Changing the nature of governance to create value.

It is an inconvenient fact that the relationship between good corporate governance and performance has turned out to be so elusive. Clearly, because many factors influence a company's performance, it is difficult to isolate statistically the role and contribution of governance. Given this fact what does good governance actually mean, and how does good governance contribute to the stock performance of public companies?

We attempt to answer these two questions by

* contrasting three different forms of corporate governance: fiduciary governance, shareholders'-rights governance, and what we call value-creating governance;

* surveying past studies on the relationship between governance and performance;

* reporting the results of our study in the Canadian context, and

* proposing a framework for a value-creating approach to corporate governance.

The fundamental issue that has plagued corporate governance since ownership became separate from management is the "agency" problem, which may be framed as follows: How does a principal (the board as the agent of the shareholders) ensure that the agent (management) works fully in the interest of the principal, given that the agent has more information about his/her own performance, has a superior knowledge about the task to be performed, has control over the information flow to the principal, and may well have interests that are divergent from those of the principal?

Succinctly put, we believe that the fiduciary form of governance, which has been widely promoted over the past few years, fails to address the agency problem.

An alternative way to cope with the agency issue is to rely on the efficiency of financial markets--in particular, the market for corporate control. Unfettered takeover activity, it is argued, would ensure the swift removal of underperforming management and incompetent boards. However, this form of governance, to which we refer as shareholders' rights, not only raises difficult policy issues in the Canadian context, but also requires a strong and abiding faith in the efficiency of markets.

We propose instead a form of governance we call value creating--a systematic effort to cope with the fundamental issues arising from the information and expertise advantage that management has over board members, as well as from the divergent interests of management and board members (as representatives of shareholders).

The Limits of Fiduciary Governance

Over the past decade, but with intensified urgency since the corporate fiascos and scandals of 2001 and 2002, standard prescriptions for good governance have shaped a new governance orthodoxy. The key tenets of this fiduciary form of governance are that:

* the board should consist of a majority of independent directors, with the standards of independence given an increasingly stringent definition;

* the positions of chairman and chief executive officer (CEO) should be separated or, as a last resort, the position of head director should be created;

* members of board committees should be drawn exclusively from the rostrum of independent directors;

* both the CEO and chief financial officer should formally sign all of the company's financial reports and vouch for their adequacy in representing fairly the company's financial condition; and

* the board should tightly control auditors and executive remuneration--particularly the level of stock options.

Other provisos usually include regular meetings of independent board members only, formal performance evaluation of board members, and mandatory shareholding by board members and senior management.

This form of governance is called "fiduciary" because its aim is to ensure that directors discharge fully their "legal duty of care." It seeks to enhance the rights of shareholders, to protect them against self-serving behaviour by management, and to act as a sort of insurance policy against managerial incompetence and misconduct.

This latter point is controversial. Indeed, while no one questions the fact that there were governance failures at Enron, WorldCom, and elsewhere, it does not follow that strict obedience to the new rules of fiduciary governance would have prevented these fiascos. Certainly, as has been pointed out before (Allaire and Firsirotu 2002; Sonnenfeld 2002), these companies seemed to comply with most of the formal prescriptions of good fiduciary governance.

No less an authority than Richard C. Breeden, court-appointed corporate monitor of WorldCom, writes in a recent report:

Despite the record of its inability or unwillingness to control Ebbers, WorldCom seemed to meet most of the governance standards of its time. Indeed, in several areas WorldCom exceeded the accepted norms of "best practice" in corporate governance, even though there was little if anything about its governance that was "good" in reality. This illustrates the fact that good governance is not achieved by simply adhering to "checklists" of recommended "best practices." (2003, 30.)

No doubt giving shareholders the benefits of representation by determined and dedicated people with no interest other than that of shareholders is, per se, a desirable goal. Over the past few years, public, institutional, and legal pressures have made boards more responsive, proactive, and vigilant than they were a few years ago and have given a great deal of substance to the formalities of governance.

Yet, although fiduciary governance provides some benefits, demonstrating a causal, statistically valid link between governance and company performance is extremely difficult. Corporate performance is affected by many different factors and management decisions and actions, good or bad, that tend to have a delayed impact on the company's performance. That has not, however, deterred academics from pursuing this elusive link with formidable tenacity. After surveying a large number of studies, Patterson concludes that significant correlations between performance and governance are infrequent and, in any case, "even when a correlation is found, it is difficult to prove causation in these types of statistical studies" (2000, 5). In another example, Dalton et al. (1998) review 54 studies on board composition and financial performance and find virtually no correlation between the two.

Amid a large, generally cautious body of work reporting negative or insignificant results, however, three studies are often quoted as providing some evidence of a relationship between good governance and performance: Millstein and MacAvoy (1998); the McKinsey surveys (Coombes and Watson 2000; McKinsey & Company 2002); and Gompers, Ishii, and Metrick (2003).

Millstein and MacAvoy 1998

In a study based on governance grades assigned to the 300 largest U.S. firms by the California Public Employees' Retirement System (CalPERS 1998), Millstein and MacAvoy (1998) find that, over the 1991-to-1995 period, companies that received a high grade had a return on capital that was 7.3 percent higher than companies that rated an average grade.

CalPERS had asked the boards of each of these 300 companies to compare their governance processes to those of General Motors, which at that time had just published 24 governance guidelines. The grades reflected the boards' own evaluation of their compliance with GM's guidelines or their intention to review their governance structures. A company rated an F if it failed to respond to the CalPERS survey. What Millstein and MacAvoy established was not really a linkage between good governance and performance but that companies that perform well are more likely to respond quickly and positively to an enquiry from a large institutional investor about their governance practices and intentions.

While admitting that their results show, at best, an association between company performance and a self-evaluation of governance practices, Millstein and MacAvoy estimate that the "difference in dollar gains for investors generated by an 'A+' company over a 'C' company over the five-year period is $6.359 billion." Such a heroic conclusion should be received with a great deal of skepticism, however, particularly in light of better-designed studies that have failed to support any such linkage.

The authors also fail to explain some inconsistent results. For instance, their entire case for claiming that companies with a high grade had a higher return on invested capital rests on a single multiple regression analysis that provides estimates of the difference between an A+ company and a C company. However, a close examination of their results also shows that a company with a B or even an F grade would be associated with a better economic performance than a company with an A grade--that is, the regression coefficient is better for an F and a B grade than it is for an A grade.

