0.011) 0.003 (0.005) 0.006 (0.005) yes yes 8.87*** (0.58)

(0.011) 0.002 (0.005) 0.005 (0.005) yes yes 8.94*** (0.59)

(0.007) 0.003 (0.003) 0.007** (0.003) yes yes 9.28*** (0.31)

(0.007) 0.003 (0.003) 0.007** (0.003) yes yes 9.32*** (0.31)

(0.035) 0.007 (0.014) 0.022 (0.014) yes yes 5.38*** (1.67)

(0.035) 0.005 (0.014) 0.021 (0.014) yes yes 5.79*** (1.65)

(0.031) 0.007 (0.013) 0.017 (0.013) yes yes 0.28 (1.48)

(0.031) 0.006 (0.013) 0.017 (0.013) yes yes 0.64 (1.46)

0.59 2.58

0.60 3.43

0.63 2.58

0.63 3.43

0.50 2.58

0.50 3.43

0.39 2.58

0.39 3.43

provided to CEOs holding the COB title is related to a higher salary (regression 2a), whereas incentive compensation is similar in both governance structures (regressions 3a and 4a). Duality does not, therefore, have any significant effect on the level and the proportion of incentive compensation. This finding is not consistent with the predictions of both the optimal contracting and the managerial power expectations. However, the interaction between duality and performance variables will provide further results. Overall, and in line with agency theory, CEO compensation is significantly and positively related to firm performance. Regressions 1a, 3a and 4a show that performance measures play an important role in the determination of total and incentive executive compensation. These results are similar to the findings of Core at al. (1999) and Fernandes et al. (2013), but contrast with the results of Croci et al. (2012). We observe that executive compensation is more strongly associated with market performance than with accounting performance. Regressions 1b, 3b and 4b introduce interactive variables, and show that the sensitivity of CEO compensation to market performance is higher in dual firms than in firms with separate structures. In particular, incentive compensation in dual firms has a significantly stronger link to stock return than that observed in firms with separate structures. These results support the optimal contracting argument and hypothesis 2a, and are consistent with previous empirical evidence reported by Dey et al. (2011). In contrast, these results do not support hypothesis 2b and therefore invalidate the managerial power argument, which predicts that powerful CEOs holding the COB title might influence the board to serve their own interests, specifically through compensation that is not tied to performance. Our findings rather suggest that CEO compensation in dual firms is more closely related to market performance in order to align the interests of CEOs with those of shareholders when the monitoring role of the board is reduced. In line with this argument, the higher pay for CEOs combining positions might be explained by the setting of a higher fixed salary to compensate for the risk incurred by the CEO. Overall, while this study does not find any evidence pointing to higher levels of incentive compensation in dual firms, its findings clearly support the optimal contracting expectation of a stronger link between pay and market performance. The relationship between accounting performance and executive compensation is not affected by the type of board structure. Panel B shows our results when leadership compensation elements are used as dependent variables. When we include COB compensation in our estimation of leadership compensation, we find that the coefficient of the Dual variable is no longer significant (regression 5a). Thus, when taking into account the cost linked to

the compensation of both CEO and COB in firms with separate positions, we find that the aggregate compensation of leadership is equivalent to the compensation of a CEO combining both titles. This result supports Hypothesis 3. Specifically, our findings indicate that the fixed salary given to CEOs in dual firms is not significantly different in comparison to the combined fixed salary of both leaders in firms with separate structures (regression 6a). These results challenge the argument that the combined structure is associated with an additional cost. Overall, the need to set additional compensation for chairperson in separate structures offsets the granting of a higher compensation for dual firm CEOs in terms of cost. We also find that leadership total and incentive compensation are significantly associated with market performance (regressions 5a, 7a and 8a). Incentive compensation is more slightly related to accounting performance. The use of interactive variables in regressions 5b, 6b and 7b makes it possible to assess the impact of duality on the relationship between leadership compensation and firm performance. The interactive variable Dual x Stock return is positive and highly significant in regressions 5b, 7b and 8b. This finding implies that the sensitivity of CEO total and incentive compensation to market performance remains greater in dual firms than in firms with separate structures, even when COB compensation is taken into account. This evidence suggests that dual firms are willing to set incentive compensation for CEOs that is closely linked to performance, thus remaining consistent with optimal contracting theory and supporting Hypothesis 4b. However, this choice does not involve a higher level of compensation in dual firms. Overall, the choice to combine CEO and COB positions leads the board to increase CEO motivation through an incentive compensation that is closely tied to market performance, but this does not incur higher costs than those observed in firms with separate structures. Regressions 7a, 7b, 8a and 8b also show that CEOs holding the COB position receive a lower level of incentive compensation in comparison with the sum of incentive compensation packages for the CEO and the COB in separate structures. In other words, our results do not support Hypothesis 4a and indeed provide evidence that the compensation packages for the CEO and the COB in firms with separate structures are different to the package set for the CEO holding the COB position in dual firms. We show that the leadership compensation is more strongly linked to market performance in dual firms, thus supporting Hypothesis 4b and optimal contracting theory and invalidating Hypothesis 4c. However, our results do not support the prediction of Hypothesis 4b that the compensation package would include more incentive compensation in dual firms: on the contrary, we find that

