MANAGERIAL ENTRENCHMENT ANDCORPORATE SOCIAL PERFORMANCE †

JORDI SURROCAUniversidad Carlos III de Madrid

Department of Business AdministrationCalle Madrid, 126

Getafe (Madrid), Spain, 28903Phone: (34) 91-624-8640Fax: (34) 91-624-9607

e-mail: jsurroca@emp.uc3m.es

JOSEP A. TRIBÓ *Universidad Carlos III de Madrid

Department of Business AdministrationCalle Madrid, 126

Getafe (Madrid), Spain, 28903Phone: (34) 91-624-9321Fax: (34) 91-624-9607

e-mail: joatribo@emp.uc3m.es

* Corresponding author

† The authors wish to thank Sustainable Investment Research International Company, and Analistas

Internacionales en Sostenibilidad (AISTM) for their helpful comments and access to the SiRi ProTM

database. We would also like to thank Diego Prior, Stathopoulos Konstantinos and one of the referees as

well as participants at Symposium of Corporate Governance & Shareholder Activism (Milan, 2007), the

Journal of Banking and Finance 30th Anniversary Conference (Beijing, 2006), the Strategic Management

Conference (Vienna, 2006), the Foro de Finanzas (Castellon, 2006) and the seminar participants at

Universitat Autònoma of Barcelona for their useful comments on earlier drafts of this paper. We also

acknowledge the financial support of the Ministerio de Ciencia y Tecnologia (grant #SEC2003-03797),

Ministerio de Educación y Ciencia (grant # SEJ2004-07877-C02-02 and grant # SEJ2006-09401), and the

Comunidad de Madrid (Grant #s-0505/tic/000230). The usual disclaimers apply.

2

MANAGERIAL ENTRENCHMENT ANDCORPORATE SOCIAL PERFORMANCE

Abstract

We examine empirically the relationships amongst managerial entrenchment

practices, social performance, and financial performance. We hypothesize that

entrenched managers may collude with non-shareholder stakeholders in order to

reinforce their entrenchment strategy; this is particularly so in firms that have efficient

internal control mechanisms. Moreover, we prove that the combination of entrenchment

strategies and the implementation of socially responsible actions have particularly

negative effects on financial performance. We demonstrate this theoretical contention

by using different proxies for entrenchment like the existence of: anti-takeover

initiatives; managerial ownership; managers’ tenure and smoothing of earnings. Our

empirical investigation uses a database comprising 777 companies, from 27 different

countries, for the period 2002-2005.

JEL Classification: G30, M14, M41.

Keywords: Corporate Governance, Stakeholder Activism, Corporate Social

Performance, Earnings Management.

3

1/ INTRODUCTION

Agency theory has contributed to spread the long held view that decisions taken

by managers often go against the interests of the organizations they are managing. In an

agency contract, the principal, investors or outsiders, transfers certain duties to the

agent; a manager, an entrepreneur or an insider. Given that they often face a situation of

limited rationality and imperfect information, the contract does not instruct the agent

adequately, on how to act when faced with circumstances that could not have been

foreseen. Such circumstances are common in the normal running of a business.

Contracts are, therefore, incomplete and, in an information asymmetric context, give the

agent the power of pursuing their own benefit against the interests of the principal.

Managers may: exert insufficient effort; find it convenient to accept overstaffing; amass

private benefits by building empires, enjoying perquisite consumption, or even by

stealing from the firm (Shleifer and Vishny, 1997).

One of the costliest manifestations of the agency problem is managerial

entrenchment (Jensen and Ruback, 1983). Managers, who place a great value on control

but own only a small equity stake, work to ensure their own job security thereby

entrenching themselves and staying on in that position even if no longer competent or

qualified to run the firm (Shleifer and Vishny, 1989).

Walsh and Seward (1990) discuss different classes of managerial entrenchment

practices to neutralize the disciplinary mechanism of a market for corporate control.

Dual-class recapitalization, poison pills, supermajority amendments, anti-takeover

amendments and golden parachutes are examples of such practices. Also, other authors

(Stulz, 1988; Morck et al. 1988; De Miguel et al., 2004) have emphasized managerial

ownership above certain levels as a takeover deterrence mechanism that promotes

managerial entrenchment. The use of these different mechanisms generates a decrease

in managerial turnover, which has also been used by different authors as a proxy for

managerial entrenchment (Denis et al., 1997 and Dahya et al., 1998). Finally, managers

may resort to income smoothing and other earnings management practices as a way of

improving their job security (Fudenberg and Tirole, 1995). This idea is supported

empirically byYeo et al. (2002).

Although the presence of anti-takeover defenses may serve to decrease the

efficiency of external control mechanisms to the point where the cost of a takeover is

prohibitive, we cannot assume that entrenched managers will be perfectly hedged

against the actions of shareholders. This is so because of strong internal corporate

governance mechanisms like the proportion of independent directors on the board

4

composition, reduce the need for the operation of the takeover market (Sundaramurthy

et al., 1997). In fact, Sundaramurthy et al., (1997) shows that the negative market

reaction to takeover defenses adopted by entrenched managers is contingent on board

configuration: the larger the proportion of independent directors, the lower the negative

market reaction after the implementation of anti-takeover measures. Sundaramurthy’s

(2000) relies on this contingency model to explain the consequences of anti-takeover

provisions on shareholders’ wealth. In such a model, both board structure and

shareholder monitoring moderate the relationship between anti-takeover provisions and

shareholder interests. In addition, Vafeas (1999) expects other internal governance

mechanisms to be effective disciplinary arrangements, in order to resolve agency

problems such as the existence of different board subcommittees.

Then, the capital markets would recognize internal control structures such as the

board of directors, ownership structures, and board subcommittees, as playing a

substitute monitoring role to the takeover market. Hence, we expect that, in the presence

of such control mechanisms, entrenched managers will reinforce their entrenchment

strategy by engaging in other types of practices so as to block existing governance

mechanisms.

In this paper, we extend the work of Cespa and Cestone (2004) and we

hypothesize that entrenched managers may collude with non-shareholder stakeholders –

i.e., employees, communities, customers, and suppliers – to protect themselves from

internal disciplining mechanisms, causing a subsequent reduction in shareholders’

wealth. We rely on two arguments to justify the manager’s commitment to follow

stakeholder-friendly behavior. Firstly, stakeholders generally accumulate certain powers

to promote or disgrace top executives (DeAngelo and DeAngelo, 1998; Hellwig, 2000;

Rowley and Berman, 2000); they may engage in costly boycotts and media campaigns

(Baron, 2001; Feddersen and Gilligan, 2001; John and Klein, 2003) or stakeholder

representatives may be present in corporate boards (Luoma and Goodstein, 1999;

Schneper and Guillén, 2004). Moreover, due to the fact that the manager retains the

confidence of stakeholders, it will be more costly for displeased shareholders to remove

him because they will have to face pressure from the non-shareholder stakeholders. The

second argument which justifies stakeholder concessions for entrenchment purposes is

that, by colluding with stakeholders, the manager reduces a firm’s attractiveness to

potential raiders. For example, generous long-term stakeholders’ concessions hinder the

raider’s ability to generate profit (Pagano and Volpin, 2005). Moreover, we argue that

markets are far from perfect and that they do not fully internalize through discounted

5

share prices the cost of such entrenchment actions. Otherwise, it would be useless for

managers to pursue these kinds of entrenchment strategies. This market inefficiency

relies on the fact that stakeholder satisfaction also has a positive impact on financial

performance (e.g., Berman et al., 1999) and that it is difficult for outside investors to

disentangle this positive effect from the negative one linked to the implementation of an

entrenchment strategy. Managers take advantage of this fact and may be inclined to use

stakeholder satisfaction for entrenchment purposes.

In order to gain support from stakeholders, entrenched managers engage in a

broad array of practices to deal with and create relationships with corporate stakeholders

and the natural environment; the so-called corporate social performance or CSP

(Waddock, 2004). Therefore, this study specifically tests the hypothesis that managerial

entrenchment practices are positively related to improvements in CSP which, in turn,

negatively affect firms’ financial performance. Moreover, we expect this relationship

will be stronger in those firms with well-developed internal corporate governance

mechanisms for controlling a manager such as those with: a large proportion of

independent directors in the board; the position of CEO and the chairman occupied by

different persons; managerial remuneration based on performance related schemes

(option-like schemes); and, especially, when such remuneration and promotion is

decided by an independent committee. Additionally, the presence of certain types of

blockholders, like the state, in concentrated ownership structures may facilitate the

manager’s monitoring. In such situations, the manager has greater difficulty in

triggering entrenchment strategies based on confrontation with shareholders like the

aforementioned poison pills and/or the limitation of shareholders’ rights rather than

strategies relying on cooperation with non-shareholders stakeholders.

To demonstrate our theoretical contention, we make use of an international

database provided by the Sustainable Investment Research International (SiRi)

Company, an international network of social research organizations that scrutinizes

firms with respect to their practices toward employees, communities, suppliers,

customers, environment, and shareholders. These data include and expand upon those of

Kinder, Lyndemberg, Domini, and Company (KLD) which has been used in several

papers studying stakeholders’ issues (Agle et al., 1999; Berman et al., 1999; Hillman

and Keim, 2001; Waddock and Graves, 1997). Our final sample comprises 717 firms

from 27 nations; hence, we can argue that it is representative of different institutional

frameworks. This is a relevant point, because the institutional environment clearly

6

shapes the relationship between a firm’s corporate governance structure and stakeholder

interests (Aguilera and Jackson, 2003; Schneper and Guillén, 2004).

Our study has antecedents in and contributes to both the agency and stakeholder

theories. In particular, our study is one of the few to investigate the association between

corporate governance mechanisms and stakeholder concessions. Previous literature on

this issue has analyzed relationships between CSP and CEO’s incentives (McGuire et

al., 2003; Coombs and Gilley, 2005), board composition and environmental

performance (Kassinis and Vafeas, 2002, 2006); frequency of takeovers and

stakeholders’ rights (Schneper and Guillén, 2004); and the influence of institutional

investor type, managerial equity, and independence of boards on CSP (Johnson and

Greening, 1999). Our article goes a step further and examines the connections among

these four elements: first, entrenchment practices (i.e. anti-takeover devices; limitation

of shareholders’ rights; managers’ stakes and tenure; income smoothing); second,

internal corporate governance mechanisms (i.e. board independence; the presence of

board subcommittees; performance-based payment schemes and ownership

concentration); third, a firm’s social performance; and last, a firm’s financial

performance. Also, this study advances the understanding of stakeholder phenomena by

providing evidence of another explanation for CSP: the entrenchment hypothesis (Cespa

and Cestone, 2004). Previous literature on stakeholder theory has provided normative,

instrumental and descriptive/empirical explanations of CSP (Donaldson and Preston,

1995). Our entrenchment argument for improving CSP falls into the instrumental strand

and develops the descriptive approach (Mitchell et al., 1997; Jawahar and McLaughlin,

2001), which suggests that the degree to which managers give priority to competing

stakeholders’ claims – the salience – is positively related to the stakeholder attributes of

power, legitimacy, and urgency. In our case, we add to Mitchell and colleagues’ (1997)

model the idea that, not only is the CEO’s perception of the aforementioned attributes

important to explain stakeholder salience, but also to managerial strategic actions such

as entrenchment practices. This study also extends agency theory from a stakeholder

perspective, by examining the role employees, communities, customers, and suppliers,

may play in exacerbating or ameliorating conflicts of interests between managers and

shareholders. In doing so, we expand the game-theoretical model of Pagano and Volpin

(2005) to include not only workers but other stakeholders (Cespa and Cestone, 2004),

and subsequently, we test empirically its main propositions. Finally, in this study we

provide further evidence of the entrenchment motives that may explain certain practices

of earnings manipulations like income smoothing (Yeo et al, 2002; Fudenberg and

7

Tirole, 1995) as well as their connection with CSP. Managers smooth earnings, as part

of an entrenchment strategy, in order to ensure the stability of cash-flow streams so that

they can satisfy the short-term interests of shareholders. However, this practice may

generate problems in the medium term. Then, in order to eliminate possible medium-

term problems that may put their jobs at risk, managers will try to connive with non-

shareholder stakeholders by satisfying the interests of this group (improving a firm’s

CSP). Hence, regardless of the entrenchment mechanism used (poison pills, limitation

of shareholder rights or income smoothing), our basic claim is that entrenchment

practices lead to improvements in a firm’s CSP, particularly in the presence of strong

internal corporate governance mechanisms.

The rest of the article is structured as follows. Section 2 summarizes the most

relevant literature akin to the objectives of this work. In Section 3 we develop the

hypotheses. Section 4 is methodological and describes the sample, variables and

empirical models to be tested. The empirical results obtained are presented in Section 5,

while some extensions are addressed in Section 6. In the final section of the article, we

lay out the main conclusions of this research and discuss the significance of our results.

2/ REVIEW OF THE LITERATURE

Managerial entrenchment

Traditionally, agency theory has dominated the analysis of corporate

governance. Its main concern is the separation of ownership and control, and the

possibility that managers (agents) take actions that hurt shareholders – principals –

(Eisenhardt, 1989). Managers may amass private benefits by building empires, maintain

costly labor, pay inflated transfer prices to affiliated entities, or simply exert insufficient

effort. The genesis of this agency problem is twofold: first, the dispersion of ownership

and, second, the reduced proportion of equity that managers hold in the firm (Jensen

and Meckling, 1976). Ownership dispersion generates free-riding problems in

monitoring and controlling the behavior of managers, while low managerial equity

holdings imply that managers enjoy the private benefits of control and only bear a

fraction of its costs. As a consequence of both forces, managers may deploy corporate

assets to obtain personal benefits like perquisite consumption.

In this context, a good governance structure is then one that is able to align the

interests of principals and agents. Shleifer and Vishny (1997) discussed the available

8

mechanisms to force agents to internalize the welfare of shareholders that are classified

according to their internal or external nature. Internal mechanisms are managerial

incentives like stock-options and other forms of performance-based payment schemes,

and control structures such as the presence of institutional blockholders, the presence of

outsiders in the board of directors, or the existence of audit/remuneration/nomination

committees. On the other hand, the market for corporate control, product market

competition or managerial labor markets are examples of external mechanisms of

corporate governance.

The need for governance mechanisms is greater when a manager’s stake is low

because managers internalize a low proportion of the agency costs. In that case, an

external governance mechanism, like the threat of a takeover, may turn out to be a

powerful method of disciplining managerial behavior because managers know that they

are at risk of being dismissed for sub-optimal corporate performance, even if they were

not responsible for these bad results). Anticipating this possibility, managers may adopt

entrenchment practices in different forms (Walsh and Seward, 1990). Then, as

managerial stake increases, agency costs become lower since managers bear a larger

share of these costs (Jensen and Meckling, 1976). This internalization of costs reduces

the incentives to trigger entrenchment strategies. Some researchers argue, however, that

increasing managerial ownership may not increase shareholder wealth; managers could

increase their ownership to a degree that would allow them to dominate the board of

directors, thereby becoming insulated from internal or external corporate governance

mechanisms such as hostile takeovers (Fama and Jensen, 1983). We adopt a synthesis of

these different views, as proposed by Morck et al (1988) and De Miguel et al (2004),

where they argue that entrenchment is found in a middle range of managerial stake.

Below the lower bound level, the monitoring of managers is so intense that there is no

entrenchment. Beyond the upper bound level, the aforementioned internalization of

entrenchment costs, by the manager, prevents these practices

Then, apart from the managerial ownership stake that may be used strategically

as an entrenchment tool for deterring takeovers (Morck et al., 1988), there are different

mechanisms available for the implementation of an entrenchment strategy that could

decrease managerial turnover (Denis et al., 1997 and Dahya et al, 1998). Among them

is the issue of common stock, with limited voting rights, which is then exchanged for a

certain number of old common shares; the repurchase of large blocks of shares from

potential acquirers without the approval of shareholders; poison pills; new security

issues; specific acquisitions and divestitures; supermajority amendments; golden

9

parachutes; and earnings smoothing. Each of these practices may reduce the efficacy of

corporate governance mechanisms to the point that the cost of a takeover exceeds its

benefits. Moreover, the implementation of these measures is not only costly on their

own but also, once the manager is immune to the threat of takeovers, he can pursue his

own interests at the expense of shareholders’ best interests (Shleifer and Vishny, 1989).

