Corporate Governance: Effects on firm performance and economic growth

CORPORATE GOVERNANCE:

EFFECTS ON FIRM PERFORMANCE AND ECONOMIC GROWTH

Summry

This document addresses corporate governance and its effect on corporate performance and economic performance.  It first recapitulates and builds on previous work undertaken by DSTI, for example, it gives a more explicit exposition of the shareholder and stakeholder models of corporate governance.  It then goes on to address some of the underlying factors that promote efficient corporate governance, and examines some of the strengths, weaknesses, and economic implications associated with various corporate governance systems.  In addition to providing data not presented in the previous work, it also provides newly available information on ownership concentration and voting rights in a number of OECD countries.  The document also provides a survey of empirical evidence on the link between corporate governance, firm performance and economic growth. Finally, several policy implications are identified.

One of the most striking differences between countries’ corporate governance systems is the difference in the ownership and control of firms that exist across countries.  Systems of corporate governance can be distinguished according to the degree of ownership and control and the identity of controlling shareholders.  While some systems are characterised by wide dispersed ownership (outsider systems), others tend to be characterised by concentrated ownership or control (insider systems).  In outsider systems of corporate governance (notably the US and UK) the basic conflict of interest is between strong managers and widely-dispersed weak shareholders.  In insider systems (notably Germany and Japan), on the other hand, the basic conflict is between controlling shareholders (or blockholders) and weak minority shareholders.

This document shows how the corporate governance framework can impinge upon the development of equity markets, R&D and innovative activity, entreprenuership, and the development of an active SME sector, and thus impinge upon economic growth.  However, there is no single model of corporate governance and each country has through time developed a wide variety of mechanisms to overcome the agency problems arising from the separation of ownership and control. The document looks at the various mechanisms employed in different systems (e.g. concentrated ownership, executive remuneration schemes, the market for takeovers, cross-shareholdings amongst firms, etc.) and examines the evidence on whether or not they are achieving what they were intended to do.  For example, one of the benefits of concentrated ownership is that it brings more effective monitoring of management and helps overcome the agency problems arising from the separation of ownership and control.  Some of the costs, however, are low liquidity and reduced possibilities for risk diversification.  While dispersed ownership brings higher liquidity it may not provide the right incentives to encourage long-term relationships that are required for certain types of investment.  Therefore, one of the challenges facing policy makers is how to develop a good corporate governance framework which can secure the benefits associated with controlling shareholders acting as direct monitors, while at the same time ensuring that they do not impinge upon the development of equity markets by expropriating excessive rents.

Introduction

At the 1998 Industry Ministerial, a new direction for industrial policy was stressed and Ministers agreed on a number of priority areas for future work, including corporate governance.  The OECD Council, meeting at Ministerial level in April 1998, also stressed the importance of corporate governance and called upon the OECD to develop a set of corporate governance standards and guidelines.  In order to fulfil this Ministerial mandate, the OECD established an Ad Hoc Task Force on Corporate Governance, consisting of representatives from national governments, other relevant international organisations and the private sector.  DSTI also participated in the Secretariat team serving the Task Force and contributed substantive input into the development of the OECD Principles on Corporate Governance, see OECD (1999a).  OECD Ministers, meeting in May 1999, endorsed the Principles developed by the Task Force and also agreed that the Principles be assessed in due course, possibly in two years time.  The OECD Council, therefore, also requested continuing analytical work in this area, see OECD (1999b).

The May 1999 Council Ministerial also called upon the OECD to study the causes of growth disparities (e.g. technological innovation, framework conditions for firm creation and growth, SMEs, etc.), and identify the factors and policies which could strengthen long-term growth performance. While macroeconomic factors certainly play a major part in the economic performances of OECD countries, governments have increasingly come to recognise that there are strong complementarities between sound macroeconomic policies and sound microeconomic foundations.  As the last decade has seen a convergence on what constitutes good macroeconomic policy the OECD countries have increasingly come to recognise that weakness in microstructures can have profound impacts on a macro level.  For example, the 1997 financial crisis in Asia was thought to be due, in part, to weaknesses in the banking sector and in corporate governance.  Countries are therefore looking towards microeconomic foundations and structures in order to enhance their economic performance.  The OECD reports on Regulatory Reform, the Jobs Study and the Principles for Corporate Governance are good examples of this new approach.  This approach is also in line with the new direction of work for the Industry Committee as set out by Industry Ministers at their 1998 OECD Ministerial meeting.

One key element of improving microeconomic efficiency is corporate governance. Corporate governance affects the development and functioning of capital markets and exerts a strong influence on resource allocation.  It impacts upon the behaviour and performance of firms, innovative activity, entrepreneurship, and the development of an active SME sector.  In an era of increasing capital mobility and globalisation, corporate governance has become an important framework condition affecting the industrial competitiveness of OECD countries.  Meanwhile, in transition economies, privatisation has raised questions about the way in which private enterprises should be governed.  It is thought that poor corporate governance mechanisms in these countries have proved, in part, to be a major impediment to improving the competitiveness of firms.  Better corporate governance, therefore, both within OECD and non-OECD countries should manifest itself in enhanced corporate performance and can lead to higher economic growth.

However, there is no single model of corporate governance.  Governance practices vary not only across countries but also across firms and industry sectors.  However, one of the most striking differences between countries’ corporate governance systems is in the ownership and control of firms that exist across countries.  Systems of corporate governance can be distinguished according to the degree of ownership and control and the identity of controlling shareholders.  While some systems are characterised by wide dispersed ownership (outsider systems), others tend to be characterised by concentrated ownership or control (insider systems).  In outsider systems of corporate governance (notably the US and UK) the basic conflict of interest is between strong managers and widely-dispersed weak shareholders.  In insider systems (notably Continental Europe and Japan), on the other hand, the basic conflict is between controlling shareholders (or blockholders) and weak minority shareholders.  However, these differences are also rooted in variations in countries’ legal, regulatory, and institutional environments, as well as historical and cultural factors.  Therefore, policies that promote the adoption of specific forms of governance should attempt to account for the product and factor market contexts, and other institutional factors, within which they are being contemplated.

The OECD Principles for Corporate Governance represent a common basis that OECD Membercountries consider essential for the development of good governance practice.  This work, on the other hand, provides an economic rationale for why corporate governance matters and explores the relationship between corporate governance, corporate performance, economic growth, and, where relevant, industry structure.  The search for good corporate governance practices in this context, therefore, is based on an identification of what works in different countries and circumstances, to discern what lessons can be derived from these experiences, and to examine the conditions for transferability of these practices to other countries.  Continued work in this area, therefore, will aim to ascertain what are the key factors that shape the effectiveness of different corporate governance mechanisms, and to determine what are the key policy adjustments that are most needed in individual systems of corporate governance.  This analytical work will also provide valuable input to the work of other Committees and Directorates, especially DAFFE, and into OECD horizontal projects.  In particular, it will provide input into the assessment of the OECD Principles in due course and to the OECD mandate in determining the underlying factors contributing to economic growth.

This paper recapitulates and builds on previous work undertaken by DSTI, see OECD (1998a).It also builds on lessons gleaned in the development of the OECD Principles for Corporate Governance.  It structures the previous DSTI work better (e.g. it gives a more explicit exposition of the shareholder and stakeholders models of corporate governance) and goes on to provide a qualitative assessment of the strengths, weaknesses and economic implications of different systems of corporate governance.  In addition to new data on ownership concentration and voting rights in a number of OECD countries, it also provides data not presented in the previous work.  It also provides a survey of empirical evidence on the link between corporate governance, firm performance and economic growth, identifying areas in which a consensus view appears to have emerged in the literature.  This work also examines areas not covered previously e.g. the markets for corporate control, the effects of executive remuneration, etc.

Section II of this paper provides an analytical framework for understanding how corporate governance can affect corporate performance and economic growth.  Section III looks at the critical differences in corporate governance systems in OECD countries.  It then goes on to provide a qualitative assessment of the strengths, weaknesses, and economic implications associated with the different systems. Section IV provides a review of the empirical evidence of the effect of corporate governance on corporate performance and economic performance, and section V concludes.  Wherever possible, we also identify those areas where policy implications emerge.

II.  Analytical Framework: The Shareholder and Stakeholders Models of Governance

Corporate governance has traditionally been associated with the “principal-agent” or “agency” problem.  A “principal-agent” relationship arises when the person who owns a firm is not the same as the person who manages or controls it.  For example, investors or financiers (principals) hire managers (agents) to run the firm on their behalf.  Investors need managers’ specialised human capital to generate returns on their investments, and managers may need the investors’ funds since they may not have enough capital of their own to invest.  In this case there is a separation between the financing and the management of the firm, i.e. there is a separation between ownership and control, see Berle and Means (1932).

Before looking at the relationship between corporate governance, firm performance, and economic growth, it is useful to have a framework with which to understand how corporate governance can affect firm behaviour and economic performance.  One of the problems with the current debate on corporate governance is that there are many different, and often conflicting, views on the nature and purpose of the firm.  This debate ranges from positive issues concerning how institutions actually work, to normative issues concerning what should be the firm’s purpose.  Therefore, in order to make sense of this debate, it is useful to consider the different analytical backgrounds or approaches that are often employed.

The term corporate governance has been used in many different ways and the boundaries of the subject vary widely.  In the economics debate concerning the impact of corporate governance on performance, there are basically two different models of the corporation, the shareholder model and the stakeholder model.  In its narrowest sense (shareholder model), corporate governance often describes the formal system of accountability of senior management to shareholders.  In its widest sense (stakeholder model), corporate governance can be used to describe the network of formal and informal relations involving the corporation.  More recently, the stakeholder approach emphasises contributions by stakeholders that can contribute to the long term performance of the firm and shareholder value, and the shareholder approach also recognises that business ethics and stakeholder relations can also have an impact on the reputation and long term success of the corporation.  Therefore, the difference between these two models is not as stark as it first seems, and it is instead a question of emphasis.

The lack of any consensus regarding the definition of corporate governance is also reflected in the debate on governance reform.  This lack of consensus leads to entirely different analyses of the problem and to the strikingly different solutions offered by participants in the reform process.  Therefore, having a clear understanding of the different models can provide insights and help us to appreciate the different sides of this debate.  An understanding of the issues involved can also provide the basis from which to identify good corporate governance practices and to provide policy recommendations.

II.1  The Shareholder Model

According to the shareholder model the objective of the firm is to maximise shareholder wealth through allocative, productive and dynamic efficiency i.e. the objective of the firm is to maximise profits. The criteria by which performance is judged in this model can simply be taken as the market value (i.e. shareholder value) of the firm.  Therefore, managers and directors have an implicit obligation to ensure that firms are run in the interests of shareholders.  The underlying problem of corporate governance in this model stems from the principal-agent relationship arising from the separation of beneficial ownership and executive decision-making.  It is this separation that causes the firm’s behaviour to diverge from the profitmaximising ideal.  This happens because the interests and objectives of the principal (the investors) and the agent (the managers) differ when there is a separation of ownership and control.  Since the managers are not the owners of the firm they do not bear the full costs, or reap the full benefits, of their actions. Therefore, although investors are interested in maximising shareholder value, managers may have other objectives such as maximising their salaries, growth in market share, or an attachment to particular investment projects, etc.

The principal-agent problem is also an essential element of the “incomplete contracts” view of the firm developed by Coase (1937), Jensen and Meckling (1976), Fama and Jensen (1983a,b), Williamson (1975, 1985), Aghion and Bolton (1992), and Hart (1995).  This is because the principal-agent problem would not arise if it were possible to write a “complete contract”.  In this case, the investor and the manager would just sign a contract that specifies ex-ante what the manager does with the funds, how the returns are divided up, etc.  In other words, investors could use a contract to perfectly align the interests and objectives of managers with their own.  However, complete contracts are unfeasible, since it is impossible to foresee or describe all future contingencies.  This incompleteness of contracts means that investors and managers will have to allocate “residual control rights” in some way, where residual control rights are the rights to make decisions in unforeseen circumstances or in circumstances not covered by the contract.  Therefore, as Hart (1995) states: “Governance structures can be seen as a mechanism for making decisions that have not been specified in the initial contract.”

So why don’t investors just write a contract that gives them all the residual control rights in the firm, i.e. owners get to decide what to do in circumstances not covered by the contract?  In principle this is not possible, since the reason why owners hire managers in the first place is because they needed managers’ specialised human capital to run the firm and to generate returns on their investments.  The “agency” problem, therefore, is also an asymmetric information problem i.e. managers are better informed regarding what are the best alternative uses for the investors’ funds.  As a result, the manager ends up with substantial residual control rights and discretion to allocate funds as he chooses.  There may be limits on this discretion specified in the contract, but the fact is that managers do have most of the residual control rights.^[1]  The fact that managers have most of the control rights can lead to problems of management entrenchment and rent extraction by managers.  Much of corporate governance, therefore, deals with the limits on managers’ discretion and accountability i.e. as Demb and Neubauer (1992) state “corporate governance is a question of performance accountability”.

One of the economic consequences of the possibility of ex-post expropriation of rents (or opportunistic behaviour) by managers is that it reduces the amount of resources that investors are willing to put up ex-ante to finance the firm, see Grossman and Hart (1986).  This problem, more generally known as the hold-up problem has been widely discussed in the literature, see Williamson (1975, 1985) and Klein, Crawford and Alchian (1978).  A major consequence of opportunistic behaviour is that it leads to socially inefficient levels of investment that, in turn, can have direct implications for economic growth.  According to the shareholder model, therefore, corporate governance is primarily concerned with finding ways to align the interests of managers with those of investors, with ensuring the flow of external funds to firms and that financiers get a return on their investment.

An effective corporate governance framework can minimise the agency costs and hold-up problems associated with the separation of ownership and control.  There are broadly three types of mechanisms that can be used to align the interests and objectives of managers with those of shareholders and overcome problems of management entrenchment and monitoring:

−	One method attempts to induce managers to carry out efficient management by directly aligning managers interests with those of shareholders e.g. executive compensation plans, stock options, direct monitoring by boards, etc.
−	Another method involves the strengthening of shareholder’s rights so shareholders have both a greater incentive and ability to monitor management.  This approach enhances the rights of investors through legal protection from expropriation by managers e.g. protection and enforcement of shareholder rights, prohibitions against insider-dealing, etc.
−	Another method is to use indirect means of corporate control such as that provided by capital

markets, managerial labour markets, and markets for corporate control e.g. take-overs.

One of the critiques of the shareholder model of the corporation is the implicit presumption that the conflicts are between strong, entrenched managers and weak, dispersed shareholders.  This has led to an almost exclusive focus, in both the analytical work and in reform efforts, of resolving the monitoring and management entrenchment problems which are the main corporate governance problems in the principal-agent context with dispersed ownership.  For example, most of this work has addressed concerns related to the role of the board of directors, stock options and executive remuneration, shareholder protection, the role of institutional investors, management entrenchment and the effectiveness of the market for take-overs, etc.

The fact is that the widely held firm, presumed in Berle and Means (1932) seminal work, is not the rule but is rather the exception.^[2]  Instead, the dominant organisational form for the firm is one characterised by concentrated ownership.  One of the reasons why we observe ownership concentration may be due, in part, to the lack of investor protection.  However, unlike the widely-held corporation where managers have most of the residual control rights with shareholders having very little power, the closelyheld corporation is usually controlled by a majority shareholder or by a group of controlling blockholders. This could be an individual or family, or blockholders such as financial institutions, or other corporations acting through a holding company or cross shareholdings.