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

G. Broye et al. / European Management Journal xxx (2017) 1e15

incentive compensation is higher in firms with separate structures. This finding can be interpreted as follows: Given the lack of significant difference in incentive CEO compensation in combined and separate structures (regressions 3a, 3b, 4a and 4b), this finding suggests that when we consider the incentive compensation set for the COB, the global incentive compensation of the CEO and the COB is higher than the incentive compensation set for the CEO alone in combined structure. In accordance with our expectations and with previous studies, CEO compensation and leadership compensation are shown to be higher in larger firms and in firms with strong growth opportunities. Ownership concentration is negatively associated with the levels of total and incentive compensation, suggesting that more direct and active monitoring provided by large shareholders involve limited compensation levels and smaller incentives. Hence, contrary to our expectations, this increased monitoring acts as a substitute for incentive compensation, rather than a complement. These results are in line with the findings of Core et al. (1999) and Van Essen et al. (2015). Conversely, the Institutional main shareholder variable has an insignificant effect on total compensation and a positive and significant effect on incentive compensation. Institutional shareholders therefore seem to favor a higher level of performance-based compensation for CEOs, probably because their large portfolio does not put them in a position to carry out strong, direct monitoring. This result is consistent with works by Croci et al. (2012) and Fernandes et al. (2013), who find that institutional shareholdings are associated with more equity-based compensation. In contrast to our expectations and to Croci et al. (2012), we find that family firms are associated with higher level of total compensation and are not associated with incentive compensation. A possible interpretation is that families over-compensate the CEOs in order to buy their loyalty, and allow them to expropriate minority shareholders (Croci et al., 2012). As expected, and in line with agency theory expectation, the CEO main shareholder variable is negatively associated with total, fixed and incentive compensation, suggesting that CEO large ownership reduces the need for any compensation contracts that encourage CEOs to act in the interest of shareholders (Core et al., 1999). Regarding board composition, a higher proportion of independent directors is associated with a higher level of total compensation. Whilst contrary to our expectations, this result is similar to those of Fernandes et al. (2013), Croci et al. (2012) and Van Essen et al. (2015). This finding may be consistent with the argument that independent directors are willing to increase CEO compensation through increased reliance on incentive compensation. We observe that the Independent directors variable is positively and significantly associated with both fixed and incentive compensation. However, the variable is not significantly associated with the ratio of CEO incentive compensation to total compensation. Independent directors can therefore be considered to effectively act to increase the level of CEO performance-based compensation, but this increase is not significant in relation to total compensation. Thus, we cannot conclude that independent boards play a significant role in the CEO incentive policy. These results are not consistent with Croci et al. (2012), who describe a positive association between board independence and the proportion of equity-based compensation to total compensation in Continental European firms. Like Croci et al. (2012), this study shows that larger boards are associated with higher total and incentive compensation. Despite potential coordination problems, large boards seem to provide more effective monitoring through more incentive compensation arrangements. Finally, we find that CEO tenure has no significant impact on CEO compensation, whereas CEO age implies a higher level of fixed salary.