Different governance structures may amplify or mitigate the consequences of

entrenchment that emerge within this setting (Walsh and Seward, 1990; Sundaramurthy

et al., 1997; Sundaramurthy, 2000): 1) the composition of boards of directors; 2) the

concentration of ownership; 3) the existence of different board subcommittees that will

take on the responsibility to put in place adequate managerial incentive schemes.

On the composition of the boards, we can argue that boards with a majority of

outside non-executive directors have all the incentive to oppose managers’

entrenchment actions (Fama and Jensen, 1983). However, this control will be

diminished if the chief executive officer (CEO) is also the chairman of the board. This

dual role of CEO can give rise to a conflict of interest preventing boards from being

effective in their monitoring and supervision tasks.

The second mechanism relates to the ownership structure. In more concentrated

ownership structures, large investors have the incentive to collect information and

monitor the management, thereby reducing managerial agency costs and preventing

entrenchment (Shleifer and Vishny, 1986). However, a second problem emerges as

large shareholders pursue their own interests and may expropriate minority shareholders

(Shleifer and Vishny, 1997). In fact, they may even collude with managers, by allowing

certain entrenchment strategies, in order to expropriate minority shareholders (Pound,

1988). Remarkably, when control is dissipated among several large investors, a decision

to expropriate minority shareholders requires the consent of a coalition of investors –

controlling coalition- a situation which dilutes the power, protects the minority and

prevents entrenchment collusion with the manager; this influences corporate

performance positively (Bennedsen and Wolfenzon, 2000). Finally, it is worth stressing

that ownership concentration as a governance mechanism is closely connected with a

firm’s financial structure as well as its dividend policy. In concentrated ownership

structures, blockholders may force managers to use debt as a financial instrument in

order to avoid the dilution of ownership concentration. Also, blockholders have enough

power to demand a large payout ratio. Both instruments, debt (Berger et al. 1997), and

dividends (Farinha, 2003) can complement (Zwiebel, 1996) or substitute (Jensen, 1986)

each other in order to constrain managerial discretion and prevent entrenchment.

10

Lastly, a firm can mitigate the problem of agency by setting up independent

committees that will decide promotions and will enforce the adequate incentive schemes

for managers (e.g. option-like compensation schemes). The Securities and Exchange

Commission (SEC) considers committees as important tools to monitor corporate

activities; it requires all of its registered companies to disclose whether or not they have

an audit committee, a compensation committee or a nominating committee. Past

evidence supported the idea that the existence and composition of these committees are

positively related to effective decision-making (Vafeas, 1999; Anderson and Reeb,

2004). However, committees under the influence of top management, such as those

composed in the majority by insiders, are more likely to harm minority shareholder

interests (Vafeas, 1999).

In summary, this analysis suggests that organizations may suffer from

managerial misbehavior and that several strong governance structures can hinder

managers from extracting private benefits by implementing entrenchment strategies

(Combs and Skill, 2003). In the next section, we propose stakeholder activism as

another control mechanism to monitor and control managers.

Stakeholder activism

Manager’s decision can have a direct impact on all stakeholder groups and,

consequently, the manager can be viewed as the stakeholders’ agent, and not just the

shareholders’ agent (Hill and Jones, 1992; Jones, 1995). Therefore, a firm is not

conceived as a bilateral relationship between shareholders and managers but as a

multilateral set of relationships amongst stakeholders (Aguilera and Jackson, 2003).

Each stakeholder has, in turn, their own interests, which generally are in conflict with

other stakeholders’. Certainly, one of the most important conflicts of interest is between

managers and all other stakeholders; this defines an amplified agency problem (Hill and

Jones, 1992). This divergence of interests is problematic because it prevents

stakeholders from maximizing their utility and may lead to the emergence of power

differentials among stakeholders. Moreover, managers may use such differentials to

further entrench their position and modify the firm’s institutional structures to their

advantage (Hill and Jones, 1992). In this context, stakeholders articulate different

responses to restrict management power.

A primary response may be to reward or punish management’s actions as a

means of influencing their behavior (Rowley and Berman, 2000). Boycotts and

lobbying are some examples of these actions (Baron, 2001; Feddersen and Gilligan,

11

2001; John and Klein, 2003). By wielding the threat of having to endure costly boycotts

and media campaigns, stakeholders exert a substantial controlling influence over firms.

Another channel of influence is via the board of directors. Luoma and Goodstein

(1999) demonstrated that some institutional characteristics shape the stakeholder

representation on boards of directors. Typically, these boards include representation

from labor, creditors, and regulatory agencies (Schneper and Guillén, 2004). Under this

control structure, managerial decisions are therefore monitored and influenced by the

presence of stakeholders’ representatives (Luoma and Goodstein, 1999). Consequently,

this particular composition of a corporate board is likely to affect organizational

outcomes and processes (Wood, 1991), in such a way that organizations have to satisfy

the needs of a wide variety of stakeholders (Jones, 1995), including both shareholders

and non-shareholders.1

Finally, the actions of stakeholders may affect the threat of a hostile takeover

and with that the own CEO replacement. In this sense, Schneper and Guillén (2004)

show that the frequency of hostile takeovers is inversely related to stakeholders’ (non-

shareholder) power, and this result may explain why countries labeled as stakeholder-

oriented like Germany or Japan are characterized by the low occurrence of hostile

takeovers. This is further evidence that, in such countries, markets behave inefficiently

and do not fully penalize those companies that satisfy stakeholders’ interests for

entrenchment purposes. Thus, incumbent managers have enough motivation to

committing themselves to socially responsible behavior, aimed at gaining stakeholders’

support. Once executives have built close relationships with key stakeholders, they may

prove difficult to remove by shareholders. Additionally, over time, executives can gain

additional power by controlling the information revealed to stakeholders, thereby

reinforcing their entrenchment position (Walsh and Seward, 1990).

1This stakeholder-centered model of corporate governance is popular in Germany and Japan (Schneper

and Guillén, 2004). In Germany, for example, there are two types of limited liability companies, the Gesellschaft mit beschräukten Haftung (GMBH) which is not listed, and the Akitiengesellschaft (AG), which is a listed company. Both types face the legal obligation of creating a dual level board: The first level is the supervisory board, or Aufsichstrat, which is entrusted to monitor the managers and, the second is the management board or Vorstand, responsible for the daily management of the firm. Remarkably, workers’ representatives and those members elected by shareholders, and who are not employees of the firm, have an equal number of seats in the supervisory board. Directors elected by the shareholders come from other commercial and financial firms which have a long-term, close relationship with the firm and, as such, they are equivalent to the external directors of US firms. It is important to stress that the stakeholder-centered model of corporate governance places greater emphasis on internal mechanisms, such as boards of directors that reflect broad stakeholder participation, in order to discipline managers.

12

3/ HYPOTHESES

In the previous sections, we have argued that the incentives for the efficacy of

the board of directors in monitoring and controlling management can be compromised

by a variety of entrenchment practices such as poison pills, shark repellents or

greenmail and/or disincentives linked to income smoothing. Also, we have explained

that the independence of the board (including the non-dual role of the CEO), the

concentration of ownership and the existence of different boards subcommittees can be

effective disciplinary devices to resolve the agency problem. Moreover, stakeholder

activism may be another corporate governance device. In fact, we argue that, under

some circumstances, stakeholders have the incentives and the power to monitor

managers closely. However, even if these internal corporate governance mechanisms

are in place, managers’ incentives to seek entrenchment can lead them to find ways to

escape these controls. Moreover, managers will have particular incentives to do so when

these latter corporate governance mechanisms are well developed. In that case they need

powerful entrenchment strategies. One of these strategies is to canvass support from

stakeholders so as to channel their efforts to the entrenched manager’s own advantage.

This strategy enjoys the benefit of diminishing the pressure from activist stakeholders

while at the same time opposing the pressure from shareholders that have more

difficulty in controlling a manager that has other stakeholders as allies. To

operationalize such behavior, managers may engage in a broad array of practices to

create and manage relationships with corporate stakeholders and the natural

environment, the so-called corporate social performance (CSP).

A particular objective of this paper is to distinguish the entrenchment

perspective of CSP from other competing explanations. Previous literature on

stakeholder theory has provided normative, instrumental and descriptive/empirical

explanations of CSP (Donaldson and Preston, 1995). The central tenet of normative

stakeholder theory is that “the economic and social purpose of the corporation is to

create and distribute increased wealth and value to all its primary stakeholders groups,

without favoring one group at the expense of others” (Clarkson, 1995: 112), in order to

ensure that each primary stakeholder group continues as a part of the corporation.

Therefore, CSP is considered to be an end. In contrast to this normative perspective,

instrumental theory conceives stakeholder management as a means to achieve the

ultimate end of the firm, i.e. marketplace success (Jones, 1995). The basic assumption

behind this theory is that the CSP may be an organizational resource that would lead to

13

more efficient or effective use of resources, which in turn has a positive impact on

corporate financial performance (Orlitzky et al., 2003).

Our entrenchment arguments for improving CSP clearly do not fit in with the

ethical/moral or instrumental arguments discussed above. Managerial self-interested

utilization of CSP is explained by the stakeholders’ power to influence the firm. As

such, our story belongs to the descriptive realm of stakeholder theory (Mitchell et al.,

1997; Jawahar and McLaughlin, 2001). According to this theory, the degree to which

managers assign priority to competing stakeholders claims – the stakeholder salience –

is positively related to the cumulative number of stakeholders attributes of power,

legitimacy, and urgency; considering power as the stakeholders’ capacity to influence

the firm’s behavior; legitimacy as the perception that, within a social system of norms,

values and beliefs, the actions undertaken by someone are desirable, proper, or

appropriate; and urgency is the degree to which a stakeholder’s claim calls for

immediate attention (Mitchell et al., 1997). Based on Schneper and Guillén’s (2004)

findings, we contemplate power as a function of each stakeholder role in corporate

governance. Then, a manager who wants to implement an entrenchment strategy will

want to be protected against the actions of powerful stakeholders. In that case, there are

two possibilities: collaboration or confrontation.

Jones (1995) and Hill and Jones (1992) supported the confrontation strategy.

Jones (1995), in deriving implications of his instrumental theory, suggested that

decreases of CSP are connected to a managerial entrenchment strategy. In a similar

vein, Hill and Jones (1992) predicted that managers will undertake strategic actions to

reduce stakeholder power – strategies that negatively affect corporate efficiency. For

example, management reduces customer power through product and/or market

diversification; community power by delocalizing the production outside the regional

boundaries; or employee power through several practices of production organization

and bureaucratic mechanisms (p. 147).

Our claim, following Cespa and Cestone (2004) is exactly the opposite, we

hypothesize that stakeholders and incumbent managers will be natural allies against

non-controlling shareholders and potential raiders, particularly when there are efficient

internal corporate governance mechanisms capable of preventing managerial

entrenchment impulses. In that case collaboration with stakeholders cannot be blocked

by shareholders easily, on the basis of a “suspicious” entrenchment strategy. This

further stimulates managers’ incentives to improve a firm’s CSP with entrenchment

intentions. Hellwig (2000) pointed out that managers set on entrenchment will find

14

allies in stakeholder sectors such as the political system, labor, the media, the judiciary,

and even in the universities, against outside shareholders. For the incumbent manager,

the advantages of an alliance with stakeholders are even more obvious given that

stakeholders dispose powerful tools for promoting or disgracing top executives.

DeAngelo and DeAngelo (1998) and Hellwig (2000) quoted several examples in which

unions, local communities, the media, or customers acted as “white squires” to block

hostile takeovers.

In addition, the implementation of expensive policies aimed at improving a

firm’s CSP reduces its attractiveness to a raider. Generous long-term contracts with

workers and suppliers, as well as long-term commitments to support environmental or

philanthropic organizations are too heavy a burden to borne by a raider (Pagano and

Volpin, 2005). These concessions, however, are not fully internalized in share prices for

two reasons. First, as stakeholder theory researchers (e.g., Berman et al., 1999) have

already demonstrated, maintaining good relationships with key stakeholders creates an

organizational resource that leads to more effective use of a firm’s resources; this has a

positive impact on financial performance. Therefore, in a context of information

asymmetries, capital markets will be unable to determine if social concessions are a

means of improving financial performance by generating a valuable organizational

resource or if they are part of an entrenchment strategy (market inefficiency). And

second, as mentioned before, social concessions related to the implementation of an

entrenchment strategy are triggered mainly in the presence of strong internal corporate

governance mechanisms. The existence of such strong internal corporate governance

mechanisms, together with the market inefficiency assumption, hinder any steep

reductions in share prices. This lower discount explains the managers’ strategy of

satisfying stakeholders’ interests as a way of reinforcing their entrenchment strategy.

Therefore, anticipating the impact of generous CSP initiatives in terms of

reductions in stakeholder activism as well as reductions in the pressure from existing

shareholders and potential raiders, has the following consequences: entrenched

managers, especially when closely monitored as a result of internal corporate

governance mechanisms, will make a commitment to follow socially responsible

behavior that attracts stakeholders’ support. Thus, our first hypothesis is:

15

Hypothesis 1. Managerial entrenchment practices have a positive impact on a

firm's social performance. Moreover, this effect is pronounced in those firms

with efficient internal corporate governance mechanisms like those with boards

that have a large proportion of independent directors; a CEO different from the

chairman of the board; different board subcommittees; a performance-based

managerial compensation scheme and concentrated ownership.

It is important to state that internal corporate governance mechanisms, while

playing a positive moderating role by controlling the connection between entrenchment

and CSP, also constrain managerial discretion. This should hinder the implementation

of expensive social responsible activities. That is, we also expect a negative direct effect

of the internal governance mechanisms on a firm’s CSP.

Types of stakeholders: employees

According to Mitchell et al. (1997), the degree to which managers give priority

to competing stakeholder claims, the so-called salience, is positively related to the

stakeholder attributes perceived by managers. In the previous section, we have

distinguished between the salience of shareholders versus other stakeholders. Now, we

explore the salience of a particular stakeholder group, the employees.

Within the managerial entrenchment strategy, workers constitute one of the

stakeholder groups that receive preferential attention by the manager because they have

the capacity to influence a firm’s behavior and, at the same time, share common

interests with incumbent managers. As a consequence it is likely that entrenched

managers, particularly those that face the pressure from efficient internal corporate

governance mechanisms, will commit themselves to giving employees more

concessions (e.g. workers’ stake)

The capacity of employees to influence decision-making, organizational

arrangements and performance outcomes is well documented in previous literature (see

for example, Scheneper and Guillén, 2004). This power is derived from political action

or legal mechanisms that are at their disposal. By political action we mean that workers

may lobby against/in favor of an incumbent CEO by demonstrating, mobilizing

politicians, appealing to the media, and constituting organized pressure groups like trade

unions (Pagano and Volpin, 2005). Several studies illustrate, for example, that

employees tend to oppose hostile takeovers (Scheneper and Guillén, 2004). In addition,

16

the employees’ power to promote or disgrace top executives is amplified when they

have institutionalized legal mechanisms at their disposal. One of the most direct means

by which employees and other stakeholders protect their individual interests in the

corporation, is through the presence of stakeholder directors on corporate boards, or

board subcommittees such as the audit, compensation, executive, and nominating

committees (Luoma and Goodstein, 1999). Furthermore, workers can directly affect the

likelihood of CEO replacement through individual share ownership.

Importantly, employees are not only a powerful stakeholder group, but are

natural allies of managers set on entrenchment. There is a vast literature that

demonstrates that hostile takeovers have negative consequences for workers (Aguilera

and Jackson, 2003). In countries with low employment protection, a hostile takeover

may result in job cuts and, sometimes, causes a worsening in the overall working

conditions. For example, successful raiders tend to renegotiate the labor contracts that

already exist, cutting wages to a minimum and stepping up monitoring to maintain

workers’ effort (Conyon et al., 2001).

