Daniel J.H. Greenwood

Home | Previous Page

Fictional Shareholders: For Whom Are Corporate Managers Trustees, Revisited

69 Southern California Law Review 1021 (1996)
Download printable (.pdf) version

Daniel J.H. Greenwood [FN*]

Copyright © 1996 University of Southern California; Daniel J.H. Greenwood

TABLE OF CONTENTS

I. INTRODUCTION

It is a commonplace of American law that corporations are fictional. The U.S. Supreme Court said so, in the first important corporate law case to come before it, Bank of the United States v. Deveaux, [FN1] and repeated it in the next, the famous contract case of Dartmouth College v. Woodward: [FN2] The corporation is simply a convenient, though misleading, way to refer to its shareholders or members. Many modern theorists agree that the corporation is a metaphor, though they have different visions of what it "really" is. [FN3]

But despite this ancient and sophisticated discourse regarding the corporation, the literature and cases have relatively little discussion of the shareholders. This omission is particularly glaring in light of the dominant paradigm of corporate law, which holds that the central task of corporate law is to lessen or eliminate the potential conflict between shareholders and corporate managers -- the so-called problem of separation of ownership and control identified in Berle and Means' seminal work [FN4] and put into its modern form by Jensen and Meckling. [FN5]

Modern cases and theory, like the older ones, assume that shareholders, unlike corporations, are not problematic. Corporations may be legal fictions, mere metaphors for underlying -- and quite different -- realities. But shareholders, it is generally assumed, are not problematic at all. They are widows in Iowa, profit-maximizing investors or -- more recently -- institutional investors, and little further discussion is needed. After all, whatever else shareholders may or may not want, every shareholder wants to make a profit and that is all that is really important for the operation of corporate law and, indeed, the corporation itself.

Virtually all the major groups of corporate law scholars today agree on the centrality of the shareholder to corporate law; all but the communitarians agree that virtually the sole task of corporate law is to ensure that managers act as agents for the shareholder owners. [FN6] This Article directly challenges the almost universally held assumption that shareholders, in the form understood by the law, are a group of human beings entitled to respect and consideration and having interests that exist independent of corporate law. I contend, rather, that corporate law theorists have missed the critical point that an agency (or trust) relationship has quite a different significance when the "principal" (or beneficiary) is a set of legally defined interests that are not under the control of any individual or group of individual human beings who could choose to redefine or act in opposition to those interests.

This Article, then, is an attempt at a careful look at the role of shareholders in corporate legal theory. Shareholders, I contend, are a legal fiction, and in many ways a far more problematic fiction than the corporation itself. Indeed, since corporate law and the market alike drive corporations to act in the interests of these fictional shareholders, the shareholder is the most important fiction of corporate law: The legally imputed characteristics of corporate shareholders are the power behind the throne of managerial autonomy, the driving force that determines the structure and functioning of our corporate system. For this reason, we need to examine the nature of our fictional shareholders more carefully: Both the successes and the failures of our system ultimately reflect the characteristics of the shareholder we have created.

Specifically, I contend that the fictional shareholder is fundamentally different from the human beings who ultimately stand behind the fiction. The law and the legally created structure of corporation and market filter out all the complexity of conflicted, committed, particularly situated, deeply embedded and multi-faceted human beings, leaving only simple, one-sided monomaniacs. Human beings have short lives, spent in particular places with particular relationships to other human beings; they constantly confront the problems of finitude and commitment. Shareholders, in contrast, are in significant senses immortal, uncommitted and universal: They are indifferent as to time and place, language and religion. They are indifferent between projects and personalities. They are understood to care deeply about one important and vital human aim -- profit maximization -- but not at all about numerous others. While the ultimate owners of the shares are specific, situated, conflicted and committed human beings, shareholders in most instances may be thought of more appropriately as a "large, fluid, changeable and changing market." [FN7]

These differences between fictional shareholders and human beings can be grouped into two broad categories, each with a distinctive impact on the society and economy we use them to create. First, like classical utilitarians and the market itself, shareholders do not take the distinction between persons seriously. [FN8] That is, in many important situations they are indifferent to distributional issues that are critical to ordinary human beings. Second, they are fundamentally inhuman because they have only one goal, profit maximization -- and, thus, need not make the compromises among conflicting goals that are the essence of human politics and life.