Another reason why ownership concentration is so prevalent as the dominant organisational form is because it is one way of resolving the monitoring problem.  According to the principle-agent model, due to the divergence of interests and objectives of managers and shareholders, one would expect the separation of ownership and control to have damaging effects on the performance of firms.  Therefore, one way of overcoming this problem is through direct shareholder monitoring via concentrated ownership.  The difficulty with dispersed ownership is that the incentives to monitor management are weak.  Shareholders have an incentive to “free-ride” in the hope that other shareholders will do the monitoring.  This is because the benefits from monitoring are shared with all shareholders, whereas, the full costs of monitoring are incurred by those who monitor.  These free-rider problems do not arise with concentrated ownership, since the majority shareholder captures most of the benefits associated with his monitoring efforts.

Therefore, for the closely held corporation the problem of corporate governance is not primarilyabout general shareholder protection or monitoring issues.  The problem instead is more one of crossshareholdings, holding companies and pyramids, or other mechanisms that dominant shareholders use to exercise control, often at the expense of minority investors.  It is the protection of minority shareholders that becomes critical in this case.  One of the issues that arises in this context is how do policy makers develop reforms that do not disenfranchise majority shareholders while at the same time protect the interests of minority shareholders.  In other words, how do we develop reforms that retain the benefits of monitoring provided by concentrated ownership yet at the same time encourage the flow of external funds to corporations, and which, in turn, should lead to dilution of ownership concentration.

Another critique of the shareholder approach is that the analytical focus on how to solve the corporate governance problem is too narrow. The shareholder approach to corporate governance is primarily concerned with aligning the interests of managers and shareholders and with ensuring the flow of external capital to firms.  However, shareholders are not the only ones who make investments in the corporation.  The competitiveness and ultimate success of a corporation is the result of teamwork that embodies contributions from a range of different resource providers including investors, employees, creditors, suppliers, distributors, and customers.  Corporate governance and economic performance will be affected by the relationships among these various stakeholders in the firm.  According to this line of argument, any assessment of the strengths, weaknesses, and economic implications of different corporate governance frameworks needs a broader analytical framework which includes the incentives and disincentives faced by all stakeholders.

II.2  The Stakeholder Model

The stakeholder model takes a broader view of the firm.  According to the traditional stakeholder model, the corporation is responsible to a wider constituency of stakeholders other than shareholders. Other stakeholders may include contractual partners such as employees, suppliers, customers, creditors, and social constituents such as members of the community in which the firm is located, environmental interests, local and national governments, and society at large.  This view holds that corporations should be “socially responsible” institutions, managed in the public interest.  According to this model performance is judged by a wider constituency interested in employment, market share, and growth in trading relations with suppliers and purchasers, as well as financial performance.^[3]

The problem with the traditional stakeholder model of the firm is that it is difficult, if not impossible, to ensure that corporations fulfil these wider objectives.  Blair (1995) states the arguments against this point of view: “The idea [...] failed to give clear guidance to help managers and directors set priorities and decide among competing socially beneficial uses of corporate resources, and provided no obvious enforcement mechanisms to ensure that corporations live up to their social obligations.  As a result of these deficiencies, few academics, policymakers, or other proponents of corporate governance reforms still espouse this model.”^[4]

However, given the potential consequences of corporate governance for economic performance, the notion that corporations have responsibilities to parties other than shareholders merits consideration. What matters is the impact that various stakeholders can have on the behaviour and performance of the firm and on economic growth.  Any assessment of the implications of corporate governance on economic performance must consider the incentives and disincentives faced by all participants who potentially contribute to firm performance.  With this in mind, the stakeholder model has recently been redefined, where the emphasis has been to more narrowly define what constitutes a stakeholder.  Therefore, the “new” stakeholder model specifically defines stakeholders to be those actors who have contributed firmspecific assets, see Blair (1995).  This redefinition of the stakeholder model is also consistent with both the transaction costs and incomplete contract theories of the firm in which the firm can be viewed as a “nexus of contracts”, see Coase (1937), Williamson (1975, 1985), Jensen and Meckling (1976), and Aoki, Gustafsson and Williamson (1990).

The “best” firms according to the “new” stakeholder model are ones with committed suppliers, customers, and employees.  This new stakeholder approach is, therefore, a natural extension of the shareholder model.  For example, whenever firm-specific investments need to be made, the performance of the firm will depend upon contributions from various resource providers of human and physical capital.  It is often the case that the competitiveness and ultimate success of the firm will be the result of teamwork that embodies contributions from a range of different resource providers including investors, employees, creditors, and suppliers.  Therefore, it is in the interest of the shareholders to take account of other stakeholders, and to promote the development of long term relations, trust, and commitment amongst various stakeholders (see Mayer, 1996).  Corporate governance in this context becomes a problem of finding mechanisms that elicit firm specific investments on the part of various stakeholders, and that encourage active co-operation amongst stakeholders in creating wealth, jobs, and the sustainability of financially sound enterprises, see the OECD Principles of Corporate Governance (OECD 1999a).

However, opportunistic behaviour and hold-up problems arise whenever contracts are incomplete and firm specific investments need to be made.  As discussed previously, one consequence of opportunistic behaviour is that in general it leads to underinvestment.  The principal-agent relationship discussed in the shareholder model is only one of the many areas in which this occurs.  Underinvestment in the stakeholder model would include investments by employees, suppliers, etc.  For example, employees may be unwilling to invest in firm specific human capital if they are unable to share in the returns from their investment, but have to bear the opportunity costs associated with making those investments.  Alternatively firms may be unwilling to expend resources in training employees if once they have incurred the costs they are unable to reap the benefits if employees, once endowed with increased human capital, choose to leave the firm. Suppliers and distributors can also underinvest in firm-specific investments such as customised components, distribution networks, etc.  In this broader context, corporate governance becomes a problem of finding mechanisms that reduce the scope for expropriation and opportunism, and lead to more efficient levels of investment and resource allocation.

According to the stakeholder model, corporate governance is primarily concerned with how effective different governance systems are in promoting long term investment and commitment amongst the various stakeholders, see Williamson (1985).^[5]  Kester (1992), for example, states that “the central problem of governance is to devise specialised systems of incentives, safeguards, and dispute resolution processes that will promote the continuity of business relationships that are efficient in the presence of selfinterested opportunism”.  Blair (1995) also defines corporate governance in this broader context and argues that corporate governance should be regarded as the set of institutional arrangements for governing the relationships among all of the stakeholders that contribute firm specific assets.

One of the critiques of the stakeholder model, or fears of participants in the reform process, is that managers or directors may use “stakeholder” reasons to justify poor company performance.  The benefit of the shareholder model is that it provides clear guidance in helping managers set priorities and establishes a mechanism for measuring the efficiency of the firms’ management team i.e. firm profitability. On the other hand, the benefit of the stakeholder model is its emphasis on overcoming problems of underinvestment associated with opportunistic behaviour and in encouraging active co-operation amongst stakeholders to ensure the long-term profitability of the corporation.

One of the most challenging tasks on the reform agenda is how to develop corporate governance frameworks and mechanisms that elicit the socially efficient levels of investment by all stakeholders.  The difficulty, however, is to identify those frameworks and mechanisms which promote efficient levels of investment, while at the same time maintaining the performance accountability aspects provided by the shareholder model.  At a minimum, this implies that mechanisms that promote stakeholder investment and co-operation should be adopted in conjunction with mechanisms aimed at preventing management entrenchment.  Stakeholder objectives should not be used to prevent clear guidance on how the firms’ objectives and priorities are set.  How the firm will attain those objectives and how performance monitoring will be determined also need to be clearly defined.

II.3 The Interaction of Corporate Governance with the Institutional and Economic Framework

There is another argument, not addressed above, that asks why should we worry about corporate governance in the first place, since product market competition should provide incentives for firms to adopt the most efficient corporate governance mechanisms.  Firms that do not adopt cost-minimising governance mechanisms would presumably be less efficient and in the long run would be replaced, i.e. competition should take care of governance.  This line of argument would oppose any external policy interventions on the grounds that at best they are unhelpful and at worst distortionary.  Rather than justifying public intervention, it says that the resolution of governance problems should be left to market participants.   Thus recent development in the managerial labour market, such as executive stock options and the market for corporate control, e.g. leveraged and management buy-outs, are seen as market responses to institutional deficiencies.^[6]

While there are likely to be important interactions between product markets and corporate governance systems, market competition alone cannot solve the market failures arising from asymmetric information, hold-up, and principal-agent problems that are at the heart of the corporate governance problem.  Market failures resulting in socially inefficient outcomes are one of the strongest arguments in favour of policy intervention.  For example, product market competition does not prevent managers from expropriating shareholders’ rents, with the consequential effect of producing sub-optimal levels of investment.  Therefore, solving this problem requires more than encouraging product market competition.

However, we should keep in mind that the effectiveness and form of different corporate governance systems may be influenced by a number of factors, including product market competition, the structure of capital and labour markets, and the regulatory and legal environments.  For example, as Mayer (1996) states, where there is limited competition in capital markets for the ownership of firms (i.e. lack of an active take-over market), product market competition may be especially needed to encourage good corporate performance.  It has also been suggested, see Petersen and Rajan (1995), that competition in financial markets makes it difficult for firms to establish long-term relationships with financial institutions.

Shleifer and Vishny (1997) argue that much of the differences in corporate governance systems around the world stem from varying regulatory and legal environments.  They maintain that the differences between corporate governance systems in OECD countries, while important, are relatively small compared to the difference between these countries and others.  For example, in less developed countries corporate governance mechanisms may be non-existent and, where they do exist, are often particularly weak and ineffective.  However, even in rich OECD countries, corporate governance problems can still act as a major impediment to economic growth.  For example, Barca (1995) and Pagano, Panetta, and Zingales (1995) claim that Italian corporate governance mechanisms are so underdeveloped as to retard the flow of external capital to firms.  Therefore, understanding corporate governance and its effects can guide policy discussions, not only on the improvements in OECD countries’ corporate governance systems, but also provide a basis for understanding the changes that may be required in other countries where corporate governance systems are severely underdeveloped.

These interactions between corporate governance, competition, and regulatory and legal environments, leads to a systems approach to governance, see Aoki (1994).  The drawback of a systemic approach to corporate governance is that it is often difficult to formulate in any precise way the interactions between different parts of an economy.  However, whenever possible and feasible, policies which promote specific forms of corporate governance should attempt to account for the interactions between governance and other institutional factors, e.g. the legal and regulatory environment, the structure of product markets, and labour and capital markets, etc.  It is important, therefore, that the governance of companies be considered in the context of the overall properties and structure of economies.

III. Corporate Governance in OECD Countries: Strengths, Weaknesses, and Economic Implications

One of the most striking differences between countries corporate governance systems are the contrasts in the ownership and control of firms that exist across countries.  Corporate governance systems can be distinguished according to the degree of ownership concentration and the identity of controlling shareholders.  While some systems are characterised by wide dispersed ownership (outsider systems), others tend to be characterised by concentrated ownership (insider systems) where the controlling shareholder may be an individual, family holding, bloc alliance, or financial institution and other corporations acting through a holding company or via cross shareholdings.  Therefore, two of the most basic conflicts that can occur in corporate governance are the conflict between a controlling manager and ‘outside’ widely dispersed shareholders, and the conflict between ‘inside’ controlling shareholders and outside minority shareholders, see Shleifer and Vishny (1997) and Becht (1997).  While the interface between management and dispersed shareholders has been an extensively studied aspect of corporate governance, the consequences of the relationship between large controlling shareholders and weak minority shareholders is less well understood.  However, it is this latter relationship that is the most common form of corporate governance found in both OECD and non-OECD countries.

Table 1 below gives some international comparisons of ownership concentrations across countries.^[7]  Until recently, lack of hard data has been a major impediment to research in this area.  The European Corporate Governance Network (ECGN) was founded in 1996 as a vehicle for encouraging comparative empirical research on corporate governance in Europe.  For the European countries, therefore, it was possible to get the most up-to-date information available based on the recent work in this area by the ECGN.^[8]   Data for the US and Japan, on the other hand, has had to rely on older studies and is the most recent available.  As can be seen from the table the average equity holding of the largest shareholder^[9]varies from 40% to 80% in most continental European countries whereas ownership concentration is substantially lower in the UK, US, Japan, and the Netherlands.  Additional information on ownership concentration is provided by Franks and Mayer (1995), who find that in Germany and France, 80% of large publicly quoted companies have a single shareholder that owns more than 25% of the stock, while in the United Kingdom, the equivalent figure is 16%.  Furthermore, the authors find that in more than half of the largest French and German firms there is a single majority shareholder i.e. a single shareholder owns more than 50% of the stock; whereas the equivalent figure for the UK is 6%.

Table 1.  International Comparison of Ownership Concentration

	Year	Number of companies	Average Largest Stake (% of equity)
Austria	1996	600 largest listed and unlisted non-financial companies 62 largest listed companies	82.2 52.4
Belgium	1995	135 listed companies	44.8
France	1996	282,322 companies listed and unlisted companies 680 listed companies only	66.2 57.9
Germany	1996	402 listed companies	55.9
Italy	1996	4173 listed and unlisted manufacturing companies 214 listed companies only	61.1 48.0
Netherlands	1996	137 listed companies	26.9 45.3a
Spain	1995	394 listed companies	38.2 47.1b
United Kingdom	1992	189 listed companies	14.4
United States	1980	457 listed non-financial companies	25.4a
Japan	1984	143 mining and manufacturing companies	33.1a
Korea	1996	30 largest chaebols	44.1c

Source: Austria: Gugler et al. (1999); Belgium: Becht et al. (1999); France: Bloch and Kremp (1999); Germany: Becht and Bohmer (1999); Italy: Bianchi et al. (1999); Netherlands: de Jong et al. (1999); Spain: Crespi-Cladera and Garcia-Cestona (1999); United Kingdom: Goergen and Renneboog (1999); United States and Japan: Prowse (1994); Korea: OECD (1998b) ^a Percentage of outstanding shares owned by the largest five shareholders. ^b Percentage of outstanding shares owned by the largest three shareholders.

^c Share of ownership held by the founding family and relatives plus those of other companies in the group.

However, in the absence of one share - one vote, data on direct ownership concentration can either under or overestimate actual control that shareholders exercise over the corporation.  This is because countries have a variety of legal devices that can be used to separate ownership (i.e. cash-flow rights) from control (i.e. voting) rights.  Box 1 below provides a taxonomy of ownership and voting power.  Dispersed ownership and dispersed voting power is associated with many small shareholders, the absence of any large shareholders, and with one share - one vote.  In this case, managers end up with substantial residual control rights but takeovers are possible.  However, it is also possible to have dispersed ownership but concentrate voting power through the use of dual class shares, golden shares, proxy votes and voting trusts, the issue of stock with contingent voting rights triggered by control disputes, and pyramiding.  This gives rise to a situation of strong controlling blockholders and weak minority owners.  Takeovers may be impossible in this case.  When ownership is concentrated and voting rights are aligned with ownership rights, minority owners are again in a very weak position.  While similar to the case where ownership is dispersed and voting power is concentrated, managers in the latter case are also weak and the majority owners retain residual control over the corporation.  However, takeovers are possible in this case since cash flow rights and voting rights are aligned.  Systems of corporate governance found in OECD countries usually fall into one of these three categories.

