by Lynn Stout
It is essential to recognize that neither contingency is met unless the board of directors wants it to be. Focusing on the firm’s financial health and legal ability to pay dividends, “retained earnings” and “profits” are accounting concepts over which directors have enormous control. Both depend not only on how much money the company brings in (earnings), but also on how much money it spends (expenses). Directors can’t always increase earnings, but they can increase expenses. If a company is raking in cash, its directors have the option of allowing accounting profits to increase. They also have the option of raising executives’ salaries, starting an on-site childcare center, improving customer service, beefing up retirement benefits, and making corporate charitable contributions. Thus, even when a company is minting money, it is the board that decides how much of the new wealth will show up in the company’s financial statements in a form that can be paid to shareholders.
Second, even when a corporation has enough profits or retained earnings to legally pay a dividend, directors are under no obligation to declare one, and often don’t. It is standard operating practice for U.S. corporations to pay only small dividends or no dividends to their shareholders, retaining the lion’s share of earnings for future projects. If this practice boosts the company’s stock price, shareholders enjoy an indirect economic benefit. But the benefit is only indirect, and vulnerable to the board’s decisions. If the board decides to run the firm in the interests of customers, employees, or executives—or simply run it into the ground—the earnings will become expenses, and stock price will decline.
This means it’s simply incorrect, as a factual matter, to describe the shareholders of a public corporation that is a going concern as the company’s residual claimants. Shareholders are only one of several groups that—at the board of directors’ discretion—are residual claimants and risk bearers in corporations, in the sense that they gain and lose as the company’s health fluctuates. When a corporation does well, its board may indeed declare bigger dividends for shareholders. But the directors may also decide, in addition or instead, to give rank-and-file employees raises and greater job security, to provide executives with a company jet, or to retain the cash so bondholders enjoy increased protection from the risk of corporate insolvency. Conversely, stakeholders suffer along with shareholders when times are bad, as employees face layoffs, managers are told to fly coach, and debtholders find their bonds downgraded. Directors use their control over the firm to reward many groups with larger slices of the corporate pie when the pie is growing, and spread the loss among many when the pie shrinks. The corporation is its own residual claimant, and it is the board of directors that decides what to do with the corporation’s residual.
The Third Mistaken Assumption:
Shareholders Are Principals and Directors Are Their Agents
Finally, a third fundamental belief associated with the principal-agent model of the corporation is that shareholders and directors are just that—principals and agents. Again, this premise is wrong.
In law, the word “principal” normally refers to someone who hires another person (an “agent”) to serve his interests. Thus the principal exists prior to, and independent of, the hiring of the agent. Yet when a corporation is formed, the first thing that must happen is that the firm’s “incorporator” must appoint a board of directors to act on the corporation’s behalf. Only after the board exists does the corporation have the power and ability to issue stock and so contract to acquire stockholders.65 Both the corporation itself and its board of directors (the supposed “agents”) must exist prior to, and independent of, the stockholders (the supposed “principals”).
Even more significant, a hallmark of agency is that the principal retains the right to control the agent’s behavior.66 Yet one of the most fundamental rules of corporate law is that corporations are controlled by boards of directors, not by shareholders.67 This does not mean that corporate law does not grant shareholders certain rights that can give them influence over boards. Indeed, shareholders have three—the right to vote, the right to sue, and the right to sell their shares. But all three rights have remarkably little practical value to shareholders seeking to make directors of public companies do their bidding and serve their interests.
Consider first shareholders’ voting rights. As a matter of law these are severely limited in scope, primarily to the right to elect and remove directors. Shareholders in public corporations have no right to select the company’s CEO; they cannot require the company to pay dividends; they cannot stop directors from squandering revenues on employee health care, charitable contributions, or executive jets; and they cannot vote to sell assets or the company itself (although they can sometimes veto a sale or merger proposed by the board). Voting procedures further limit the shareholder franchise. Delaware law, for example, assumes only directors have authority to call a special shareholders’ meeting, and shareholders who wait for the annual meeting to try to elect or remove directors must pay to solicit their own proxies. Perhaps most importantly, shareholder activism is the classic example of a “public good.” In a public firm with widely-dispersed share ownership, shareholders’ own “rational apathy” raises an often insurmountable obstacle to collective action. In the words of corporate law guru Robert Clark, a cynic could conclude that shareholder voting in a public company is “a mere ceremony designed to give a veneer of legitimacy to managerial power.”68
What about shareholders’ right to sue corporate officers and directors for breach of fiduciary duty if they fail to maximize shareholder wealth? As we saw in Chapter 2, here too, shareholders’ rights turn out to be illusory. Executives and directors owe a fiduciary duty of loyalty to the corporation that bars them from using their corporate positions to enrich themselves at the firm’s expense. But thanks to the business judgment rule, unconflicted directors remain legally free to pursue almost any other goal. Directors can safely donate corporate funds to charity; reject profitable business strategies that might harm the community; refuse risky projects that benefit shareholders at creditors’ expense; fend off hostile takeover bids in order to protect the interests of employees or the community; and refuse to declare dividends even when shareholders demand them.69 Contrary to the principal-agent model, shareholders in public companies cannot successfully sue directors simply because those directors place other stakeholders’ or society’s interests above shareholders’ own.