The McKinsey Surveys

In 2000 and again in 2002, McKinsey & Company conducted a "global investor opinion survey." Based on responses from more than 200 institutional investors worldwide, the authors conclude that investors are prepared to pay a significant premium for companies that are well governed. Their 2000 survey (Coombes and Watson 2000) indicated that investors would pay a premium of 18 percent for "U.S. and U.K." companies, and 12-to-14 percent for "North American" companies, according to the 2002 survey (McKinsey & Company 2002). That premium is much larger in other parts of the world--for instance, more than 30 percent for companies in eastern Europe and Africa, according to the 2002 survey.

Table 1 presents the questions from the McKinsey surveys. These questions may appear a bit "leading" and rather too black and white. For instance, the combination of "minority of outside directors" and "directors own little or no stock" is unusual in practice. In any case, the oft-quoted premium "for well governed companies" from these McKinsey studies must be set in its proper context. It refers to a very specific situation, somewhat artificial, where an investor is faced with the stark choice between companies "A" and "B', as described in Table 1.

From the results of both McKinsey surveys, we can conclude:

* In a context of widespread abuses and very low standards of governance, investors will seek and reward good fiduciary governance. There are, however, diminishing returns to fiduciary governance. A World Bank study of the governance practices of public pension funds worldwide finds that "[t]he relationship [between governance and performance] is highly non-linear, with rapidly diminishing (but positive) marginal returns to governance levels" (Impavido 2002, 3).

* In developed countries such as the United Kingdom, the United States, and Canada, once basic fiduciary rules of governance have been established, the real benefits come from, as Coombes and Watson (2000) put it, "identifying innovative ways to raise governance standards further," which is the essential theme of our paper.

* The hypothesis that financial markets in developed economies will give added value for good fiduciary governance practices--irrespective of a company's current economic performance--remains to be tested formally.

Gompers, Ishii, and Metrick 2003

A third widely cited study showing that good governance leads to better returns for shareholders is that of Gompers, Ishii and Metrick (2003). In their study, however, good governance has very little to do with fiduciary governance. In fact, Gompers, Ishii, and Metrick's concept of governance refers to what we and others have called "shareholders' rights governance." In this case, a company is well "governed" if there are no impediments to its takeover by a qualified bidder--that is, if companies have not adopted anti-takeover measures or, if they are domiciled in states that restrict takeovers, they have chosen to remove themselves from the protection afforded against takeovers.

According to Gompers, Ishii, and Metrick, the following provisions are indicative of poor governance:

* anti-greenmail provisions;

* blank-cheque preferred stock;

* business-combinations laws;

* "constituency" statutes;

* a classified or staggered board;

* "poison pills" or "shareholders'-rights" provisions;

* golden parachutes;

* supermajority requirements and "control-share" acquisition statutes, and

* unequal voting, or dual classes of shares.

(See the Appendix for a short description of these measures.) Indeed, that such measures have a negative impact on shareholders' wealth receives broad empirical support in the United States (see, among others, Subramanian 2002).

Gompers, Ishii, and Metrick undertake a fairly exhaustive study of the relationship between performance (measured by "Tobin's Q"--essentially, the ratio of a company's market value to its book value) and a "Governance Index" (made up of the absence or presence of impediments to corporate takeovers) for some 1,500 U.S. companies from September 1990 to December 1999. The lower the company's score on the Governance Index, the better the governance, as it indicates the relative ease with which the company could be taken over through a hostile acquisition.

The authors then classify the top 10 percent of companies in their index as "dictatorships" and the bottom 10 percent as "democracies." The 10 largest dictatorships, in terms of their market capitalization in 1990, were: GTE, Waste Management, General Re, Limited Inc., NCR, K Mart, United Telecommunications, Time Warner, Rorer, and Woolworth. The 10 largest democracies were: IBM, Wal-Mart, E.I. Du Pont de Nemours, Pepsico, American International Group, Southern Company, Hewlett Packard, Berkshire Hathaway, Commonwealth Edison, and Texas Utilities.

Gompers, Ishii, and Metrick then proceed to construct and manage an artificial portfolio of stocks that includes the shares of the roughly 150 companies (weighted by their market cap) in the bottom 10 percent of their Governance Index (remember, the lower the score, the better the governance). Their investment strategy also calls for selling short the shares of all the companies in the top 10 percent of the Governance Index (that is, selling immediately shares that have been borrowed in the hope of replacing them at a future date with shares bought at a much lower price).

The authors claim that had such an investment strategy been carried out throughout the 1990s, it would have realized an "abnormal" annual return of 8.5 percent,(1) a well-publicized number that is fast becoming a standard quotation in speeches on how good governance will improve performance.

Clearly, however, the "governance" in this particular research has little in common with the various measures subsumed under the standard fiduciary prescriptions for good governance. What Gompers, Ishii, and Metrick (2003) have shown, with all due reservations, is that a relationship may exist between performance and the extent to which a company is available for easy, unfettered takeover, and the extent to which it operates as a "shareholder democracy" rather than as a "managerial dictatorship."

Fiduciary Governance and the Performance of Canadian Companies 

To assess the relationship between fiduciary governance and performance in a Canadian context, we used the governance ratings of the Globe and Mail's Report on Business (ROB), published on October 7, 2002. In doing so, we had to assume that the scores in the ROB survey correlate with "good" fiduciary governance practices. Indeed, the ROB has done as good a job as any of assembling in a rating instrument key suggestions and prescriptions for good fiduciary governance.

The ROB scale uses four components:

* board composition (relating to, for example, the independence of board and committee members), worth up to 40 points out of 100;

* shareholding and compensation issues (relating to, for example, the stock ownership of board members and the CEO), worth up to 23 points;

* shareholders' rights issues (relating to, for example, the existence of a staggered board, the dilution effect of options, and multiple classes of shares), worth up to 22 points, and

* disclosure issues (for example, governance guidelines, information on the relatedness of directors, the frequency of meetings), worth up to 15 points.

Our Methodology

The ROB rates the governance of the 270 Canadian companies listed in the Standard & Poor's/Toronto Stock Exchange (S&P/TSX) index, with scores ranging between 36 and 96 (out of 100). To increase the probability of capturing significant relationships, however, we eliminated from the database companies with less than $300 million in market capitalization (as of February 24, 2003), since their governance practices and performance vary so much that they ensured nonsignificant results. Our sample of companies thus consisted of the 177 largest Canadian companies, spread across 14 different industries.