It is also possible to have concentrated ownership but dispersed voting power.  This can be accomplished through the use of voting caps (i.e. restrictions on voting rights of large share positions) which can be used to prevent large shareholders from exercising control.  For example, although a shareholder may own 40% of the shares, there is a restriction that allows only 10% of the shares to be voted.  This effectively gives rise to a situation of concentrated ownership but where the means for performing direct monitoring are not available due to the dispersion in voting power.  Therefore, although this type of system affords some protection to minority shareholders from voting rights restrictions, concentrated ownership but dispersed voting power has mainly disadvantages from a corporate governance perspective.  This is because not only does it insulate management from effective direct monitoring by the owners, but it also protects them from more indirect means such as hostile take-overs.  For example, when cash-flow and control incentives are misaligned there are few means of intervention available to remove inefficient management i.e. take-overs are difficult, also resulting in “strong managers and weak owners”. This system also results in low liquidity, low diversification possibilities for investors, and a high cost of capital.  Therefore this system has all the consequential problems of monitoring found in outsider systems but none of its advantages.  While an individual firm might chose this ownership structure, this governance structure is the exception rather than the norm.  For these reasons, as a system of corporate governance, this structure is not really observed in practice.

Box 1.  Taxonomy of Ownership and Voting Power

	Dispersed Voting Power	Concentrated Voting Power
Dispersed Ownership	Many small shareholders	Many small shareholders
	One share/One vote	Voting power concentrated in the hands of
		blockholders via dual class shares, golden
		shares, proxy votes, voting trusts, etc.
	Implications:	Implications:
	“Strong Managers, Weak Owners”	“Strong Voting Blockholders, Weak
	Takeovers are possible	Minority Owners”
		Takeovers are impossible
Concentrated Ownership	Large shareholders	Large Shareholders
	Voting power of ownership diluted via	Voting rights aligned with ownership rights
	capped voting	via one share/one vote, or concentrated
		via separation devices.
	Implications:	Implications:
	“Strong Managers, Weak Owners”	“Weak Managers, Weak Minority Owners,
	Takeovers are difficult	Strong Majority Owners”
		Takeovers are possible

Table 2 below provides a comparison of voting power concentration in a number of OECD countries.  As can be seen from the table, blockholdings in Continental Europe are considerably higher than in the United States and United Kingdom.^[10]  The median largest voting stake in listed companies in most European countries is over 50%.  This suggests that voting control by a large blockholder is the rule rather than the exception in Europe.  Furthermore, in no European country is the median largest shareholder small enough to fall below the 5% disclosure threshold.  In addition, in the United States over 50% of companies have a largest shareholder who holds less than 5% of the shares, while there are virtually no such companies in Austria and Germany.

Table 2.    Comparison of Voting Power Concentration in Listed Industrial Companies: Size of Largest Ultimate Voting Block (%)

	Number of companies	Median largest voting block	Mean largest voting block
Austria	50	52.0	54.1
Belgium	121 BEL20	50.6 45.1	41.2 38.3
France	CAC40	20.0	29.4
Germany	374 DAX30	52.1 11.0	49.1 17.3
Italy	216	54.5	48.0
The Netherlands	137	43.5	42.3
Spain	193	34.2	40.1
United Kingdoma United States:	250	9.9	14.4
NYSE NASDAQ	1309 2831	0b 0b	3.6 3.4

Source:  Becht and Roell (1999) and recent  studies by the ECGN can be found on http://ecgn.org ^a Random sample of 250 listed companies. ^b Below the 5% disclosure threshold.

Not only do patterns of ownership concentration and control differ dramatically across countries, but the identity of owners also differs dramatically.  Some comparisons of the distribution of share ownership for a number of OECD countries can be found below in table 3.  In outsider systems of corporate governance, typical of the US and UK, ownership is primarily associated with institutional investors, however, individual ownership is more prevalent in the US than in the UK.  This is not true of insider systems, where ownership is in the hands of either other corporations or controlling investors, see Mayer (1996).  For example, in the rest of Europe, Japan, and Korea, cross-shareholdings amongst firms are commonplace, relationships with banks are strong, and large family holdings often dominate institutional investors.

Table 3.  Ownership of common stock in selected OECD countries Per Cent at year end

Financial Sector of which:	US (1996)	Japan (1994)	Germany (1996)	France (1994)	UK (1994)	Italy (1994)	Sweden (1996)	Australia2 (1996)	Korea (1996)
	46	44	30	8	68	8	30	37	26
Banks & other Financial Institutions	7	281	10	4	10	5	1	4	12
Insurance Companies and Pension Funds	28	161	12	2	50	3	14	25	6
Investment funds	12	–	8	2	8	–	15	8	8
Non-financial Sector of which:	54	56	70	92	32	92	70	63	74
Non-financial enterprises	–	24	42	58	1	25	11	11	21
Individuals	49	24	15	19	21	50	19	20	34
Public authorities	–	1	4	4	1	8	8	–	7
Foreign	5	7	9	11	9	9	32	32	12
Total	100	100	100	100	100	100	100	100	100

Note :  Due to rounding, the figures may not add up to the total.

Pension funds in Japan are managed by trust banks and insurance companies.  Division between banks and insurance companiesare estimated.  No data are available on the extent to which mutual funds own shares.  Securities houses do manage such funds.

These companies are included under other financial institutions.

Australian figures are for end September 1996.

Source : Japan: OECD (1996), Economic Surveys; France and Italy: OECD (1997), Economic Surveys, France; Korea: OECD (1998b), Economic Surveys; Australia: OECD (1998c), Economic Surveys; US, UK, German, and Sweden: OECD (1998d), Financial Market

Trends, no. 69

In addition to differences in ownership concentration and the identity of owners across OECD countries, significant regulatory differences also exist.  Corporate governance is affected by the legal and regulatory framework in which firms operate and vice versa i.e. there is a two-way causal relationship between ownership structures and the legislative environment.  For example, stock exchange regulations regarding dual class-shares and takeover codes, which require firms to make full tender offers once they have acquired a certain percentage of equity, can impinge on ownership structure.  In a system with strong protection of shareholder rights the expropriation of minority shareholders is limited.   Investors anticipating higher returns are ready to pay more for shares.  This, in turn, can induce controlling shareholders to reduce their stakes or give up control, leading to dispersed ownership.  Hence, differences in corporate governance systems around the world strongly reflect differences in regulatory and legal environments.^[11]

On the one hand, although the legal and regulatory environment affects corporate governance, it is also the case that legal rules and regulations are also, in part, the outcome of different corporate governance systems.  For example, systems with dispersed ownership may have a stronger need for regulations that protect shareholder rights.  For example, as ownership structure in the US has become more dispersed, the legislative environment has adapted to the particular needs arising from dispersed ownership.  And many European countries are adapting their legislative environments, in particular the strengthening of minority shareholder protection, in response to abuses by controlling shareholders that can arise in their systems of corporate governance.

Differences in countries’ systems of corporate governance with respect to ownership concentration, the identity of owners, and the regulatory and legislative framework, all have important implications for both firm performance and economic performance.  As noted above, the main agency problems in outsider systems stem from the conflicts of interest between managers and dispersed shareholders, while insider systems generally have large blockholders who exercise control over management.  In the latter case, therefore, the main conflict of interest is between controlling blockholders and weak minority shareholders.  These differences are associated not only with the degree of monitoring and control which owners exercise, but also with differences in the degree of commitment and trust which exist amongst stakeholders.  For example, the identity of owners can affect firm performance through the incentives they provide for various stakeholders to make firm-specific investments.  This in turn can impact upon the structure of industry and underlying economic performance.  For example, ownership by other firms, cross shareholding, and pyramiding can influence the behaviour of firms in product markets.^[12]The remainder of this section examines some of these potential channels of influence and looks at the various strengths and weaknesses of outsider and insider systems of corporate governance.  Section IV, on the other hand, examines the available empirical evidence on the link between corporate governance and performance.

III.1 Outsider Systems of Corporate Governance

“Outsider” systems, typical of the United States and the United Kingdom, are characterised by relatively widely dispersed share ownership and high turnover.  These systems tend to foster a more open and equitable distribution of information and place a stronger emphasis on the protection of shareholders rights and, in particular, those of minority investors.  Systems that protect minority shareholders discourage active corporate governance since they give rise to an absence of concentration of ownership.  Since strong minority shareholder protection is also associated with an active stock market, the corporate governance frameworks in the US and UK are designed to promote stock market activity.  For example, minority investors are normally afforded a high degree of protection in securities law, and the requirements for disclosure tend to be relatively stringent.

Regulation in outsider systems has traditionally been structured to strike a balance between providing adequate shareholder protection whilst, at the same time, allowing investors to assume risks as they see fit.  This requires a framework that emphasises the need for reliable and adequate information, so that investors are able to make informed investment decisions.  For example, regulation has traditionally been structured to provide information to, and create relative equality among, investors regarding access to information.  For this purpose, disclosure requirements are fairly stringent and there is a strong emphasis on the protection of shareholder rights.  For example, elaborate rules prevent groups of shareholders from communicating and sharing information among themselves without making the information available to all shareholders.  In addition, the legal framework supports the rights of shareholders to control the company and in many cases the board and management are explicitly accountable to shareholders.  Theoretically, shareholders (through the use of their voting rights) have the power to select members of the board and to vote upon key issues facing the company.  In practice, however, the fragmentation of ownership is a serious barrier to the actual exercising of such control.

The promotion of financial markets is also important for sectoral development.  Industry sectors that rely on external funding are favored in outsider systems, where there is strong protection of minority shareholders and more transparency.  Another important aspect of an active equity market is that this also encourages innovative activity, entrepreneurship, and the development of a dynamic small and mediumsized enterprise (SME) sector.  For example, venture capital and business angels are vital in supporting innovative activity and entrepreneurial talent, and these are intimately linked with stock markets since this provides investors with both an exit mechanism and liquidity.  If we believe that active financial markets are linked with economic development, then regulations that promote stock market activity may provide one of the underlying sources for economic growth.  For example, the protection of minority shareholders, which is linked to the development of stock markets, is critical to promoting innovative activity, entrepreneurship, and the development of sectors that rely on external funding.

The corporate governance framework in outsider systems also favours the use of public capital markets.  The importance of equities and corporate bonds as a source of long-term finance is an important feature of outsider systems.  Debt financing by banks tends to be short term and banks tend to maintain arm’s length relationships with the corporate sector.  For this reason, debt-equity ratios are relatively low. Capital markets in outsider systems also play a much greater role in influencing the behaviour of key parties.  The monitoring of management relies largely on the discipline of capital markets, which is thought to serve as a particularly effective device for disciplining managerial behaviour.  For example, share prices are likely to fall whenever management fails to maximise shareholder value, exposing the company to the threat of a take-over bid and the removal of inefficient management.  However, liquid stock markets, strict trading rules, and adequate disclosure of information (all characteristics associated with outsider systems) are necessary in order for the market for corporate control to act as effective disciplining device.

Some of the advantages of dispersed ownership, therefore, include enhanced liquidity of stocks and, consequently, better risk diversification possibilities for investors.  However, according to the principal-agent model, one would expect the separation of ownership and control to have damaging effects on the performance of firms.  These difficulties are even greater when ownership is dispersed.  Not only do the interests and objectives of managers and shareholders diverge, but the incentives to monitor management are also particularly weak in this case.  This is because the benefits from monitoring are shared with all shareholders, whereas the full cost of monitoring are incurred by those who monitor. Therefore, shareholders have an incentive to “free ride” in the hope that other shareholders will do the monitoring on their behalf.  With dispersed ownership, “exit” instead of “voice” are the consequences,^[13]and managers end up with considerable discretion and the possibility to extract private benefits.  For example, managers may use their discretion to maximise firm size rather than profits; to hoard cash flow rather than pay it out in the form of dividends; to pay themselves excessive salaries; or to entrench and protect themselves from indirect means of corporate control.  Because of this, outsider systems are often associated with a lack of managerial accountability.

An effective corporate governance framework can limit the scope for managerial discretion.

Therefore, the corporate governance, legal, and regulatory frameworks in outsider systems have developed in response to the particular problems arising not only from the separation of ownership and control, but also from the diffuse nature of share ownership.  Many of the reforms or practices that have arisen aim at addressing weaknesses in monitoring; at strengthening managerial accountability; and at aligning the objectives of managers more closely with those of shareholders.  In addition to a strong disclosure regime and an emphasis on shareholder rights, examples include monitoring by institutional investors or pension funds, well-functioning markets for corporate control, and executive remuneration and compensation packages aimed at encouraging managers to maximise shareholder value, etc.

III.1.1 The role of financial institutions

Share ownership in the UK, and increasingly in the US, is characterised by the domination of institutional investors (see table 3 above).  In the last forty years, both the US and the UK have seen a sharp rise in the proportion of equity held by financial institutions, coupled with the declining role of individuals in direct ownership.  Rising from less than 10% of all equities in the 1950’s to over 40% today, institutional investors have become the largest owners of equity in both countries.  Over the same period, the proportion of shares directly held by individuals has fallen from about 90% to around 50% in the US, and from about 50% to around 20% in the UK.^[14]  The reasons for this trend can be attributed to the tax incentives extended by governments to collective schemes and the wide growth of mutual funds and unit trusts as a result of the advantages of wider diversification, professional management, and lower execution costs relative to direct share ownership.  A further factor explaining this phenomenon has been the increasing tendency for companies to issue shares directly to institutional investors in the primary market, as their funding requirements have risen in recent years.

The high degree of share ownership held by financial institutions is a major factor in the way corporate governance is exercised in the UK and US.  Box 2 below provides an overview of the role of financial institutions in the US corporate governance framework.  Share ownership by institutional investors in the UK is also highly concentrated.  For example, the top 25 institutional investors in the UK control more than 40% of the value of shares held by all institutional investors.  The largest shareholders of listed companies are often the institutional investors and, as a consequence, many of the major companies have broadly similar ownership profiles.  The increased monitoring by institutional investors is seen as an improvement in the way corporate governance is exercised in the US and UK, since this is addressing one of the major weaknesses of outsider systems.  While in the US competition in the market can to some extent provide a benchmark for fund performance, there are still concerns as to who monitors the fund managers.  The high level of concentration amongst institutional investors in the UK may only serve to exacerbate this problem.

Box 2. The role of financial institutions in corporate governance in the US

Unlike most of its leading competitors, financial institutions in the US have long been

restricted by tougher legal and regulatory constraints, resulting in a more fragmented

financial sector.  As a result, commercial banks’ equity holdings are especially low.

More than just restricting their share ownership, legislation also hinders banks in their

client advisory capacity.  For instance, where a client of a bank becomes insolvent, the

legal doctrine of ‘equitable subordination’ downgrades the bank’s claim on the firm’s

assets if it can be shown that they exerted a significant influence on the activities of the

firm.  Therefore, in comparison with other countries, commercial banks in the US play

only a passive role in the monitoring of listed companies.

Mutual funds have also become increasingly significant and they now account for

approximately 12% of total stocks.  However, their activities are also restricted from a

corporate governance perspective, thus limiting their effectiveness as institutional

monitors.

Of the different types of financial institutions in the US, it is the pension funds that have

fostered the greatest upsurge in institutional monitoring.  Given the use of indexing as an

investment strategy, public employee retirement funds in the mid-1980’s realised that the

only alternative to further enhance fund value entailed boosting corporate performance.