11

5.3. Determinants of COB compensation Table 3 presents the determinants of COB compensation from the sub-sample of firms with separate positions. The highest VIF value in these regressions is 3.45, indicating that there is no problem of multicollinearity. The analysis of COB compensation enables us to explore our primary results in more depth. We find that the chairperson's total and incentive compensation are not associated with firm performance. This result is not consistent with Hypothesis 5, and fails to support the prediction that the chairperson's compensation contracts are designed to motivate COBs to maximize the performance of the firm. As previously mentioned, very few French chairpersons are granted stock options or restricted shares. Our finding further demonstrates that firms do not attempt to align the interests of COBs and shareholders by using incentives closely linked to performance. We also find evidence that the level of total and incentive COB compensation is positively and significantly related to firm size, but is not associated with growth opportunities. Regarding the ownership structure variables, we find that Log largest shareholder and Institutional main shareholder are negatively associated with COB total compensation, and that Family firm and Institutional main shareholder are significantly and negatively related to COB incentive compensation. Overall, these results suggest that ownership concentration, and specifically family and institutional blockholders, are associated with lower compensation and incentives for the COB. We previously predicted that these investors would press for more incentives to be given to the COB in order to ensure the efficient monitoring of the CEO. On the contrary, these results suggest a substitution effect, and are consistent with the view that active monitoring by controlling shareholders reduces the need to offer the chairperson high levels of compensation. Similarly, we find that CEO ownership is negatively associated with total and incentive compensation. These results support the expectation that the COB will receive less incentive to monitor effectively when there is little risk of agency conflicts occurring. This study also finds that the proportion of independent directors is not significantly associated with the level of total and incentive COB compensation. This result is not consistent with the expectation that independent directors will press for higher levels of pay and incentive compensation for the COB to monitor management more effectively. Board size has also no incidence on the level of total and incentive COB compensation. Overall, these results suggest that the COB compensation is more impacted by the type of ownership structure than by board composition. CEO tenure is positively related to COB total and incentive compensation, whereas CEO age has a negative impact on the incentive compensation of the COB and a positive impact on his/her fixed salary. These results do not allow us to clearly conclude that potential CEO power, as estimated through CEO tenure and CEO age, systematically results in giving less incentives to the COB to achieve effective monitoring. The age and tenure of the COB have no impact on his/her total and incentive compensation, although COBs with a longer tenure receive higher levels of fixed compensation. However, the chairperson's compensation is significantly influenced by his/her previous position: COBs who served as the CEO of the company before chairing the board received significantly higher levels of total and incentive compensation. The firm therefore values the skills acquired through this previous experience. Inversely, the COB founder variable is weakly and negatively associated with total compensation, and is not related to COB incentive compensation. Our results also provide evidence that the COB main shareholder variable has no impact on COB compensation. When the COB sits on the compensation committee, he/she

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

12

G. Broye et al. / European Management Journal xxx (2017) 1e15

Table 3 COB compensation determinants. This table reports coefficient estimates using logarithm of COB compensation as dependent variables. Standard errors (in parentheses) are corrected following the method of White (1980). ***, ** and * indicate significance at the 1%, 5% and 10% level, respectively. COB compensation (N ¼ 233)

ROA Stock return Log assets Market to book Log largest shareholder Family firm Instit. main shareholder CEO main shareholder Independent directors Board size CEO tenure CEO age COB tenure COB age COB previous CEO COB main shareholder COB founder COB in committee Industry fixed effects Years fixed effects Constant Adj. R2 Max VIF

Log Total compensation

Log Fixed salary

Log Incentive compensation

Log Relative incentive

0.016 (0.009) 0.20 (0.18) 0.39*** (0.12) 0.021 (0.057) 0.65*** (0.24) 0.46 (0.33) 0.68** (0.29) 1.54*** (0.40) 0.97 (0.71) 0.014 (0.047) 0.036** (0.017) 0.0011 (0.016) 0.043 (0.028) 0.004 (0.011) 0.93*** (0.26) 0.38 (0.26) 0.65* (0.38) 0.56** (0.24) yes yes 6.49*** (2.04)

0.032* (0.018) 0.32 (0.21) 0.047 (0.15) 0.012 (0.060) 0.30 (0.18) 1.28*** (0.30) 0.74** (0.30) 0.12 (0.44) 1.34* (0.71) 0.014 (0.062) 0.028 (0.022) 0.077*** (0.019) 0.11*** (0.032) 0.006 (0.012) 0.45 (0.31) 0.34 (0.35) 0.16 (0.41) 0.73*** (0.27) yes yes 3.36 (2.48)