Then, for each company, we computed scores for both economic performance and market performance using measures that have been commonly used in previous studies of the relationship between governance and performance (see Table 1).(2) The measures of economic performance based on the latest financial information available as of April 2003 were:

* the company's compounded annual sales growth over the past five years (Sales CAGR);

* the company's return on invested capital for the past 12 months (ROIC);

* a Value-Creation Index (VCI), defined as the company's ROIC divided by its weighted average cost of capital (WACC), and

* measure of economic value added (EVA), defined as [(ROIC minus WACC) times capital invested]. That EVA was then expressed as a percentage of invested capital at the beginning of the 12-month period (% EVA).

The measures of market performance were:

* Tobin's Q (or its common approximation), defined as the market value of the company's equity (as of February 24, 2003) plus its long-term debt divided by the book value of its equity plus its long-term debt;

* %MVA, which is the company's market value added over four years, as a ratio of the company's market value four years ago, translated into a compounded annual growth rate, and

* a variable called "delta," which can be explained as follows: To test the hypothesis that investors would pay a premium for good fiduciary governance (and assuming that investors would view a high ROB rating as an indication of such governance), we computed the regression line between Tobin's Q and VCI for each of the 14 industries. The notion is straightforward: perhaps good governance has no impact on internal performance measures such as VCI = ROIC/WACC), but, for any given level of VCI, the market (as measured by Tobin's Q may pay a premium for a well-governed company; therefore, the deviations, or "deltas," from the straight regression line between VCI and Q should capture that relationship. (See Figure 1 for an illustration of delta.) If that relationship exists, companies with high governance scores should be more likely to exhibit positive deltas, and companies with low governance scores should be more likely to have negative deltas.

The Results

We can draw several conclusions from the overall results of these performance measures for the 177 companies (see Table 2).

First, companies earning a governance grade of A were three to four times larger (as measured by equity market value) than those with lower grades. Second, the revenues of grade A companies grew much more slowly over the past five years than those of companies with lower governance grades. Third, as measured by the compounded annual rate of growth in market value added (%MVA)--which is the best measure from an investor's perspective--companies that rated an F outperformed all other companies, closely followed by companies with a C grade. Fourth, according to the delta measure, a governance grade of A does not appear to lead to a market value that is systematically higher than would have been predicted on the basis of the company's performance (VCI).

All in all, one cannot read too much into these raw results, as industry factors and large variations within each group may explain these differences. In fact, even though we carried out many different multivariate regression analyses that included industries, company size, and the four components of the ROB governance scores as independent variables, the analyses failed to establish any significant relationship between any of the seven performance measures and the ROB governance scores. (These results are reported fully in Allaire and Firsirotu 2003.)

On the other hand, the coefficient for the ROB component of "board composition" revealed a tendency, although not significant, to come out negative--that is, the higher the score for board independence, the poorer the performance! Such a result may be only a statistical aberration, but we single it out because previous empirical work in the United States also shows a negative relationship between board independence and performance (see Agrawal and Knoeber 1996; Yermack 1996; Klein 1998). The authors of two particularly thorough studies (Bhagat and Black 1999, 2001) conclude:

Contrary to conventional wisdom, we find evidence of a negative relationship between [performance and] the degree of board independence (proxied by an "independence" variable that equals the proportion of independent directors minus the proportion of inside directors). These results are driven by poor performance at firms with supermajority-independent boards. Firms with an independence level of 0.4 or higher (which corresponds, for a typical eleven-member board with one affiliated director, to eight or nine independent directors and only one or two insiders) perform worse than other firms. We find no strong correlation between board independence and firm performance for other firms. (1999, 27-8.)

Controlled Ownership versus Widely Held Ownership

In 51 of the 177 large Canadian companies in our sample, a single shareholder, or a few related shareholders, had voting control, often through their holdings of a class of shares with multiple votes. Using the same performance measures as described above, as well as the four components of the ROB's governance scale, Table 3 presents the results for companies with such "controlled" ownership compared with companies with widely held ownership. At this level of aggregation, the table shows that:

* both types of companies were similar in size as measured by their market value;

* revenues (Sales CAGR) of controlled-ownership companies grew faster than those of widely held companies over the past five years;

* by all measures of "internal" performance (ROIC, VCI, %EVA), controlled-ownership companies performed better than widely held companies;

* evidence from measures of "external" performance (Tobin's Q, %MVA, and delta) is mixed--for example, controlled-ownership companies performed better on %MVA but widely held companies did better on the delta variable; again, however, the differences between the mean scores of the performance variables were not statistically significant, and

* mean governance scores are statistically different between the two groups of companies, with the largest difference recorded for "board composition." Surprisingly, the difference between the two groups is relatively small for the "shareholder rights" component, but that result may reflect the way the ROB scale is constructed, as well as the inclusion of such variables as the "dilution effect of stock options" and the "re-pricing of options" under "shareholder rights."

Again, we undertook a large number of multivariate regression analyses to assess the impact of ownership type on company performance (see Allaire and Firsirotu 2003), but in none of the regressions was this variable a significant factor. Our results indicate that:

* a good part of the observed differences between the means of the two groups disappear when industry effects are taken into account, and

* the within-group variations are large relative to the between-group differences.

The Performance of Canadian Companies Listed on the NYSE 

Of the 177 Canadian companies used in our sample, 69 were also listed on the New York Stock Exchange (NYSE). Given the different legal and regulatory context of the NYSE compared to that of the S&P/TSX, we wondered whether significant differences existed between the companies listed on the NYSE and those that were not, again using the same performance and governance variables as before. Table 4 presents the mean results for the two groups of companies, the most interesting of which are that:

* in terms of market capitalization, the NYSE-listed companies were more than three times as large as those not listed there;

* the mean differences between the two groups on performance scores were not statistically significant, and

* except for the "shareholder rights" variable, the governance scores between the two groups were statistically different, but the differences were generally smaller than was the case for the controlled-ownership versus widely held ownership groups of companies.

Our regression analyses also indicate that whether or not a company was listed on the NYSE was not a statistically significant factor in explaining its performance.

Conclusions on Fiduciary Governance

All in all, our results are in line with those of most previous empirical studies. Whatever the measures of performance, whatever the sophistication of the analyses, we found that the "quality" of governance as measured by the ROB rating scores had no statistically significant relationship to performance. Given the large number of factors that influence company performance and the variable time lags between any action and its impact on performance, these results are not surprising.

Heightened public scrutiny and insistent monitoring by institutional investors have spurred board members to a high level of vigilance and stiffened directors' resolve to assert their authority over management. Such medicine, in an appropriate dose, may act as a prophylactic against egregious behaviour by management and may be helpful in restoring a healthy trust in public corporations. Fiduciary governance, once it has reached an acceptable level, may have a minimal impact on the performance of any particular company, but its most useful, albeit transient, role may be to rekindle the investing public's willingness to entrust its savings to the stock market.