Finally, the right to sell shares sometimes can protect a disgruntled individual investor who wants to express her un-happiness with a board by “voting with her feet.” But when disappointed shareholders in public companies sell en masse, they drive down share price, making selling a Pyrrhic solution. An important exception to this rule arises in the case of hostile takeovers, where a public company’s shareholders may have a collective opportunity to transform the company into a private firm by selling their shares to a single buyer who, because she does not face collective action problems, can remove an uncooperative board cheaply and quickly. During the 1970s and early 1980s, as the Chicago economists’ arguments began to gain steam and changes in the banking industry made hostile takeover bids more feasible, it appeared that just such a lively “market for corporate control” might develop. But a series of quick legal reactions soon brought most hostile takeovers to a halt. These include the passage by almost every state of some form of antitakeover statute; the invention of the “poison pill” antitakeover defense by uber-corporate lawyer Martin Lipton; and the practical reversal of the Delaware Supreme Court’s 1986 Revlon ruling (which at first seemed to require boards to maximize shareholder wealth) by cases decided only a few years later.70 The end result is that the economic and governance structure of public corporations continues to insulate boards of directors from dispersed shareholders’ command and control in ways that make it impossible to fit the square peg of the public corporation into the round hole of the “standard” principal-agent model.
So Why Embrace the Principal-Agent Approach
?
It thus turns out that, when examined more closely, all three basic assumptions about corporate structure typically associated with shareholder value thinking—the assumptions that the shareholders own the corporation, that they are its residual claimants, and that they are principals who hire directors as their agents—are factually incorrect. This raises the question of why, as Hansmann and Kraakman observed in 2001, we have seen a “rapid convergence on the standard shareholder-oriented model as a normative view of corporate structure and governance.”71 If shareholders are not really owners, residual claimants, or principals in corporations, why should we want to run corporations as if they are?
To the extent an answer to this fundamental question is found in the literature on corporate theory, the answer seems to be that shareholder primacy is believed to be desirable because it is thought to offer the best solution to the agency cost problem described by Jensen and Meckling. After all, directors and executives are only human. If given a broad range of discretion to run firms in the interests not only of shareholders but also stakeholders and possibly even society at large, they might be tempted to use their autonomy to serve themselves. As Frank Easterbrook and Daniel Fischel described the argument in 1991, “a manager told to serve two masters (a little for the equity holders, a little for the community) has been freed of both and is answerable to neither.”72 As Mark Roe of Harvard Law School put it more recently, shareholder value maximization may be the best rule of corporate governance because “a stakeholder rule of managerial accountability could leave managers so much discretion that managers could easily pursue their own agenda, one that might maximize neither shareholder, employee, consumer, nor national wealth, but only their own.”73
In Jensen’s and Meckling’s terms, director discretion leads to agency costs. And as Jensen and Meckling also argued, agency costs can be reduced when an alert principal exists to measure and monitor the agent’s performance. To the current generation of corporate experts and business leaders, it seems obvious that shareholders should be that principal. It also seems obvious that if we focus on shareholder value and especially on share price in measuring corporate performance, it becomes harder for managers to claim that they are doing a good job for the firm, when in fact they are merely doing well for themselves.
But this is all in theory. If agency costs are really as large an economic drain on corporations as shareholder primacy advocates assume, and if changing corporate governance rules to make boards more accountable to shareholders and more focused on increasing shareholder wealth is really as effective a solution, we should see evidence of this in the business world. Adopting shareholder value maximization as corporate goal should improve corporate performance.
And it is here that shareholder primacy theory finds itself most vulnerable. We have seen how, as a descriptive matter, shareholder primacy ideology is inconsistent with both corporate law and with the real economic structure of public corporations. Next we shall see how it is inconsistent with the empirical evidence as well.
CHAPTER 4
How Shareholder Primacy Gets the Empirical Evidence Wrong
Chapters 2 and 3 explored how shareholder primacy thinking is neither required by law nor consistent with the real economic structure of public corporations. Nevertheless, the law permits companies to embrace the goal of shareholder value if they elect to do so. A corporation could, for example, mandate shareholder primacy in its charter (although as we have just seen, virtually no public corporation does so).