In addition to the previous arguments, managers may be interested in colluding

with employees, not only to gain their support, but also to reduce a firm’s attractiveness

to any potential raiders who may be particularly interested in companies with efficient

internal corporate governance mechanisms. In such cases, employment policy is likely

to be used to deter hostile takeovers, particularly in a context of inefficient financial

markets that are more likely to be found in non-Anglo-Saxon countries. In these

countries, social concessions to workers are more commonplace and difficult to reverse.

Given this scenario, generous long-term wage contracts reduce the raider’s ability to

generate profit after the takeover and, therefore, make the acquisition less attractive

(Pagano and Volpin, 2005).

These arguments suggest the following hypothesis:

Hypothesis 2. Managerial entrenchment practices have a positive influence on

employees’ satisfaction, this effect is pronounced in those firms with efficient

internal corporate governance mechanisms like those with boards that have a

large proportion of independent directors with; a CEO different from the

chairman of the board; different boards subcommittees; managerial

performance-based compensation mechanism and concentrated ownership.

17

Performance analysis

The instrumental approach is an important perspective of stakeholder theory

(Donaldson and Preston, 1995). It advocates the formulation and implementation of

processes that satisfy stakeholders because they control key resources and suggests that

stakeholder satisfaction, in turn, will ensure the long-term survival and success of the

firm (Freeman, 1984; Hillman and Keim, 2001). Accordingly, stakeholders that own

resources relevant to the firm’s success will be more willing to offer their resources to

the extent that their different claims and needs are fulfilled (Strong et al., 2001).

Therefore, under this approach we expect that stakeholder satisfaction leads to higher

commitment, greater effort, and, ultimately, to superior performance (Hosmer, 1994;

Stevens et al., 2005). Thus, stakeholder management has strategic value from a “means

to an end” perspective (Berman, et al., 1999), which is opposed to the intrinsic value of

the normative approach.

However, consistent with our previous propositions, we argue that when

managers implement entrenchment practices, improvements in CSP may complement

such practices and reinforce the (negative) effect of entrenchment on shareholders’

value. Note that entrenchment decreases the efficiency of external control mechanisms,

the hostile takeover, which reduces the pressure on managers and affects detrimentally a

firm’s financial performance. Some studies like Walsh and Seward (1990), or more

recently, Sundaramurthy (2000) show that capital markets react negatively to the

adoption of these practices. Walsh and Seward (1990) also reached the same conclusion

using accounting-based measures of performance.

Importantly, Sundaramurthy et al. (1997) and Sundaramurthy (2000) suggest

that the strength or weakness of internal monitoring mechanisms, such as the board

structure and the ownership concentration, moderate the relationship between anti-

takeover provisions and shareholders’ wealth. Remarkably, we have suggested in

previous hypotheses, that a manager trying to insulate himself from internal monitoring

mechanisms may also follow a generous policy of social concessions. Therefore, these

managerial concessions to stakeholders, especially in the context of existing internal

governance mechanisms that are efficient, should play a moderating role in the

connection between the implementation of entrenchment practices and financial

performance. McWilliams and Siegel (2001) termed these types of concessions as

discretionary CSP and pointed out that it is negatively related to shareholders’ wealth.

Therefore, we expect this moderating role to be negative. Stated formally:

18

Hypothesis 3: Managerial entrenchment practices when combined with social

concessions, have a negative impact on financial performance. This is

particularly evident when social concessions are triggered in a context where

there are efficient internal governance mechanisms.

Finally, we expect that this result also holds when we focus on the specific

dimension of social performance; workers’ satisfaction. This is so because, on the one

hand, social concessions to workers are particularly costly and, on the other hand, they

strongly reinforce the entrenchment position of the manager before shareholders given

the saliency of these stakeholders to achieve the firm’s success. This stakeholder

characteristic is particularly attractive in a context where efficient governance

mechanisms aimed at preventing entrenchment are present. In such situations, the

reinforcement of social concessions, mainly to salient stakeholders like workers, will be

needed to complement an entrenchment strategy. Hence, the negative impact of

entrenchment on performance will be more evident when combined with workers’

satisfaction:

Hypothesis 4: Managerial entrenchment practices, when combined with social

concessions to workers, have a particularly negative impact on financial

performance, especially in a context where there are efficient internal

governance mechanisms

4. METHODS

4.1. Sample and Variables

Our sample is composed of 777 industrial firms from 27 different countries that

are included at least once in the 2002-2005 SiRi PRO TM database.2 This is compiled by

the Sustainable Investment Research International Company (SiRi) – the world’s largest

company specializing in socially responsible investment analysis, based in Europe,

North America, and Australia. SiRi comprises eleven independent research institutions,

such as KLD Research and Analytics Inc. in the USA or Centre Info SA in Switzerland.

They provide detailed profiles of the leading international corporations. Companies are

analyzed according to their reporting procedures, policies and guidelines, management

2 The distribution of firms among the countries with the largest proportion is as follows: US (29.5%); UK

(15.17%); Switzerland (10.8%); France (8%); Germany (6.4%); Japan (5.4%); Netherlands (3.6%), Spain

(3.2%), Italy (3%).

19

systems, and key data. This information is extracted from financial accounts, company

documentation, international databases, media reports, interviews with key

stakeholders, and ongoing contact with management representatives.

The firm’s rating contains 199 information items that cover all major stakeholder

issues such as community involvement, environmental impact, customer policies,

employment relations, human rights issues, activities in controversial areas (e.g.

alcohol), supplier relations, and corporate governance. We complement these data on

corporate responsibility with financial data from 2000-2005, extracted from OSIRIS.

This is a database compiled by Bureau van Dijk (BvD) and provides information on

financials, ownership, earnings, and stock data for the world’s publicly traded

companies from over 130 countries. Also, it provides current and historical information

on 38,000 companies worldwide, including several thousand unlisted and de-listed

companies. Finally, in order to compute one of the variables of performance (the

abnormal returns), we use information provided by Bloomberg on the MSCI world

index, in US dollars (monthly data).

4.2. Measures

Corporate social performance (CSP) is notoriously difficult to operationalize

(Aupperle et al., 1985) because it is a multi-dimensional construct (Carroll, 1979) that

captures a wide range of items, ideally, one for each relevant stakeholder (Waddock and

Graves, 1997). Until quite recently, many studies used variables associated with only

one stakeholder (Wood and Jones, 1995) as a proxy for the multidimensional construct

of CSP as it relates to stakeholders. However, with the emergence of data that covers

numerous stakeholders, particularly the KLD data, some of this problem has been

corrected (see, e.g., Hillman and Keim, 2001; Berman et al., 1999; Waddock and

Graves, 1997).

In the present study, we have used the SiRi PRO TM ratings, which include

research fields similar to the KLD data and associated with stakeholders. Five research

fields are devoted to measuring the level of a firm’s responsibilities to its stakeholders:

community, customers, employees, environment, and vendors and contractors. Another

section provides an overview of firms’ corporate governance practices. However, we

have excluded this part from our measure of CSP, because in our study we focus on the

degree of satisfaction of the non-shareholders’ stakeholders. Next, each one of these

research fields is evaluated separately and rated on scales ranging from 0 (worst) to 100

(best). This is the score of each stakeholder. Finally, each stakeholder score is weighted

20

according a registered methodology developed by SiRi. These weights are uniform

within the sector the firm belongs to. Finally, the corporate social performance

indicator used in this study is the corresponding SiRi measure defined as the weighted

sum of non-shareholders stakeholder scores (the scores of community, customers,

employees, environment, and vendors and contractors), using the SiRi’s weights.3

Workers’ satisfaction (Workers): We approach this issue using another variable

provided by SiRi. In this case, the SiRi analysts assess the level of a firm’s

responsibilities to its employees, building an aggregate score for this particular

stakeholder group. This score is an aggregation of 37 indicators that cover different

aspects of the firm’s involvement in workers’ issues. These indicators are grouped into

four broad areas: the level of a firm’s transparency or disclosure of information on

worker’s issues; the existence of corporate policies and principles for employees; the

importance of management procedures; and the level of dispute in their relations with

employees.

Performance. This is captured by three different proxies in order to provide

more robust results. First, the return on assets (ROA), which is the ratio of earning

before interest and tax to total value of assets. Second, Tobin’s q, which is defined on a

log scale as the ratio of the sum of equity value plus a firm’s liabilities to the total

assets. Last, we use the abnormal returns that are computed through a single factor

model, using as a factor the MSCI world index in US dollars. More specifically, we

have detracted the expected monthly returns derived from a CAPM model using a

window of 5 years (monthly data) from the firm’s equity return in the last month of the

corresponding year.

Managerial entrenchment. As discussed in the theoretical section, managers

intending to insulate themselves from external monitoring may follow several

entrenchment practices. In our case we use different measures of entrenchment provided

by the SiRi PROTM database that approximates the: existence of anti-takeover devices;

limitation of shareholders’ voting rights; existence of multiple classes of stock with

different voting rights; managers’ stake and tenure.4 More specifically, Anti-takeover is

3 Visit www.siricompany.com for more information. An example of a detailed profile can be found at

www.centreinfo.ch/doc/doc_site/SP-Novartis-06.pdf. For more information on the SiRi ProTM database,

visit the webpage www.ais.com.es/ingles/productos/derivados.htm#1.4 In the robustness section, we introduce another measure that we explain there based on income

smoothing.

21

a dummy that it is equal to 1 if the firm has implemented any of the following takeover

measures: voting caps, increased voting rights over time, restrictions on board

appointment rights, and poison pills. Shareholders_Rights is measured by a 3-point

Likert scale; the highest value of this item corresponds to the situation in which major

controversies have impact on the rights and treatment of shareholders, for example,

governance arrangements that affect detrimentally the interests of shareholders; insider

trading scandals involving company directors; or major conflicts of interest among

board members. The intermediate value indicates the existence of controversies, both

major and minor, but where the company has taken credible steps to resolve the

problem; and finally, the lowest value indicates that the SiRi analyst did not find any

information on controversies involving shareholders of the company. The variable

OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes

of stock with different voting rights and 0 otherwise. Concerning managerial

ownership, the literature shows that this variable can also be used as an entrenchment

device. Stulz (1988) presents a model where high managerial ownership can preclude

the possibility of a takeover. Also, along this line, Weston (1979) reports that firms

where the insiders’ stake is larger than 30% have never been acquired by a hostile

takeover. Morck et al (1988), McConnel and Servaes (1990) and recently De Miguel et

al. (2004) show the existence of a non-linear relationship between managerial

ownership and firm performance, where entrenchment appears in the middle of the

range for managerial stake. We follow the specification shown in De Miguel et al.

(2004) and we estimate Tobin’s q in terms of managerial ownership; its quadratic term

and cubic terms. In the definition of managerial ownership we follow Farinha (2003)

and we include the holdings of pension funds, bearing in mind that managers can

exercise some influence in the composition of such a stake. As controls we also

incorporate size, leverage and investment, as defined below. The results show that the

relationship between Tobin’s q and managerial ownership is decreasing in the range

between 17% and 69.8%.5 Hence we characterize this entrenchment region with a

dummy (Managown_Entrench) that is equal to 1 when managerial stake is in this

region; and 0, otherwise. Finally, in some specifications, we have used, instead of

managerial ownership, a more direct measure of entrenchment that is also positively

related to managers’ holding; this is the manager’s tenure. The larger the manager’s

5 In De Miguel et al. (2004) the range found using Spanish data is between 30% and 70%. This is not

surprising given the large ownership concentration of Spanish firms.

22

stake, the less likely is the possibility of their replacement and the longer the managerial

tenure. More specifically, Manager_tenure is defined as a dummy that takes the value

of 1 if the directors’ term in office (including managers) is more than three years, and 0

otherwise. We choose this threshold because Fredrickson et al (1988) show that a

disproportionately large number of CEOs have tenure that lasts three years or less.

Then, a CEO with tenure larger than three years is a potential candidate to become a

manager that follows entrenchment practices.

In addition to providing results using each indicator separately, we also

aggregate the scores for these five indicators. The resulting score is labeled as

Entrenchment. For the sake of robustness, this variable has also been computed using

the principal components of the aforementioned indicators and the results found remain

the same, qualitatively. Finally, in some specifications, we define a variable

DΔEntrenchment which is equal to 1 when variations of entrenchment are larger than

the mean of the sector for the corresponding year and country; and 0, otherwise.

Internal corporate governance mechanisms: According to our framework, we

control for the strength of internal corporate governance. We use different variables to

approach this issue: the independence of the board of directors, the separation of the

roles of CEO and chairman, the existence of board subcommittees, the existence of a

board performance evaluation system and the presence of large shareholders.

Board_Independence is SiRi’s Likert-type variable that takes three different values

contingent on the percentage of independent directors with respect to the mean value of

the sector. The highest value corresponds to the situation in which a majority of non-

executive directors are considered independent; the intermediate value indicates that

50% or less of non-executive directors are independent; and when the information

disclosed by the company does not allow us to determine the share of independent non-

executive directors, the firm receives the lowest value.6 Non-dual_CEO is a SiRi’s

dummy variable that it is equal to 1 when the chairman is not the CEO, and 0 otherwise.

6 According to the SiRi questionnaire, a director is not considered independent when assessing business

and other relationships may result in a conflict of interest. SiRi guidelines are very precise on the

possibility of conflicting of interests. This precludes, among other things a director that has held an

executive position within the company group; and/or is a director or employee of a company in which the

company has a notifiable holding; and/or is, or was recently, employed by, or on the board, of a

significant customer or supplier to the company; and/or has had an association with the company of more

23

Audit_Committee, Nomination_Committee, and Remuneration_Committee are 3

dummy variables obtained from Siri; each one receiving the value 1 if such a committee

exists with independent members. Due to the high correlation among these variables,

we define Control_Committee as a 4-point Likert scale in which 0 represents the

absence of any one of these committees and 3, the joint presence of audit, nomination,

and remuneration committees.

Performance_Evaluation is another SiRi’s Likert-type scale variable that can

take three different values; it takes the value of 1 when two conditions are met: there is

a board performance evaluation system, and there are no controversies on executive

payments like payment unrelated to performance; the existence of golden parachutes;

the repricing of options, or excessive pension benefits. This variable has a value of zero

when one of the previous conditions is not met and, finally, its value is equal to -1 if

none of the previous two conditions is satisfied.

Another mechanism for internal control is the role played by large shareholders.

We studied this issue by employing measures of stakes in the hands of government, and

ownership concentration. State_Ownership is the percentage of ownership in the hands

of the State; and Ownership_Concentration is the stake of the three largest

blockholders, where we have excluded the manager’s stake in order to avoid

confounding effects with such a variable that is used as a proxy of entrenchment. 7

Finally, in order to study the existence of differential effects of entrenchment on

CSP contingent on the aforementioned corporate governance mechanisms, we define

the following variables that characterize situations with strong internal corporate

governance mechanisms. DBoard_Independence as a dummy that it is equal to 1 if the

variable of Board_Independence is larger than the mean of the sector for the

corresponding year and country; and 0 otherwise. Following the same logic we define

DNon-dual; DControl_Committee; DPerformance_Evaluation, DState_Ownership and

DOwnership_Concentration.

Control variables: We control for financial structure, dividends, size, firms’

age, performance, investment, growth opportunities, industry, country and year. To

control for the financial structure, we use the variable Debt that measures the gearing of

than nine years; and/or is related through blood, marriage or equivalent to other directors or advisers to

the company. 7 Alternatively, we have used the stake of the largest shareholder and that of the five largest and the

results remain qualitatively the same.

24

the company. This is defined as the ratio of non-current liabilities plus loans, to

shareholders funds. Dividend is the long-term dividend policy. This is characterized by

the sector average pay-out ratio times the firm’s after-tax profits. We have used this

sectoral component in order to reduce potential endogeneity problems given that

stakeholder satisfaction has a clear impact on the proportion of funds transferred to

shareholders (pay out ratio). Size is the fixed-asset value on a log scale; and Age is the

number of years of the company’s existence. We include the aforementioned return on

assets (ROA) as a control for financial performance in the specifications of CSP as

dependent variable. Investment is the ratio of fixed assets to total assets. Growth is equal

to 1 when the rate of increase in sales is larger than the value for the corresponding

sector and year, and 0 otherwise. Intangibles is the ratio of intangible assets to total

fixed assets. Finally, we introduce temporal and sectoral dummies. The control of

countries is addressed through four dummies that capture the origin of their legal codes.

