The belief that governance best practices lead to superior firm performance is widespread. But as academic research and this article demonstrate, most studies prove that there is no link between governance and performance. Nor is there proof that the highly desirable director independence has a positive impact on firm performance.
During the bull market of the 1990s, the American model of corporate governance was heralded as the most successful in the world at creating value. Indeed, corporate law scholars Henry Hansmann and Reiner Kraakman predicted, in a 2000 paper provocatively titled “The End of History for Corporate Law,” that global corporate governance would converge around the U.S. shareholder-oriented model as a result of its exemplary record at creating value. (see www.ssrn.com; January, 2000).
The corporate scandals that began in October 2001 with the collapse of Enron and that continue to the present day have shaken investors’ faith in the capital markets and the efficacy of existing corporate governance practices in promoting transparency and accountability. The Conference Board’s Commission on Public Trust and Private Enterprise remarked in January 2003 that “[t]he events of the last year suggest that, in many instances, [the] compact among shareowners, boards, and management has been significantly weakened, diminishing the trust investors and the general public have in our system of corporate governance.”
Congress and regulators responded to this crisis of confidence by imposing new corporate governance requirements on public companies. For their part, investors started to take corporate governance issues more seriously. Moody’s Investor Services announced plans to incorporate governance assessments into credit ratings. To date, these and other measures have been premised on the assumption that corporate governance affects financial performance in some way. As an empirical matter, however, that proposition is far from settled. Indeed, researchers disagree on the existence and strength of the relationship between various corporate governance features and performance.
This article summarizes the results of studies that attempt to correlate corporate governance with firm performance. Because the literature is so vast, this article will address only governance issues relating to the board of directors and takeover defenses, which have received the bulk of the attention from researchers, and are considered particularly important by institutional investors.
The board of directors
A large proportion of the regulatory changes have focused on boards of directors. This is unsurprising, given the critical functions performed by the board and its key committees—audit, compensation and nominating/governance. Many prominent witnesses testified before Congress and the national stock exchanges that captive boards were a major contributing factor in the corporate scandals, and that reforms designed to increase the independence and competence of boards were crucial to restoring investor confidence. For example, former SEC chairman Arthur Levitt testified that “company boards often fail to confront management with tough questions”; he recommended toughened independence standards. The Sarbanes-Oxley legislation marked the first time federal legislation imposed an independence requirement on public company boards, requiring companies to have an audit committee composed exclusively of independent directors and defining independence more narrowly than previous regulatory provisions. Sarbanes-Oxley also prohibited new company loans to directors. Proposed changes to the listing standards of the New York Stock Exchange and Nasdaq would require a majority of directors on listed company boards to be independent, as well as all directors on key committees, and would strengthen the independence definition.
Some individual companies have taken additional steps to increase the independence of their boards. For example, Bankruptcy Examiner Richard Breeden, to prevent “the cronyism that existed in WorldCom’s past,” recommended that the post-bankruptcy board be made up exclusively of fully independent directors, with the exception of the CEO; that no compensation, consulting agreements, or payments of any kind to directors be permitted other than board and committee retainers; and that related party transactions involving board members be prohibited. (Although recent reforms accelerated the adoption of independence reforms and gave “best practices” the force of law, over the past three decades boards have been moving toward greater independence: In 1973, the average U.S. corporate board had five inside members; by 1998, that number had dropped to two. (Lynne L. Dallas, “Developments in U.S. Boards of Directors and the Multiple Roles of Corporate Boards,” Univ. of San Diego School of Law Public Law and Legal Theory Research Paper No. 48, Jan. 31, 2002; available on www.ssrn.com).
Enhanced director independence is intuitively appealing. Common sense tells us that a director with ties to a company or its CEO would find it more difficult to turn down an excessive pay request, challenge the rationale behind a proposed merger or bring to bear the skepticism necessary for effective monitoring. Despite the consensus that has developed around director independence, however, there is limited evidence that the so-called “résumé independence”-defined as the absence of familial, employment and financial relationships between directors and the company – promoted through both the reforms to date and codes of governance best practice actually improves corporate performance.
With one exception, empirical studies have not found a statistically significant positive relationship between the degree of board independence and better financial performance. In a recent study, Sanjai Bhagat of the University of Colorado at Boulder and Stanford’s Bernard Black found no correlation between the degree of board independence and four measures of firm performance, controlling for a variety of other governance variables, including ownership characteristics, firm and board size, and industry. (Journal of Corporate Law, Winter 2002). Bhagat and Black did find that poorly performing firms were more likely to increase the independence of their board. Earlier meta-analyses similarly failed to find a correlation between board independence and various measures of corporate performance. One study, by Paul MacAvoy and Ira Millstein, did find a positive relationship between a novel measure of independence and a measure of firm performance related to economic value added (EVA™).