At the same time, the insistent focus on "independent" directors, so dominant in the current reformist drive, may have deleterious effects. The ascendancy of legalistic and ill-informed "governors" with little upside benefits and great downside risks, urged to assert their authority over management, may instill in the whole corporation a timid, cautious, risk-averse, bureaucratic style of management.

Fiduciary governance, beyond its most basic contribution and its debatable "insurance policy" function, may well prove unsatisfactory, for it fails to address the fundamental problem of corporate governance in most circumstances: the generalized and persistent superiority of knowledge, information, access, and company-specific expertise that management enjoys over independent board members.

Shareholders'-Rights Governance

Shareholders'-rights governance is concerned with the extent to which the control of the corporation is "on the market." Here, the quality of governance is measured by the relative absence of any impediment to, or defences against, takeovers. From this vantage point, as Gompers, Ishii, and Metrick (2003) show, there are strong indications that an unfettered market for takeovers best serves shareholders' interests, since markets are better than any system of fiduciary governance at eliminating the incompetent and disciplining the complacent and the greedy. Free markets for control create value through swift and radical actions to change the ownership and management of poorly performing companies.

Yet, even in the United States, that freest of free-market economies, state governments, spurred by the wave of leveraged buyouts and hostile takeovers in the 1980s and 1990s, have enacted measures to moderate and discipline the free market for corporate control.

Thirty-one U.S. states (including Delaware) have "business combination" statutes that impose a moratorium, usually from three to five years, on transactions--such as merger, liquidation, and sale of assets--that a successful bidder can implement unless the target company's board has approved the transaction. Thirty-one states also have "constituency" statutes that allow (and that in two states oblige) directors legally to consider other stakeholders--such as employees, the local community, and suppliers--in assessing an offer to acquire control of a company. These "constituency statutes" provide directors with a legal basis (and protection) for rejecting a takeover even though the transaction might have been in the interest of shareholders.

In addition, 27 states have "control-share-acquisition" statutes that prevent a bidder from voting its shares beyond a specified threshold (typically between 20 percent and 50 percent) unless a majority of "disinterested" shareholders vote to allow the bidder to exercise the voting rights of its control stake. And 25 states have "pill-validation" statutes that endorse the use of "poison pills" against a hostile bidder.

U.S. companies have been enthusiastic in taking advantage of such anti-takeover provisions. As Gompers, Ishii, and Metrick (2003) show, of the roughly 1,500 firms in their study, by 1998, 88 percent had adopted "blank-cheque preferred shares" provisions, 59 percent had classified or staggered boards, and 55 percent had adopted poison pills, In all, fully 90 percent of these companies were subject to state-enacted restrictions on "business combinations." Of course, a company may generally opt out of these state-enacted restrictions.

In the European Union, where stakeholders other than the shareholders have more influence on corporations than in the United States, a number of countries have prevented takeovers of their "strategic" industries by issuing (or retaining after privatization) a "golden share" that effectively gives governments or their agencies control in the event of an impending change of ownership. The European Court, however, recently declared these golden shares anticompetitive and in breach of the EU's charter; in response, European governments are now rushing to find other means of achieving the same goals.

Shareholders" Rights and Governance in Canada

In Canada, restrictions on takeovers exist in a number of sensitive industries, such as banking, telecommunications, and the media. Furthermore, the widespread practice of dual or multiple share structures with different voting rights has been the most effective measure against takeovers, enabling large minority shareholders to control a majority of the votes and making a takeover impossible without their assent to the transaction.

Such practices do not, however, produce a "managerial dictatorship" whereby anti-takeover measures are used to entrench management or give it free rein to pursue its interests to the possible detriment of shareholders. Rather, what we have in Canada could be described as "minority shareholder dictatorships," where the concern is the actual or potential divergence of interest between minority shareholders and other shareholders, particularly with respect to such issues as takeovers, decisions on CEO succession, and executive compensation when minority shareholders are also members of management.

Although we found no statistically significant differences in performance between controlled-ownership and widely held companies, nor does the market indicate clearly an overwhelming preference for one or the other as measured by the ratio of market value to book value, many shareholders nevertheless believe in the intrinsic value of the "one-share, one-vote" principle and are urging companies to get rid of dual classes of shares.

This stance, though defensible on grounds of financial market efficiency--for example, an unconstrained market for control should trigger actions to replace underperforming management and boards--raises important policy issues.

Is control by minority shareholders--who ought to have the same long-term, wealth-creation interests as other shareholders--better than anti-takeover measures designed to entrench management? Perhaps, but anti-takeover measures are easier to challenge and remove. Is it possible for minority controlling shareholders to pursue objectives that are not in the interest of other shareholders? Certainly, but evidence that such behaviour is more frequent and more flagrant in controlled corporations than in widely held ones is less than compelling. The most scandalous corporate misdeeds of the past few years, with one exception (Adelphia), have occurred in widely held corporations.

Institutional shareholders may believe that the elimination of different classes of shares would give them the clout to prevent the adoption of anti-takeover measures designed to entrench management, thus opening the door to enhanced shareholder value through increased takeover activity. It should be pointed out, though, that the market prices these investors paid for shares in these "controlled" companies reflected this well-known constraint on their takeover. Furthermore, a free market for corporate control, combined with a relatively weak Canadian dollar, could quickly transform the Canadian corporate landscape in ways that would not be politically acceptable. Canadian governments, at either the federal or the provincial level, are unlikely to stand by while foreign buyers take over significant Canadian companies in large numbers. If Pennsylvania can act to protect "its" companies from being taken over by "foreigners" from New York, one can easily imagine the political leverage of such an issue in Canada.

Indeed, eliminating all multiple-vote, dual-class shares from the Canadian corporate scene could well lead, not to a free market for corporate control, but to the enactment of regulations and the establishment of additional government bodies to oversee and approve takeover transactions, particularly when the bidder is a foreign entity. We have been down that road before, and it is not one that would enhance the efficiency of Canadian capital markets or create much value for Canadian shareholders. The venerable "stakeholders" model of the corporation, which went underground in North America during the 1980s and 1990s, could well emerge with renewed political vigour.

Value-Creating Governance

The relationship between a company's board (as principal) and its management (as agent) is, at its core, deeply asymmetrical, no matter how remarkable the general business experience of the board's independent members. It is puzzling, therefore, that the popular programs to improve corporate governance avoid dealing with this "agency" problem, from which flow most problems of corporate governance. The fiduciary proposals discussed thus far, for example, do little to correct this fundamental flaw of corporate governance. Paying board members more money, urging them to own more shares of the company, or allowing them to hire outside consultants at the company's expense attempt to address the issue but serve only as poultice on a sore.