The consequences for corporate law are also twofold: First, in the eyes of the law and corporate management, shareholders are all the same. As a result, managers are given relatively clear direction without any need to pierce the cacophony of inconsistent demands from conflicted and conflicting individuals. Corporate management is therefore far easier than political management. This simplicity, however, is based on an illusion -- the conflicts do not disappear merely because the law presumes that shareholders are above them.

Second, the actual owners of the shares are irrelevant to corporate law: Neither the interests nor the desires of the people behind the shares count. Because managers manage on behalf of a fictional principle rather than a human principal, corporations are a strange, driven kind of institution -- neither managers nor anyone alse has the ultimate authority to stop the institution from acting out its logic to the fullest.

This Article proceeds as follows. First, I explain what I mean by calling shareholders fictional and outline in more detail the basic characteristics of the legally determined fictional entity. Second, I illustrate some ways in which the fictional shareholder imposes its will on the corporation -- here, I follow the current consensus that the conflict between managers and shareholders has been resolved in favor of shareholder control, but with a twist, since I view the corporations not as controlled by human owners but rather as run in the largely legally defined interests of fictional creations. Neither those legally defined interests nor their fictional holders can be mapped in any simple way onto an underlying group of human beings. Finally, I explore the consequences of having our largest institutions run in the interests of a legal fiction and offer some preliminary suggestions regarding areas in which an institution run in the interests of fictional shareholders will be similar to, or different from, one run by or in the interests of human beings.

Berle and Means' classical corporate theory and their leading contemporary critics agree that corporate law should strive to organize corporations so that managers act in the interests of shareholders. In the modern jargon, corporate law should seek to reduce the agency costs inherent in the separation of ownership and control. In contrast, this Article's analysis suggests that the agency metaphor is deeply misleading. Since shareholders are a legal fiction rather than living, breathing human beings in their full richness, they are not principals in any ordinary sense. The corporation, then, is not usefully understood as a more or less perfect agent acting more or less responsibly on behalf of its principal. Rather, corporate law creates a corporate entity that may behave distinctly differently from the ways in which any (or all) of the human participants would behave were they free from legal constraint.

For corporate theory, this shift in perspective is of enormous importance. If the corporation's shareholders cannot be identified with human citizens of the political community, then even the most sophisticated proof that the "genius" of American law forces corporations to act in shareholder interests [FN9] cannot demonstrate that corporate actions reflect the will or interest of any citizen or group of citizens. Rather, the corporation becomes an independent actor in our polity and economy. [FN10] Because the fictional shareholder is fundamentally different from any human being -- even human beings who own shares -- a corporation acting in shareholder interests will act quite differently from the way its supposed principals would have it act.

Corporate freedom, it follows, need not necessarily promote human freedom, as most theorists have assumed; [FN11] nor can corporations easily be assimilated to the private side of the great public/private divide in liberal theory. [FN12] Rather, corporations -- even when the market and law work as they are supposed to, even without considering deviations from the norms of shareholder control or competitive markets -- belong closer to the governmental side, as elaborate human creations, meant to promote human happiness but potentially taking on a life and mission of their own. [FN13]

The corporate system we have created generates a conflict that is not reducible to either of the classic conceptions: It is not a class conflict, as that term is understood in either the Marxist or sociological traditions, and it is not the agency conflict with which so much of corporate law is concerned. It is, instead, more closely related to the problem of government as understood by the classic liberal theorists: a human institution which may often and in predictable ways cease to serve the limited (if essential) purposes for which it was formed.

In short, we have created an institution for a specific purpose and put it on a sort of automatic pilot, so that it continues to pursue the preset goals whether or not they continue to be useful. Corporate law succeeds because it is single-minded, and fails because it lacks a principle of moderation or any significant countervailing power.

II. UNDERSTANDING SHAREHOLDERS: THE THEORY OF THE FICTIONAL SHAREHOLDER

A. Distinguishing People Who Own Shares From Shareholders

Shareholders are a convenient and sometimes extremely misleading metaphor that can prevent us from seeing the real rights and responsibilities at issue. Plenty of people own shares -- indirectly, virtually every American with a pension plan or an insurance policy (which is to say virtually every American with any assets at all) is the beneficial owner of some stock. [FN14] But shareholders as understood by the law are neither a class or social group in our society nor a random collection of people who happen to own shares.