0.014 (0.020) 0.26 (0.37) 0.90*** (0.17) 0.002 (0.094) 0.35 (0.23) 0.80** (0.39) 1.28*** (0.43) 2.13*** (0.52) 1.56 (1.05) 0.081 (0.072) 0.10** (0.041) 0.071*** (0.024) 0.029 (0.056) 0.008 (0.018) 1.32*** (0.42) 0.019 (0.40) 0.89 (0.59) 1.42*** (0.36) yes yes 1.11 (3.48)

0.012 (0.009) 0.18 (0.16) 0.33*** (0.069) 0.009 (0.039) 0.074 (0.089) 0.67*** (0.16) 0.54*** (0.17) 0.75*** (0.23) 0.17 (0.45) 0.060* (0.030) 0.042** (0.019) 0.030*** (0.011) 0.021 (0.021) 0.001 (0.007) 0.51*** (0.16) 0.081 (0.17) 0.36 (0.23) 0.48*** (0.15) yes yes 3.73** (1.44)

0.55 3.29

0.44 3.29

0.60 3.29

0.56 3.45

receives less total and incentive compensation but more fixed salary. An interpretation of this result is that the COB receives more fixed pay for his/her contribution to the committee, but does not use his/her position in the committee to push for more incentive compensation. 6. Discussion This research examines the relationship between board structure choice and leadership compensation by considering the chairperson's compensation contract. To our knowledge, this is the first study to take the COB's compensation into account when addressing the implications of board leadership structure in terms of potential costs. The paper aims to contribute to the literature investigating the costs and benefits of duality choice, with a focus on the specific French context. The study of this context sheds light on an institutional environment characterized by concentrated ownership and weak protection of investors, and by a “mixed” board structure model allowing some freedom in board structure selection. Consistent with the previous literature on duality, we first examined the relationship between leadership duality and CEO

compensation. Specifically, this paper questioned the validity of the predictions of optimal contracting theory versus managerial power theory in the French context (Dey et al., 2011; Van Essen et al., 2015). Our results show that the choice of a combined structure is associated with higher CEO compensation. Overall, our findings tend to support optimal contracting theory, meaning that the CEO compensation contract is used as a means to reduce agency conflicts. We suggest that board structure is likely to play a key role in the protection of minority shareholders’ interests and avoid their expropriation through excessive CEO compensation. This finding suggests that combined structures entail additional costs for firms. However, our study extends previous research by taking COB compensation into account and focusing on leadership compensation. Our paper concurs with the analysis made by Brickley et al. (1997) or Dey et al. (2011), who recommend investigating the costs and benefits of alternative leadership structures in greater detail. We argue that it is important to consider COB compensation and provide new insights into the financial cost of separating titles. Our results show that overall, the cost linked to leadership compensation is equivalent in firms with a combined or a separate structure, because the separation of these two positions entails COB compensation arrangements. Specifically, we suggest

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

G. Broye et al. / European Management Journal xxx (2017) 1e15

13

research has largely documented that the national legal and institutional environment affects the degree of complementarity or substitutability among different firm-level governance mechanisms (Schiehll and Martins, 2016), and their specific costs (Doidge et al., 2007). Since this study draws on French firms from an external context with weak investor protection and high ownership concentration, it may not be possible to generalize its conclusions to other environmental contexts. We might expect the compensation cost of duality to be higher in the Anglo-American context, which is characterized by higher reliance on incentives for the CEO and by more dispersed ownership. Another limitation of this study is that it focuses on the compensation cost of separating titles, but does not address other types of costs. Aguilera et al. (2008) argue that the implementation of corporate governance practices may have several types of associated costs, such as costs of compliance, opportunity costs, proprietary costs, or reputational costs. Furthermore, while focusing on the financial cost of separating titles, our study overlooks potential benefits, such as improved corporate performance (Dey et al., 2011; Krause & Semadeni, 2013). This paper also has limitations regarding the choice of variables used to explain compensation elements. Although we controlled for some characteristics of CEOs and COBs in our models, we did not include critical personal characteristics associated to the human and social capital of leaders (Combs & Skill, 2003; Damiani & Ricci, 2014; Hambrick & Mason, 1984; Harris & Helfat, 1997). This issue requires a thorough analysis in the French context given the specific educational background of CEOs, the particular path of access to top executive positions via prestigious “grandes
ecoles” and civil service “corps”, and the high levels of social capital of elite CEOs (Goyer & Jung, 2011). Additionally, this study focused solely on CEO and COB characteristics, and did not consider the influence of board members, or top management team specificities such as socialpsychological factors, on the pay setting process (Fredrickson, Davis-Blake, & Sanders, 2010).