U.S. corporate governance guru Ira Millstein acknowledges the fundamental character of the issue, but says only that his "suggestions can be boiled down simply to this: diligent independent directors, properly led, informed and assisted, can circumscribe the agency problems" (2002). This is all very well, but not very helpful in practice.

To tackle this most fundamental issue, we propose the concept of "value-creating" governance, an approach grounded in solid theoretical foundations,(3) in the observations and prescriptions of the most astute and experienced practitioners of corporate governance, and in the corporate practices of large corporations diversified in unrelated sectors. Our suggestions are based not only on the well-developed framework of agency theory but also, and more practically on the ways through which some "premium-diversified" companies--such as General Electric, Emerson, and United Technologies--have managed to create value through their "strategic" or "internal" governance practices.(4)

The corporate offices of these diversified companies face the same kinds of agency issues as board members do: How can the corporate centre create value by ensuring that the management of each autonomous operating company (whether a group, division, or subsidiary) acts to maximize the economic value creation of its unit? How can the autonomy of operating managers be reconciled and combined with the need for adequate corporate control and the ability of the corporate office to make key decisions, to allocate capital, and to steer the overall company in the proper direction? These companies have had to think long and hard about such issues, find the governance processes to achieve their objectives, and set up the proper system of checks and balances.

We have carefully mapped how these processes and frameworks actually work in diversified companies that use value-creating governance (see Allaire and Firsirotu 1993), and we can give some structure and order to the issue by examining what we call the "four pillars of governance':

* pillar I: the legitimacy and credibility of principals (that is, board members);

* pillar II: the strategy process and dialogue;

* pillar III: the quality of financial and strategic information, and

* pillar IV: the compensation and incentive system.

The Legitimacy and Credibility of Principals

Board members are the agents of shareholders and the principals of management. As such, they must possess the attributes of legitimacy and credibility. Indeed, having legitimate and credible principals is the sine qua non--the necessary, but not sufficient, condition--for value-creating governance. Without this pillar of governance, the other three pillars tend to become ineffective, largely pro forma processes. It is, however, a common error of the governance debate to skip considerations of legitimacy and credibility and focus instead on the independence of board members and on improving the other three pillars.

Legitimacy

The standing of board members with shareholders and management depends on their legitimacy: By what criteria were the candidates for board membership selected? How much of their own wealth is at risk in the company? Did parties with important stakes in the company select them?

The reason most boards fail the test of legitimacy, either partly or totally, originates in the slow transformation of corporate ownership from large individual shareholders to myriad fragmented small holders and passive funds. As a result, the election of board members has shifted gradually to a perfunctory ritual controlled entirely by management. Of course, boards have always tended to be stacked with friends, cronies, and public figureheads, and under the circumstances, it is difficult to imagine how a more "democratic" process would have worked.

Times have, however, changed. Now, institutional investors often collectively own half or more of the public equity of a company, and they are vocal and pro-active on matters of governance. Yet, in the current calls for governance reforms, the rallying cry--the panacea for this most fundamental of governance flaws--seems to be "independence."

Legendary investor and curmudgeon Warren Buffett, chairman of Berkshire Hathaway, in his 2002 letter to the company's shareholders, writes:

The current cry is for "independent" directors. It is certainly true that it is desirable to have directors who think and speak independently--but they must also be business-savvy, interested and shareholder-oriented....Over a span of 40 years, I have been on 19 public-company boards (excluding Berkshire's) and have interacted with perhaps 250 directors. Most of them were "independent" as defined by today's rules. But the great majority of these directors lacked at least one of the three qualities I value. As a result, their contribution to shareholder well-being was minimal at best and, too often, negative. These people, decent and intelligent though they were, simply did not know enough about the business and/or care enough about shareholders to question foolish acquisitions or egregious compensation. (2002, 17.)

Bhagat and Black, after an extensive and exhaustive review of past empirical research on the relationship between board independence and long-term company performance, write:

A priori, it is not obvious that independence (without knowledge or incentives) leads to better director performance than knowledge and strong incentives (without independence). Maybe the optimal board has some knowledgeable, incentivized inside directors, and some independent directors -- who might thereby become better informed, and could also be better incentivized than many independent directors are today. (2001, 265.)

If a board member is truly, immaculately independent, does he or she then become, ipso facto, "legitimate'? No! Legitimacy means that board members represent and defend interests of shareholders either because they are, themselves, important investors in the company or have been directly selected by investors. Berkshire Hathaway, for example, would receive the lowest possible score for the independence of its board, given that its seven directors include two of its executives, Warren Buffett and Charles Munger, Buffett's wife and son, a partner in the law firm that does substantial work for the company and a director, Walter Scott, Jr., who co-invested with Berkshire Hathaway to acquire a $1.5 billion business!

Grudgingly, under the legal gun of the Sarbanes-Oxley Act of 2002, Buffett has committed to adding new independent members to the board. However, in complying with the new rules, Buffett will "select directors who have huge and true ownership interests [in Berkshire Hathaway] (that is, stock that they or their family have purchased...), expecting those interests to influence their actions to a degree that dwarfs other considerations such as prestige and board fees." (Buffett 2002, 19).

Bhagat, Carey, and Elson, in a study of the relationship between directors' stock ownership and management turnover, agree with Buffett's concern, concluding:

First, there was a significant correlation between the amount of stock owned by individual outside directors and firm performance....Second, and more important for the analysis, the greater the dollar value of the individual outside director's equity holdings in the enterprise, the more likely a disciplinary-type CEO turnover in a poorly performing company would exist. (1999, 15.)

Minimum stock-holding requirements for board members are a step toward enhanced legitimacy; in most cases, however, the amounts so invested represent but a tiny fraction of the board member's wealth. In the current acrimonious and litigious climate, board members have a great deal at risk, but all of it negative: their reputation may be damaged and their personal wealth may come under attack should their directors' liability coverage prove insufficient or deficient. These downside risks certainly promote exemplary "fiduciary" governance, but also perhaps a style of governance that is legalistic, punctilious, and hesitant.

Accordingly, we strongly suggest that any investor, or collection of investors, with a sizeable stake in a company (say, 5 percent or more) should be allowed to propose candidates for board membership. There could be a minimum holding period (two years for instance) before a shareholder would be allowed to participate in the nomination process.