Methodological differences may explain, at least in part, the difference between the MacAvoy and Millstein results, on the one hand, and the findings of the other studies. The studies that found no relationship categorized directors using publicly available disclosures from proxy statements regarding relationships between directors and companies. These disclosures have been criticized by shareholder activists as incomplete because they do not show ties between directors and individual company officers (especially the CEO), which have the same, if not greater, potential to compromise objectivity. They also do not contain data on board interlocks resulting from common membership on boards of private companies, universities, foundations and other organizations. The most notorious example is the board of Enron: If ties through charitable and other institutions had been disclosed in the proxy statement and a narrower definition of independence employed, three of Enron’s ostensibly independent directors would have been considered non-independent, including the chairman of the compensation committee.
MacAvoy and Millstein, by contrast, measured board independence using responses to a survey by the California Public Employees’ Retirement System. Those responses, which also provided information on board procedures and thus were broader than simply resume independence, produced grades. The grades were combined with the authors’ own assessments of the boards’ professionalism based on board behavior. Thus, the MacAvoy & Millstein study captured more board characteristics than other studies, and the performance effect may have been generated by one or more of those other factors. It is important to note that MacAvoy and Millstein, whose study is often grouped with others evaluating the effect of independence, couch the debate in terms of a “professional” board, not just one whose members have formal independence from the company.
The general inability of researchers to correlate board independence with firm performance may also reflect a more fundamental issue. The oversight tasks that were the central focus of the post-Enron independence reforms-restraining compensation, ensuring the integrity of financial reporting, and nominating new directors, to name a few-are all important parts of the board’s monitoring role. Monitoring requires directors to challenge and negotiate against management and, in extreme cases, terminate the CEO’s employment. It thus should come as no surprise that studies have found, for example, that companies whose boards exhibit greater formal independence have higher levels of executive turnover.
However, monitoring is only one of a board’s roles. Directors also contribute to a firm’s success by advising senior management, channeling outside resources to the firm and relating to stakeholders such as communities and employees. Researcher Jill Fisch has theorized that while independence may improve directors’ performance as monitors, it may be counterproductive with respect to managerial tasks. Similarly, Lynne Dallas has argued that the board’s multiple roles, and the differing significance of independence with respect thereto, support the use of a dual board structure.
The lack of a relationship between board independence and performance across a large number of companies does not, however, mean that increased independence is not the right prescription for any individual company’s board. When a board is falling short of fulfilling its monitoring duties-for example, where executive compensation is excessive in relation to performance or where the integrity of financial controls has been compromised in the past-increasing the number of independent directors on the board could reduce the likelihood of the problem recurring. But increased board formal independence should not be viewed as a panacea for poor corporate performance or board failure.
Takeover defenses, which impede the sale of a company without its board’s consent, are widely used among U.S. corporations. These devices, especially classified boards and poison pills, are frowned upon by institutional shareholders as they insulate boards of directors from being accountable to shareholders. None of the post-Enron regulatory reforms directly addressed takeover defenses, probably because attention was focused on the integrity of the financial statements and the role of the board of directors, and because state courts-especially those in Delaware-have created a substantial body of law regarding the deployment of defenses. But these issues continue to play a key role in how institutional shareholders evaluate corporate governance.
There is no shortage of empirical studies regarding the relationship between corporate performance and (1) the presence or adoption of a particular takeover defense and (2) company scores on an index that measures the presence of shareholder-disempowering corporate governance features (nearly all of which qualify as takeover defenses). The results of these studies are inconsistent and suggest that more research on the interaction between defenses, as well as the relationship between defenses and other firm characteristics, would be useful.
Studies of individual defenses
The earliest studies on the relationship between takeover defenses and firm performance examined the stock price effects in a short window following announcements by companies that they had adopted a particular defense. These “event studies” measured the reaction during an interval of a few days following the announcement, and have produced mixed results. Many such studies focused on poison pill adoption, with a few finding a modest negative effect in subsamples but most finding no statistically significant impact.
Others looked at the effect on stock prices of charter and bylaw amendments (often called “shark repellents”) designed to prevent or slow down hostile takeovers. The results of these studies were weak and inconclusive. Because event studies say more about the opinions investors hold about a particular defense than about the way the defense actually operates, differing shareholder perceptions about the defense, as well as about the information signaled by its adoption, may play a large part in explaining the indefinite results of these studies.