that incentives for the chairperson are necessary to avoid potential agency conflicts between shareholders and the chairperson. However, we have no evidence that this incentive compensation is related to firm performance. By evaluating the specific cost associated with leadership compensation, we conclude that the two types of structure are equivalent in terms of compensation costs. Overall, our study highlights that this dimension must be taken into consideration when studying the trade-offs inherent to board structure choice. This paper has also implications for cross-national governance research (Aguilera et al., 2008; Schiehll & Martins, 2016; Van Essen et al., 2013). Our study highlights the necessity to consider the institutional context in which board structure models can be selected and compensation policy for both CEOs and COBs can be shaped. Even if the role played by the incentive compensation could be considered weakened by the high ownership concentration that is inherent to the French context, this study reveals that this mechanism is still used in France, with CEO compensation arrangements that are linked to market performance. Our results also suggest that board structure significantly impacts compensation policy in France. Our results also evidence the interactions between compensation incentives, board composition and ownership structure. In the French context we find some similar complementary and substitution effects to those previously reported in the Anglo-Saxon framework. Our findings have practical implications for regulators and practitioners who publicly debate the benefits and costs of duality, and commonly advocate the separation of titles as the best practice to ensure “good” corporate governance. This study sheds new light on the outcomes of board structure choice by highlighting an under-researched aspect associated with the financial cost of duality for firms. We believe that this research might be useful in helping firms to understand the different implications of their board structure choice. It may also help regulators in the European Union to identify the strengths and weaknesses of alternative models given the recent movement toward greater freedom of choice about board structure among member states (Belot et al., 2014). Specifically, our findings can be of use to practitioners seeking a better understanding of the interdependences between board composition and the compensation practices implemented for both the CEO and the COB. This analysis also provides regulators and practitioners with new insights into the determinants of COB pay within the French business context, where the exact role played by the chairperson and his/her level of pay is a current source of controversy. The French market authority recommends the removal of variable and equity-based compensation for chairpersons (AMF, 2015). This position may seem surprising in view of literature demonstrating that the incentive compensation of directors positively impacts firm performance. The AMF recommendation, however, is primarily based on an insufficient grasp of the precise role played by French chairpersons, and a lack of clarity about how this role is linked to the level of compensation. Companies rarely communicate about the specific role played by the chairpersons and to what extent their contribution has an influence on the strategic direction and performance of the firm. It would therefore be worthwhile to clarify the respective roles and responsibilities of the CEO and the COB (Proxinvest, 2013; Saint-Geours, 2009). Overall, this study might encourage firms to consider how the COB contributes to strategymaking and firm performance, and to set compensation that is sufficiently attractive to achieve this.

This paper focuses on compensation cost, a specific dimension of the consequences of separating the CEO and COB roles in firms within the French national context. Further research is needed to better understand the global effect of duality from a multidimensional perspective in the light of the overall costs and benefits of splitting the roles. Most importantly, further research is required to estimate the overall implications of duality across countries. Cross-country analysis should provide a more context-dependent understanding of outcomes associated with the choice of board structure. Previous literature provides evidence that institutional settings considerably affect these outcomes. For instance, the board model adopted in a particular institutional context influences the advisory role of the board, the board's ability to exchange information and knowledge resources, and the extent to which boards can contribute to the strategy-making process (Adams & Ferreira, 2007; Heyden, Oehmichen, Nichting, & Volberda, 2015). It is now necessary to investigate how institutional aspects also affect the role played by the chairperson and his/her contribution to the firm e from both a monitoring and an advisory perspective. Future research might also benefit from the use of our framework in different external settings to assess the influence of national contexts on COB compensation policy and the cost of duality.