Rather, shareholders are treated as if their entire identity were their shareownership, as if their sole goal in life were to maximize the risk- adjusted present value of the future income stream represented by those shares and as if they had no competing interests that might, even occasionally, warrant taking an action not designed to improve "shareholder" value. [FN15] The world is a wide and diverse place, and I imagine such single- minded people must be out there somewhere. But they are a scarce and mysterious species, as hard to find as gremlins in caves or golems in old shuls.

Shareholders are a legal fiction in a very precise sense. The law [FN16] demands that corporate directors and managers manage the corporation in the interests of the shareholders and the corporation. But by "shareholder interests" the law does not mean the interests -- let alone the will -- of the actual people who are the beneficial owners of the shares (or, in our increasingly institutional stock market, the people who are the ultimate beneficiaries of the legal entities that own the shares). The actual people are not consulted; they have only primitive, indirect and ineffective means of letting their perceived interests or actual will be known. No owner of shares ever negotiates a contract with or submits instructions to the directors or managers. Nor does the board act like a sort of Benthamite neutral observer examining the life situations of the actual people out there and determining that, whether they know it or not, their interests require some action or other.

Rather, the law creates a simplified and fundamentally inaccurate image of a hypothetical shareholder and then requires that the interests of this nonexistent person be the focus of corporate efforts. [FN17] It is the interests of a fictional person whose sole interest is the shares it owns that is the focus of legal and corporate efforts to promote "shareholder" interests -- in effect, the shareholder is reduced to the shares.

I call this shareholder fictional because it is a coherent story, an essentially complete and unified being, lacking the complexity and contrasting commitments of real, human, people. When corporations are seen as owned by the fictional shareholder, the struggles of the corporate world seem to be a simple novel about a central character with one driving force, one story to tell and -- although this has not often enough been remarked -- one fatal flaw.

The fictional shareholder is also fictional because of a specific falsehood: the ideological belief that shareholders, as they are understood in the law and the marketplace, can be identified with specific individual human beings, and therefore, that defending shareholder rights is the same as defending human rights. I hope to show, in contrast, that while shareholder rights may often promote human rights and that while shareholder gains are often human gains, the correlation is by no means complete or automatic. In this sense, shareholders are fictions in precisely the same way that corporations are called legal fictions: Sometimes one must look through or around the legally visible entity to understand the human relations that are affected by it.

It is fictional, also, because this one-sidedness means that the shareholder's story can be written by outsiders; unlike true stories, it need not be written by or with the participation of the participants themselves. [FN18] We -- whoever we are -- can know what the shareholder wants, without knowing anything about the person behind the shareholder. The Fischer Separation Theorem tells us that shareholder risk preferences do not matter; the simplification of the fiction excludes almost every other value. Like Thomas Hobbes' man in the state of nature, shareholders have no names, no particular needs, no specific individuality, place or generation. But if the life of man in the state of nature is "solitary, nasty, poore, brutish and short," and primarily motivated by a drive to power and glory, [FN19] the life of the fictional shareholder is an endless pursuit of present discounted future returns; a "life" that is indeed solitary, and nasty on many understandings, but thin rather than poore, and eternal rather than short. The similarity is that all shareholders are alike, and philosophers (or business managers) may argue about how best to serve their interests without ever needing to consult with the subject of the debate.

I do not mean to suggest that the shareholder is fictional in the sense of nonexistent. On the contrary, shareholders are legally created entities that exist in the strong sense that they determine much of our lives. Indeed, one variant of the fictional shareholder -- the portfolio investor -- is among the most powerful influences in the economic marketplace, and in some sense can be understood as the financial market itself. For both better and worse, fictional shareholders control much of the direction of our economy.

Nor do I mean to suggest that the motivations of fictional shareholders are entirely different from those of the underlying owners. After all, presumably most owners of shares would prefer that their shares be worth more rather than less, ceteris paribus. This Article is concerned with those instances where ceteris is not paribus. The problem is not that fictional shareholders always or even often make the wrong decision, but that they make no decision at all: The fiction of the one-sided shareholder hides the tradeoffs that must be made in life from the view of those who must make them.