We recognize that such a reform would not be possible without changes in corporate laws. However, a rather simple legal change, cumulative voting, which is already allowed under several Canadian jurisdictions, would open the door to more legitimate representation on boards. Small investors would also benefit from such a change by making it possible and plausible for them, or for associations representing them, to propose candidates for boards and to exercise their cumulative voting rights in favour of such candidates.

At any rate, the issue of board member selection and election is now squarely on the table. The SEC recently issued a staff report that examines several proposals to enhance the legitimacy of the nomination process for board membership (SEC 2003), including our proposal described above. Some companies have already and voluntarily adopted nominating procedures that give far more say to shareholders in deciding who is to represent their interests.

Credibility

Board members must not only be legitimate; they must also be credible. They must show evidence of knowledge and expertise pertinent to the corporation, not just general business credentials, accumulated over the years in industries that may have little in common with the company on whose board they sit.

Assuming a board member is legitimate and has solid business experience, he or she, on joining a board, must then acquire credibility with management. Achieving a reasonable level of credibility requires an important investment of time and intellect early on to acquire a good understanding of the company's strategic and competitive issues, the sources of its economic value, the quality of its leadership at various levels, its management values, its key drivers of share value, and so on. This requirement is thus far more demanding than the typical orientation session for new board members.

Board members will have achieved credibility with management when the latter believes that dialogue with the board is adding value, providing insight and useful advice, and when management realizes that directors will quickly spot, and vociferously react to, presentations that are factually incomplete, self-serving, or selectively designed to force assent.

Obviously, "independence" is no guarantee of credibility--it may, in fact, be the opposite. Hence, the conclusion, drawn from much good empirical evidence, that the "optimal" board would include a good mix of interested, credible, but not necessarily independent members, and independent outsiders, legitimate and eager to acquire credibility. The board should describe the profile of expertise and experience relevant to the company that would give the board enhanced credibility. The nomination process, however determined, should seek to close the gap between the target profile and current board membership.

The Other Pillars

Only legitimate and credible board members can give full effectiveness to the other three pillars of governance. Discussions of strategy, the control and monitoring of performance, and the design of incentive systems take on a different intensity and quality when carried out by legitimate and credible principals, and offer a great opportunity for shareholder value creation.

To achieve this potential, however, boards must deal effectively with another key aspect of the agency problem: information asymmetry. A board, even one that is made up of legitimate and credible principals, needs to ensure that it has access to reliable, valid, and timely information. The board must, therefore, design the information system it needs for its purpose, one that includes, of course, the standard information required for fiduciary purposes, but also includes strategic information that is particularly relevant to the type of business the board is governing. For instance, the board should have access to regular and independent assessments of the company's competitive position and its clients' level of satisfaction, as well as results of surveys on employees' levels of satisfaction and turnover.

The board should insist on a strategy process that includes discussions on strategic orientation early on, and directors should be on-side before management proceeds to prepare its strategic plan. Too often, directors get to review the "final" product only very late in the game--a polished, multicoloured slideshow production designed to impress and awe the audience, not to generate a lively discussion.

In short, various measures could be taken to reduce somewhat management's enormous information advantage. By focusing on key indicators of organizational, market, and financial performance and ensuring it receives reliable and timely information on these indicators from independent sources, the board could promote a healthy sharing of information.

Finally, an effective compensation and incentive system for management is fundamental to value-creating governance. It is also very tricky to design. It must deal with a host of issues, such as the proper balance of rewards between short-term and longer-term performance; the calibration of overall compensation to relevant markets; safeguards against tampering with performance measures and indicators; and the weight of internal measures (such as ROIC, EVA) versus external performance scores (such as the indexed stock price and market value added.)5

Conclusion

After a period of a general slackening of standards, the benefits of a tightening up of the fiduciary aspects of governance are undeniable. We argue, however, that the current spate of suggestions to improve governance might prove unsatisfactory, even counterproductive.

The obsessive quest for board independence--a sort of philosopher's stone of governance--could have unintended and undesirable consequences. The pursuit of unfettered shareholders' rights to sell companies to the highest bidder, albeit desirable from the strict vantage point of market efficiency, most likely would hit the brick wall of Canadian political realities.

Value-creating governance, on the other hand, offers arrangements that differ from the current orthodoxy. It seeks to resolve the fundamental dilemma that has plagued corporate governance since the days of Berle and Means (1932): the issues of legitimacy and credibility of board members as well as the information advantage of management over board members.

To address these issues, we propose:

* a fundamental review of nominating process for board membership; cumulative voting for the election of board members to enhance the representative character of boards;

* an intensive program to train and educate board members in the strategic, economic, and financial specifics of the company;

* an information system designed by and for directors to provide them with relevant, reliable, and timely financial and strategic information;

* an on-going dialogue between the board and management on key strategic issues, including early discussions of strategic orientations, and

* a much more thorough design of compensation systems to address issues that are rarely dealt with in the standard approach to management compensation.

The implementation of these measures would require a change in the way board members are selected and in the way boards operate. Over the past few years, "fiduciary" governance has gone through a sorely needed process of revitalization. We believe that, while retaining the positive aspects of that effort, it is now time to move forward toward a form of governance aimed at creating genuine value for shareholders.

Appendix: Definitions of Anti-Takeover Measures(6)

Anti-greenmail: Greenmail refers to the agreement between a large shareholder and a company in which the shareholder agrees to sell his stock back to the company, usually at a premium, in exchange for the promise not to seek control of the company for a specified period of time. Anti-greenmail provisions prevent such arrangements unless the same repurchase offer is made to all shareholders or unless the transaction is approved by a majority of shareholders. These measures are thought to discourage accumulation of large blocks of stock because one source of exit for the stake is closed, but the net effect on shareholder wealth is unclear. Five U.S. states have specific anti-greenmail laws, and two others have "recapture-of-profits" laws that enable companies to recapture profits, which is viewed as a version of anti-greenmail laws (albeit a stronger one).

Blank-cheque preferred stock: Preferred stocks over which the board of directors has broad authority to determine voting, dividend, conversion, and other rights. Although it can be used to enable a company to meet changing financial needs, this measure can also be used to implement poison pills or to prevent takeover by placement of the stock with friendly investors.

Business combination laws: Laws that impose a moratorium on certain kinds of transactions (for example, asset sales, mergers) between a large shareholder and the firm for a period usually ranging between three and five years after the shareholder's stake passes a pre-specified (minority) threshold.

Constituency statutes: Legal provisions authorizing (and, in two U.S. states, obliging) boards of directors to consider constituencies other than shareholders in evaluating a takeover proposal.