6.1. Limitations

7. Conclusion

Our study has several limitations. Comparative governance

6.2. Future research

Our study contributes to the literature investigating duality

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

14

G. Broye et al. / European Management Journal xxx (2017) 1e15

choice and enriches comparative corporate governance research by providing a contextualized study on how the choice of duality affects leadership compensation. Our findings show that COB compensation is likely to influence the cost of duality, and this dimension should be considered in future research concerning the advantages and disadvantages of splitting the positions of CEO and COB. Acknowledgements The authors would like to thank the editor, Sabina Siebert, the associate editor, Mariano Heyden, and the two anonymous reviewers for their insightful comments and suggestions. References Adams, R., & Ferreira, D. (2007). A theory of friendly boards. Journal of Finance, 62, 217e250. AFEP/MEDEF. (2013). Code de gouvernement d’entreprise des soci
et
es cot
ees. Aguilera, R. V., Filatotchev, I., Gospel, H., & Jackson, G. (2008). An organizational approach to comparative corporate governance: Costs, contingencies, and complementarities. Organization Science, 19(3), 475e492. Aguilera, R. V., & Jackson, G. (2003). The cross-national diversity of corporate governance: Dimensions and determinants. Academy of Management Review, 28(3), 447e465. Anderson, C. A., & Anthony, R. N. (1986). The new corporate directors: Insights for board members and executives. Wiley.
des Marche
s Financiers. (2014). Rapport de l’AMF sur le gouvernement Autorite
mune
ration des dirigeants. d’entreprise et la re
des Marche
s Financiers. (2015). Rapport de l’AMF sur le gouvernement Autorite
mune
ration des dirigeants. d’entreprise et la re Baber, W., Janakiraman, S., & Kang, S.-H. (1996). Investment opportunities and the structure of executive compensation. Journal of Accounting and Economics, 21, 297e318. Baliga, R., Moyer, C., & Rao, R. (1996). CEO duality and firm performance: What's the fuss? Strategic Management Journal, 17, 41e43. Bebchuk, L., & Fried, J. (2003). Executive compensation as an agency problem. Journal of Economic Perspectives, 17(3), 71e92. Bebchuk, L., & Fried, J. (2004). Pay without performance: The unfulfilled promise of executive compensation. Cambridge: Harvard University Press. Belot, F., Ginglinger, E., Slovin, M., & Sushka, M. (2014). Freedom of choice between unitary and two-tier boards: An empirical analysis. Journal of Financial Economics, 112, 364e385. Boyd, B. (1994). Board control and CEO compensation. Strategic Management Journal, 15(5), 335e344. Boyd, B. (1996). Determinants of US outside director compensation. Corporate Governance: An International Review, 4(4), 202e211. Brickley, J., Coles, J., & Jarrell, G. (1997). Leadership structure: Separating the CEO and chairman of the board. Journal of Corporate Finance, 3, 189e220.
mune
ration des dirigeants et gouvernance des Broye, G., & Moulin, Y. (2010). Re
es. Finance Contro ^le Strat
entreprises: Le cas des entreprises françaises cote egie, 13(1), 67e98.
mune
ration des pre
sidents non exe
cutifs Broye, G., & Moulin, Y. (2014). La re
pend-elle de leur capital humain? Management International, 19(1), 204e218. de Capezio, A., Shields, J., & O'Donnell, M. (2011). Too good to be true: Board structural independence as a moderator of CEO pay-for-firm-performance. Journal of Management Studies, 48(3), 487e513. Chhaochharia, V., & Grinstein, Y. (2009). CEO compensation and board structure. Journal of Finance, 64(1), 231e261. Combs, J. G., & Skill, M. S. (2003). Managerialist and human capital explanations for key executive pay premiums: A contingency perspective. Academy of Management Journal, 46(1), 63e73. Conyon, M., & Peck, S. (1998). Board control, compensation committees, and top management compensation. Academy of Management Journal, 41(2), 146e157. Cordeiro, J., & Veliyath, R. (2003). Beyond pay for performance: A panel study of the determinants of CEO compensation. American Business Review, 21(1), 56e66. Cordeiro, J., Veliyath, R., & Romal, J. (2007). Moderators of the relationship between director stock-based compensation and firm performance. Corporate Governance: An International Review, 16(6), 1384e1393. Core, J. E., Guay, W. R., & Larcker, D. F. (2003). Executive equity compensation and incentives: A survey. Economic Policy Review, 9(1), 27e50. Core, J. E., Holthausen, R. W., & Larcker, D. F. (1999). Corporate governance, chief executive officer compensation, and firm performance. Journal of Financial Economics, 51(3), 371e406. Croci, E., Gonenc, H., & Ozkan, N. (2012). CEO compensation, family control, and institutional investors in Continental Europe. Journal of Banking & Finance, 36(12), 3318e3335. Cyert, R. M., Kang, S. H., & Kumar, P. (2002). Corporate governance, takeovers, and top-management compensation: Theory and evidence. Management Science, 48(4), 453e469.