This Article, then, is about the separation of the shareholder from the owners of the shares. For the law and the market, the shareholders are simple, unidimensional, time-indifferent, fungible, uninterested and disinterested -- uncommitted to any particular place, project or community. The fictional shareholder is truly cosmopolitan (in the negative sense of the word): deracinated, lacking religion, language, nation or civilization. It is fully mobile, able to leap regional and, as the Mexicans have discovered recently, national boundaries in a single bound. In many ways, the fictional shareholder resembles the "homo economicus" of the economists more than any human being. Putting our economy in the hands of the fictional shareholder has many of the positive and negative effects of unbridled capitalism: the tremendous creativity of constant change and the tremendous unsettlingness of constant change.

B. The Owners of the Shares

The facts about the ownership of publicly traded stock in the United States are fairly well known. About half the publicly held stock is held institutionally -- principally by pension funds, insurance companies, mutual funds, bank trust funds and endowment funds. [FN20] Most of this institutional ownership is, in turn, on behalf of identifiable individual human beings: the beneficiaries of pension funds, the policy holders of mutual insurance companies, the stockholders of mutual funds. Some of it is more difficult to see as held on behalf of specific people: Who is, for example, the ultimate beneficiary of Harvard University's endowment? [FN21]

The indirect ownership of this institutionally owned stock is fairly broad. Virtually all American households own a car and carry automobile insurance. A large percentage of Americans own their own homes and carry homeowners insurance. Many Americans hold life insurance. Virtually all of these people -- clearly the ones who hold their insurance through mutual companies, and arguably the rest as well -- are indirect beneficiaries of insurance company stock holdings. In addition, a significant number of Americans have pension plans or 401(k) plans; all of the former and most of the latter group are also indirect holders of stocks. Finally, about twenty percent of American households hold stock mutual funds. [FN22] Thus, it seems safe to assert that a significant proportion of Americans are indirect stockholders or closely related to such stockholders. [FN23]

Indirect ownership of stock, however, has few of the attributes of common law fee simple property: Berle and Means' claim that shareholders have few rights and little control of the publicly held corporation is even more true with respect to the indirect ownership of stock. Direct stockholders in publicly owned corporations retain three basic rights of ownership: the right to sell the stock (and thus, acting collectively, the company), the right to vote for directors and on certain narrowly defined issues presented to them by the directors, and the right to receive a proportional share of any dividend issued by the directors. Indirect shareholders generally have even fewer ownership rights. Pension beneficiaries, for example, generally lack all three of the basic shareholder rights: As a rule, they are barred by law from making decisions regarding the sale or voting of the stock held by their fund and have no right to (present) distribution of the dividends declared by the board of the indirectly "owned" corporation. [FN24] Similarly, open-ended mutual fund shareholders lack even a legal right to annual elections to the mutual fund's board of directors or to sell the fund to another manager (although they do have the right, absent in corporations, to withdraw their capital), and have essentially no rights with respect to the firms whose shares are held by the fund, other than to a pro rata share of any dividend distribution the firm may make. [FN25]

Direct individual ownership of stock in the United States is quite concentrated, in the sense that a very small number of individuals holds an extremely high percentage of the individually owned stock, [FN26] but quite dispersed, in the sense that close to half of Americans own some stock directly. Since the typical direct stock owner has an extremely small stock holding, it appears that most stock owners (direct and indirect) hold only tiny amounts of stock, and even tinier amounts of any particular corporation's stock, whether measured absolutely or in comparison to the stock owner's total wealth and earning capacity.

People who own shares, directly or indirectly, then, look like most of us: Indeed, they are most of us. Their interest in the corporations whose stock they own is likely to be quite limited. It is limited economically, in that their interest in any one corporation, or even in the stock market as a whole, is likely to be a limited part of their total wealth (for most Americans, even in the upper half of the wealth distribution, earned income -- wages and salary -- is vastly more important than all financial assets combined, let alone any particular indirectly held stock investment). It is limited, too, in that they are unlikely to know much about the investment -- in the case of institutional stock holdings, now the largest part of the market, the ultimate human beneficiaries of those stock holdings are unlikely even to know which corporations' stock they indirectly "own." The interest of owners of stock in the underlying corporations is limited in other senses as well: People have interests other than in the maximization of their wealth, many of which are likely to be more salient at any given time than any particular stock holding.

C. Fictional Shareholders

The shareholder encountered in the law and practice of business organizations is quite different from this picture of the actual "owners."