Another set of studies has examined the extent to which takeover defenses affect the premiums paid to target shareholders in completed transactions. Companies stress that takeover defenses can be used to give the board the breathing room and leverage to negotiate a better deal for shareholders. Several studies seem to support this notion, finding that target shareholders of companies adopting poison pills prior to receiving bids receive higher premiums. However, John Coates has argued that these studies should not be relied upon to show causation, pointing out that the larger premiums found in friendly as well as hostile deals are inconsistent with the pill acting as causative agent. He has theorized that some other factor-such as a mismatch between share price and intrinsic firm value, or a greater likelihood, that adopting firms are poorly managed-accounts for the study results.
More recently, a study by Lucian Bebchuk and two colleagues found that classified boards do not lead to the receipt of higher premiums by target company shareholders. The authors reviewed data regarding all hostile bids made from 1996 through 2000 and identified target companies with “effective staggered boards” (ESBs)-classified or staggered boards at companies where other measures are in place to prevent circumvention.
The study found that the average final premium for completed bids at targets with ESBs was 8 percent higher than premiums at companies without ESBs; that difference, however, was not statistically significant. Bebchuk’s study also found that shareholders of ESB targets that were not acquired did not recoup the value of the foregone takeover premium through increases in share value in the first 30 months of remaining independent. In response to criticism of the study,
Bebchuk conducted a small follow-up study that evaluated potential benefits of a classified board in negotiated transactions, which found that shareholders in targets with ESBs did not receive higher premiums. The flip side of acquisition premiums is deterrence-a defense may increase the premium paid to one company while deterring a bidder from trying to acquire another company. Measuring such deterrence presents methodological challenges, because companies adopting defenses may be unusually likely to be the target of a bid. Most studies regarding poison pills have not found evidence of deterrence, although one early study provided evidence that firms with pills in place before a bid was made were more likely to defeat bids than those without pills. All but one study of shark repellents found no evidence of deterrence.
As the preceding discussion makes clear, it is difficult to draw conclusions about the financial impact of individual takeover defenses, given the many other economic and governance factors operating on companies. Indeed, Coates (who was also a co-author of the Bebchuk study) urges that the interaction among defenses is more important than the presence of any individual defense.
Such interactions have begun to draw serious scholarly attention. In 2001, Andrew Gompers and two colleagues attracted a great deal of attention when they announced the results of a study that had found a significant positive relationship between companies’ scores on a multi-factor governance index and long-term corporate performance. The study, which was published in the Quarterly Journal of Economics in early 2003, constructed an equal-weighted governance index for 1,500 companies reflecting 24 variables relating to shareholder rights, almost all of which involved takeover defenses.
Companies were ranked from those with the highest scores, which were least empowering of shareholders, to those with the lowest scores, which were most empowering. The authors found that an investment strategy that sold short the firms in the highest decile and bought those in the lowest decile would have earned abnormal returns of 8.5 percent per year during the period from September 1990 to December 1999. The study also found a statistically significant negative relationship between scores on the governance index and Tobin’s Q.
A more recent study by K.J. Martijn Cremers and Vinay Nair expanded on Gompers’ work and highlighted a potential weakness of the original study. The Cremers and Nair study investigated the interaction between the “external governance” factors examined by Gompersbasically, the vulnerability of a company to hostile takeover-and “internal governance” mechanisms, which are measured in two ways: the percentage share ownership by institutional blockholders, defined to be an institutional shareholder with equity ownership greater than 5 percent, and the percentage of share ownership by public pension funds. They found that high external governance was correlated with significant abnormal returns only when measures of internal governance were also high, and that high internal governance generated such returns only in the presence of high external governance. Interestingly, after changing the cutoff on the Gompers index for firms with low external governance from 14 to 13, and extending the study period by two years (through 2001), the authors found that external governance was not significantly correlated with abnormal returns. This study, then, highlights the complexity of the governance/performance relationship and its sensitivity to study design.
The empirical research on takeover defenses does not support blanket approval for, or condemnation of, their use. Instead, it suggests that firm-specific factors such as the share ownership profile, and the inclination of the board to use defenses for entrenchment rather than bargaining power, are as important as the specific defenses employed. Shareholders and others seeking to evaluate the governance risk posed by defenses must incorporate an assessment of management and board incentives, such as stock ownership, as well as the track record of the board in managing conflicts of interest.
Since the wave of corporate scandals began, a consensus has developed around the importance of good corporate governance to individual companies and the U.S. economy as a whole. Companies are under more pressure than ever before to adopt governance best practices and to convince investors that their governance is responsible. The easy course may be simply to adopt a one-size-fits-all model, and there are features-such as independent board committees-that make sense across the board. But as the academic research shows, there is no governance “magic bullet,” and no substitute for thoughtful, contextual analysis.