Damiani, M., & Ricci, A. (2014). Managers' education and the choice of different variable pay schemes: Evidence from Italian firms. European Management Journal, 32(6), 891e902. Desender, K. A., Aguilera, R. V., Crespi, R., & Garcia-Cestona, M. (2013). When does ownership matter? Board characteristics and behavior. Strategic Management Journal, 34(7), 823e842. Deutsch, Y. (2007). The influence of outside directors' stock-option compensation on firms' R&D. Corporate Governance: An International Review, 15(5), 816e827. Dey, A., Engel, E., & Liu, X. (2011). CEO and board chair roles: To split or not to split? Journal of Corporate Finance, 17(5), 1595e1618. Doidge, C., Karolyi, G. A., & Stulz, R. M. (2007). Why do countries matter so much for corporate governance? Journal of Financial Economics, 86(1), 1e39. Expert Corporate Governance Service (ECGS). (2012). Directors' pay 2011: An international comparison. Faccio, M., & Lang, L. H. (2002). The ultimate ownership of Western European corporations. Journal of Financial Economics, 65(3), 365e395. Faleye, O. (2007). Does one hat fit all? The case of corporate leadership. Journal of Management & Governance, 11, 239e259. Fama, E., & Jensen, M. (1983). Separation of ownership and control. Journal of Law and Economics, 26, 301e326. Fernandes, N., Ferreira, M., Matos, P., & Murphy, K. (2013). Are US CEOs paid more? New international evidence. Review of Financial Studies, 26(2), 323e367. Finkelstein, S. (1992). Power in top management teams: Dimensions, measurement, and validation. Academy of Management Journal, 35(3), 505e538. Finkelstein, S., & D'Aveni, R. (1994). CEO duality as a double-edged sword: How boards of directors balance entrenchment avoidance and unity of command. Academy of Management Journal, 37(5), 1079e1108. Firth, M., Fung, P., & Rui, O. (2007). How ownership and corporate governance influence chief executive pay in China's listed firms. Journal of Business Research, 60, 776e785. Florou, A. (2005). Top director shake-up: The link between chairman and CEO dismissal in the UK. Journal of Business Finance & Accounting, 32(1), 97e128. Fraile, I. A., & Fradejas, N. A. (2014). Ownership structure and board composition in a high ownership concentration context. European Management Journal, 32(4), 646e657. Fredrickson, J. W., Davis-Blake, A., & Sanders, W. M. (2010). Sharing the wealth: Social comparisons and pay dispersion in the CEO's top team. Strategic Management Journal, 31(10), 1031e1053. Goyer, M., & Jung, D. K. (2011). Diversity of institutional investors and foreign blockholdings in France: The evolution of an institutionally hybrid economy. Corporate Governance: An International Review, 19(6), 562e584. Grabke-Rundell, A., & Gomez-Mejia, L. R. (2002). Power as a determinant of executive compensation. Human Resource Management Review, 12(1), 3e23. Hambrick, D. C., & Mason, P. A. (1984). Upper echelons: The organization as a reflection of its top managers. Academy of Management Review, 9(2), 193e206. Harris, D., & Helfat, C. (1997). Specificity of CEO human capital and compensation. Strategic Management Journal, 18(11), 895e920. Heidrick, Struggles (2011). European corporate governance report 2011. Hermalin, B. E., & Weisbach, M. S. (1998). Endogenously chosen boards of directors and their monitoring of the CEO. American Economic Review, 88(1), 96e118. Heyden, M. L., Oehmichen, J., Nichting, S., & Volberda, H. W. (2015). Board background heterogeneity and exploration-exploitation: The role of the institutionally adopted board model. Global Strategy Journal, 5(2), 154e176. Hill, C., & Phan, P. (1991). CEO tenure as a determinant of CEO pay. Academy of Management Journal, 34(3), 707e717. Jensen, M. (1993). The modern industrial revolution, exit, and the failure of internal control systems. Journal of Finance, 48(3), 831e880. Jensen, M., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305e360. Jensen, M., & Murphy, K. (1990). Performance pay and top-management incentives. Journal of Political Economy, 98, 225e264. Karabadse, A., & Karabadse, N. (2007). The return of the chairman. Business Strategy Review, Winter, 60e65. Krause, R., & Semadeni, M. (2013). Apprentice, departure, and demotion: An examination of the three types of CEO-board chair separation. Academy of Management Journal, 56(3), 805e826. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor protection and corporate governance. Journal of Financial Economics, 58(1), 3e27. Leblanc, R. (2005). Assessing board leadership. Corporate Governance: An International Review, 13(5), 654e666. Linck, J. S., Netter, J. M., & Yang, T. (2008). The determinants of board structure. Journal of Financial Economics, 87(2), 308e328. Mehran, H. (1995). Executive compensation structure, ownership, and firm performance. Journal of Financial Economics, 38(2), 163e184. Murphy, K. J. (1999). Executive compensation. Handbook of Labor Economics, 3, 2485e2563. Proxinvest. (2013). La r
emun
eration des dirigeants des soci
et
es du SBF (Vol. 120). Roberts, J. (2002). Building the complementary board. The work of the Plc chairman. Long Range Planning, 35, 493e520. Ryan, H. E., & Wiggins, R. A. (2004). Who is in whose pocket? Director compensation, board independence, and barriers to effective monitoring. Journal of Financial Economics, 73, 497e524.
phalisme en question. Saint-Geours, J. P. (2009). Gouvernement d’entreprise: Le bice Lettre de l’IFA, 19, 16e17. Schiehll, E., & Martins, H. C. (2016). Cross-national governance research: A