We follow La Porta et al. (1988) and we distinguish four types of countries according to

where the country’s company/commercial law is derived from (i) British common law,

(ii) French civil law, (iii) German civil law and (iV) Scandinavian civil law.

4.3. Empirical Analysis

In our empirical application, we rely on two basic specifications that we estimate

using robust regressions, one explaining variations in CSP and the other explaining

variations in performance. In order to explain a firm’s CSP as well as workers’

satisfaction and test Hypotheses 1 and 2, we consider the following basic specification

(that also includes temporal, sectoral and the legal-origin country dummies):

1 1 2 3

4 5 6

7 8 9 10

11

Entrenchment Control_Committee

Non-dual_CEO + Board_Independence Performance_Evaluation

State_Ownership + Ownership_Concentration Debt Dividend

Si

it it it

it it it

it it it it

CSP

12 13 14 15 16ze Age ROA Investment Growth Intangiblesit it it it it it it

[1]

Where represents changes in the variable between one period and the

previous one. In analyzing the possible differential effect of entrenchment strategies, we

conducted further estimations of specification [1] by breaking the variable

Entrenchment into its basic components: Anti-takeover, Shareholders_Rights,

OneShare_OneVote; Managown_Entrench and Manager tenure. Also, to study the

possible moderating effect of the variables of governance, we crossed the

25

aforementioned variable of significant variation in entrenchment (D Entrenchment)

times the aforementioned dummies that characterize those situation where the

mechanisms of corporate governance are well developed. This generates the following

variables: DBoard_Independence*D Entrenchment; DNon-dual*D Entrenchment ,

DControl_Committee*D Entrenchment; DPerformance_Evaluation*

D Entrenchment; DState_Ownership*D Entrenchment and

DOwnership_Concentration*D Entrenchment. We focus on significant variations of

entrenchment because we expect that these variations will appear mainly as a reaction to

the existence of strong internal control mechanisms. It is in these situations where we

expect the stronger impact on a firm’s CSP (Hypothesis 1).

It is important to stress that by estimating in differences, we eliminate the

unobservable heterogeneity that may be potentially correlated with the independent

variables. For example, the intrinsic characteristics of the manager should condition a

firm’s CSP policy and at the same time may be connected to the intensity of the

entrenchment actions employed, as well as with the governance characteristics of the

firm. Additionally, as we have explained in the theoretical section, we expect that

pressure from different stakeholders is connected to internal corporate control

mechanisms. That is, an endogeneity problem in specification [1], which it is not

directly connected to the unobservable heterogeneity, may exist perfectly. We tackle

this problem by advancing the dependent variable by one period.

In order to test Hypothesis 2, we have modified slightly the previous

specification by substituting the variable of CSP with that of Workers’ satisfaction

(Workers).

The second specification is aimed at explaining the variations in the corporate

financial performance. The basic specification is as follows (that also include temporal,

sectoral and legal-origin country dummies):

1 1 2 3 4

5 6 7

8 9 10

Entrenchment Entrenchment_

Control_Committee Non-dual_CEO + Board_Independence

_ State_Ownership + Ownership_Concent

it it it it

it it it

it it

Performance CSP CSP

Perform Evaluation

11 12 13 14 15 16

17 18

Debt Dividend Size Age ROA Investment

Growth Intangibles

it

it it it it it it

it it it

[2]

From this specification, it is possible to test Hypotheses 3 and 4 for whether a

variation in entrenchment when combined with a variation in CSP (Hypothesis 3) or in

workers’ satisfaction (Hypothesis 4) have further negative effects on financial

26

performance. This test requires the inclusion of an interaction term between CSP and

entrenchment practices, the variable ΔEntrenchment_ΔCSP for testing Hypothesis 3 or

the equivalent variable ΔEntrenchment_ΔWorkers for testing Hypothesis 4. Also,

consistent to what we have mentioned in the theoretical part, we expect the

complementarity between both variables (entrenchment and CSP) will appear mainly in

a context with efficient governance mechanisms where managers set on entrenchment

will reinforce their strategy with the implementation of social responsible actions. In

such situations, we expect intensive managerial entrenchment complemented with

improvements in socially responsible actions. Thus, in some specifications (columns 4

and 5 of Tables 5B and 5C) we use the interactive variable DΔEntrenchment_ΔCSP that

crosses the aforementioned dummy DΔEntrenchment, which is equal to 1 when

variations of entrenchment are larger than the mean of the sector for the corresponding

year and country, with a variable of variations in CSP (ΔCSP). Following the same

logic, we define the variable DΔEntrenchment_ΔWorkers. Then, the complementarity

hypotheses would be supported if the coefficient of such a variable is negative.

5. RESULTS

Table 1 provides the descriptive statistics of the relevant variables of the model8,

while Table 2 displays the correlation matrix. On inspection of the correlation matrix,

we find that variations in CSP are positively correlated with those of entrenchment.

More specifically, variations in CSP are positively correlated with increases in anti-

takeover initiatives, with deterioration of shareholders’ rights and changes in managerial

ownership. Remarkably, these increases in CSP and workers’ satisfaction are also

negatively correlated with variations in different internal control mechanisms like the

existence of independent audit, nomination, and remuneration committees or the

implementation of performance evaluation compensation schemes. Also, debt and

dividends hinder improvements in a firm’s social performance. We interpret these

8The number of observations is not the same for all variables because our sample is an incomplete panel

data. There are 542 observations for which we have information on all variables. Once we take into

account differences, as we do in the regression analysis, and eliminate some outliers using the method of

Hadi (1992, 1994), this number reduces to 302 observations for the specifications of CSP, and 245 for the

specifications of financial performance. The lower number of outliers in the specification of CSP is

explained in terms of the persistence in the values of this variable along time. Finally, in the specification

of Table 6 that uses information on accruals, the number of observations reduces to 231 and 207

respectively, depending on the measure used.

27

results as evidence that when entrenchment is more difficult, as a result of an increase in

the aforementioned internal control mechanisms, increases in CSP are also less likely.

However, we show in the next table that, if firms are able to improve significantly their

entrenchment strategy within a framework where internal governance mechanisms are

well developed, although unlikely, this is accompanied by significant increases in a

firm’s CSP.

----------------------------------------Insert Tables 1 and 2 about here----------------------------------------

Table 3 summarizes the regression analysis of specification [1], whereby we test

the effect of managerial entrenchment practices on CSP (all the variables are

normalized). Further, we break the variable Entrenchment into its five basic

components: Anti-takeover, Shareholders_Rights, OneShare_OneVote,

Managown_Entrench and Manager_tenure.

----------------------------------------Insert Table 3 about here

----------------------------------------

This table shows that variations in entrenchment measures in period t have a

positive impact on variations in CSP in period t+1 (column 1). This is particularly

evident for anti-takeover measures as well as for those actions that limit the one vote-

one share principle, which are standard initiatives triggered by an entrenched manager

as we have explained in the theoretical section. Also, when managers’ stake moves into

the “entrenchment region” (between 17% and 69.8%), it has a positive impact on

variations in CSP. This also applies to managerial tenure (see column 2). A second

result shows that when entrenchment is more difficult, due to an increase in the number

of independent control committees or the existence of a performance evaluation system,

there is a negative impact on variations in CSP (significant at 1% level). This reflects

the complementarity idea between entrenchment and CSP that this paper relies on.

Moreover, once we look at the coefficient of the interactive variables, we find that

variations in entrenchment have a particular positive impact on variations in CSP in

those scenarios where entrenchment is more difficult ex-ante. Mainly, when the number

of control committees with independent members is larger than the mean for the sector,

year and country, as well as when the performance evaluation systems are more

developed than the average for the corresponding sector, year and country. In such

28

cases, corporate governance is more developed and, if the manager is set on

entrenchment, he complements this strategy with increases in CSP. This conforms to

Hypothesis 1, where collusion with non-shareholder stakeholders is more likely in those

scenarios where shareholders can control the manager closely.

In addition, we find that a generous dividend policy hinders the implementation

of policies aimed at satisfying non-shareholder stakeholders. Finally, consistent with the

formulation of the instrumental stakeholder theory (Freeman, 1984), we find that

intangible assets have a positive impact on CSP.

Concerning the results when we focus on workers’ satisfaction (see Table 4),

they are, in essence, the same as those for the overall score of CSP. This conforms to

Hypothesis 2. However, there is a difference once we focus on the specific

entrenchment policies that we study (anti-takeover measures, limitation of shareholders’

rights and the existence of different classes of shares). In particular, the results for

workers’ satisfaction are only significant when managers implement anti-takeover

measures. This is in accordance with Pagano and Volpin (2005), where wage

concessions – that improve workers’ satisfaction- are described as anti-takeover

initiatives. We can explain the lack of a significant effect from the limitation of

shareholder rights because workers may well have shares in the firm. In fact, the non-

significant sign that we obtain may be the result of compensating the positive effect due

to managerial concessions to workers with the negative effect due to the limitation of

shareholder rights.

Finally, it is worth stressing that the presence of the state as a blockholder

stimulates practices aimed at improving a firm’s CSP and workers’ satisfaction, in

contrast with other corporate governance mechanisms. We can argue that the state may

be less interested in preventing entrenchment practices if they are associated with

improvements in stakeholders’ satisfaction. This is so because satisfied stakeholders,

principally workers, may eventually become constituent members of the current party in

government and, firms with state participation have all the incentives to treat them well.

----------------------------------------

Insert Table 4 about here

----------------------------------------

To analyze the ex-post consequence of implementing entrenchment initiatives

combined with increases in CSP, we show in Tables 5A, 5B and 5C, the results of

estimating specification [2] using different measures of performance. In Table 5A, we

29

use variations in ROA; in Table 5B, variations in Tobin’s q; and in Table 5C, variations

in the abnormal returns derived from a single factor model.9

----------------------------------------

Insert Table 5A about here

----------------------------------------

Once we focus on the estimations for ROA, we find that variations in

entrenchment have a negative impact on variations in ROA. Also, in columns 2 and 3,

we obtain that these variations combined with variations in CSP and/or workers’

satisfaction also have a negative impact on variations in ROA as a measure of financial

performance. This result also holds when we focus on relative variation of ROA

compared with the mean of the sector (column 4 and 5). That is, there is a

complementary relationship between both variables: when there is an increase in social

performance, the marginal effect of the entrenchment on the returns on assets is more

negative.

----------------------------------------

Insert Table 5B about here

----------------------------------------

The results of estimating the variations on Tobin’s q and variations in the

abnormal returns (Tables 5B and 5C respectively) conform to those using ROA.

However, there is a difference given that the complementary connection between both

variables - CSP and entrenchment - is only shown when the variation in entrenchment is

high enough (larger than the mean for the sector, year and country –

DΔEntrenchment=1). We have argued previously that this corresponds to situations of

large internal control, where we expect that the large variations in entrenchment trigger

the kind of socially responsible policies that reinforce the power of the manager. This

generates a decrease in performance. In an unreported estimation, we investigated this

issue in more depth and crossed the variable of entrenchment with a dummy that is

equal to 1 when there has been a significant change in CSP (larger than the mean of the

sector and year). The result shows that only in such situations, can entrenchment have a

9 Given the endogeneity between market measures of performance and ownership structure (Demsetz and

Villalonga, 2001), we have detracted from the variable of entrenchment the component on managerial

ownership in those estimations that use market measures of performance (Tables 5B and 5C). Note that

one of the components of entrenchment (Managown_Entrench) is defined in terms of a decreasing

relationship between managerial ownership and Tobin’s q.

30

negative impact on performance. Hence, it seems that using market measures of

performance, the complementarity between entrenchment and CSP only holds in

extreme situations that correspond to significant variations in these variables. Only in

such cases do markets interpret negatively the variations of CSP as being connected to

an entrenchment strategy. Finally, this argument also works when we focus on

variations in workers’ satisfaction.

Concerning the alternative market performance measures (abnormal returns), the

results also conform to those using the Tobin’s q. Then, there is a substantial reduction

in performance when variations in entrenchment are connected to variations in CSP.

----------------------------------------

Insert Table 5C about here

----------------------------------------

Given the previous set of results, we provide support for the thesis regarding the

existence of an entrenchment motive to justify substantial variations in CSP. When

managerial control mechanisms are well developed, entrenchment is less likely.

However, when it does occur, it is so intensive as to also trigger socially responsible

actions. This is a reinforcing mechanism, given that managerial collusion with non-

shareholder stakeholders allows them to channel stakeholders’ power against

shareholders’ power. In this situation there is a clear negative impact on financial

performance.

6/ ROBUSTNESS

6.1/ Income Smoothing

In order to investigate the robustness of our results, in line with our theoretical

arguments, we maintain that earnings manipulation, and income smoothing in

particular, is an additional entrenchment mechanism. Managers smooth earnings as a

natural entrenchment strategy in order to ensure a stable stream of profits that will

satisfy shareholders (Fudenberg and Tirole, 1995). However, although earnings

manipulations improve financial performance in the short-term, they damage the

medium-term interests of shareholders. The manager anticipates this fact and has all the

incentives to trigger entrenchment initiatives. Within this setting, we characterize a

situation where we expect to find, according to our theory, an increase in CSP. This is a

way to test the robustness of our results.

31

To test this contention, we use two alternative measures of income smoothing.

First, we approach this variable through the correlation between changes in accruals and

changes in cash flow (Incsmooth1), where

Accruals CA Cash CL STD DEP , with ΔCA is the change in current

assets; ΔCash is the change in cash; ΔCL is the change in current liabilities; ΔSTD is

the change in debt included in current liabilities; and DEP is the depreciation and

amortization. The second measure (Incsmooth2) is the ratio of net income before

extraordinary items to the standard deviation of cash-flows. In columns 1 and 2 of Table

6 we use the first measure while in columns 3 and 4 we use the second measure.

----------------------------------------

Insert Tables 6 about here

----------------------------------------

From this table, we obtain that those firms that smooth earnings show increases

in CSP as well as in workers’ satisfaction. It is important to stress that this is not due to

the increase in the short-term financial performance due to earnings manipulation,

because we have controlled by a performance variable through the ROA. Our

explanation, which supports our basic theory, is that income smoothing is connected

with entrenchment practices that may further stimulate improvements in CSP. This

reinforces the basic claim of the paper.10

6.2 Expropriation of minority shareholders

We have conducted an additional analysis to investigate if the changes of CSP

and in workers’ satisfaction are explained in terms of entrenchment practices or they are

a strategy implemented by large shareholders to expropriate minority shareholders. In

the corporate governance literature (e.g. Shleifer and Vishny, 1997), the presence of

blockholders is contemplated to have an ambiguous effect on firm’s financial

performance; on the positive side, blockholders diminish the entrenchment possibilities

of managers, which impacts positively on performance; but, on the negative side, large

shareholders may expropriate minority shareholders thereby reducing the market price

of shares. One strategy that blockholders may follow to expropriate minorities is the

10 We have conducted an alternative analysis (available upon request) where earnings manipulation is

approached through the discretionary accruals obtained as the difference between the accruals and the

expected ones given by the models of Jones (1991), Dechow et al. (1995) and Kothari et al (2005)

respectively. The results are consistent with those found using the proxies for income smoothing.

32

overinvestment in socially friendly policies (Barnea and Rubin, 2006). By

implementing certain social programs, blockholders receive the full benefits associated

with CSP, but only bear a portion of the costs to implement such policies (proportional

to their stakes). This association between ownership concentration and CSP found

support in some recent studies (Barnea and Rubin, 2005; Neubaum and Zahra, 2006).

Keeping in mind this idea, it is important to distinguish the changes of CSP that

are connected to the entrenchment practices from those that may be explained in terms

of the implementation of expropriating strategies. To do so, in Table 7, we present the

results of the basic specification [1], once we distinguish firms according to their

ownership concentration. This is proxied by the stake of the three largest blockholders.