Classified board: One in which the directors are placed into different classes and serve overlapping terms. Since only part of the board can be replaced each year, an outsider who gains control of a corporation may have to wait a few years before being able to gain control of the board. This provision may also deter proxy contests, since fewer seats on the board are open each year.

Golden parachutes: Severance agreements that provide cash and non-cash compensation to senior executives upon a triggering event such as termination, demotion, or resignation following a change in control. They do not require shareholder approval.

Poison pills: Securities that provide their holders with special rights in the case of a triggering event such as a hostile takeover bid. If a deal is approved and the bidder proceeds, the pill is triggered. In this case, typical poison pills give the holders of the target's stock other than the bidder the right to purchase stock in the target company at a steep discount, making the target unattractive or diluting the acquirer's voting power. The early adopters of poison pills also called them "shareholders' rights" plans, ostensibly since they give current shareholders the "rights" to buy additional shares, but more likely as an attempt to influence public perceptions. A raider-shareholder might disagree with this nomenclature.

Supermajority requirements for approval of mergers: Charter provisions that establish voting requirements for mergers or the other business combinations that are higher than the threshold requirements of U.S. state law. They are typically 66.7, 75, or 85 percent, and often exceed attendance at the annual meeting. This category includes both the companies with this provision and those incorporated in U.S. states with a "control-share acquisition" law. These laws require a majority of disinterested shareholders to vote on whether a newly qualifying large shareholder has voting rights. In practice, such laws work much like supermajority requirements.

Unequal voting rights/dual class of shares: Provisions that limit the voting rights of some shareholders and expand those of others. Under time-phased voting, shareholders who have held the stock for a given period of time are given more votes per share than recent purchasers. Another variety is the substantial-shareholder provision, which limits the voting power of shareholders who have exceeded a certain threshold of ownership. Of course, this measure also covers situations where the company has issued different classes of shares with unequal number of votes attached to each category

Table 1: McKinsey Surveys of the Willingness to
Pay a Premium for a Well-Governed Company

Suppose you are considering investing in the following companies, A
and B, in the same country.
Past performance has been virtually identical and future market
potential appears to be similar for
both companies. However, they differ in board governance practices. B
has put in place "good"
board governance practices.

Company A: "Poor" governance           Company B: "Good" governance

* minority of outside directors;       * majority outside directors;

* outside directors have financial     * outside directors are truly,
  ties with management;                  independent, no ties
                                         with management;

* directors own little or no stock;    * directors are significant
                                         shareholdings;

* directors compensated only           * material proportion of
  with cash;                             directors' pay is
                                         stock-related;

* no formal director evaluation        * formal director evaluation;
  process;

* very unresponsive to investor        * very responsive to investor
  request for information on             requests for information on
  governance issues.                     governance issues.

Questions:

* "In those countries for which you are the key investment
  decisionmaker, would you be
  willing to pay more for company B's stock compared to A's?"

* "If yes, what percentage premium do you estimate you would be willing
  to pay for
  B's stock?"

Source: McKinsey & Company 2002.

Table 2: Performance and Governance Grades

    Governance      Equity        Sales       ROIC       VCI
      Grade         Market        CAGR        (%)
                    Value         (%)
                 ($ billions)

      A (85)         4.84          4.54        7.30      1.09
      B (75)         1.87         12.54        6.10      0.97
      C (65)         1.08         11.70        6.40      0.93
      D (54)         1.36         20.66        7.70      1.13
      F (44)         1.19         10.99        7.71      1.17

       %EVA         Tobin's        %MVA      Delta N      N
                       Q                       (%)

       0.64          1.22         -11.25     -13.92       29
      -0.15          1.62           3.72       2.47       32
      -0.52          1.38          11.26       5.44       34
       0.82          1.27         -11.80     -31.49       42
       1.11          1.27          14.27     -16.42       40

Note: n = 177; all results are the median of their distribution.

(a) Globe and Mail ROB ratings, classified as follows:

A = score of 80 and above;

B = score of 70 to 79;

C = score of 60 to 69;

D = score of 50 to 59;

F = score of 49 and below.

Table 3: Performance in Controlled Ownership and Widely Held
Ownership Companies

Mean              Equity       Sales      ROIC      VCI
Score             Market       CAGR       (%)
                  Value        (%)
               ($ billions)

                      Unweighted Mean Scores

Controlled          3.77         19        7.3       1.19
(n=51)

Widely held         3.94       15.1        6.5       0.94
(n=126)

                    %EVA       Tobin's     %MVA      Delta
                                  Q

                         Unweighted Mean Scores

Controlled
(n=51)               0.0        1.58       1.21     -0.049

Widely held
(n=126)             -0.6        1.61       0.55      0.042

Mean            Board       Compensation   Share-   Disclosure   Total
scores        Composition                  holder
                                           Rights

                           Globe and Mail Governance Scores

Controlled
(n=51)           19.4           11.7         15.1      5.6        51.8

Widely held
(n=126)          28.0           13.5         18.1      7.4        67.0

Table 4: Performance of NYSE-Listed and Nonlisted Companies

Mean              Equity       Sales      ROIC      VCI
Score             Market       CAGR       (%)
                  Value        (%)
               ($ billions)

                              Unweighted Mean Scores

Listed on
NYSE (n=69)        6.80        13.5        6.8      0.96

Non Listed
NYSE (n=108)       2.03        18.1        6.7      1.04

Mean              %EVA         Tobin's    %MVA      Delta
Score                            Q

Listed on
NYSE (n=69)       -2.4          1.59      -4.09     0.038

Non Listed
NYSE (n=108)       0.8          1.62       3.83     0.002

Mean            Board       Compensation  Share-    Disclosure   Total
scores        Composition                 holders
                                          Rights

                          Globe and Mail Governance Scores

Listed in
NYSE (n=51)       28.2          14.3       17.7        8.0        68.2

Non Listed
(n=108)           23.8          12.2       16.9        6.2        59.0

Table 5: The Four Pillars of Governance

Pillar I                        Pillar II

Legitimacy and credibility of   Strategy process,
principals (board members)      planning, and dialogue

Legitimate: who do you          As principal do you get
represent? What do you          to review and approve
have at stake?                  the strategic planning
                                process?
Credible: do you
understand this                 Does the process
business, its key drivers       include an early
of value, its strategic         discussion of
issues? Did you invest          orientations with the
time and intellect to           board before the
gain the respect of             strategic plan is
management for your             finalized?
understanding and
insights about the              Is there sufficient time
business? A necessary,          allocated to strategic
but not sufficient,             review and discussion?
condition for good
governance; weakness            Are key strategic issues
here turns the other            reviewed and discussed
three pillars into              with management at
proforma exercises.             regular board meetings
                                throughout the year?