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007

G. Broye et al. / European Management Journal xxx (2017) 1e15 systematic review and assessment. Corporate Governance: An International Review, 24(3), 181e199. Shleifer, A., & Vishny, R. W. (1997). A survey of corporate governance. The Journal of Finance, 52(2), 737e783. Tosi, H., Werner, S., Katz, J., & Gomez-Mejia, L. (2000). How much does performance matter? A meta-analysis of CEO pay studies. Journal of Management, 26(2), 301e339. Tuggle, C. S., Sirmon, D. G., Reutzel, C. R., & Bierman, L. (2010). Commanding board of director attention: Investigating how organizational performance and CEO duality affect board members' attention to monitoring. Strategic Management Journal, 31(9), 946e968. Van Essen, M., Engelen, P. J., & Carney, M. (2013). Does “good” corporate governance help in a crisis? The impact of country- and firm- level governance mechanisms in the European financial crisis. Corporate Governance: An International Review, 21(3), 201e224. Van Essen, M., Otten, J., & Carberry, E. J. (2015). Assessing managerial power theory: A meta-analytic approach to understanding the determinants of CEO

15

compensation. Journal of Management, 41(1), 164e202. Waelchli, U., & Zeller, J. (2013). Old captains at the helm: Chairman age and firm performance. Journal of Banking & Finance, 37, 1612e1628. Weisbach, M. S. (2007). Optimal executive compensation versus managerial power: A review of Lucian Bebchuk and Jesse Fried's pay without performance: The unfulfilled promise of executive compensation. Journal of Economic Literature, 45(2), 419e428. Westphal, J. D., & Zajac, E. J. (1995). Who shall govern? CEO/board power, demographic similarity, and new director selection. Administrative Science Quarterly, 40, 60e83. White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroscedasticity. Econometrica, 48(4), 817e838. Yang, T., & Zhao, S. (2014). CEO duality and firm performance: Evidence from an exogenous shock to the competitive environment. Journal of Banking & Finance, 49, 534e553. Yermack, D. (2004). Compensation, retention, and reputation incentives for outside directors. Journal of Finance, 59, 2281e2308.

Please cite this article in press as: Broye, G., et al., The cost of CEO duality: Evidence from French leadership compensation, European Management Journal (2017), http://dx.doi.org/10.1016/j.emj.2017.01.007