Therefore, institutions like CalPERS (California Public Employees Retirements System)

have played an important role in removing the former CEOs of IBM and General Motors.

However, owing to their closer commercial ties, other pension funds such as employee

sponsored plans, have been less vociferous on corporate governance issues.

The many constraints on institutions holding large blocks of shares in individual

companies, as well as the legal protection afforded to investors, has served to encourage

dispersed ownership.  On the other hand, they have also served to limit the significance of

concentrated ownership in the governance process and to discourage long-term

relationships between financial institutions and the corporate sector.

III.1.2 The role of the board of directors

In addition to the control exercised by institutional investors, another relatively low-cost monitoring device can be found in the board of directors.  The board plays a major role in the corporate governance framework.  The board is mainly responsible for monitoring managerial performance and achieving an adequate return for shareholders, while preventing conflicts of interests and balancing competing demands on the corporation.  When necessary, the board also has the authority to replace the management of the corporation.  For example, if management is under-performing, then the board can replace the current management with new, presumably more efficient, management that will maximise the firm’s profits.  The board is also responsible for reviewing key executive and board remuneration.  Box 3 below provides an example of some of the recommendations that have arisen in the UK during the 1990’s regarding the role of the board in corporate governance.

Box 3. The role of the board in corporate governance in the UK

The Cadbury and Greenbury committees, set up in the early 1990’s following a series of

high profile corporate scandals and collapses, have been credited with being the driving

force behind many recent advances in the corporate governance framework in the United

Kingdom.

In the UK, the board of directors tends to be made up of an equal number of executive

and non-executive directors, the proportion of non-executive directors increasing since

the recommendations of the Cadbury committee in 1992.  Non-executive directors play a

key role in exercising control, and in this regard the Cadbury code recommends a

minimum membership of three on each company board.  The proportion of companies

with combined Chairman and CEO roles, a feature linked to a number of corporate

collapses, has also declined in the same period.

The recommendations of the Cadbury and Greenbury committees were also critical to the

widespread establishment of board committees relating to the audit and strategy of the

corporation.  Audit committees were proposed in order to develop the monitoring role of

non-executive directors.  The recommendations suggest that audit and strategy

committees have a minimum membership of three and exclude all executive directors.

These committees have been almost universally adopted, although some smaller

companies have been slower to respond.

In order for boards to effectively fulfil their monitoring role they must have some degree of independence from management.  While the emphasis in outsider systems is on independence, in reality there is the very serious problem that, like management, the board too can become entrenched.  This is particularly the case when board members are compensated for their activities, and are themselves responsible for overseeing executive and board remuneration.  And while there is a trade-off between compensation that attracts high quality individuals as non-executive board members, this also provides incentives to serve on a large number of boards.  This in turn can interfere with performance, since service on too many boards reduces the monitoring ability of board members.  Even if regulations were to limit the number of board positions that can be held, when you have dispersed ownership, agency problems arising from the separation and ownership and control still exist.  Although the board in theory should represent the interests of shareholders and the company, in practice they often become part of the management of the corporation.  Therefore, in outsider systems characterised by weak owners, board members, like management, can easily become entrenched.  Because of these problems, there is still a widely held perception of the board as a relatively weak monitoring device.

III.1.3 The role of market mechanisms

The market for corporate control is perhaps a much more effective disciplinary device than either monitoring by institutional investors or by the board of directors.  Capital markets in outsider systems play a key role in influencing the behaviour of participants in the corporate governance framework. As mentioned above, when then management of a firm is inefficient or failing to maximise shareholder value, this exposes the company to the threat of a take-over bid, with the consequential removal of inefficient management.  In the UK there has been an average of over 200 mergers and acquisitions per year over the last decade, compared with an average of about 50 in Germany.  While up until now the market for corporate control has not been a key feature of insider systems of corporate governance, this is gradually beginning to change, as mergers and acquisition activity is increasing and hostile takeovers are becoming more common.  The US in particular has an active market for corporate control as witnessed by its active market in mergers and acquisitions, including a significant number of hostile take-overs.  For example, in the mid-1980s alone, the value of US mergers totalled approximately one trillion dollars, representing 40% of average annual market capitalisation.  However, the extent of hostile take-overs in the US may be somewhat overstated, with only 172 successful bids between 1985 and 1989.  Nevertheless, the mere threat of a take-over may be enough to act as an effective disciplining mechanism and to diminish the motivation for managerial opportunism.

The intensity of the mergers and acquisitions market is not in and of itself evidence of a powerful disciplinary device at work.  Take-overs can be prompted by rent seeking behaviour, empire building, and tax minimisation, as much as from a desire to boost efficiency levels.  Therefore, it is not always correct to assume that a buoyant market for corporate control reflects the true extent of corporate monitoring.  Even if this were not the case, it should not be forgotten that with legal, advisory, and financing costs constituting on average 4% of the purchase price, this is a particularly expensive way of aligning the interests of management with those of shareholders.  Nevertheless, the mechanism is effective and should not be inhibited.

Product market competition can to some extent act to reduce the scope for managerial inefficiency and opportunism.  This is because there are fewer rents to be expropriated when markets are competitive.  Competition also provides a benchmark by which the performance of the firm can be judged when compared with the performance of other firms in a similar sector.  However, the effects of product market competition are slow to act, forcing inefficient companies into bankruptcy only after a long period of time has elapsed, by which point most shareholder value has been eradicated.  Therefore, bankruptcy legislation, by influencing the claims and control of different investors in the event of insolvency, also plays an important role in corporate governance.

III.1.4 Short-termist behaviour

While some of the advantages of dispersed ownership include enhanced liquidity of stocks and better risk diversification possibilities for investors, critics also argue that the focus in this type of a system can be excessively short-term, reducing overall investment to a level lower that is considered efficient. Although an active market for corporate control can act as a disciplining device on managerial behaviour, in an economy characterised by frequent take-overs, long-term commitments between stakeholders may be more difficult to sustain.  Not only may this reduce overall investment by providing weak incentives for stakeholders to provide firm-specific investments, but it may also create biases in the type of investment projects undertaken, and there are tradeoffs between short-run benefits and long-run impacts on performance.

The heavy reliance in outsider systems on financial markets may encourage managers to focus excessively on projects with short term payoffs even when this is to the detriment of long term corporate performance.  On the other hand, special financial devices have developed in outsider systems (e.g. NASDAQ in specialised capital markets) which may be more effective forms of risk financing for longterm R&D projects than traditional intermediaries in many insider banking systems of corporate governance.  However, projects with longer-term payoffs, such as basic research, may still be undervalued as a result of stock market myopia.  The lack of an adequate measurement for intangible assets also favours applied or targeted types of research at the expense of more exploratory research.  On the other hand, more competitive markets for finance and corporate control can lead to tighter monitoring of research activities by company managers, with a more careful selection of projects and strengthened cost control.  This in turn can lead to increased efficiency in applied research.

Overall, increased competition in product and capital markets, while affecting the level of R&D only marginally, has impacted on the pattern of R&D, shifting research away from basic, exploratory research, towards more applied and visible activities, see OECD (1998e).  This impacts upon innovation, technological development and long-term economic growth since the purpose and contribution of basic research is to increase the pool of knowledge required for applied research, which may be hurt in the long run as a consequence.^[15]  For example, according to a survey of American companies, the average length of research projects has decreased from 21.6 months in 1991 to 16.7 months in 1996 (R&D Magazine, 1997). A reduction in government funding, which favours long-term research, also has the effect of reinforcing market pressures towards more applied R&D.

The question, therefore, is to what extent this shift toward applied research has occurred at the expense of long-term growth potential, which depends primarily on science and basic research.  For example, this shift results in a gradual weakening of potential technological opportunities, eroding the basis for innovation, as well as the potential gains from technology diffusion -- eventually hampering longterm economic growth and job creation.  While a shift in R&D expenditures towards short-term, highpayoff projects may have the effect of spurring higher productivity growth in the short to medium turn, any expansion thus triggered should be of a transitory nature.  Therefore, policy makers in systems of corporate governance which tend to be more short-termist in nature, should pay particular attention to the potential consequences of reducing government funding for basic research.  This is further underpinned by the reported perception of firms that the exploratory component of their research activities needs to be funded by government due to a low private rate of return.^[16]

By rejecting projects in which returns fail to satisfy investor demand for more rapid payoffs, it does not necessarily follow that managers are inefficient or guilty of short-sightedness.  Managers in this case are merely following the dictates of the market.  However, executive compensation plans geared to align the interest of managers with those of shareholders may serve to exacerbate short-termist behaviour. For instance, when managerial remuneration is due largely to stocks or stock options, managers have an incentive to maximise short-term results in order to increase their own compensation.  Managers can maximise the total benefit to themselves by engaging in projects that maximise short-term shareholder value and then cash in on their stocks while moving to an executive position with another corporation.  At the same time they benefit from the reputation effects of having increased shareholder value in their previous firm of employment.  Therefore, while executive remuneration consisting largely of shares or stock-options would seem to align the interests of managers with those of shareholders, they may also create some of the wrong incentives.  This would also seem to concur with the view that executive pay packages in the UK and US are becoming excessive.  This, in turn, has coincided with a move by shareholders in these systems to limit executive remuneration, or at least to have it more closely linked to corporate performance.

III.2 Insider Systems of Corporate Governance

“Insider” systems typical of Europe (except UK), Japan and Korea, are characterised by concentrated ownership or voting power and a multiplicity of inter-firm relationships and corporate holdings.  This is very different from the structure of companies in the US and UK, that are characterised by dispersed outside shareholdings (see tables 1, 2 and 3 above).  Holding companies, banks, other nonfinancial corporations, and familial control are dominant features of insider systems.  This includes close relationships with banks, cross-shareholdings (both horizontal and vertical), and pyramidal structures of corporate holdings.  The significance of a pyramidal structure is that it allows shareholders at the top of the pyramid to exercise control in disproportion to their actual holdings.  Cross-shareholdings, pyramiding, dual-class shares, proxy votes, and voting trusts can all help shareholders extend their control at relatively low cost.  Institutional investors such as pension funds, mutual funds, and insurance companies also tend to play a much smaller role in corporate governance than is the case in outsider systems.

The advantage of concentrated ownership or concentrated voting power is that it can overcome the problems with the monitoring of management that are associated with dispersed ownership.  This is because when cash flow rights and control rights are aligned, majority shareholders now have both the incentive and the power to monitor management.  And with dispersed ownership but concentrated voting power it is controlling blockholders who have an incentive to engage in active monitoring.  This is because, with concentrated voting power (or ownership) controlling blockholders and majority shareholders obtain a large fraction of the benefits from monitoring, and the concentrated voting rights gives them the necessary power to influence the decision making process.  The basic conflict, therefore, that arises with insider systems of corporate governance is between controlling shareholders (or blockholders) and outside minority shareholders i.e. “strong voting blockholders, weak minority owners” or “weak managers, weak minority owners, strong majority owners”.

With dispersed ownership but concentrated voting power, management entrenchment is also a possibility, especially if the blockholders are managers themselves.  Morck et. al. (1988) define management entrenchment as the situation where a “manager who controls a substantial fraction of the firm’s equity may have enough voting power or influence to guarantee his employment with the firm at an attractive salary.”^[17]  The possibility of management entrenchment is one of the arguments often used in support of the one-share/one-vote principle.  This is not to say that management entrenchment does not arise in outsider systems.  The weak monitoring incentives associated with dispersed ownership can also lead to management entrenchment.  But at least with one-share/one-vote take-overs are possible, whereas with dispersed ownership and concentrated voting power, take-overs become impossible.

Although concentrated voting power has the advantage of increased monitoring, and in principle increased firm performance, the controlling owner also has an incentive to extract private benefits. Concentrated ownership or voting power raises the possibility that large blockholders or majority shareholders collude with management at the expense of small shareholders.  One of the consequences of rent extraction by controlling shareholders is that it raises the cost of equity capital as minority shareholders demand a premium on shares issued.  This problem may become particularly acute when small investors do not have enough legal rights to secure a return on their investment.  In this case, ownership concentration and voting power concentration can become detrimental, since small investors avoid holding shares and the flow of external capital to firms is severely impeded, see Shleifer and Vishny (1997), La Porta et. al. (1997), and Barca (1995).

The problem of rent extraction is particularly severe in the case of small shareholders of listed companies that belong to pyramidal structures e.g. holding companies.  When there is a difference between cash flow rights and voting rights, the incentives for extracting private benefits are much stronger and collusion between managers and blockholders is more likely.  While in the US and UK managers should in principle maximise shareholder value, in Continental Europe managers are often forced to maximise “blockholder value”, and this does not necessarily maximise minority shareholders’ returns.  In this case, minority owners can also be expropriated though intra-group transfers by blockholders that control the group.  For example, blockholders may have managers pursue objectives that are more profitable to them by diverting resources to other companies owned by the blockholders, see Becht (1997) and Barca (1997). Blockholders could also agree to vote favourably on management sponsored proposals and be compensated by side payments.  These incentives arise because blockholders only bear a fraction of the costs of these payments by foregoing the dividend payments associated with their cash flow rights, but receive the full benefits associated with any side payment.

One of the consequences of rent extraction in insider systems is the lack of liquidity in secondary markets as investors withhold funds, and a lack of opportunities for risk diversification as a consequence of illiquid markets.  Capital markets in insider systems therefore tend to be much less well developed than those found in outsider systems.  On the other hand, concentrated ownership not only increases the incentives for monitoring, with presumably positive benefits for firm performance, but it also encourages more long-term relationships and commitment amongst stakeholders.  This, in turn, can also impact upon firm performance, increasing profitability in the long run.  Therefore, although capital markets in insider systems tend to be much less developed, the long-term nature of relationships in insider systems provides incentives that encourage a greater investment in firm-specific assets.

III.2.1 The role of banks in corporate governance

Long-term relations with financial institutions can affect the performance of the corporate sector. Differences in corporate governance systems are thought to influence the cost of capital and the availability and type of financing available to firms.  For example, stock market capitalisation as a percentage of GDP in insider systems is normally lower than that found in outsider systems, see table 4 below.  If the development of financial markets is linked with economic development then this can impinge upon economic growth.  However, this may not matter if there are other sources of financing available to the corporate sector. For this reason, in insider systems there is a much greater emphasis on banks as providers of external finance and debt/equity ratios are typically higher.

Unlike the arms length relationships between banks and corporate clients found in outsider systems, banks in insider systems tend to maintain more complex and longer term relationships with the corporate sector, see box 4.  In particular, the German and Japanese systems of corporate governance are characterised by long-term relationships with banks which are thought to encourage bank financing, whereas firms in the US and UK benefit from high levels of equity capital.  The benefits of a ‘bank-based’ system are that banks perform important monitoring and screening functions.  The close relationships between banks and client firms in insider systems provides greater access to firm-specific information, and is thought to be a factor contributing to lower risk premiums, thus lowering the overall cost of capital faced by firms.  For example, two of the principle assertions as to the merits of ‘bank-based’ systems are that it reduces asymmetric information problems enabling banks to supply more external finance to firms at a lower cost, and thus increasing investment; and that it increases monitoring, thus ensuring firms are run more efficiently.