----------------------------------------

Insert Table 7 about here

----------------------------------------

Remarkably, we have found that the impact of our proxy of entrenchment on

variations in CSP and on workers’ satisfaction (last two columns) has only positive

effects in those firms where ownership is not concentrated (lower than the mean of the

sector). In these firms –columns 2 and 4-, we expect that there is no expropriation to

minority shareholders. This allows ensuring that the effect found on variation in CSP

and in workers’ satisfaction is explained exclusively in terms of entrenchment and not

in terms of expropriation. This gives further support to our results.

6.3/ Separating the sample by leverage and by the degree of financial

markets efficiency

As a robustness check, we conduct two additional analyses. First, we split the

sample into those firms whose leverage is larger than the mean of their corresponding

sector and country. The results, reported in the first two columns of Table 8 show that

the effect of entrenchment on CSP is only visible for low-leverage firms. We can argue

that debt constrains the available resources which then hinders investments in socially

responsible activities as well as the implementation of entrenchment policies. Hence, in

more leveraged firms, we should obtain, as we do, a weaker connection between

entrenchment and CSP.

----------------------------------------

Insert Table 8 about here

----------------------------------------

33

A second analysis is performed by comparing firms whose country of origin has

an Anglo-Saxon legal code, to those whose countries have a German, or French-origin

legal code. We have argued previously that financial markets are not fully efficient and

do not fully internalize the costs of implementing costly socially responsible actions

linked to entrenchment. This fact, among other things, explains the use of socially

responsible actions as an entrenchment device, even in the presence of external

corporate governance mechanisms. A consequence of this is that the results connecting

entrenchment and CSP should be more evident in countries with less developed

financial markets. As La Porta et al (1998) shows, countries with French and German-

origin legal codes have less developed financial markets compared with Anglo-Saxon

ones. Consistently, we find that entrenchment only explains our measure for CSP in

countries whose legal codes have French or German origins. Note also, that countries

with less developed external corporate control mechanisms are also those countries with

more developed internal corporate control mechanisms. In such cases, according to our

Hypothesis 1, a much stronger impact of entrenchment on CSP should be expected. This

is exactly what we find.

7. CONCLUSIONS

In this paper, we investigate the effect of the implementation of entrenchment

strategies on socially responsible actions. Entrenchment practices are approached by the

existence of: anti-takeover devices;, the limitation of shareholders’ voting rights; the

existence of multiple classes of stock with different voting rights; managerial stake;

managers’ tenure and the implementation of income smoothing policies. Our basic

premise following Cespa and Cestone (2004) is that the manager may be controlled by

shareholders – externally through the financial markets and internally through the board

of directors – as well as by the activism of different stakeholders that have enough

power to shape the firm’s decision-making. In such a scheme, entrenchment strategies

aimed at hindering the actions of shareholders cannot be implemented, unless

accompanied by other measures tailored to neutralize stakeholders’ pressure. In this

case, there are two possibilities: confrontation with stakeholders (a strategy is similar to

an entrenched manager at odds with shareholders); or collusion with stakeholders in

order to satisfy their interests. Our basic claim is that an entrenched manager will

choose the collusion strategy especially in those situations where internal corporate

governance mechanisms are well developed and there is little slack for a manager to set

on entrenchment. In such cases, the collusion with non-shareholder stakeholders not

34

only will it tackle the pressure from stakeholders but, more importantly, it allows

channeling the salience of these stakeholders against agents -shareholders- who intend

to replace the manager.

We tested this claim by looking at the connection between variations in different

entrenchment proxies and variations in the scores of corporate social performance

(CSP) as well as on workers’ satisfaction. We find that there is a clear positive impact

by the former on the latter, especially in those firms with: efficient internal corporate

governance mechanisms like the existence of independent control committees and the

implementation of performance evaluation schemes. This confirms the main theoretical

contention of the paper where a firm’s CSP is an integral part of a manager’s definition

of an entrenchment strategy.

Additionally, we prove the robustness of this contention by using an alternative

proxy of entrenchment that is the implementation of earnings manipulation practices

based on income smoothing (Fudenberg and Tirole, 1995). Our results are fully

consistent with the theoretical claim.

A second type of analysis looks at the consequences of implementing

entrenchment strategies on financial performance; we find that its negative impact on

financial performance is more pronounced when combined with the implementation of

socially responsible actions. Moreover, we have shown, using market measures of

performance that this result is particularly true for the large variations of entrenchment

that we expect as the reaction to the existence of stringent internal governance

mechanisms. In such a situation, although entrenchment is less likely, we argue that

when a manager is set on entrenchment he implements an intensive strategy that it is

complemented with the development of socially responsible policies. This is further

evidence that entrenched managers heed stakeholder satisfaction not only as a

consequence of stakeholder activism, but also as a way of reinforcing their

entrenchment strategy against the shareholders.

Remarkably, our result also holds true when we focus on one of the non-

shareholder stakeholders, the workers. We argue that these stakeholders are amongst the

most powerful and, consequently, the entrenched manager should pay particular

attention to looking after their interests.

Finally, we have shown that the results are more pronounced in countries with

less efficient financial markets (civil-law countries) that are also those countries with

more developed internal corporate control mechanisms. In such countries, managers can

take advantage of that inefficiency and over-invest in socially responsible activities in

35

order to complement an entrenchment strategy, given that market prices do not fully

reflect the cost of implementing these activities. Also, we discard other explanations

linked to the implementation of expropriating policies in order to explain the increases

in social performance. We have proved that the increases in CSP linked to entrenchment

only appear when ownership concentration is low. Expropriation is not found within

such a framework; therefore, the increases in CSP are due, exclusively, to entrenchment.

7.1 Implications

This work forms a bridge between the corporate governance literature and

stakeholder theory. According to this latter line of research, the management of

stakeholders is a good way of improving financial results (Jones, 1995; Donaldson and

Preston, 1995), whereas corporate governance emphasizes the difficulty in reconciling

the demands of a wide set of stakeholders (Jensen, 2001; Tirole, 2001). We show that

trying to satisfy different stakeholders’ interests, independently of their salience, may

have bad consequences on performance when combined with the implementation of

entrenchment policies. Then, it is not a good policy to give the managers any leeway in

determining the degree of satisfaction of non-shareholder stakeholders because

managers may define a socially responsible policy strategically to complement an

entrenchment strategy. Furthermore, we find that the existence of strong internal

corporate governance mechanisms like independent boards or external committees for

different control issues is not a guarantee that CSP may not be used in this pervasive

way. In fact, the linkage between CSP and entrenchment is stronger for those firms with

strong corporate governance mechanism. Then, how does one deal with such a

problem?

There are different possibilities. First, following Cespa and Cestone (2004), if a

firm’s CSP may be used as an integral part of an entrenchment strategy, then, some

form of governance mechanism that hinders managerial discretion on social issues is

needed. A possibility is to regulate social issues in order to avoid overinvestment in

socially responsible actions. Undoubtedly, mandatory accounting practices to reflect

these issues on the balance sheet may be a first step in this direction. A second way to

prevent entrenchment problems, especially when involving other stakeholders, is to

transfer ownership to this group. For example, if workers also have shares in a firm, it

will not make sense for the manager to implement simultaneously an entrenchment

strategy of confrontation with shareholders and another of collusion with other

36

stakeholders, like workers because the interests of the latter will also be aligned with

those of shareholders. Paradoxically, the interests of shareholders are better defended by

transferring part of their powers to other stakeholders.

7.2 Future research

A natural extension of our work will be to focus on more specific dimensions of

a firm’s CSP in order to study exactly who the most relevant stakeholders are for

managers who implement an entrenchment strategy. We suggest in the paper that one of

the most important groups consists of the workers, but a further analysis of this idea is

necessary. Another extension consists of investigating, in greater depth, the connection

between ownership structure and the entrenchment motive behind some socially

responsible activities. The type of blockholder and its social sensitivity is expected to

have a significant influence on the reaction of stakeholders to the manager’s

implementation of entrenchment practices with a social dimension. Finally, a further

investigation on different institutional contexts may also be of interest given the

significant differences that exist in top-management orientation across countries. In

Anglo-Saxon countries, managers are more inclined to satisfy shareholders’ interests,

while in Continental Europe and Japan, managers have traditionally been more sensitive

to the development of long-term relationships with employees, banks, and suppliers.

The investigation of this as well as other issues is left for future research.

37

REFERENCES

Agle B.R., Mitchell, R.K. and Sonnenfeld, J.A. (1999), “Who matters to CEOs? An

investigation of stakeholder attributes and salience, corporate performance, and

CEO values”, Academy of Management Journal, 42(5), pp. 507-525.

Aguilera R.V. and Jackson, G. (2003), “The cross-national diversity of corporate

governance: Dimensions and determinants”, Academy of Management Review,

28(3), pp. 447-465.

Anderson, R.C. and Reeb, D.M. (2004), “Board composition: Balancing family

influence in S&P 500 firms”, Administrative Science Quarterly, 49, pp. 209–237.

Aupperle, K., Carroll, A. and Hatfield, J. (1985), “An empirical examination of the

relationship between corporate social responsibility and profitability”, Academy of

Management Journal, 28(2), pp. 446-63.

Barnea, A. and Rubin, A. (2006), “Corporate Social Responsibility as a Conflict

Between Shareholders” (March 10, 2006). EFA 2006 Zurich Meetings Available

at SSRN: http://ssrn.com/abstract=686606

Baron, D.P. (2001), “Private politics, corporate social responsibility, and integrated

strategy”, Journal of Economics and Management Strategy, 10, pp. 7-45.

Bennedsen, M. and Wolfenzon, D. (2000), “The balance of power in closely held

corporations”, Journal of Financial Economics, 58, pp. 113-139.

Berger, P., Ofek, E. and Yermack, D. (1997), “Managerial entrenchment and capital

structure decisions”, Journal of Finance 52, pp. 1411-1438.

Berman, S., Wicks, A., Kotha, S. and Jones, T. (1999), “Does stakeholder orientation

matter? The relationship between stakeholder management models and firm

financial performance”, Academy of Management Journal 42(5), pp. 488-506.

Carroll, A.B. (1979), “A three-dimensional conceptual model of corporate social

performance”, Academy of Management Review, 4(1), pp. 497-505.

Cespa, G. and Cestone, G. (2004), “Corporate Social Responsibility and Managerial

Entrenchment”, CEPR Discussion Paper No. 4648.

Clarkson, M. (1995), “A stakeholder framework for analyzing and evaluating corporate

social performance”, Academy of Management Review, 20, pp. 92-117.

Combs, J.G. and Skill, M.S. (2003), “Marginalist and human capital explanations for

key executive pay premiums: A contingency perspective”, Academy of

Management Journal, 46(1), pp. 63-73.

38

Conyon, M.J., Girma, S., Thompson, S. and Wright, P.W. (2001), “Do hostile mergers

destroy jobs?”, Journal of Economic Behavior and Organization, 45, pp. 427-440.

Coombs, J.E., and Gilley, K.M. (2005) “Stakeholder management as a predictor of CEO

compensation: main effects and interactions with financial performance”,

Strategic Management Journal, 26, pp. 827-840.

Dahya, J., A. Lonie and Power, D. (1998), “Ownership structure, firm performance and

top executive change: an analysis of UK firms”, Journal of Business Finance and

Accounting 25, pp. 1089-1118.

De Miguel, A., Pindado, J., and de la Torre, C. (2004), “Ownership structure and firm

value: New evidence from Spain”, Strategic Management Journal, 25(12), pp.

1119-1207.

DeAngelo, H. and DeAngelo, L. (1998), “Ancient redwoods and the politics of finance:

The hostile takeover of the Pacific Lumber Company”, Journal of Financial

Economics, 47(1), pp. 3-53.

Dechow, P., Sloan, R. and Sweeney, A. (1995), “Detecting Earnings Management”, The

Accounting Review, 70, pp. 193-225.

Demsetz, H., and Villalonga, B. (2001), "Ownership Structure and Corporate

Performance" Journal of Corporate Finance 7 (3), pp. 209-233.

Denis, D., Denis, D. and Sarin, A. (1997), “Ownership structure and top executive

turnover”, Journal of Financial Economics 45, pp. 193-221

Donaldson, T.L. and Preston, L.E. (1995), “The stakeholder theory of the corporation:

Concepts, evidence, and implications”, Academy of Management Review, 20(1),

pp. 65-91.

Eisenhardt, K.M. (1989), “Agency theory: an assessment and review”, Academy of

Management Review, 14, pp. 57-74.

Fama, E. and Jensen, M. (1983), “Separation of ownership and control”, Journal of Law

and Economics, 26(2), pp. 301-25.

Farinha, J., (2003), “Dividend policy, corporate governance and the managerial

entrenchment hypothesis: an empirical analysis”, Journal of Business Finance and

Accounting, 30 (9&10), pp. 1173-1209.

Feddersen, T. and Gilligan, T. (2001), “Saints and markets: Activists and the supply of

credence goods”, Journal of Economics and Management Strategy, 10, pp. 149-

171.

Fredrickson. J. W. Hambrick, D. C. and Baumrin. S. (1988), “A model of CEO

dismissal”, Academy of Management Review, 13, pp. 255-270.

39

Freeman, R.E. (1984). Strategic management: A stakeholder approach. (Pitman: Boston,

MA).

Fudenberg, D. and Tirole, J. (1995), “A Theory of Income and Dividend Smoothing

Based on Incumbency Rents”, Journal of Political Economy, 103, pp. 75-95. DOI:

10.1086/261976

Hadi, A. S. (1994), “A Modification of a Method for the Detection of Outliers in

Multivariate Samples”, Journal of the Royal Statistical Society 56(2), pp. 393-396.

Hadi, A. S. (1992), “Identifying Multiple Outliers in Multivariate Data”, Journal of the

Royal Statistical Society 54(3), pp. 761-771.

Hellwig, M. (2000). “On the economics and politics of corporate finance and corporate

control”. In: Vives, X. (Ed.), Corporate Governance: Theoretical and Empirical

Perspectives. Cambridge University Press, Cambridge, England, pp. 95–136.

Hill, C.W. and Jones, T.M. (1992), “Stakeholder-agency theory”, Journal of

Management Studies, 29, pp. 131-154.

Hillman, A.J. and Keim, G.D. (2001), “Shareholder value, stakeholder management,

and social issues: What’s the bottom line?”, Strategic Management Journal,

22(2), pp. 125-139.

Hosmer, L.T. (1994), “Strategic planning as of ethics mattered”, Strategic Management

Journal, 15, pp. 17-34.

Jawahar, I.M. and McLaughlin, G.L. (2001), “Toward a descriptive stakeholder theory:

An organizational life cycle approach”, Academy of Management Review, 26(3),

pp. 397-414.

Jensen, M.C. (2001), “Value maximization, stakeholder theory, and the corporate

objective function”, Journal of Applied Corporate Finance, 14, pp. 8-21.

Jensen, M., (1986), “Agency costs of free cash-flow, corporate finance and takeovers”,

American Economic Review 76, pp. 323-329.

Jensen, M. and Meckling, W. (1976) “Theory of firm: Managerial behavior, agency cost

and capital structure”, Journal of Financial Economics, 3, pp. 305-360.

Jensen, M. and Ruback, R. (1983), “The market for corporate control: The scientific

evidence”, Journal of Financial Economics, 11, pp. 5-50.

John, A. and Klein, J. (2003), “The boycott puzzle: Consumer motivations for purchase

sacrifice”, Management Science, 49(9), pp. 1196-1209.

Jones, J. (1991) “Earnings Management during Import Relief Investigations”, Journal

of Accounting Research, 29, pp. 193-228.

40

Jones, T.M. (1995) “Instrumental stakeholder theory: A synthesis of ethics and

economics”, Academy of Management Review, 20(2), pp. 404-437.

Johnson, R.A. and Greening, D.W. (1999), “The effects of corporate governance and

institutional ownership types on corporate social performance”, Academy of

Management Journal, 42(5) pp. 564-576.