But as shareholder primacy advocates often point out, there are less-extreme strategies that companies also could adopt to make directors and executives more eager to embrace shareholder value as a goal. For example, a company might encourage its directors to focus more exclusively on shareholder interests by ensuring they are “independent” (that is, not also employed as executives by, or doing business as creditors or suppliers with, the firm). Another strategy that keeps boards attentive to public shareholders’ demands is to make sure the company has only one class of equity shares with equal voting rights, not a “dual class” equity structure where there is a second class of shares, typically held by managers or other insiders, with greater voting power. Yet a third way to give shareholders greater influence over boards is to remove “staggered board” provisions that typically allow shareholders to elect only one-third of the members of the board in any given year, thus making it easier for dissident shareholders to try to replace the entire board in a single proxy voting season. Similarly, removing anti-takeover defenses like “poison pills” makes it harder for the board to fend off a hostile takeover bid at a premium price and so also makes directors more attentive to keeping share price high.
If shareholder primacy ideology is correct, companies that come closer to the ideal of the standard shareholder-oriented model by adopting these sorts of internal governance rules and structures should show superior economic performance compared to those that do not, including increased profits, greater growth, and—most importantly and most obviously—higher share prices and returns to investors. This observation raises an exciting possibility. We don’t need to rely on theorizing to determine if shareholder value thinking is best. We can test ideology with real data.
Testing the Shareholder Value Thesis: No Clear Results
Many modern finance economists and legal scholars have attempted just this project. Legal and economic journals are full to bursting with papers that examine the statistical relationship between various measures of corporate performance and supposedly “shareholder friendly” elements of corporate governance like director independence, a single share class, or the absence of staggered boards and poison pills. Dozens of empirical studies test the supposed superiority of the shareholder-oriented firm. There remains a notable shortage of reliable results showing that shareholder primacy actually works better.
Consider two of the most popular types of empirical tests, cross-sectional analyses that compare the performance of corporations with shareholder friendly governance structures against more manager-oriented companies, and event studies that look at what happens when firms adopt particular shareholder primacy “reforms.” In both cases, the basic technique is to test the statistical relationship between some element of internal governance (a staggered board, dual classes of shares with different voting rights) and some measure of corporate performance (typically share price but sometimes other measures like operating income or “Tobin’s Q,” the ratio between the book value of the company’s assets and the value of the company’s shares in the eyes of the stock market).
The result of all these empirical tests? Confusion. For example, one recent paper surveyed the results of nearly a dozen empirical studies of what happens when companies have multiple share classes. It concluded that some studies found that dual class structures had no effect on performance, some found a mild negative effect, and some a mild positive effect.74 Moreover, at least one study found that multiple share classes greatly improved performance—exactly the opposite of what the standard shareholder primacy model would predict.75
Similarly, there seems to be no reliable connection between various measures of corporate performance and the percentage of independent directors on a board.76 Statistical analyses of the effects of poison pill and staggered board antitakeover defenses also have produced mixed results,77 as have studies of the effects of compensating directors with shares.78 One study of the performance of U.S. financial institutions during the 2008 credit crisis found that the stock prices of companies that came closer to the shareholder primacy ideal actually did worse.79
The lack of empirical support for the supposed superiority of the shareholder-oriented model has not gone unnoticed.80 Influential corporate scholar Roberta Romano of Yale Law School has denounced some shareholder-oriented governance reforms as “quack corporate governance.”81 In an important survey paper coauthored with Sanjai Bhagat of the University of Colorado and Brian Bolton of Whittemore business school, she
concludes that “the empirical literature investigating the effect of individual governance mechanisms on corporate performance has not been able to identify systematically positive effects and is, at best, inconclusive.”82 The U.S. Court of Appeals for the District of Columbia Circuit recently handed down a similarly scathing critique of a Securities Exchange Commission (SEC) decision to impose on public companies a “proxy access” rule that gave certain shareholders seeking to nominate and elect their own candidates to the board the right to use corporate funds to send proxy solicitations to their fellow shareholders. In striking down the rule, the court noted that the evidence on the benefits of proxy access was at best mixed, and failed to support the SEC’s conclusion that making it easier for shareholders to nominate board candidates “will result in improved board and company performance and shareholder value.”83
Fishing with Dynamite: Why Individual Company
Performance Isn’t the Right Metric
The remarkable lack of a reliable empirical connection between shareholder-oriented governance practices and better corporate performance at the level of the individual corporation, taken on its own, should make us hesitate mightily before assuming that corporate law “reform” will produce better results. But the evidence of a link is even weaker than it appears. This is because most empirical studies focus only on how governance changes influence economic performance at the level of the individual company, typically measured over a few days or at most a year or two.84 These studies may be looking in the wrong place and at the wrong time period. It is not only possible, but probable, that individual corporations can use strategies to “unlock shareholder value” that have the effect of increasing the wealth of certain investors at certain times, while perversely reducing aggregate shareholder wealth over the long term.