Table 4. Market capitalisation of listed domestic equity issues

As per cent of GDP at year-end

Australia (Assoc. of SE)	1975	1980	1985	1990	1993	1994	1995	1996
	22	40	37	37	71	67	70	80
Austria	3	3	7	17	16	16	14	15
Belgium Canada	15	8	26	33	37	36	37	44
(Toronto and Vancouver)	30	45	45	43	61	59	66	86
Denmark	11	8	26	30	31	34	33	41
Finland	–	–	–	17	28	39	35	49
France	10	8	15	26	36	34	32	38
Germany (Assoc. of SE)	12	9	29	22	24	24	24	28
Greece	–	–	–	–	–	–	14	19
Ireland	–	–	–	–	–	–	40	49
Italy1	5	6	14	14	15	18	19	21
Japan	28	36	71	99	68	77	69	66
Korea	–	–	–	43	42	50	40	29
Mexico	–	–	–	16	50	31	32	32
Netherlands	21	17	47	42	58	67	72	95
New Zealand	–	–	39	20	56	53	53	56
Norway	–	–	16	23	24	30	30	36
Spain	32	8	12	23	25	25	27	33
Sweden	3	10	37	40	58	66	75	95
Switzerland2	30	42	91	69	114	109	129	136
Turkey	–	–	–	–	20	17	12	17
United Kingdom United States	37	38	77	87	122	114	122	142
(NYSE, Amex and Nasdaq)3	48	50	57	6	81	75	98	114
Italy – All Italy on a net basis since 1985. Switzerland – only Zurich through 1990. United States – including foreign shares in 1975.

Source: OECD(1998), Financial Market Trends, February.

On the other hand, the emergence and survival of new firms is strongly affected by the possibility and cost of obtaining finance.  In insider systems, characterised by small and illiquid public capital markets and the absence of venture capital markets, new firms and SMEs may find it very difficult to obtain equity financing.  Therefore, the dominant financing pattern for firm start-ups and small firms in insider systems implies a heavier reliance on debt financing than found in outsider systems.  This is a serious problem for new firms, since they have no established track record or long term relationship with the financial sector. Banks, in this case, tend to be too conservative in their lending policies.  This is because banks face an asymmetric risk when assessing new start-ups.  For example, in the worst case scenario the bank can lose all the credit it has extended to a new firm, but should the venture succeed the best the bank can hope for is to be fully repaid, including accrued interest.  In this way, the bank is excessively exposed to downside risk.  Therefore, the absence of an active equity market and a heavy reliance on debt financing, both characteristics of insider systems, can impinge upon the development of a vibrant and thriving SME sector.

Box 4. The role of financial institutions in insider systems of corporate governance

In many of the Continental European countries, commercial banks play a leading role in the

governance of the corporate sector.  Banks tends to be powerful, independent, and mostly private

institutions.  The universal banking system has also enabled them to dominate all facets of financial

intermediation, which in turn has resulted in capital markets remaining considerably less developed

than in outsider systems.  Since information is often shared between the bank and its corporate

client, bank based systems rely on confidentiality.  This runs counter to the requirements of strong

public disclosure associated with outsider systems.

The German system of corporate governance is the classic example of a bank-based system.  The

German tradition has been for each firm to have a house bank that takes responsibility for most of

the financial transactions of the company.  The banks tend to hold considerable equity portfolios

themselves, and are often seen as representing all shareholders.  In addition, banks often name

representatives to the boards and exercise considerable leadership and control in the company with

which they have a relationship.

Many insider systems are also characterised by a pattern of interlocking share ownership among

groups of financial and non-financial companies.  In the Japanese context, control is exerted

through keiretsu structures comprised of groups of financial and industrial companies, including

suppliers and purchasers.  A crucial feature of the Japanese system is the emergence of banks as a

significant element of corporate governance, where banks exercise their control over companies in

the group through a combination of shareholding and lending activities.  Therefore, the main

external control over managers is by the main bank, although this is rarely the largest stakeholder

in the company.  The largest stakeholder usually being another non-financial corporation.

In other countries, such as Korea, ownership and control tends to be characterised by a small

number of ‘founding’ families of entrepreneurs on the one hand, and a pervasive state on the other.

The government in the past has been able to exert immense pressure on the banking sector and

direct much of their lending activities, often on the basis of political connections rather than a

proper evaluation of risk.  Coupled with a tendency for chaebol owners to focus on growth at the

expense of overall profitability, this led to inadequate risk assessment in the expansion and

diversification of chaebols.  The prevalence of cross-shareholdings and intra group debt guarantees

between group companies only served to exacerbate this problem.

However, long-term relationships with banks can also reduce biases that might favour investments that generate short-term improvements in performance, see von Thadden (1995).  For example, a lower cost of capital, by lowering the discount rate applied to the future stream of profits, encourages more long-term investment.  This, in turn, can provide firms with a competitive advantage. The low cost of funds is thought to have contributed to the relatively high level of domestic investment that has been a strong feature of the Japanese economy in the past.  Therefore, long-term investments without an immediate payoff, such as R&D and innovation, are more likely to be undertaken in corporate governance systems that are based on long-term relationships with financial institutions.  While banks in general may have a lower tolerance to risk than shareholders, the close nature of the relationship tends to reduce asymmetric information and risks associated with uncertainty.  However, the overall effect of these opposing influences makes it difficult to predict which system of corporate governance is more likely to promote innovative activity.

The role of financial institutions in financing failing companies is another important distinction between different countries corporate governance systems.  Asymmetric information problems become particularly important in the refinancing of failing firms.  The failure of creditors to be able to distinguish between firms with good or bad prospects during periods of financial difficulties, can result in premature liquidations.^[18]  Insider systems, with their closer relationships between banks and the corporate sector, can mitigate some of these information problems and may result in a more efficient allocation of resources. Furthermore, competition in financial markets in outsider systems may also undermine the development of long term relations between firms and financial institutions.  For example, the willingness of banks to provide rescue finance for failing firms may hinge on the expectation that these investments will yield long term returns.  However, when there is competition in financial markets and firms are free to shift to the lowest cost supplier of finance once they are out of financial distress, then there is little incentive for banks to provide rescue funds to the corporate sector.  Banks in insider systems also play a much more important role in the restructuring of poorly performing firms.  Mayer (1996) states that restructuring of poorly performing firms by Japanese banks is an important feature of the country’s financial system; and that financial institutions (e.g. banks, pension funds, and life assurance companies) in the US and UK often intervene too late in corporate restructuring.

III.2.2 Long-term relationships: intra-group holdings and commitment

The focus in insider systems is much more on building long term relationships with many of the contractual partners of the firm, thereby encouraging greater trust, loyalty and commitment among the various stakeholders.  It is thought that the continuous nature of relations between these groups, including banks, the workforce, contractors and clients, promotes a greater investment in firm-specific assets. Whereas, outsider systems of widely dispersed ownership, where shareholders can walk away from relations with other stakeholders without suffering any costs, makes it difficult to sustain trust and commitment and do not provide the right incentives to encourage firm-specific investment, see Mayer (1996).

The importance of commitment and trust for firm performance can also vary by industry sector or type of productive activity, see Williamson (1985), Mayer (1996) and Maher (1997). For example Mayer (1996) states that they are “particularly important where productive activity depends on the involvement of and investment by a large number of stakeholders.  Complex manufacturing processes, which require several different supplier and purchaser arrangements, may be particularly dependent on ownership patterns that promote commitment and trust.  Ownership patterns are also relevant to activities that may require firm specific investments by employees in training and acquisition of skills.  Incentives to undertake such investments may require commitments by employers to long-term employment and promotion policies within the firm.”^[19]   Long-term relations and commitment are therefore particularly important in high technology industries or activities with high asset specificity.  In these industries it is the firm-specific investments made by various stakeholders, rather than the flow of external funds, that impinges upon the performance of the firm.^[20]

Apart from the concentration of ownership, the identity of owners can also have major economic implications.  Where ownership of a firm is held by other corporations with which the firm has strong trading relations e.g. suppliers or buyers, this provides incentives for firms to make relationship-specific investment, see Williamson (1985).  Ownership by one corporation in another effectively reduces transaction costs and ‘hold-up’ problems associated with incentives to engage in opportunistic behaviour. In this situation, economic stakeholders, therefore, have a greater incentive to invest in relationshipspecific investment.  On the other hand, complex patterns of ownership and large shareholdings can also impede the restructuring of firms and industries, and diminish the adaptability of firms to respond to changing circumstances and globalisation.  The role of the Korean chaebols in connection to the 1997 crisis in Asian markets provides one example amongst many, see OECD (1999c).

One of the problems with complex patterns of ownership and cross-shareholdings is that insiders need not always own an outright majority of a company in order to maintain control.  For example, pyramid structures allow dominant insiders to exercise control over a group with only a small share of the total outstanding equity of the firm.  Pyramids perform an important role in shaping the corporate governance framework in Italy.  Widespread equity linkages in Italy have made it possible for around 150 core groups to exert control over almost 6500 firms.^[21]  In Italy these equity linkages tend to be vertical and unidirectional i.e. at each level of the pyramid, firms own a control stake in firms belonging to the immediately lower level.  On the other hand, the controlling entity is provided with a wide access to risk capital without jeopardising control.  The potential liability of the controlling shareholder/blockholder is limited.  For example, in the event of bankruptcy the controlling shareholder is only liable for the indirect, often minority, stake held in the firm.  Therefore, pyramidal groups by providing a wide access to capital have some of the benefits associated with dispersed ownership, while minimising the agency costs associated with the monitoring of management.  However, this type of group structure is particularly prone to expropriation and a lack of transparency.

In contrast, in Japan equity linkages are normally bi-directional.  While unidirectional linkages ensure stable control over all firms belonging to the pyramidal group, inter-group cross-shareholdings on the other hand facilitate strategic alliances.  Cross shareholdings, whereby corporations with mutual business interests (or belonging to the same group) intentionally hold each other’s shares, can be used to create a significant core of insiders.  In Japan, these cross holding are concentrated among the keiretsu, made up of large corporations across a wide range of industries, with banks at their centre.  Large portions of stocks are held under this type of an arrangement, often with the understanding that the shares will not be sold without the understanding of the issuer.  Through these linkages the strategic and operational decisions of the firm are overseen by the holding company.

Complex patterns of ownership, including both horizontal and vertical arrangements, serve to protect both the group and lower level holdings from hostile take-overs.  Long-term relationships between firms, cemented by cross-shareholdings, may restrict the possibilities for a transfer of share ownership.  In addition, the prevalence and stability of cross-shareholdings in insider systems, makes it difficult to purchase a significant portion of shares in a company.  As a result, the market for corporate control in insider systems is likely to be less well developed than in outsider systems.  The available data suggests that this is indeed the case.  For example, mergers and acquisitions activity in Japan is only marginal.  In Germany hostile take-overs have been virtually absent and in 1988, for example, take-overs were only half that level of that found in the UK.  However, this is gradually changing and takeover activity, both in Japan and in continental Europe, is on the increase, see OECD(1999b).  While on the increase, relative to the US and UK, takeover activity in insider systems remains small in comparison.

Moreover, the absence of an effective market for corporate control may also impede the development of an international production base, and may prevent firms from entering through domestic acquisitions.  In Germany, for example, the limited importance of listed joint-stock companies, the role of cross-shareholdings between partner firms, and the important role of employee representation on company boards, all make it very difficult for outsiders to buy German firms.  At a time of globalisation this may be increasingly costly to firms.

While long term relationships between firms and suppliers, and the subsequent sharing of information, can help promote efficiency gains in terms of costs and quality, the subsequent increase in industry concentration may weaken the overall level of competition in product markets.  The complex patterns of cross-ownership that often arise between related companies in insider systems, and which result in large corporate groups, can also result in collusive behaviour, and has similar implications for competition policy.  This suggests that insider systems of corporate governance need to pay particular attention to strengthening product market competition.

III.3 A Convergence in Systems?

The increasing globalisation of capital markets and liberalising of international trade seem to have created an environment in which differences in corporate governance are becoming less severe.  For example, while family companies often predominate in Australia, Canada, and New Zealand, this pattern of concentrated ownership coexists with a strong recognition of shareholders rights, as well as the importance of greater transparency in the corporate sector.  And in outsider systems there has been a growing appreciation of the powerful monitoring incentives associated with concentrated ownership. Increasingly, institutional investors and pension funds in the United Kingdom and United States are becoming active participants in the corporate governance of firms in which they hold substantial holdings. Venture capital markets and second tier markets such as NASDAQ have also developed to accommodate outside financing towards closely held firms.

Convergence forces at work in both types of system are primarily a result of globalisation of financial markets.  There is also growing evidence that firms are adopting corporate governance arrangements that international investors appear to value.  Firms, and in particular large multinational firms, are increasingly adopting the best practices of existing systems in an effort to improve corporate efficiency and to attract external capital funds.  In addition, the interests of international investors, coupled with the capital requirements of major firms expanding abroad, has led to a number of firms seeking a listing on foreign stock exchanges.  This change in the method of financing is having a major impact on corporate governance.  Raising capital though foreign stock exchanges, where shareholders are more concerned about firm risks, is very different from a bank-based system, where the banks are more interested in the risk of default.  Therefore, in many insider systems of corporate governance, the increasing importance of foreign investors as a source of capital for listed companies, is raising the demand for more transparency and minority shareholder protection.

The agency problems that arise from the separation of ownership and control raises the need for a corporate governance framework which strengthens managerial accountability and encourages managers to maximise profits, rather than pursue their own objectives.  In addition, a good corporate governance framework needs to protect minority shareholders from rent extraction by either managers or controlling shareholders while at the same time encouraging efficient investment by stakeholders.  The means by which this is attained varies widely across countries and, even within a single country, across industrial sectors.  Each country has through time developed a wide variety of capital market mechanisms and financing arrangements, legal and regulatory frameworks, and other mechanisms to address these agency problems.  This is exemplified by the current divergence in corporate governance structures found in OECD countries.

The full implications of recent developments are hard to predict, but there appears to be an overall trend towards a degree of convergence in governance and financing patterns, with outsider systems adopting some of the features of insider systems, and vice versa.  However, the extent of the divergences between systems, which are historically contingent and rooted in cultural, historical and legal differences, suggests that complete convergence is unlikely.  Furthermore, these different systems of corporate governance are converging from different directions.  Therefore, the instruments through which improvements will be attained, and the policy actions that are called for, are also different.

IV. Corporate Governance and Performance: The Empirical Evidence

The previous section showed that the patterns of ownership and control and, thus, the systems of corporate governance varied considerably amongst OECD countries.  Ultimately, what is important is whether or not these different corporate governance arrangements and, in particular, differences in ownership and control affect corporate performance or economic growth.  If at the end of the day, corporate governance has no impact on performance, then it is not clear why policy makers should concern themselves with this topic.  However, both the analytical framework in section II and the discussion in section III above, showed that there are a number of potential channels of influence through which governance can affect performance.  For example, these differences are associated not only with the degree of monitoring and control which owners exercise, but also with the incentives they provide for investment, innovation, and entrepreneurial activity.  As this section will show, the available empirical evidence also suggests that corporate governance does affect performance and is thus an important framework condition for the industrial competitiveness of OECD countries.

The remainder of this section will summarise some of the empirical findings regarding the impact of corporate governance on performance.  In particular, it will look at the effects of different types of ownership structure on firm performance.  In addition, it will examine the available evidence regarding the conflicts that arise between different types of shareholders and, in particular, the possible detrimental effects of dominant shareholders.  Other questions this section addresses include what is the available evidence on entrenchment or rent extraction on the part of management, and does the existence of an active take-over market actually work as a mechanism to correct managerial inefficiencies and affect performance?  It will also examine the available evidence on whether or not executive compensation packages are effective in aligning the interest of managers with those of shareholders and, thereby, increase firm performance.