Kassinis, G. and Vafeas, N. (2002) “Corporate boards and outside stakeholders as

determinants of environmental litigation”, Strategic Management Journal, 23, pp.

399-415.

Kassinis, G. and Vafeas, N. (2006), “Stakeholder pressures and environmental

performance”, Academy of Management Journal, 49(1), pp. 145-159.

Kothari, S.P., Leone, A.J. and Wasley, C.E. (2005), “Performance Matched

Discretionary Accrual Measures”, Journal of Accounting and Economics, 39, pp.

163-197.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A. and Vishny, R. (1998), “Law and

Finance,” Journal of Political Economy 106, 1113-1155.

Luoma, P. and Goodstein, J. (1999), “Stakeholders and corporate boards: institutional

influences on board composition and structure”, Academy of Management

Journal, 42(5), pp. 553-563.

McConnell, J. J., and Servaes, H. (1990), “Additional evidence on equity ownership and

corporate value”, Journal of Financial Economics, 27, pp. 595-612.

McGuire, J., Dow, S., and Argheyd, K. (2003), “CEO incentives and corporate social

performance”, Journal of Business Ethics, 45, pp. 341-359.

McWilliams A, Siegel D. (2001) “Corporate social responsibility: A theory of the firm

perspective”, Academy of Management Review 26(1), pp. 117-127.

Mitchell, R.K., Agle, B.R. and Wood, D.J. (1997) “Toward a theory of stakeholder

identification and salience: Defining the principle of who and what really counts”,

Academy of Management Review, 22(4), pp. 853-886.

Morck, R., Shleifer, A. and Vishny, R. (1988), “Managerial ownership and market

valuation”, Journal of Financial Economics 20, pp. 293-315.

Neubaum, D.O. and Zahra, S.A. (2006), “Institutional ownership and corporate social

performance: The moderating effects of investment horizon, activism, and

coordination”, Journal of Management, 32, pp. 108-131.

Orlitzky, M., Schmidt, F.L. and Rynes, S.L. (2003), “Corporate social and financial

performance: A meta-analysis”, Organization Studies, 24(3), pp. 403-441.

41

Pagano, M. and Volpin, P. (2005) “Managers, workers, and corporate control”, The

Journal of Finance, 60, pp. 841-868.

Pound, J. (1988), “Proxy contests and the efficiency of shareholder oversight”, Journal

of Financial Economics, 20, pp. 237–265.

Rowley, T. and Berman, S. (2000), “A brand new brand of corporate social

performance”, Business and Society, 39(4), pp. 397-418.

Schneper, W.D. and Guillén, M.F. (2004). “Stakeholder rights and corporate

governance: A cross-national study of hostile takeovers”, Administrative Science

Quarterly, 49(2), pp. 263-295.

Shleifer, A. and Vishny, R.W. (1997), “A survey of corporate governance”, Journal of

Finance, 52(2), pp. 737-83.

Shleifer, A. and Vishny, R. (1989), “Management entrenchment: The case of manager-

specific investments”, Journal of Financial Economics, 25, pp. 123-140.

Shleifer, A. and Vishny, R. (1986), “ Large shareholders and corporate control”,

Journal of Political Economy, 94(3), pp. 461-88.

Stevens, J.M., Steensma, H.K., Harrison, D.A. and Cochran, P.L. (2005) “Symbolic or

substantive document? The influence of ethics codes on financial executives'

decisions”, Strategic Management Journal, 26, pp. 181-195.

Strong, K.C., Ringer, R.C., and Taylor, S.A. (2001), “The ruled of stakeholder

satisfaction (timeliness, honesty, empathy”, Journal of Business Ethics, 32(3), pp.

219-230.

Stulz, R. (1988), “Managerial control of voting rights: financing policies and the market

for corporate control”, Journal of Financial Economics 20, pp. 25-54.

Sundaramurthy, C. (2000), “Antitakeover provisions and shareholder value

implications: A review and a contingency framework”, Journal of Management,

26(5), pp. 1005-1030.

Sundaramurthy, C., Mahoney, J.M. and Mahoney, J.T. (1997), “Board structure,

antitakeover provisions, and stockholder wealth”, Strategic Management Journal,

18, pp. 231-245.

Tirole, J. (2001), “Corporate governance”, Econometrica 69(1), pp. 1-35.

Vafeas, N. (1999), “The nature of board nominating committees and their role in

corporate governance”, Journal of Business Finance and Accounting, 26(1), pp.

199-225.

Waddock SA. (2004), “Parallel universes: Companies, academics, and the progress of

corporate citizenship”, Business and Society Review, 109(1), pp. 5-42.

42

Waddock, S.A. and Graves, S.B. (1997), “The corporate social performance-financial

performance link”, Strategic Management Journal, 18(4), pp. 303-319.

Walsh, J.P. and Seward, J.K. (1990), “On the efficiency of internal and external

corporate control mechanisms”, Academy of Management Review, 15(3), pp. 421-

458.

Weston, J. F. (1979), “The tender takeover”, Mergers and Acquisitions, 15, pp. 74-82.

Wood, D.J. (1991), “Corporate social performance revisited”, Academy of Management

Review, 16(4), pp. 691-718.

Wood, D.J., and Jones, R.E. (1995), “Stakeholder mismatching: A theoretical problem

in empirical research on corporate social performance”, International Journal of

Organizational Analysis, 3(3), pp. 229-267.

Yeo, G. H. H., Tan, P. M.S., Ho, K. W., and Chen, S. (2002), “Corporate Ownership

Structure and the Informativeness of Earnings”, Journal of Business Finance &

Accounting, 29 (7&8), pp. 1023-1040. doi:10.1111/1468-5957.00460.

Zwiebel, J. (1996), “Dynamic Capital Structure Under Managerial Entrenchment”,

American Economic Review 86 (5), pp. 1197-1215.

43

TABLE 1: Descriptive Statistics

Variable Obs Mean Std. Dev. Min MaxCSP 1703 48.075 13.753 3.516 80.14Workers 1703 53.792 15.308 0 96.875Anti-takeover 1304 0.609 0.488 0 1OneShare_OneVote 1304 0.748 0.434 0 1Shareholders_Rights 1304 0.926 0.262 0 1Managerial_Ownership 711 4.144 15.219 0 100Managown_Entrench 711 0.048 0.214 0 1Workers_Ownership 711 0.44 4.14 0.00 85.10Manager_Tenure 1304 0.379 0.485 0 1Entrenchment 655 1.339 0.705 0 4ROA 2632 4.936 9.038 -207.470 69.690Tobin’s_q (log) 1406 -0.083 1.384 -3.050 8.127Abn_Returns 2394 -0.004 1.337 -23.049 41.443Control_Committee 1304 1.352 1.403 0 3Non-dual_CEO 1304 0.889 0.314 0 1Board_Independence 1304 0.814 0.366 0 1Performance_Evaluation 1304 0.243 0.472 -1 1State_Ownership 711 2.698 9.405 0 89.65Ownership_Concentration 711 24.349 20.571 0 100Debt 959 111.645 148.686 0 974.68Dividend 810 959344 2781009 0 24800000Size 980 15700000 29300000 31822 392000000Age 4116 69.628 21.732 1 339Investment 980 0.603 0.534 0 5.314Growth 4116 0.040 0.196 0 1Intangibles 909 0.324 0.248 0 1

TABLE 2: Correlation Matrix

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 211 ΔCSP 1.0002 ΔWorkers 0.718*3 ΔROA 0.063 0.0784 ΔEntrenchment

0.134* 0.087 -0.0455 ΔAnti-takeover

0.210* 0.241* 0.010 0.427*6 ΔOneShare_OneVote

0.119* 0.089 0.055 0.416* 0.0657 ΔShareholders_Rights

0.056 -0.023 -0.003 0.419* -0.073 0.0368 Managown_Entrench

0.162* 0.152* 0.019 -0.155* -0.008 0.013 -0.0909 Manager_Tenure

-0.092 -0.094 -0.070 0.360* -0.307* -0.046 0.115* 0.02110 ΔControl_Committee

-0.244* -0.284* -0.076 0.008 -0.323* -0.009 0.007 -0.047 0.285*11 ΔNon-dual_CEO

-0.084 -0.033 -0.011 -0.132* -0.049 -0.047 -0.081 -0.041 -0.171* -0.06812 ΔBoard_Independence

0.040 -0.068 -0.061 0.072 -0.025 0.001 0.127* -0.027 0.043 -0.036 0.1034*13 ΔPerformance_Evaluation

-0.239* -0.255* -0.034 0.031 -0.242* -0.040 0.047 -0.010 0.233* 0.417* -0.078 -0.02414 ΔState_Ownership

0.085 0.145* 0.015 -0.129* 0.081 -0.099 -0.028 -0.008 -0.149* -0.091 -0.030 -0.014 -0.01115 ΔOwnership_Concentration

0.035 0.056 0.059 -0.095 0.001 -0.061 0.065 0.155* 0.122* 0.061 0.044 0.011 0.092 0.02216 Debt

-0.106* -0.037 -0.042 -0.148* 0.046 0.058 -0.091 0.051 -0.237* -0.033 0.016 -0.077 -0.085 -0.048 0.05017 Dividend

-0.327* -0.255* -0.114* 0.064 -0.085 -0.122* 0.148* -0.048 0.227* 0.228* -0.253* 0.027 0.216* 0.025 -0.049 -0.09318 Size

-0.333* -0.219* -0.018 -0.030 -0.026 -0.078 0.085 -0.084 0.039 0.191* -0.181* 0.033 0.157* 0.009 0.021 0.294* 0.601*19 Age -0.182* -0.024 -0.020 -0.073 0.010 0.056 -0.001 -0.035 -0.079 -0.023 -0.073 -0.037 -0.027 -0.096 -0.036 0.117* 0.040 0.03720 Investment

0.013 -0.092 0.039 0.049 0.064 -0.022 -0.053 -0.008 0.012 -0.067 -0.080 -0.085 0.011 0.025 -0.111* -0.102* 0.015 -0.136* -0.114*21 Growth 0.055 -0.006 0.026 -0.058 -0.057 0.1161* -0.045 0.119* -0.155* -0.016 -0.045 -0.056 -0.044 -0.191* -0.065 0.035 -0.037 -0.030 0.110* -0.0522 Intangibles

0.378* 0.310* 0.1737* 0.014 0.011 0.063 0.018 0.077 -0.010 0.034 -0.035 -0.007 0.014 0.093 0.072 -0.128* -0.152* -0.176* -0.133* -0.07 0.006* Significant at 1% level.

TABLE 3: The impact of Entrenchment on Corporate Social Performance (CSP)Table 3 reports the results of conducting robust regressions on the variations (Δ) of CSP. The variation of the dependent variable is led by one period. The variable CSP is the score provided by SiRi for non-shareholder stakeholders’ degree of satisfaction. Anti-takeover is a dummy that it is equal to 1 if the firm has implemented any anti-takeover measures. OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes of stock with different voting rights. Shareholders_Rights is a variable that takes three values (0, 0.5, 1) depending on the degree of limitation of shareholders’ voting rights (see text for details).Managown_Entrench is a dummy that is equal to 1 when managerial stake has a value between 17% and 69.8%. In this region there is a negative relationship between managerial stake and Tobin’s q (see text for details). Manager_Tenure is defined as a dummy that takes the value of 1 if the directors’ (including managers) term in office is more than three years old. Then, Entrenchment=Anti-takeover + OneShare_OneVote + Shareholders_Rights + Managown_Entrench + Manager_Tenure. Control_committee is the sum of the following three variables: Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committee is a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committee is a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value for the sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of the performance evaluation system (see text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake held by the three largest blockholders. The crossed variable DControl_Committee*DΔEntrenchment is the result of multiplying DΔEntrenchment (that is equal to 1 when variations of entrenchment are larger than the mean for the sector for the corresponding year and country), with a dummy DControl_Committee that is equal to 1 if Control_Committee is larger than the value for the sector in the corresponding year and zero otherwise. Following the same pattern, we define DNon-dual_CEO*DΔEntrenchment; DBoard_Independence*DΔEntrenchment; DPerformance_Evaluation* DΔEntrenchment; DState_Ownership* DΔEntrenchment; DOwnership_Concentration *DΔEntrenchment. Debt is the ratio of non-current liabilities plus loans to shareholders’ funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Age is the number of years of the company’s existence, ROA is the EBITDA to the total assets. Investment is the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is larger than the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangible assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).ΔAnti-takeover 0.959** (1.840) 1.043** ( 1.820)ΔOneShare_OneVote 0.578** (1.960) 0.636** (2.130)ΔShareholders_Rights 0.737 (1.480) 0.737 (1.460)Managown_Entrench 0.780** (2.200)Manager_Tenure 0.347** (1.900)ΔEntrenchment 1.004*** (2.310) 1.017** (1.980)ΔControl_Committee -1.446** (-2.200) -1.831***(-2.790) -1.964*** (-2.970) -2.006*** (-3.020)ΔNon-dual_CEO -0.570 (-1.550) -0.474 (-0.880) -0.543 (-1.460) -0.537 (-1.220)ΔBoard_Independence 0.434 (0.940) 0.425 (1.150) 0.459 (0.980) 0.637 (1.290)ΔPerformance_Evaluation -2.096*** (-3.390) -2.151*** (-3.620) -2.078*** (-3.470) -2.076*** (-3.470)ΔState_Ownership 0.684*** (3.180) 0.731*** (3.220) 0.668*** (2.880) 0.688*** (3.130)ΔOwnership_Concentration 0.007 (0.020) 0.114 (0.270) 0.265 (0.600) 0.401 (0.930)DControl_Committee*DΔEntrenchment 1.425** (2.090)DNon-dual_CEO*DΔEntrenchment -0.135 (-0.260)DBoard_Independence*DΔEntrenchment 0.431 (0.880)DPerformance_Evaluation*DΔEntrenchment 0.920** (2.020)DState_Ownership*DΔEntrenchment -0.604 (-1.050)DOwnership_Concentration*DΔEntrenchment 0.208 (0.450)Debt -0.454 (-0.700) -0.587 (-0.600) -0.376 (-0.560) -0.634 (-0.880)Dividend -2.212*** (-3.000) -2.361*** (-3.040) -2.244*** (-2.980) -2.402*** (-3.120)Size 0.024 (0.030) -0.091 (-0.050) -0.038 (-0.050) -0.079 (-0.100)Age -0.181 (-0.720) -0.190 (-0.450) -0.162 (-0.630) -0.096 (-0.370)ROA 0.196 (0.280) 0.456 (1.180) 0.475 (0.660) 0.510 (0.710)Investment 0.550 (0.770) 0.782 (1.170) 0.438 (0.620) 0.800 (1.030)Growth 0.689*** (2.570) 0.727*** (2.950) 0.728*** (2.770) 0.729*** (2.830)Intangibles 2.448*** (4.690) 2.575*** (5.490) 2.433*** (4.470) 2.686*** (4.750)Intercept 1.147 (0.810) -1.625 (-0.80) -1.801 (-0.820) -0.296 (-0.160)R2

48.31% 48.51% 47.83% 49.03%Fitness of the model (F test) 15.85 (0.000) 14.41 (0.000) 14.40 (0.000) 8.74 (0.000)Number of observations 302 302 302 302***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values