Pillar III                      Pillar IV

Quality of financial and        Compensation and
strategic information           incentive system

Is financial information        Has the board set
reliable, valid, and            compensation
timely?                         principles and practices
                                that are optimal for this
Are significant                 particular company's
accounting judgments            shareholders?
and treatments well
understood? Has the             Are management
impact of alternative           incentives linked to
treatments been                 genuine value-creation?
assessed? Does the
board have access to            Does the compensation
reliable and                    system keep an
independent                     appropriate balance
information on                  between short-term and
competitive position,           longer-term economic
on client assessment of         performance?
the company? Are
capital investment
budgets and acquisition
proposals thoroughly
reviewed?

The authors wish to acknowledge the excellent support provided for this research by Jean-Francois Fremaux, project director at the J. Armand Bombardier Chair, Universite du Quebec a Montreal.

(1) It is not clear whether the authors adjusted this "abnormal" return for the transaction costs and higher risks that such an investment strategy would entail.

(2) For a more detailed description of these measures, see Allaire and Firsirotu (2003).

(3) See, for example, Jensen and Meckling (1976); Allaire and Firsirotu (1990, 1993); Moldoveanu and Martin (2001); and Monks and Minow (2001).

(4) See, for example, Welch (2001) for a description of General Electric's "operating system."

(5) The design of executive compensation systems is too complex to be covered adequately here. Readers may want to consult Allaire (2003) for a discussion of compensation issues and a set of suggestions for a new compensation system.

(6) The following definitions are taken from Subramanian (2002) and Gompers, Ishiii, and Metrick (2003).

References

Agrawal, Anup, and Charles R. Knoeber. 1996. "Firm Performance and Mechanisms to Control Agency Problems between Managers and Shareholders." Journal of Financial and Quantitative Analysis 31 (3): 377-97.

Allaire, Yvan. 2003. "Fair Wages for an Honest Day's Work: Executive Compensation for the Post-Enron Era." J. Armand Bombardier Chair working paper. Universite du Quebec a Montreal. January.

Allaire, Yvan, and Mihaela Firsirotu. 1990. "Strategic Plans as Contracts." Long Range Planning 23 (1): 102-15.

--. 1993. L'entreprise strategique: penser la strategie. Montreal: Gaetan Morin Editeur.

--. 2002. "La 'bonne' et la 'mauvaise' gouvernance." La Presse (Montreal), October 21.

--. 2003. "Corporate Governance and Performance: The Elusive Link." J. Armand Bombardier Chair working paper. Universite du Quebec a Montreal. April.

Berle, Adolf, and Gardiner Means. 1932. The Modern Corporation and Private Property. New York: Commerce Clearing Housing.

Bhagat, Sanjai, and Bernard Black. 1999. "The Uncertain Relationship between Board Composition and Firm Performance." Business Lawyer 54 (May): 921-63.

* 2001. "The Non-Correlation between Board Independence and Long Term Firm Performance." Journal of Corporation Law 27: 231-73.

Bhagat, Sanjai, Dennis C. Carey, and Charles M. Elson. 1999. "Director Ownership, Corporate Performance, and Management Turnover." Business Lawyer 54 (May): 885-919.

Bredeen, Richard C. 2003. "Restoring Trust: Corporate Governance for the Future of MCI, Inc." Report to the Hon. Jed S. Rakoff, United States District Court for the Southern District of New York. August.

Buffett, Warren. 2002. "Letter to Shareholders." Omaha, Neb.: Berkshire Hathaway.

CalPERS (California Public Employees Retirement System) 1998. Corporate Governance Core Principles and Guidelines. Sacramento, Cal.: CalPERS.

Coombes, Paul, and Mark Watson. 2000. "Three Surveys on Corporate Governance." McKinsey Quarterly (4): 75-7.

Dalton, Dan R., Catherine M. Daily, A.E. Ellstrand, and J.L. Johnson. 1998. "Meta-Analytic Reviews of Board Composition, Leadership Structure, and Financial Performance." Strategic Management Journal 19: 269-90.

Gompers, Paul, Joy Ishii, and Andrew Metrick. 2003. "Corporate Governance and Equity Prices." Quarterly Journal of Economics 118 (1): 107-55.

Hirschman, Albert O. 1970. Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States. Cambridge. Mass.: Harvard University Press.

Impavido, Gregorio. 2002. "On the Governance of Public Pension Fund Management." Working paper. Washington, DC: World Bank.

Jensen, Michael C., and William H. Meckling. 1976. "Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure." Journal of Financial Economics 3: 305-60.

Klein, April. 1998. "Firm Performance and Board Committee Structure." Journal of Law

McKinsey & Company. 2002. Global Investor Opinion Survey on Corporate Governance. New York.

Millstein, Ira M. 2002. Testimony to the Oversight Hearing on Accounting and Investor Protection Issues. U.S. Senate Committee on Banking, Housing, and Urban Affairs, February 27. Washington, DC.

Millstein, Ira M., and Paul MacAvoy. 1998. "The Active Board of Directors and Performance of the Large Publicly Traded Corporation." Columbia Law Review 98 (5): 1283-99.

Moldoveanu, Mihnea, and Roger Martin. 2001. "Agency Theory and the Design of Efficient Governance Mechanisms." Working paper. Joseph L. Rotman School of Management, University of Toronto. February 2.

Monks, Robert A.G., and Nell Minow. 2001. Corporate Governance, 2nd ed. Cambridge, Mass.: Blackwell Publishing.

Patterson, D. Jeanne. 2000. The Link between Corporate Governance and Performance: Year 2000 Update. Conference Board Research Report 1276-00-RR. New York: Conference Board..

SEC (U.S. Securities and Exchange Commission). 2003. "Review of the Proxy Process Regarding the Nomination and Election of Directors." Staff Report. Washington, DC. July.

Sonnenfeld, Jeffrey A. 2002. "What Makes Great Boards Great." Harvard Business Review, September.

Subramanian, Guhan. 2002. "The lnfluence of Antitakeover Statutes on Incorporation Choice: Evidence on the 'Race' Debate and Antitakeover Overreaching." University of Pennsylvania Law Review 150 (6): 1795-1873.

Welch, Jack. 2001. Jack: Straight from the Gut. New York: Warner Books.

Yermack, David. 1996. "Higher Market Valuation of Companies with a Small Board of Directors." Journal of Financial Economics 40 (2): 185-211.