IV.1 Ownership concentration and firm performance

The ownership and control of firms are pronounced and vary dramatically across OECD countries.  Therefore, one of the questions that arises when considering whether or not corporate governance affects performance includes whether or not owner-controlled firms are more profitable that manager-controlled firms?  A priori it is not clear whether or not concentrated ownership and control will improve performance.  On the one hand, concentrated ownership by providing better monitoring incentives should lead to better performance.  On the other hand, it might also lead to the extraction of private benefits by controlling blockholders at the expense of minority shareholders.  These issues are central to the debate surrounding corporate governance practices, particularly since concentrated holdings are the primary means of control in so many countries around the world.

Therefore, one question to ask is whether or not the agency problem arising from the separation of ownership and control is a serious one and does concentrated ownership effectively overcome these problems?  The principle-agent model suggests that managers are less likely to engage in strictly profit maximising behaviour in the absence of strict monitoring by shareholders.  Therefore, if owner-controlled firms are more profitable than manager-controlled firms, it would seem that insider systems have an advantage in that they provide better monitoring which leads to better performance.  The vast majority of empirical studies, it turns out, do seem to favour the beneficial effects of enhanced monitoring as a result of higher ownership concentration.

Gugler (1999) provides a comprehensive survey of empirical studies of the effects of ownership concentration on corporate performance, beginning with the seminal work of Berle and Means (1932) to more recent work by Leech and Leahy (1991), Prowse (1992), Agrawal and Knoeber (1996), and Cho (1998).  Based primarily on studies from the US and UK, he finds that although the results are ambiguous, the majority of studies find that “owner-controlled” firms significantly outperform “manager-controlled” firms.  Firms are usually classified as owner controlled if there is a single block of equity exceeding 5 or 10 percent.  The dependent variable used in these studies were are all proxies for the performance of the firms as measured by net income/net worth, rate of return on equity or Tobin’s Q, or the riskiness of returns.  Although a number of studies find no significant difference between the two, the number of studies that find that manager-controlled firms outperform owner-controlled firms is negligible.

On balance, therefore, the empirical evidence is supportive of the hypothesis that large shareholders are active monitors in companies, and that direct shareholder monitoring helps boost the overall profitability of firms.  This result is also borne out by studies of managerial turnover.  For example, Franks and Mayer (1994) find a larger turnover of directors when large shareholders are present, again indicating that large shareholders are active monitors.  It seems, therefore, that the beneficial effects of direct monitoring, and a better match between cash flow and control rights, more than outweigh the costs of low diversification opportunities or rent extraction by majority owners.

Although the evidence points to a greater role and fewer restrictions for large shareholders, the policy implications of these results should be viewed with caution.  One of the problems with the numerous studies that exist on the effects of ownership concentration on corporate performance is that they are all based on US or UK samples of firm data.  There are some exceptions however, notably, Round (1976, Australian), Thonet and Poensgen (1979, German), Jacquemin and Ghellinck (1980, French), and Prowse (1992, Japanese).  For example, Thonet and Poensgen (1979) found that for a sample of listed German manufacturing firms, manager-controlled firms significantly outperform owner-controlled firms in terms of profitability, but that owner-controlled firms had higher growth rates.  Jacquemin and Ghellinchk (1980), using French firm data, found no differences between familial and non-familial controlled firms. Prowse (1992) also does not find any relationship between ownership concentration and profitability in Japanese companies.

Although these few studies are not conclusive evidence, they do highlight that the policy conclusions of the results based on US and UK data are not necessarily transferable to other countries.  In the US and UK, levels of ownership concentration are low relative to other countries.  Therefore, although it may be true that more direct shareholder monitoring, fewer restrictions for large shareholders, and more ownership concentration improves performance in the US and UK, this may not be the case in countries where ownership concentration is already relatively high.  Once concentration levels reach very high levels then it is not clear that more monitoring will continue to improve things and may actually work in the opposite direction.  In fact, empirical studies for both the US and UK suggest that at low levels of concentration, performance^[22] increases as concentration increases, but then declines as concentration levels keep increasing, see Morck, Shleifer and Vishny (1988), McConnell and Servaes (1990), Wruck (1989), and Franks, Mayer and Renneboog (1995).

We must also keep in mind that corporate governance structures are not static but dynamic in nature.  Recent changes in corporate governance in the US and UK include the increased importance of institutional investors, which also leads to reduced scope for management discretion, and can act as a substitute for large shareholder monitoring.  Furthermore, transferring the “separation of ownership and control” view of monitoring as suggested by the large body of primarily US and UK economic studies to Continental Europe, Korea, and Japan can be misleading because the economic decision agents are very different from the US/UK scene.  Holding companies, the state, banks, other non-financial corporations, and familial control are dominant features in the rest of Europe, Korea, and Japan.  Different owners will have different objectives, and it is highly likely that the identity of owners will matter for firm performance.  For example, managers of corporations under governmental or quasi-governmental control are likely to have different incentives and will, therefore, behave differently to managers of corporations in the private sector.  For this reason, ownership concentration and the identity of owners should be viewed as variables that exert a simultaneous, but different, influence on firm performance.

In addition, Roe (1994) states that the low ownership concentration in the US compared to other countries may be the result of policies initiated by controlling managers that discourage large holdings e.g. anti-takeover devices.  This implies, that for the US at least, that managers are strong relative to shareholders and that management entrenchment is a serious problem.  Therefore, policy makers in outsider systems like the US and UK should pay particular attention to the negative effects of mechanisms that are often employed by management that inhibit the market for corporate control.  Direct monitoring is just one of many devices that can be used to reduce conflicts between managers and shareholders. Therefore, where direct shareholder monitoring is weak, such as in outsider systems, policy makers need to pay particular attention to the incentive effects created by other control mechanisms such as take-overs or managerial remuneration schemes.

Furthermore, whether or not owner-controlled firms outperform manager-controlled firms, may depend not only on the initial levels of ownership concentration, but also on the industry in question.  A study by Zeckhauser and Pound (1990) finds that whether or not owner-controlled firms outperform manager-controlled firms does indeed depend on the type of industry.  They find that the superior performance of owner-controlled firms holds in industries with relatively low asset specificity (e.g. machinery and paper products), but there was no difference in industries with high asset specificity (e.g. computers).^[23]  This suggests that the nature of the firm’s investment and production decisions influence the asymmetry of information between principal and agent.  For example, in industries where outside monitoring is particularly difficult, such as is the case in high asset specificity industries, large shareholders are less effective in overcoming agency problems.  In this case, additional control mechanisms may be required (e.g. the number of outside directors on the board, the managerial labour market, markets for corporate control, managerial pay schemes, etc.).

The finding that owner-controlled firms are more profitable than manager-controlled firms is also consistent with the life-cycle model of the firm.  Corporate governance failures and bad investment decisions are less likely in the early stages of a firm’s life, a time when investment opportunities are in abundance and equity stakes are concentrated.  However, as firms grow and mature, not only does ownership become more diluted, but investment opportunities also fall short of the cash flow available. Therefore, as firms grow and mature, this provides greater incentives for the increasingly unmonitored management to expropriate rents.  The dilution and dispersion of equity stakes in this case, implies that as firms mature effective corporate governance mechanisms become increasingly important in assuring firm performance.

Finally, several studies also stressed the importance of the effects of product market competition on managerial behaviour.  For example, significant differences in the performance of owner-controlled versus manager-controlled firms are more likely in markets where the scope for managerial entrenchment is much higher i.e. in monopolistic or oligopolistic market structures.  This is because in competitive markets there are fewer rents for management to expropriate, hence, ownership structure is less likely to affect firm performance.  Weaknesses in corporate governance are, therefore, more likely when firms are both management-controlled and when firms exert a high level of market power.  One of the policy implications arising from these findings is that developments in corporate governance legislation should be developed in conjunction with competition and anti-trust policy.

IV.2 Dominant shareholders and the expropriation of minority shareholders

The presence of large shareholders improves the supervision of management and, thereby, enhances firm performance.  However, there is the very serious problem that controlling blockholders or majority shareholders can use the firm for their own private benefit, expropriating rents at the expense of minority shareholders and other stakeholders.  This ex-post expropriation by controlling shareholders is likely to lead to sub-optimal levels of investment by minority investors and other stakeholders.  After all, unless the targets of expropriation are adequately protected, they have a reduced incentive to provide firmspecific investments in the first place.  In addition, when minority investors are less willing to provide equity finance, this can lead to illiquid stock markets and reduced diversification possibilities for investors. Therefore, given the consequential impacts for innovative activity, entrepreneurship and economic growth, it is important to know whether or not expropriation is a problem.

Direct evidence on measuring the extent of expropriation of rents, by either large shareholders or controlling blockholders (which may include management), is difficult to obtain.  Therefore, empirical studies have attempted to measure this in an indirect way.  If control is associated with the ability to expropriate private benefits, then the market should value “control”.  In this case, we would expect to see controlling shares trading at a premium.  For example, if shares with superior voting rights trade at a premium, which is not accounted for by other factors^[24], then this is taken as evidence for significant private benefits of control that may come at the expense of minority shareholders.

In surveys of corporate governance, Shleifer and Vishny (1997) and Gugler (1999) find that the empirical evidence suggests that control is valued, which would not be the case if controlling blockholders or large shareholders received the same benefits as other investors.  For example, Barclay and Holderness (1989, 1992) find that in the US, large blocks of equity trade at a substantial premium to the post-trade price of minority shares, and that on average these blocks trade at a 20% premium.  This supports the hypothesis that purchasers of the block of shares that may have a controlling influence receive private benefits.  Other studies, taking a different approach, also support this hypothesis by comparing the price of shares that have identical dividend rights but differential voting rights.  For the US, DeAngelo and DeAngelo (1985), Jarrell and Poulsen (1988), and Zingales (1995) find that shares with superior voting rights trade at a premium, but that this premium is small.  However, Zingales (1995) finds that the premium rises sharply in situations where control is contested, again implying that controlling blockholders receive private benefits at the expense of minority shareholders.

Evidence from other countries, where concentrated ownership is the norm, suggests that expropriation of private benefits by controlling blockholders at the expense of other stakeholders is a major problem.  While Rydqvist (1987) and Bergstrom and Rydqvist (1990) find a relatively low premium for Sweden, studies on Israel, Italy, and Switzerland find substantial premiums.  For example, while Rydqvist (1987) finds a 6.5% average voting premium for Sweden, Levy (1982) finds a 45.5% premium in Israel, Horner (1988) finds a 20% premium for Switzerland, and Zingales (1994) finds a 82% voting premium on the Milan Stock Exchange.  The large voting premium in Italy suggests high private benefits of control, and Zingales (1994) and Barca (1995) suggest that managers in Italy divert profits to themselves at the expense of non-voting shareholders.  Zingales also measures the average proportion of private benefits to be around 30% of firm value.

The empirical evidence on Israel and Italy, therefore, suggests that agency costs can be very large in some countries.  Zingales (1994) conjectures that these private benefits of control are so large in Italy because the legal system is ineffective in preventing exploitation by controlling blockholders.  This suggests that the development (or strengthening) of policies aimed at protection of minority shareholders may be particularly needed in countries with relatively weak corporate governance or legal systems that enable such expropriation to take place.  For example, systems characterised by pyramidal groups can further exacerbate the problem, since small shareholders can also be expropriated through inter-group transfers.^[25]  However, even if one instituted corporate governance policies with strong protection of minority shareholders, it is likely the policies will be ineffective if legal systems remain weak.  Therefore, what is required is a systemic approach, which demands that problems in the legal and regulatory structure be address simultaneously with changes in corporate governance policies.

Indirect evidence has also been obtained from empirical studies relating firm performance to insider ownership, i.e. to manager or director holdings.  Morck et. al. (1988), for a sample of US firms, find a positive relationship between board ownership and firm performance in the 0-5% ownership range, a negative relationship between 5-25%, and a positive influence of management ownership beyond the 25% level.  Morck et. al. conclude that their findings are consistent with the hypothesis that at first “convergence of interests” effects dominate and managers have greater incentives to maximise firm value as their ownership stakes rise, yet their stakes are too small to become entrenched.  As their ownership stakes rise, then “management entrenchment” outweighs “convergence of interests”, but beyond 25% “convergence of interests” effects dominate again.  McConnell and Servaes (1990) and Belkaoui and Pavlik (1992) obtain similar results.

On the other hand, using US data, Holderness and Sheehan (1988a,b) do not find any differences, and Asquith and Wizman (1990) find only small - if any - transfers between stakeholders.  Their results would imply a rejection of the hypothesis that expropriation of minority shareholders is the primary reason for majority ownership.  However, both the strong evidence and level of expropriation found in other countries, suggest that this conclusion is only valid for the US.  It could well be the case that strong protection of minority investors in the US may prevent controlling shareholders from expropriation by other shareholders.  The evidence suggests that in many other countries expropriation remains a major problem.

Since expropriation by controlling shareholders can deter minority investors, the result is often a small and illiquid public equity market.  This would explain why, in countries where expropriation is a major problem, capital markets remain underdeveloped relative to the US and UK.  La Porta et. al. (1997) ranked investor protection according to a number of dimensions of shareholder and creditor protection.^[26]The authors find that investors are best protected in English common law countries, are somewhat protected by German and Scandinavian codes, and are most vulnerable in countries of French origin. They find that good shareholder protection is one determinant of liquid equity markets.  Table 4 in section III also shows that stock markets play a much larger role in countries like the UK and US, where investor protection is high, than in Continental Europe countries, such as Austria, Italy, Spain and Germany, where minority shareholder protection is weaker.  This is because when expropriation is limited by law minority investors anticipate higher returns and are ready to pay more for shares.  This can induce controlling shareholders to reduce their stakes or give up control, which in turn, leads to more liquid markets and dispersed ownership.

Small and illiquid equity markets need not be a concern if other sources of finance (internal and external) still lead to an optimal level and mix of investment.  However, it is not clear whether debt finance is an adequate substitute for equity finance for all types of investment.  This is particularly the case in industries where there is a great deal of asset specificity, and where uncertainty and risks are high e.g. high technology industries.  There is evidence to suggest that R&D investment might be adversely affected when the main source of outside finance is debt.  In addition, the higher risk and greater asymmetry of information associated with R&D projects in comparison to standard investments in physical capital stock especially inhibits new high-tech firms from obtaining bank loans, see Gugler (1999).  Figure 1 below depicts the relationship between stock market capitalisation and R&D spending for 14 OECD countries in 1994.  A regression of the effect of the ratio of stock market capitalisation to GDP on the R&D/GDP ratio shows that a 10% point “deepening” of a country’s stock market increases the R&D/GDP ratio by about 0.12 percentage points.  This result is also confirmed in studies by Long and Malitz (1985) and Bradley et. al. (1984), who find a negative correlation between leverage and R&D activity.  It would seem therefore that countries with liquid equity markets tend to invest more in R&D activity and high-tech firm start-ups. This finding has implications for a country’s future productivity and growth.