46

TABLE 4: The Impact of Entrenchment on Workers’ SatisfactionTable 4 reports the results of conducting robust regressions on the variations of workers’ satisfaction. The variations (Δ) of the dependent variable is lead by one period. The variable Workers is the score provided by SiRi of the degree of workers’ satisfaction. Anti-takeover is a dummy that it is equal to 1 if the firm has implemented any anti-takeover measures. OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes of stock with different voting rights. Shareholders_Rights is a variable that takes three values (0, 0.5, 1) depending on the degree of limitation of shareholders’ voting rights (see text for details). Managown_Entrench is a dummy that is equal to 1 when managerial stake has a value between 17% and 69.8%. In this region there is a negative relationship between managerial stake and Tobin’s q (see text for details). Manager_Tenure is defined as a dummy that takes the value of 1 if the directors’(including managers) term in office is more than three years old. Then, Entrenchment=Anti-takeover + OneShare_OneVote + Shareholders_Rights+ Managown_Entrench + Manager_Tenure. Control_committee is the sum of the following three variables: Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committeeis a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committee is a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value for the sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of the performance evaluation system (see text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake held by the three largest blockholders. The crossed variable DControl_Committee*DΔEntrenchment is the result of multiplying DΔEntrenchment (that is equal to 1 when variations of entrenchment are larger than the mean for the sector for the corresponding year and country), with a dummy DControl_Committee that is equal to 1 if Control_Committee is larger than the value for the sector in the corresponding year and zero otherwise. Following the same pattern, we define DNon-dual_CEO*DΔEntrenchment; DBoard_Independence*DΔEntrenchment; DPerformance_Evaluation* DΔEntrenchment; DState_Ownership* DΔEntrenchment; DOwnership_Concentration *DΔEntrenchment. Debt is the ratio of non-current liabilities plus loans to shareholders’ funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Age is the number of years of the company’s existence, ROA is the EBITDAto the total assets. Investment is the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is largerthan the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangible assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).ΔAnti-takeover 1.371** (1.950) 1.593*** (2.310)ΔOneShare_OneVote 0.627 (1.220) 0.610 (1.200)ΔShareholders_Rights 0.258 (0.320) 0.108 (0.140)Managown_Entrench 0.880*** (2.440)Manager_Tenure 1.232** (2.040)ΔEntrenchment 1.153** (1.830) 1.377** (1.840)ΔControl_Committee -2.722*** (-2.990) -3.365*** (-3.500) -3.348*** (-3.470) -3.162*** (-3.370)ΔNon-dual_CEO 0.146 (0.280) 0.311 (0.590) 0.258 (0.480) 0.043 (0.070)ΔBoard_Independence -1.028* (-1.690) -1.003* (-1.680) -1.128** (-1.870) -1.293** (-2.090)ΔPerformance_Evaluation -2.132*** (-2.860) -2.103*** (-2.820) -2.394*** (-3.190) -1.980*** (-2.520)ΔState_Ownership 1.003*** (3.160) 1.044*** (3.150) 0.990*** (3.000) 0.967*** (2.970)ΔOwnership_Concentration 0.343 (0.840) 0.407 (0.990) 0.581 (1.350) 0.756** (1.830)DControl_Committee*DΔEntrenchment 5.815*** (6.530)DNon-dual_CEO*DΔEntrenchment -0.144 (0.180)DBoard_Independence*DΔEntrenchment -0.497 (-0.670)DBoard_Performance*DΔEntrenchment 1.252* (1.650)DState_Ownership*DΔEntrenchment 0.122 (0.150)DOwnership_Concentration*DΔEntrenchment -0.269 (-0.410)Debt -1.073 (-1.300) -0.729 (-0.830) -0.595 (-0.670) -0.781 (-0.870)Dividend -2.666** (-2.080) -2.926*** (-2.250) -2.434** (-2.000) -2.662** (-2.190)Size 0.487 (0.550) 0.478 (0.540) 0.320 (0.360) 0.621 (0.730)Age 0.543* (1.640) 0.572* (1.710) 0.717** (2.140) 0.612* (1.740)ROA 0.715 (0.670) 1.054 (0.960) 1.154 (1.090) 0.776 (0.720)Investment -0.881 (-0.990) -0.965 (-1.090) -1.543 (-1.580) -1.075 (-1.130)Growth 0.315 (0.880) 0.522 (1.470) 0.416 (1.170) 0.434 (1.270)Intangibles 2.793*** (4.070) 2.919*** (4.280) 2.873*** (4.160) 2.800*** (4.070)Intercept -3.382*** (-2.330) -8.157*** (-2.700) -4.254* (-1.710) -4.731** (-1.850)R2 38.72 39.36 38.04 40.67

Fitness of the model (F test) 10.22 (0.000) 8.59 (0.000) 10.03 (0.000) 6.43 (0.000)

Number of observations 302 302 302 302***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values

47

TABLE 5A The Role of CSP connecting Entrenchment and Performance (ROA)Table 5A reports the results of conducting robust regressions on the variations (Δ) of entrenchment and CSP as well as their interaction on variations in a firm’s financial performance. The variations (Δ) of the independent variables are taken between period t and period t-1. The dependent variable is the ROA (columns 1, 2, 3) which is defined as the ration of earnings before interests and taxes to total assets. In columns 4 and 5, we detract from this variable the mean for the corresponding year sector and country (RROA). The variable of CSP is the score provided by SiRi once we have detracted the component on corporate governance. The variable of Workers is the score provided by SiRi for the degree of workers’ satisfaction. Entrenchment=Anti-takeover + OneShare_OneVote + Shareholders_Rights + Managown_Entrench + Manager_Tenure. Anti-takeover is a dummy that it is equal to 1 if the firm has implemented any anti-takeover measures. OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes of stock with different voting rights. Shareholders_Rights is a variable that takes three values (0, 0.5, 1) depending on the degree of limitation of shareholders’ voting rights (see text for details). Managown_Entrench is a dummy that is equal to 1 when managerial stake has a value between 17% and 69.8%. In this region there is a negative relationship between managerial stake and Tobin’s q (see text for details). Manager_Tenure is defined as a dummy that takes the value of 1 if the directors’(including managers) term in office is more than three years old. ΔEntrenchment_ΔCSP is the product of variations of entrenchment times variations in CSP. ΔEntrenchment_ΔWorkers is the product of variations of entrenchment times variations in workers’ satisfaction. Control_committee is the sum of the following three variables: Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committee is a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committee is a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value for the sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of the performance evaluation system (see text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake held by the three largest blockholders. Debt is the ratio of non-current liabilities plus loans to shareholders’ funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Ageis the number of years of the company’s existence, ROA is the is the earnings before interests and taxes to the total assets. Investment is the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is larger than the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangible assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).

Dependent Variable ΔROA (t+1) ΔROA (t+1) ΔROA (t+1) ΔRROA (t+1) ΔRROA (t+1)

ΔEntrenchment -0.566** (-1.790) -0.726** (-2.210) -0.816** (-2.160) -0.620** (-1.90) -0.703** (-1.89)ΔCSP 0.324 (0.500) 0.139 (0.200) 0.422 (0.610)ΔWorkers -0.016 (-0.330) 0.010 (0.200)ΔEntrenchment_ΔCSP -0.831*** (-2.350) -0.704** (-1.99)ΔEntrenchment_ΔWorkers -0.962** (-2.190) -0.852** (1.950)ΔControl_Committee -0.201 (-0.360) -0.332 (-0.610) -0.243 (-0.450) -0.228 (-0.420) -0.240 (-0.460)ΔNon-dual_CEO -0.211 (-0.740) -0.047 (-0.180) -0.233 (-0.770) -0.055 (-0.200) -0.180 (-0.610)ΔBoard_Independence 0.001 (0.000) 0.119 (0.210) -0.133 (-0.210) -0.198 (0.715) -0.007 (-0.010)ΔPerformance_Evaluation 0.094 (0.260) -0.036 (-0.100) 0.463 (1.200) 0.028 (0.080) 0.329 (0.900)ΔState_Ownership 0.204 (0.650) -0.146 (-0.530) 0.125 (0.380) -0.097 (-0.360) 0.066 (0.210)ΔOwnership_Concentration 0.349 (1.030) 0.304 (0.940) 0.375 (1.100) 0.320 (0.980) 0.395 (1.160)Debt -0.556 (-0.790) -0.443 (-0.770) -1.220** (-1.810) -0.343 (-0.610) -0.590 (-0.880)Dividend -5.255*** (-4.36) -3.211*** (-2.800) -5.065*** (-3.980) -3.093*** (-2.81) -5.082*** (-4.02)Size 4.355*** (3.130) 4.002*** (2.940) 4.081*** (2.950) 4.085*** (3.060) 4.645*** (3.240)Age 0.048 (0.320) -0.083 (-0.480) 0.023 (0.130) -0.060 (0.723) 0.040 (0.240)Investment 0.542 (0.850) 0.355 (0.660) 0.628 (0.960) 0.617 (1.090) 0.761 (1.170)Growth 0.131 (0.800) 0.215 (1.120) 0.218 (1.240) 0.251 (1.350) 0.211 (1.230)Intangibles 0.754** (2.070) 0.752** (2.190) 0.913** (2.230) 0.674** (2.030) 0.669 (1.770)Intercept 6.269*** (8.550) 6.491*** (9.250) 5.009*** (4.960) 0.377 (0.540) -0.292 (-0.300)R2

14.70% 16.07% 17.11% 14.17% 15.09%Fitness of the model (F test) 4.65 (0.000) 3.93 (0.000) 2.43 (0.000) 1.57 (0.061) 1.87 (0.013)Number of observations 245 245 245 245 245***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values

48

TABLE 5B The Role of CSP connecting Entrenchment and Performance (Tobin’s q)Table 5B reports the results of conducting robust regressions on the variations (Δ) of entrenchment and CSP as well as their interaction on variations in a firm’s financial performance proxied by Tobin s q. The variations (Δ) of the independent variables are taken between period t and period t-1. The dependent variable ΔPerformance is approached through the Tobin s q which is defined as the ratio of the sum of equity value plus a firm’s liabilities to the total assets taken on a log scale. The variable of CSP is the score provided by SiRi once we have detracted the component on corporate governance. The variable of Workers is the score provided by SiRi for the degree of workers’ satisfaction. Entrenchment=Anti-takeover + OneShare_OneVote + Shareholders_Rights + Managown_Entrench + Manager_Tenure. Anti-takeover is a dummy that it is equal to 1 if the firm has implemented any anti-takeover measures. OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes of stock with different voting rights. Shareholders_Rights is a variable that takes three values (0, 0.5, 1) depending on the degree of limitation of shareholders’ voting rights (see text for details). Managown_Entrench is a dummy that is equal to 1 when managerial stake has a value between 17% and 69.8%. In this region there is a negative relationship between managerial stake and Tobin’s q (see text for details). Manager_Tenure is defined as a dummy that takes the value of 1 if the directors’ (including managers) term in office is more than three years old. ΔEntrenchment_ΔCSP is the product of variations ofentrenchment times variations in CSP. ΔEntrenchment_ΔWorkers is the product of variations of entrenchment times variations in workers’ satisfaction. In columns 4 and 5, we substitute these latter variables by variables DΔEntrenchment_ΔCSP and DΔEntrenchment_ΔWorkers that are the result of multiplying DΔEntrenchment (that is equal to 1 when variations of entrenchment are larger than the mean for the sector for the corresponding year and country), by variables ΔCSP and ΔWorkers respectively. Control_committee = Nomination_Committee + Remuneration_Committee +Audit_Committee, where Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committee is a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committee is a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value forthe sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of the performance evaluation system (see text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake held by the three largest blockholders. Debt is the ratio of non-current liabilities plus loans to shareholders’ funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Age is the number of years of the company’s existence, ROA is the is the earnings before interests and taxes to the total assets. Investment is the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is larger than the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangible assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).Dependent Variable ΔTobin´s q (t+1) ΔTobin’s q (t+1) ΔTobin’s q (t+1) ΔTobin’s q (t+1) ΔTobin’s q (t+1)ΔEntrenchment -0.082** (-2.07) -0.071** (-1.850) -0.068** (-1.860) -0.081** (-2.2) -0.068** (-1.90)ΔCSP -0.120* (-1.680) -0.087 (-1.360) -0.058 (-0.93)ΔWorkers -0.001 (-0.230) 0.000 (-0.110)ΔEntrenchment_ΔCSP 0.001 (0.020)ΔEntrenchment_ΔWorkers -0.036 (-0.960)DΔEntrenchment_ΔCSP -0.051** (-1.9)DΔEntrenchment_ΔWorkers -0.062*** (-2.37)ΔControl_Committee 0.087** (1.870) 0.090** (2.010) 0.071* (1.630) 0.059 (1.26) 0.048 (1.040)ΔNon-dual_CEO -0.022 (-0.800) -0.004 (-0.140) -0.013 (-0.490) -0.036 (-1.32) -0.012 (-0.460)ΔBoard_Independence 0.113 (1.380) 0.023 (0.600) 0.032 (0.890) 0.011 (0.3) 0.026 (0.700)ΔPerformance_Evaluation 0.051 (1.120) 0.055 (1.260) 0.048 (1.150) 0.076* (1.78) 0.043 (1.040)ΔState_Ownership 0.033 (1.160) 0.046 (1.530) 0.051* (1.750) 0.020 (1.08) 0.058** (1.980)ΔOwnership_Concentration 0.007 (0.270) 0.021 (0.810) 0.010 (0.420) 0.002 (0.06) 0.017 (0.700)Debt -0.362*** (-5.22) -0.410*** (-6.050) -0.346*** (-5.25) -0.313*** (-6.81) -0.343*** (-5.220)Dividend 0.156 (1.550) 0.255** (2.100) 0.130 (1.460) -0.004 (-0.08) 0.113 (1.260)Size -0.100 (-0.780) -0.137 (-1.080) -0.058 (-0.520) 0.010 (0.28) -0.038 (-0.340)Age -0.035* (-1.780) -0.029 (-1.550) -0.032* (-1.730) -0.028* (-1.59) -0.024 (-1.280)Investment 0.200*** (2.57) 0.148** (1.990) 0.198*** (2.660) 0.127*** (2.45) 0.173*** (2.310)Growth 0.044** (2.010) 0.048** (2.250) 0.042** (2.140) 0.046** (2.26) 0.041** (2.090)Intangibles -0.047 (-1.160) -0.037 (-0.960) -0.061* (-1.590) 0.011 (0.28) -0.039 (-1.030)Intercept 1.714*** (4.150) 1.000*** (6.910) 1.692*** (4.190) 1.032*** (5.11) 1.636*** (4.030)R2

61.16% 61.22% 62.44% 61.40% 62.35%Fitness of the model (F test) 13.60 (0.000) 16.22 (0.000) 14.72 (0.000) 15.04 (0.000) 14.66 (0.000)Number of observations 245 245 245 245 245***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values

49

TABLE 5C The Role of CSP connecting Entrenchment and Performance (Abnormal returns)Table 5C reports the results of conducting robust regressions on the variations (Δ) of entrenchment and CSP as well as their interaction on variations in a firm’s financial performance proxied by the abnormal returns. These are computed through a single factor model, using as factor the MSCI world index in US dollars (see text for details). The variations (Δ) of the independent variables are taken between period t and period t-1. The variable of CSP is the score provided by SiRi once we have detracted the component on corporate governance. The variable of Workers is the score provided by SiRi for the degree of workers’ satisfaction. Entrenchment=Anti-takeover + OneShare_OneVote + Shareholders_Rights + Managown_Entrench + Manager_Tenure. Anti-takeover is a dummy that it is equal to 1 if the firm has implemented any anti-takeover measures. OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes of stock with different voting rights. Shareholders_Rights is a variable that takes three values (0, 0.5, 1) depending on the degree of limitation of shareholders’ voting rights (see text for details). Managown_Entrench is a dummy that is equal to 1 when managerial stake has a value between 17% and 69.8%. In this region there is a negative relationship between managerial stake and Tobin’s q (see text for details). Manager_Tenure is defined as a dummy that takes the value of 1 if the directors’ (including managers) term in office is more than three years old. ΔEntrenchment_ΔCSP is the product of variations of entrenchment times variations in CSP. ΔEntrenchment_ΔWorkers is the product of variations of entrenchment times variations in workers’ satisfaction. In columns 4 and 5, we substitute these latter variables by variables DΔEntrenchment_ΔCSP and DΔEntrenchment_ΔWorkers that are the result of multiplying DΔEntrenchment (that is equal to 1 when variations of entrenchment are larger than the mean for the sector for the corresponding year and country), by variables ΔCSP and ΔWorkers respectively. Control_committee = Nomination_Committee + Remuneration_Committee + Audit_Committee, where Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committee is a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committee is a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value for the sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of the performance evaluation system (see text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake held by the three largest blockholders. Debt is the ratio of non-current liabilities plus loans to shareholders’ funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Age is the number of years of the company’s existence, ROA is the is the earnings before interests and taxes to the total assets. Investment is the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is larger than the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangible assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).