Overall, the empirical evidence therefore supports the view that there are potential conflicts of interest between dominant shareholders and other stakeholders in the firm, and that there are detrimental effects associated with expropriation.  The source of this conflict is not just the structure of ownership but is also due, in part, to the legislative framework which protect investor rights.  While Barclay and Holderness (1992) argue against legal provisions that may inhibit or postpone the exercise of control by acquirers of large blocks, such prescriptions seem inappropriate in a continental European context where the protection of minority shareholders is much weaker.  Recent legislative changes in some European countries are indeed aimed at correcting the expropriation of rents by controlling shareholders and at moving in the direction of enhanced minority shareholder protection.  These legislative moves include preventing acquirers of large blocks from voting their shares until approved by other shareholders, or preventing these blockholders from purchasing the remaining shares for a specified price.  For example, in Italy, Belgium, Denmark, and France, laws were passed that every acquisition of more than 30% of the equity of the firm be followed by a tender offer to all voting shares at the same price.

The challenging task facing policy makers is to design corporate governance frameworks that secure the benefits of large shareholders as effective monitors of management whilst preventing them from extracting excessive private benefits of control.  A framework that effectively protects minority shareholders from expropriation will encourage the development of equity markets since small investors are then willing to invest in companies’ stocks.  For example, on the one hand, non-voting stock is both a low cost instrument for remaining in control and raising outside equity capital.  On the other hand, it is also a separation device that implies diluted incentives and facilitates shareholder expropriation.  Therefore, the best policy response seems to be to allow (but restrict) the issuance of non-voting stock, while at the same time provide minority shareholders with the means and power to claim due compensation if expropriation arises.  High standards of disclosure and transparency also help ensure the type of environment in which small shareholders feel comfortable investing in equity markets.  Disclosure requirements for pyramidal groups, for the structure of ownership and voting rights, and for legal separation devices, should be mandatory and enforcement should be strict, see the policy recommendations of the OECD Principles of Corporate Governance (1999a).

Figure 1. Stock market capitalisation and the R&D/GDP ratio in 1994

Stock market capitalization and the R&D/GDP ratio

Stock market capitalization (% of GDP)

Source:  Gugler (1999)

IV.3 The market for corporate control and firm performance

It is often asserted that takeovers are a mechanism for removing the management of poorly performing firms.  Takeovers act as a powerful disciplining tool since they enable anyone who identifies an underperforming company to buy a controlling interest and to reap any gains associated with transferring control from an inefficient management to an efficient one.  In market based systems of corporate governance, such as the US and the UK, takeovers play a key role in the disciplining management. However, when we look at takeover activity outside the US and UK we find that there are very few hostile takeovers.  In continental Europe, for example, capital markets are much smaller and controlling shareholders act as monitors, and there tends to be little reliance on the market for corporate control as a disciplining device.  Therefore, some of the questions that the empirical literature attempts to address are to what extent takeovers are a good substitute for direct shareholder monitoring to control management and what are the efficiency consequences of takeovers?

However, the efficiency of takeovers as a disciplining mechanism is a controversial and somewhat unresolved issue.  There is the problem that too few hostile takeovers appear to be value improving, and some even appear to result in outright reductions in value.  There are several reasons why not all takeovers will serve to enhance firm value.  Firstly, the main objective of the bidder need not necessarily be profit maximisation.  For example, managerialist theories of the firm emphasise the size and growth objectives of managers, and one of the fastest ways to increase firm growth is to acquire another firm.^[27]  Secondly, although there may be improvements in productive efficiency, takeovers may adversely affect dynamic efficiency by reducing firm-specific investment by various stakeholders.  For example, if stakeholders anticipate that takeovers increase the probability that they will be expropriated ex-post they may provide sub-optimal levels of firm-specific investment ex-ante.^[28] Thirdly, an active takeover market may only aggravate excessively short termist behaviour that can impinge on innovative activity and dynamic efficiency.^[29]

There are two approaches to analysing the effects of takeovers.  One approach is to examine the share prices of the target and acquiring firms around the announcement date of the takeover.  In reviewing the empirical evidence on takeovers that adopts this approach, the following facts have emerged.  The vast majority of empirical studies seem to find that target shareholders, on average, earn positive returns from tender offers.  For example, Jarrell and Poulsen (1987a) using a large sample of successful tender offers in the US, estimate the premia to average 19% during the 1960’s, 35% in the 1970’s, and 39% in the 1980’s. Higson and Elliot (1998) find similar results for the UK during the 1975-1990 period, where returns to target shareholders averaged 37% and were positive in 88% of the cases.  Many other studies confirm these results.^[30]  On the other hand, for bidder shareholders the returns, on average, are close to zero.  For example, for results on bidder returns in the US, see Bradley et. al. (1984), and for the UK, see Franks and Harris (1989).  Overall, the empirical evidence indicates that it is target shareholders that, on average, are the winners of takeover contests.

Another approach to analysing the effects of takeovers is to look at the post-takeover performance of the merged group.  The desirability of takeovers from a social point of view is a matter of whether the observed takeover gains for target shareholders represent real gains or just reflect transfers of wealth from one economic agent to another.  Mergers and takeovers benefit society and are socially desirable as long as efficiency gains are observed.  These gains can be due to the disciplining of inefficient management, as well as gains from economies of scale and scope, to the savings of transaction costs or other synergies.  If takeovers are a means of resolving problems associated with inefficient management, or with other efficiency gains, than the ex-post performance of the merger group should be better than the weighted average of the ex-ante performance of the acquiring and target firm prior to the takeover.

The vast majority of studies find no significant improvement in firm performance following a merger.  In fact, most find a negative impact on performance, particularly in the case of unrelated or conglomerate mergers.  On the other hand, hostile takeovers, do not exhibit the same deterioration in performance as mergers.^[31]  While some studies show small, but significant, post-hostile-takeover returns, others find no significant efficiency gains.^[32]  While the evidence on hostile takeovers is more positive, some authors express doubt that the takeover process is an effective means for disciplining management, see Ravenscraft and Scherer (1987) and Franks and Mayer (1996).  In this case, an active market for corporate control in the US and UK does not necessarily reflect the extent to which managerial inefficiencies are being corrected.  These results, therefore, raise serious concerns regarding the strong reliance on hostile takeovers as the primary means of monitoring management in outsider systems.

These results also raise concerns regarding the motives behind many mergers.  Since the efficiency gains of takeovers are, at best, quite low and target shareholders are receiving average premia of around 30% to 40%, takeovers seem to be primarily motivated by other objectives rather than with the disciplining of management.  Franks and Mayer (1996) found that there was little evidence that takeovers in the US and UK were motivated by poor performance prior to the takover bid and, hence, are primarily motivated by other objectives, such as changes in corporate strategy, or rent seeking behaviour.  Tax motives, in particular, have long been suspected as a contributor to merger and acquisition activity, and the available evidence seems to support this, for example, see Auerbach and Reishus (1988), Lehn and Poulsen (1987), and Bhagat et. al. (1990).

Since efficiency gains are small, the vast majority of studies find that target shareholder’s gains come primarily at the expense of other stakeholders, labour in particular.  While most of these studies find that employees losses mainly come in the form of layoffs, reduced wages, or lower employment and wage growth, Pontiff et. al. (1990) found that over 10% of takeovers involve pension fund reversions, accounting for 10% to 13% of takeover premia.^[33]  The fact that labour looses from takeovers can have long-term implications for investment, by reducing stakeholders’ incentives to provide firm-specific investment.  In addition, takeovers can also have (positive or negative) externalities not captured by the firm’s stakeholders (including shareholders).  For example, Shliefer and Summers (1988) point out that the acquisition of Youngstown Sheet and Tube in 1970 resulted in more that a doubling of bankruptcies in Youngstown and a plummeting of house prices.  Therefore, the sudden redeployment of corporate assets, which can occur in some takeovers, can have major impacts on whole cities and regions.

While the evidence does not seem to support the hypothesis that takeovers act to rectify managerial inefficiencies, this does not mean that policy makers need not worry about mechanisms that inhibit the market for corporate control.  Even if there is no evidence that takeovers actually improve firm performance, this does not necessarily imply that the market for corporate control does not work as a disciplining device.  This is because the mere threat of a takeover may serve this function.  Therefore, the fact that we do not observe takeovers that are motivated with the disciplining of management but, instead, are motivated by other objectives, does not mean that the market for corporate control is not an effective disciplining device, since it is the threat that serves as the mechanism.  Therefore, serious concern has been expressed recently regarding the use of defensive or anti-takeover mechanisms, particularly those that do not require shareholder approval, and which inhibit the market for corporate control.  These measures include poison pills, greenmail^[34], or state anti-takover amendments^[35].  The late 1980’s witnessed a sharp increase in the number of states in the US adopting anti-takeover legislation and firms adopting poisons pills.  For example the percentage of listed firms that were protected by such legislation increased from 38% in 1987 to 77% in 1988.^[36]

In fact, the empirical evidence supports the view that policy makers should worry about the use of measures that inhibit the market for corporate control.  Nearly all of the studies that look at the welfare effects of anti-takeover measures find that these measures reduce shareholder value.  For example, Jarrell and Poulsen (1987b) report significant negative stock-price effects associated with supermajority amendments; OCE (1984) find negative stock price effects associated with greenmail; and Malatesta and Walkling (1988) and Ryngaert (1988) find that poison pills harm target shareholders by reducing the probability of takeovers.  More recently, Bertrand and Mullainathan (1999) examined the effects of state anti-takeover legislation, in particular, business combination (BC) legislation^[37] and find that these laws raised annual wages by 1% to 2%.  Their results imply that takeovers do act to discipline management since they reduce the scope for managerial discretion i.e. uncontrolled managers that are protected by antitakeover legislation pay higher wages.^[38]  The authors conclude that BC, by impose a moratorium on specified transactions (e.g. sale of assets, mergers, and business relationships) between the target firm and the acquiring firm, unless the board votes otherwise, are effectively giving the board the power to block hostile takeovers and help to entrench management.  Jensen (1988) also points out that these defensive measures, by inhibiting the market for corporate control, raise serious concerns since they are effectively transferring critical control decisions from owners to managers, in particular, the right to make decisions regarding the firing of management.

In summary, although the evidence taken from actual takeovers does not seem to support the hypothesis that takeovers are an effective mechanism for disciplining poor management, evidence based on the use of anti-takeover amendments, suggests otherwise.  This implies that it is the threat of takeover, rather than actual takeovers, that seems to act as an effective device.  Mechanisms that inhibit the market for corporate control should therefore be viewed with apprehension.  Furthermore, even if takeovers served only to correct instances of managerial inefficiency, legal, advisory and financing costs total, on average, 4% of the purchase price.  This is a particularly expensive way of aligning the interests of management with those of shareholders.  Franks and Mayer (1996) conclude that there is a tradeoff between different methods of correcting managerial failure.

While takeovers on the one hand may lead to lower levels of managerial inefficiencies, they may also come at the expense of long-term firm-specific investment.  Takeovers, therefore, should be viewed in conjunction with other control devices.  For example, if one thinks that hostile takeovers are harmful to the economy, but if substitutive relationships exist, one could strengthen these alternative mechanisms while not constraining hostile takeovers by regulation.  For example, Fama and Jensen (1983) and Kini et. al. (1995) find that takeovers can serve as a substitute for outsider directors i.e. they are a good control mechanism when there are few outside directors on the board.  Brickley and James (1987) found that in states where takeovers are more restricted, the number of outside directors and ownership concentration are effective substitutive mechanisms for monitoring.  Schranz (1993) also finds substitutability between ownership concentration and takeovers as disciplining devices.  In conclusion, although there are concerns regarding the effectiveness of hostile takeovers as a disciplining device, the evidence seems to suggest that the market for corporate control should be allowed to function without restriction.  However, to rely on hostile takeovers as the only, or even the main, disciplinary device is not optimal either.

IV.4 Managerial compensation and firm performance

When other direct control mechanisms do not function very well, there may be a need for  special incentives that induce managers to act in the interests of shareholders i.e. to maximise profits. This is often undertaken through the design of executive remuneration packages.  In practice this usually involves tying managerial compensation to the performance of the firm, in the form of salaries, bonuses, and stock options.  In this way, managerial wealth is subjected to at least some of the same risks to which shareholders and the firm are exposed.  Therefore, linking managerial compensation to firm performance has been adopted in many countries as a way of aligning the interests of managers with those of shareholders.  However, managerial compensation has become a hotly debated issue, particularly in the US and UK, where the last 10 years have seen an explosion in the level of managerial pay.

The emphasis on executive remuneration packages varies from one country to another and the optimality of contingent performance based compensation depends on whether or not other direct monitoring alternatives are available.  In Japan and Germany, for example, executive pay tends to be considerably lower than in the US and UK.  This may reflect the closer relationship between controlling shareholders and managers that exists in insider systems.  For example, in insider systems of corporate governance there is less scope for managerial discretion and there are fewer informational asymmetries between managers and owners.  Kole (1997) provides empirical evidence supporting this hypothesis.  He finds that for 371 Fortune 500 firms in 1980, that if there is a family representative either in management or on the board of directors, the probability of adopting an equity-based compensation plan is significantly reduced.  Conyon and Leech (1993) also find that the level of director pay is lower in companies that have a higher share ownership concentration or are defined as owner-controlled.  In addition, the prevalence of longer term contracts in Japan, coupled with the fact that most managers are promoted internally, means that reputation may provide a sufficient deterrent to managers without the need for explicit incentive contracts.  Therefore, it is not surprising that managerial compensation tends to be lower in insider systems, such as Germany and Japan, than that found in the US and UK.

While the empirical evidence confirms the substitutive effects between direct monitoring by owners and compensation incentives, board monitoring or monitoring by institutional investors may also substitute for direct shareholder monitoring.   In theory at least, the use of these other mechanisms should also reduce the level of pay-incentives needed to align managers’ incentives with those of shareholders.  In practice, however, board members become like management and agency costs are expected.  Mehran (1995) finds empirical evidence to support this view.  He finds that the presence of outside directors, rather than decreasing the level of executive remuneration, actually increases the percentage of equity-based compensation.  Separating the roles of chairman and CEO is often proposed as a way of preventing boards from becoming entrenched like management and, in principle, should increase accountability.  However, Conyon and Leech (1993) found no evidence that separating the roles of chairman and CEO had any effect on executing compensation levels.  And as regards whether or not monitoring by institutional investors has a substitutive effect with compensation incentives, Cosh and Hughes (1997) do not find any evidence that institutional holdings in the UK alter the level of executive remuneration or the pay-performance relationship.  Therefore, while direct shareholder monitoring is a good substitute for compensation incentives, the evidence suggest that the board and monitoring by institutional investors, on the other hand, are relatively weak monitoring devices and not a good substitute for direct monitoring.

If the use of incentive devices is effective, then this should manifest itself in a positive relationship between managerial compensation and firm performance.  Excluding stock options, current evidence indicates that sensitivities of pay to performance are quite small.  Murphy (1985), Coughlin and Schmidt (1985) and Barro and Barro (1990) all find pay-performance elasticities in the range 0.10 to 0.17, suggesting that a 10% rise in firm profitability leads to a 1% to 1.7% rise in CEO compensation consisting of salary plus bonus.  However, Hall and Liebman (1997) suggest that previous sensitivity measures ignored changes in the value of stock and stock options, which account for virtually all of the sensitivity. From a sample based on 478 large US companies from 1980 to 1994, the authors obtain similar results as the rest of the empirical literature regarding the sensitivity of performance to salary-plus-bonus (e.g. an elasticity of 0.2).  They show that the driving force behind the pay-performance relationship is the use of both stock and stock options.  When the authors include stock and stock options, they find a mean elasticity of 4.5, suggesting a 10% rise in performance leads to a 45% increase in CEO remuneration.

However, Murphy (1998) finds that most of this increase is attributable to a general increase in the stock market and that there is little evidence that higher pay-performance sensitivities lead to higher stock performance.

Some studies have looked at the relationship between managerial compensation and firm’s sales. Baker et. al. (1988) find the elasticity of executive annual-salary-plus-bonus with respect to sales is in the 0.20 to 0.25 range and is relatively uniform across firms, industries, and time periods.  While this finding is consistent with value maximisation if larger firms employ better qualified and better paid CEOs, it is also an indication that managers may not be behaving in an optimal way.  In fact, a number of studies find that company size, and changes in size, are much more significant determinants of executive pay than measures of shareholder value, see Main et. al. (1994), Conyon and Leech (1993), and Gregg et. al. (1993).  Since one of the fastest ways to increase firm size is to acquire another firm, this empirical evidence suggests that managers have an incentive to engage in mergers and acquisitions, which may not be in the best interest of the corporation, but would have the effect of increasing their remuneration.

According to agency theory, remuneration contracts are efficient if the level of compensation is linked to aspects of performance over which managers have some control.  Otherwise, executives would not have any incentive to engage in significant effort to increase firm performance since they know they will be compensated regardless of the performance of the firm.  For example, it does not make sense for owners to compensate managers for events that are beyond their control e.g. better firm performance due to a general rise in the stock market.  Nor do owners want to penalise (fire) managers due to negative outcomes that are not the fault of the manager e.g. poor performance due to a recession.  However, it is harder for an executive manager to claim that the company has performed poorly due to general market conditions if other benchmark companies are performing well. This says that contracts, in order to be efficient, should relate compensation to rises in relative performance e.g. the performance of industry peers or direct competitors.

While the use of relative comparisons makes managerial incentives more effective and remuneration contracts more efficient, the empirical evidence concerning relative performance effects is rather negative.  For example, Murphy (1985), Barro and Barro (1990), and Hall and Liebman (1997) find that, in general, managerial compensation does not depend upon the relative performance of the firm.  One of the reasons why this may be the case is that, under current accounting rules, stock options with a performance target have to be counted as a cost against profits, while options with a fixed exercise price do not reduce current earnings.  One policy recommendation, therefore, is to harmonise accounting rules concerning stock options with a variable and a fixed exercise price.  Boards in this case would no longer be discouraged from linking stock options to performance.

Given the importance in outsider systems of stock and stock options as part of overall managerial compensation particular attention needs to be paid to the design of executive remuneration packages.  In the UK, the Cadbury Committee recommended that there should be a remuneration committee that determines the level of executive compensation, and that this committee should consist mainly of nonexecutive directors.  While this suggestion is valuable, Ezzamel and Watson (1997) find there are asymmetric adjustments in executive pay levels i.e. executives that were relatively underpaid in the previous period receive pay increases, whereas executives who were overpaid were not subjected to downward adjustments.  This is consistent with the complaint of many shareholder groups that remuneration committees, rather than strengthening the pay-performance relationship, have the effect of bidding up executive pay.

These empirical findings raise serious concerns regarding not only the efficiency of executive remuneration contracts, but also the motives behind them.  For example, when agency problems are severe and monitoring of management is very weak, executive remuneration, unless efficiently designed, can become another vehicle for managerial expropriation of rents.  Because boards of directors set compensation contracts, there is a serious concern that boards align themselves more with management than with weak dispersed shareholders.  Empirical evidence supporting this hypothesis is obtained by Yermack (1997), who finds that the timing of stock option awards coincides with favourable movements in company stock prices.^[39]  This suggests that managers who become aware of impending improvements in corporate performance may be influencing the board of directors to award more equity-based-performance pay.  Furthermore, the practice of re-pricing stock options when share prices fall, makes a complete mockery out of the notion that options are designed to align the interest of managers with those of shareholders.^[40]

Executive compensation plans may also serve to exacerbate short-termist behaviour.  For example, by contributing to managers’ incentives to increase short-term shareholder value, the spread of share options may be contributing to a temporary over-evaluation of equities and may be distorting the economy.  There are serious concerns regarding the current practice of management to issue debt and use the proceeds to buy back equity, thereby increasing the firm’s share price.  While in theory how a firm is financed should make no difference to its value, very high levels of corporate debt make a firm more vulnerable to bankruptcy in the event of a downturn.

Overall, the empirical evidence seems to indicate that these contracts, rather than aligning the interests of managers and shareholders, are enabling managers to expropriate some of the rents from shareholders.  These abuses are particularly likely when share ownership is fragmented and alternative control mechanisms are weak.  A study undertaken by the accountancy firm PwC suggests that many companies in the UK were failing to achieve a link between pay and performance.  For example, they found that for a sample of 270 quoted companies, only 7 firms put their remuneration report to a shareholder vote, and only 5% of companies disclosed even in broad terms how performance measures relate to long-term company objectives.^[41]

All of this does not suggest that the use of share and share options should be restricted,^[42] but rather that they should be more closely linked to performance.  The fundamental question, however, is whether or not shares and share options are achieving what they were intended to do i.e. aligning the interest of managers with those of shareholders.  The evidence in this regard is quite disappointing, and given the inherent conflict of interest involved, the question arises whether the board of directors should set key elements of managerial contracts at all.  The empirical evidence in this regard suggests that it should be shareholders themselves that ultimately decide on the level and shape of executive remuneration packages.  This is also in line with recent recommendations by the UK government.^[43]

V.    Conclusions

Corporate governance affects the development and functioning of capital markets and exerts a strong influence on resource allocation.  In an era of increasing capital mobility and globalisation, it has also become an important framework condition affecting the industrial competitiveness and economies of Member countries.  This paper set out to further develop our understanding of corporate governance and its affect on corporate performance and economic performance.  In doing so, it addresses some of the underlying factors that promote efficient corporate governance, and examines some of the strengths, weaknesses and economic implications associated with various corporate governance systems.  It also provided a survey of empirical evidence on the link between corporate governance, firm performance and economic growth, identifying areas in which a consensus view appears to have emerged in the literature and areas in which further research is still needed.

One of the most striking differences between countries’ corporate governance systems is the difference in the ownership and control of firms that exist across countries.  There are tradeoffs between ownership concentration and voting power concentration.  Systems of corporate governance can be distinguished according to the degree of ownership and control and the identity of controlling shareholders. In ‘outsider’ systems (notably the US and UK) of corporate governance the basic conflict of interest is between strong managers and widely dispersed shareholders.  In ‘insider’ systems (notably Continental Europe and Japan), on the other hand, the basic conflict is between controlling shareholders (or blockholders) and weak minority shareholders.

There is no single model of good corporate governance, and both insider and outsider systemshave their strengths, weaknesses, and different economic implications.  Furthermore, the effectiveness of different corporate governance systems is influenced by differences in countries’ legal and regulatory frameworks, and historical and cultural factors, in addition to the structure of product and factor markets. Corporate governance mechanisms and their effectiveness also vary depending on industry sectors and type of productive activity.  For example, in industries characterised by high asset specificity (such as many high-tech industries), monitoring is more difficult and different mechanisms may be required in order to improve firm performance.  Identifying what constitutes good corporate governance practice, and under what circumstances, is a difficult task.  The challenge, therefore, is not only to identify the strengths and weaknesses in each individual system or group of systems, but also to identify what are the underlying conditions upon which these strengths and weaknesses depend.

The benefits of concentrated ownership are that it brings more effective monitoring of management and helps to overcome agency problems.  However, the costs associated with concentrated ownership are low liquidity and reduced possibilities for risk diversification.  Dispersed ownership brings higher liquidity, which can be vital for the development of innovative activity.  On the other hand, it does not encourage commitment and long-term relationships that might be required for certain types of investments.  For example, when corporations are owned and controlled by each other, this can reduce transaction costs and incentives to engage in opportunistic behaviour.  Stakeholders, therefore, have a greater incentive to invest in relationship specific investment.  On the other hand, this can also reduce the level of product market competition.

Since equity markets are important for R&D and innovative activity, entrepeneurship, and the development of an active SME sector, corporate governance has an underlying impact on economic growth and development.  Therefore, one of the main challenges facing policy makers is how to develop a good corporate governance framework which can secure the benefits associated with controlling shareholders acting as direct monitors, while at the same time, ensuring that they do not expropriate excessive rents at the expense of other stakeholders.  The empirical evidence to date seems to suggest that this is indeed a problem and that protection of minority shareholders is critical to the development of equity markets. Therefore, policy makers in insider systems need to pay particular attention to developing corporate governance frameworks that will not hinder the development of active equity markets.

Policy recommendations should attempt to account for the interactions between corporate governance and the institutional framework in the particular country.  The search for good practice should be based on an identification of what works in defined countries, to discern what broad principles can be derived from these experiences, and to examine the conditions for transferability of these practices to other countries.  As this document has demonstrated, not only will different improvements be called for in different systems but these improvements will also depend upon the factors determining the effectiveness of different systems.  For example, some systems and circumstances may have a need for policy adjustments in other areas such as strengthening product market competition or removing distortions in corporate governance mechanisms created by other regulations (e.g. rules regarding the use of fixed or variable price stock options).  Continued work in this area, therefore, will continue to build on the work developed in this document and will aim to specify what are the crucial improvements needed in different systems and in different situations.  This work will not undertake any specific surveys or country specific reviews.  It will be based on a thematic approach and will continue to draw on empirical work already done outside and within the Organisation.  Wherever possible, it will also draw on work undertaken by other Directorates particularly as regards outreach activities on corporate governance.

---

[1] .                See Shleifer and Vishny (1997), p. 741.

[2] .                See, for example, Shleifer and Vishny (1997) and Berglof (1997).

[3] .                Mayer (1996), p. 11.

[4] .                Blair (1995), p. 203.

[5] . Zingales (1997) also defines a corporate governance system in the spirit of Williamson (1985) as a “complex set of constraints that shape the ex-post bargaining over the quasi rents generated in the course of the relationship”.

[6] .                See Keasey, Thompson, and Wright (1997), p. 3.

[7] . Data on ownership concentration should be viewed with caution since disclosure requirements in most OECD countries relate not to ownership but to control.

[8] . Information and recent work undertaken by the ECGN, including the data used in this paper, can also be found on their web site at http://www.ecgn.org.

[9] . Largest shareholder does not imply a single individual, but rather the equity holdings of a single entity. This could be the holdings of an institutional investor or pension fund, a bank, or another firm, or familial holdings, etc.

[10] . Figures for the German DAX 30 and French CAC40 are atypical.  For example, there is a minimum turnover requirement for membership of the DAX30, leading companies to widen their shareholder base in order to remain in the index.  Similarly, the CAC40 is an unrepresentative group of the largest and most liquid companies on the exchange.

[11] .              See La Porta et. al. (1998) and Shleifer and Vishny (1997).

[12] . There is a vast literature, which this work on corporate governance does not address, on how vertical and horizontal relationships can impinge upon market competition.

[13] . With dispersed ownership, if shareholders are unhappy, they will just sell their shares (exit) since they do not have the incentive, or the means available to them (voice), to engage in direct shareholder monitoring.

[14] . Overall individual share ownership has increased i.e. the percentage of the population that holds equities has increased in both countries.  However, these shares tend to be held indirectly.

[15] . However, since in many fields basic research feeds directly into industrial applications providing measurable returns, not all basic research has been downgraded e.g. biotechnology or computer science.

[16] .              See R&D Magazine (1997).

[17] .              Morck et. al. (1988), p. 294.

[18] .              See Mayer (1996), p. 19.

[19] .              Mayer (1996), p. 24.

[20] . Microsoft corporation is a good example of this i.e. it is the promotion and ability to retain human capital within the firm, rather than external funds, that impacts upon the firm’s performance.

[21] .              See OECD (1995), Economic Survey of Italy.

[22] .              Performance is measured by Tobin’s q.

[23] . High asset specificity implies that investment is specific to the industry (or the firm) and has relatively few alternative uses.  The opposite is true of low asset specificity investment.

[24] . In some cases, non-voting stock have a preferred dividend.  Therefore, the observed premium must be what is left over after having accounted for other differences in classes of shares, and which is then attributable to the superior voting rights, i.e. the premium associated with control.

[25] . For example, in a case study of IRI, Zingales found that IRI sold its majority stake in Finsiel to STET (which is controlled by IRI as well), at too high a price.

[26] . Some of the dimension include one-share/one-vote, if proxy by mail is allowed, cumulative voting, percentage of shares need to call an extraordinary shareholders meeting, anti-director rights, mandatory dividends, etc.

[27] .               See Mueller (1985).

[28] .              See Shleifer and Summer (1988).

[29] . However, Jarrell et. al. (1988) state that no empirical evidence has been found to support this theory.  For example, a study by the SEC’s Office of the Chief Economist (1985) found that there was no difference in the probability of takeover depending on whether or not firms had high R&D expenditures.

[30] . See Jensen and Ruback (1983), Jarrell et. al. (1988), DeAngelo et. al. (1984), or Franks and Mayer (1996), just to name a few.

[31] . Firm performance is usually taken as some measure of profitability, although a number of studies looked at changes in productivity or growth rates following the merger/takeover.

[32] . Rather than list the vast number of studies that find small, but positive results and those that find no efficiency gains, see Gugler (1999) for a comprehensive survey of the empirical results on hostile takeovers.

[33] . The fact that companies do not have to share surpluses in their pension plans with their workers, even if the workers contributed to the plans through money withheld from their paychecks, raises a whole set of corporate governance issues that are beyond the scope of this paper.

[34] . Greenmail occurs when the management of the target firm ends the hostile takeover threat by repurchasing the hostile suitor’s block of target stock at a premium.

[35] . For example, supermajority amendments which increases the minimum approval required for mergers and other important control transactions, or dual-class recapitalisations i.e. the creation of classes of equity with differential voting rights.  Other amendments include: control share acquisition, which require other shareholders to approve the voting rights of a large shareholder; fair price, which require paying a ‘fair price’ for all shares tendered; and business combination, which imposes a moratorium of 3-5 years on specified transactions between the target and raider, unless the board votes otherwise.

[36] .              See Comment and Schwert (1995).

[37] . Ryngaert and Netter (1988) and Schumann (1988) also present evidence on the negative effects of state anti-takeover amendments.

[38] . There are several reasons why managers may prefer to pay higher wages.  Manager may care more about having high-quality workers and lower turnover than owners do.  They might care more than owners about improving relations with employees since they are the ones who endure the worker’s complaints and enjoy their company.  Also managers may dislike bargaining.

[39] . In particular, he finds that in the 50 days immediately following CEO option awards, stock prices experience an average cumulative abnormal return of more than 2%.

[40] . When share prices fell in the late summer of 1998, many firms re-priced their options just in time so that executives enjoyed massive gains when the market rebounded, see The Economist, 7 August 1999.

[41] .              See the Financial Times, 20 July, 1999.

[42] . For example, the use of stock options can reduce the costs of starting a company since new firms can hire employees with the promise of future benefits through the use of share options.  In this way, new high-tech firms with no assets need not be penalised since they can now compete with more established firms and attract high-quality employees by offering options.

[43] . For example, the UK government has suggested that this could be done by either requiring companies to put their remuneration report to an annual shareholder vote, or by allowing shareholders to put forward resolutions on remuneration to annual general meetings.