Dependent Variable ΔAbn. Returns (t+1) ΔAbn. Returns (t+1) ΔAbn. Returns (t+1) ΔAbn. Returns (t+1) ΔAbn. Returns (t+1)

ΔEntrenchment -0.004*** (-1.970) -0.004** (-1.950) -0.005*** (-2.370) -0.004** (-1.810) -0.004***(-2.040)ΔCSP -0.004 (-1.310) -0.004 (-1.320) -0.005 (-1.370)ΔWorkers 0.000 (0.590) -0.000 (-1.13)ΔEntrenchment_ΔCSP 0.000 (-0.150)ΔEntrenchment_ΔWorkers -0.002 (-0.810)DΔEntrenchment_ΔCSP -0.003** (-1.910)DΔEntrenchment_ΔWorkers -0.003** (-2.130)ΔControl_Committee 0.001 (0.450) 0.001 (0.420) -0.001 (0.210) -0.001 (-0.630) 0.000 (0.22)ΔNon-dual_CEO 0.000 (0.110) 0.000 (0.100) 0.000 (0.200) -0.001 (-0.390) 0.000 (0.060)ΔBoard_Independence 0.003* (1.710) 0.003* (1.660) 0.004 (0.870) 0.003 (0.770) 0.002 (1.450)ΔPerformance_Evaluation -0.004** (-1.930) -0.004** (-1.840) -0.003 (-1.180) -0.003 (-1.360) -0.005** (-2.390)ΔState_Ownership 0.001 (0.410) 0.001 (0.430) 0.000 (0.100) 0.001 (0.480) 0.001 (0.800)ΔOwnership_Concentration 0.001 (1.100) 0.001 (1.040) 0.001 (0.620) 0.001 (0.560) 0.001 (1.120)Debt -0.006* (-1.600) -0.005* (-1.570) -0.006* (-1.560) -0.007** (-1.930) 0.006** (-1.910)Dividend -0.014*** (-2.260) -0.013** (-2.180) -0.020*** (-3.540) -0.013*** (-2.290) -0.017*** (-2.790)Size 0.006 (0.890) 0.006 (0.860) 0.009 (1.210) 0.008 (1.230) 0.011* (1.670)Age 0.000 (-0.470) 0.000 (-0.460) -0.001 (-0.760) -0.001 (-0.640) 0.000 (-0.360)Investment 0.002 (0.520) 0.002 (0.530) 0.000 (0.040) 0.003 (0.880) 0.001 (0.290)Growth 0.001 (1.280) 0.001 (1.270) 0.001 (1.160) 0.002 (1.530) 0.001 (1.040)Intangibles 0.000 (0.200) 0.000 (0.170) 0.000 (0.230) 0.001 (0.420) 0.000 (0.410)Intercept 0.011*** (2.800) 0.010*** (2.750) 0.009*** (1.920) 0.007** (1.810) 0.008*** (2.03)R2

21.32% 21.45% 22.05% 21.96% 23.92%Fitness of the model (F test) 3.32 (0.000) 3.11 (0.000) 3.04 (0.000) 2.46 (0.000) 3.59 (0.000)Number of observations 245 245 245 245 245***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values

50

TABLE 6. Income Smoothing as proxy for entrenchmentTable 6 reports the results of conducting regressions on the variations (Δ) for income smoothing-based earnings management, as well as for different governance mechanisms, on the variation (Δ) in a firm’s CSP (led by one period). The variations (Δ) in the independent variables are taken between period t and period t-1, while those for the dependent variables between period t+1 and t. We use two measures for income smoothing. Incsmooth1 is the correlation between changes in accruals and changes in cash flow (columns 1 and 2). Incsmooth2 is the ratio of net income before extraordinary items to the standard deviation of cash-flows (columns 3 and 4). Control_committee is the sum of the following three variables: Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committee is a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committeeis a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value of the sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of a performance evaluation system (see text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake held by the three largest blockholders. Debt is the ratio of non-current liabilities plus loans to shareholders funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Age is the number of years of the company’s existence, ROA is the earnings before interests and taxes to the total assets. Investment is the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is larger than the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangibles assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).Dependent Variable ΔCSP (t+1) ΔCSP (t+1) ΔCSP (t+1) ΔCSP (t+1)Incsmooth1 1463.548*** (1.96) 2088.766** (1.84)Incsmooth2 1.112*** (4.43) 0.954** (1.95)ΔControl_Committee -2.103*** (-2.61) -2.922***-(2.4) -1.841** (-2.02) -2.382* (-1.710)ΔNon-dual_CEO -0.491 (-1.16) 0.322 (0.51) -0.310 (-0.68) 0.423 (0.630)ΔBoard_Independence -0.123 (-0.220) -1.571** (-2.20) 0.136 (0.190) -0.892 (-0.880)ΔPerformance_Evaluation -2.296*** (-3.320) -2.869*** (-3.20) -2.544*** (-3.24) -3.088*** (-2.830)ΔState_Ownership 0.436** (1.960) 0.818** (2.110) 0.436* (1.680) 0.848** (1.950)ΔOwnership_Concentration 0.425 (0.950) 0.61 (1.250) 0.385 (0.840) 0.643 (1.250)Debt -1.364* (-1.740) -0.377 (-0.300) -2.220*** (-2.500) -1.974* (-1.690)Dividend -2.373*** (-3.000) -2.879*** (-2.300) -2.316*** (-2.66) -2.927** (-2.110)Size -0.022 (-0.03) 0.338 (0.360) 0.058 (0.080) 0.514 (0.510)Age -0.08 (-0.290) 0.709** (1.810) -0.139 (-0.490) 0.513 (1.250)ROA 0.101 (0.120) 1.157 (0.910) -0.107 (-0.110) 1.224 (0.890)Investment 0.653 (0.850) -2.012** (-1.800) 1.386 (1.260) -1.234 (-0.980)Growth 0.741*** (2.46) 0.44 (1.06) 0.812*** (2.560) 0.444 (0.960)Intangibles 2.571*** (4.170) 2.962*** (3.450) 2.396*** (3.980) 2.623*** (2.870)Intercept 81.763** (1.960) 116.900** (1.820) -1.294 (-0.450) -10.584*** (-2.670)R2

41.73% 36.98% 46.14% 39.34%Fitness of the model(F test) 9.55 (0.000) 7.92 (0.000) 11.70 (0.000) 8.18 (0.000)Number of observations 231 231 207 207***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values

51

TABLE 7. The Expropriating EffectTable 7 reports the results of conducting regressions on the variations (Δ) in entrenchment, as well as different governance mechanisms, on the variation (Δ) in a firm’s CSP (led by one period) and workers’ satisfaction (last two columns). The variations (Δ) in the dependent variables are led by one period. Columns 1 and 3 (2 and 4) focus on those firms such that the stake of the three largest blockholders is larger (lower) than the mean of the sector for the corresponding year –concentration = 1 (0)- The variable CSP is the score provided by SiRi for non-shareholder stakeholders’ degree of satisfaction. Workers is the score of workers’ degree of satisfaction provided by SiRi. Entrenchment=Anti-takeover + OneShare_OneVote + Shareholders_Rights + Managown_Entrench + Manager_Tenure. Anti-takeover is a dummy that it is equal to 1 if the firm has implemented any of the following anti-takeover measures: voting caps; increased voting rights over time; restrictions on board appointment rights; poison pills; and 0 otherwise. OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes of stock with different voting rights and 0 otherwise. Shareholders_Rights is a variable that takes three values (0, 0.5, 1) depending on the degree of limitation of shareholders’ voting rights (see text for details). Managown_Entrench is a dummy that is equal to 1 when managerial stake has a value between 17% and 69.8%. In this region there is a negative relationship between managerial stake and Tobin’s q (see text for details). Manager_Tenure is defined as a dummy that takes the value of 1 if the directors’ (including managers) term in office is more than three years, and 0 otherwise.Control_committee is the sum of the following three variables: Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committee is a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committee is a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value of the sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of a performance evaluation system (see the text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake of the three largest blockholders. Debt is the ratio of non-current liabilities plus loans to shareholders funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Age is the number of years of the company’s existence, ROA is the earnings before interest and taxes to the total assets. Investment is the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is larger than the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangibles assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).

Large ownership concentration

Low ownership concentration

Large ownership concentration

Low ownership concentration

Dependent variable ΔCSP (t+1) ΔCSP (t+1) ΔWorkers (t+1) ΔWorkers (t+1)ΔEntrenchment 1.242 (1.370) 1.082** (2.040) 0.501 (0.410) 1.460** (1.940)ΔControl_Committee -1.149 (-1.040) -2.651*** (-3.020) -1.611 (-1.040) -3.493*** (-2.78)ΔNon-dual_CEO -0.213 (-0.280) -0.695 (-1.500) 1.359 (1.440) -0.582 (-0.880)ΔBoard_Independence -0.914 (-0.800) 1.102** (2.020) -1.925* (-1.670) -0.799 (-1.130)ΔPerformance_Evaluation -1.931 (-1.490) -1.947*** (-2.750) -3.895** (-2.150) -1.630* (-1.720)ΔState_Ownership 0.875*** (2.390) 0.471 (1.310) 0.924* (1.720) 0.925** (2.000)ΔOwnership_Concentration 0.457 (0.610) -0.131 (-0.210) 0.762 (0.990) 0.082 (0.110)Debt -1.084 (-0.970) -0.081 (-0.090) -1.047 (-0.870) -1.189 (-0.900)Dividend -1.716** (-2.150) -3.631*** (-3.420) -2.113* (-1.670) -4.667*** (-2.880)Size -1.028* (-1.580) 1.212 (1.370) 0.046 (0.050) 1.481 (1.280)Age 0.180 (0.390) -0.493* (-1.590) 0.568 (0.920) 0.547 (1.160)ROA 0.217 (0.170) 0.696 (0.680) 1.698 (1.010) 0.836 (0.500)Investment 0.376 (0.180) -0.121 (-0.170) -5.692*** (-2.43) -0.797 (-0.770)Growth 0.975** (2.260) 0.569* (1.580) 0.280 (0.440) 0.636 (1.370)Intangibles 1.465 (1.200) 2.630*** (3.810) 0.333 (0.250) 3.526*** (3.900)Intercept 7.311 (1.300) -1.252 (-0.790) -11.621* (-1.650) -1.759 (-0.770)R2

52.33% 52.67% 55.70% 41.28%Fitness of the model (F test) 3.50 (0.000) 7.06 (0.000) 4.01 (0.000) 4.46 (0.000)Number of observations 110 192 110 192

***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values

52

TABLE 8: ROBUSTNESSTable 8 reports the results of conducting regressions on the variations (Δ) of CSP lead by one period. Columns 1 (2) focus on those firms such that their leverage is larger (lower) than the mean for the corresponding sector and country. In columns 3 and 4 the sample is separated in terms of the legal origin of the country codes (La Porta et al., 1998). In column 3, we focus on those countries with English-origin legal codes, while in column 4, we focus on those countries with French, German or Scandinavian-origin legal codes. The variable CSP is the score provided by SiRi for non-shareholder stakeholders’ degree of satisfaction. Workers is the score of workers’ degree of satisfaction provided by SiRi. Entrenchment=Anti-takeover + OneShare_OneVote + Shareholders_Rights + Managown_Entrench + Manager_Tenure. Anti-takeover is a dummy that it is equal to 1 if the firm has implemented any of the following anti-takeover measures: voting caps; increased voting rights over time; restrictions on board appointment rights; poison pills; and 0 otherwise. OneShare_OneVote is a dummy that it is equal to 1 if the company has multiple classes of stock with different voting rights and 0 otherwise. Shareholders_Rights is a variable that takes three values (0, 0.5, 1) depending on the degree of limitation of shareholders’ voting rights (see text for details). Managown_Entrench is a dummy that is equal to 1 when managerial stake has a value between 17% and 69.8%. In this region there is a negative relationship between managerial stake and Tobin’s q (see text for details). Manager_Tenure is defined as a dummy that takes the value of 1 if the directors’ (including managers) term in office is more than three years, and 0 otherwise. Control_committee is the sum of the following three variables: Nomination_Committee is a dummy that it is equal to 1 if there is a nomination committee with independent members and zero otherwise. Remuneration_Committee is a dummy that it is equal to 1 if there is a remuneration committee with independent members and zero otherwise. Audit_Committee is a dummy that it is equal to 1 if there is an audit committee with independent members and zero otherwise. Non-dual_CEO is a dummy that it is equal to 1 when the chairman is not the CEO and zero otherwise. Board_Independence is a variable that takes three different values (0, 0.5 and 1) contingent on the percentage of independent directors with respect to the mean value of the sector. Performance_Evaluation takes three different values (1, 0, -1) depending on the degree of development of a performance evaluation system (see the text for details). State_Ownership is the stake in the hands of the state. Ownership_Concentration is the stake of the three largest blockholders. Debtis the ratio of non-current liabilities plus loans to shareholders funds. Dividend is the sector average pay-out ratio times the firm’s after-tax profits. Size is the fixed-asset value on a log scale; Age is the number of years of the company’s existence, ROA is the EBITDA to the total assets. Investmentis the ratio of fixed assets to total assets. Growth is equal to 1 when the rate of increase in sales is larger than the value for the corresponding sector and year, and 0 otherwise. Intangibles is the ratio of intangibles assets to total fixed assets. Finally, we introduce temporal, sectoral and four dummies that capture the legal origin of country codes (British common law; French civil law; German civil law and Scandinavian civil law).

Leverage=1 Leverage=0 Anglosaxon Non-AnglosaxonDependent variable ΔCSP (t+1) ΔCSP (t+1) ΔCSP (t+1) ΔCSP (t+1)ΔEntrenchment 0.747 (1.050) 1.637*** (2.750) 1.171 (1.420) 1.339** (1.840)ΔControl_Committee -1.257 (-1.160) -2.457*** (-2.680) -1.318 (-1.150) -0.381 (-0.450)ΔNon-dual_CEO -0.365 (-0.520) -0.194 (-0.430) -0.580 (1.100) -0.880 (-1.360)ΔBoard_Independence 0.952 (1.210) 0.048 (0.080) -0.621 (-0.780) 1.174 (1.450)ΔPerformance_Evaluation -2.738*** (-2.960) -1.366* (-1.670) -3.019*** (-2.660) -0.698 (-0.850)ΔState_Ownership 0.820** (2.100) 0.830*** (2.650) -0.529 (-0.230) 0.839** (1.980)ΔOwnership_Concentration -0.392 (-0.640) 0.473 (0.940) 0.831 (1.130) 0.130 (0.280)Debt 1.606** (1.970) -0.183 (-0.080) -0.107 (-0.070) -2.396* (-1.640)Dividend -2.401** (-1.810) -2.393*** (-2.970) -2.067 (-1.010) -3.737 (-1.250)Size 0.687 (0.830) -1.002 (-1.400) -2.624 (-0.840) -4.621* (-1.730)Age -0.222 (-0.650) 0.138 (0.360) -0.445 (-0.900) -0.404 (-1.220)ROA 0.465 (0.300) 0.223 (0.270) 1.367 (0.950) -2.002 (-1.210)Investment -0.266 (-0.340) 0.588 (0.450) -0.484 (-0.360) 1.592 (0.920)Growth 0.569 (1.080) 0.575** (1.930) 0.929** (2.160) 0.701* (1.630)Intangibles 2.131*** (2.320) 2.344*** (3.010) 2.366*** (3.120) 2.196*** (2.520)Intercept -0.857 (-0.440) 4.636 (1.390) 7.646 (0.900) 0.884 (0.610)R2

53.60% 53.49% 53.22% 53.60%Fitness of the model (F test) 5.11 (0.000) 5.88 (0.000) 5.19 (0.000) 6.35 (0.000)Number of observations 142 160 129 173***p-value 0.01, ** p-value 0.05, *p-value 0.10. In parentheses the p-values