1. The Puzzle of the Uniform Shareholder

Here is the puzzle. If shares are held directly or indirectly by half the American population, and corporations act as agents of their shareholders, then we should expect two things: First, that corporate America would reflect the full diversity of human America, and second, that shareholder elections and other methods of determining what exactly it is that the shareholders are directing their agents to do would reflect in some fashion the wide range of human American politics.

We should see publicly held corporations that are as deeply dedicated to particular projects, or products, or places, or ideologies, or religions, or ways of working and lifestyles as many Americans are. We should see Berle and Means corporations that decide that profits are not the most important thing -- just as many Americans put other values ahead of wealth maximization, some corporations should reduce their income in order to better serve needs of children, family, art, leisure time, status or religion. We should see Democratic corporations, Republican corporations, stamp collector corporations and the like.

We do not. Instead, public corporations have a rather narrow range of styles and interests. Few of them even claim to have Time Inc.'s dedication to a particular corporate culture, and even those that do seem able to throw it off -- generally in the direction of corporate normality -- as easily as IBM abandoned lifetime employment or the American Can Company abandoned the can business.

There are some importantly idiosyncratic corporations out there. But by and large the idiosyncratic ones are the ones that do not fit the Berle and Means model. Either they are privately held, and thus able to vary all the basic structures of corporate law, [FN27] or they are dominated by founders who exploit the looseness of corporate law to treat a public company as if it were still private. We do not see public companies dropping their blandness to reflect the diversity of America. [FN28]

Similarly, we do not see the hotly contested shareholder elections one might expect if corporations were reflecting a diverse shareholder body. Americans, after all, have radically different views on issues such as the proper place of religion or art in public spaces, the relative importance of working conditions as opposed to, say, quality control, the proper approach to environmental quality, the proper relationship between stability and progress, the importance of growth versus the importance of economic equality, the degree to which public goods, such as education, culture or urban spaces, should be publicly financed and other issues that corporations must confront in their everyday management. So there ought to be battles among the shareholders to determine what the corporation will look like, just as there are such battles in every political unit in our country.

Far from the battles of the Kulturkampf one sees in elections to the House of Representatives or local school boards, however, corporate elections look like something out of the former Soviet Union. If management nominees or positions in a publicly held corporation's shareholders meeting win by less than eighty- five percent, it is front-page news for the Wall Street Journal, generally followed shortly thereafter by a palace revolt among the very directors management nominated. Shareholders, unlike American citizens, appear to have a general will in John Jacques Rousseau's sense: Without debate or discussion, by independent voting, they agree on what is best for the collectivity, consistently arriving at a single answer. Or perhaps the better metaphor is the Leninist one: Like Lenin's proletariat, shareholders follow the leadership of the vanguard that knows, without false consciousness, what their true interests are.

2. The Puzzle of the Missing Agency Rights

Another set of oddities seems to challenge the very notion of shareholders as principals of the corporation: Shareholders have few or none of the rights that agency law grants to principals. This is no secret; why, then, do courts and scholars continue to use the agency metaphor?

The Restatement states that an agency relationship is characterized by two special traits. The agent, who acts on behalf of the principal, is subject to the principal's control. [FN29] And the principal has an unlimited right to terminate the agent at any time. [FN30]

In sharp contrast, Delaware courts never tire of repeating that the board of a Delaware corporation has original, undelegated power to manage the corporation. [FN31] The board may make virtually every decision on its own authority; only a few decisions must be ratified by the shareholders. Even in those relatively unusual circumstances where shareholder approval is required, shareholders generally have no right to initiate action. They can vote only at specified times for specified purposes; subject to a few exceptions, the board controls their agenda. [FN32] Shareholders, to be sure, have the right to present proper proposals at the annual meeting. [FN33] But state law generally bars most proposals ordering the directors to take particular actions: That would be a breach of the directors' fiduciary duty to act in the best interests of all shareholders. [FN34] Federal law has been even more restrictive, denying even to purely advisory proposals access to the proxy machinery necessary to make proposals meaningful if, inter alia, the "proposal deals with a matter relating to the conduct of the ordinary business operations of the registrant"; this rule has been applied -- though not consistently -- to bar shareholders from expressing to management their opinions regarding employee health benefits, compensation policies, workplace management, racial discrimination, hiring and firing practices, labor relations, conditions of employment, EEO compliance, affirmative action, a company policy discriminating against homosexuals and so on. [FN35]

Even the most important decision from a shareholder perspective -- whether to sell the corporation as a whole -- must be made in the first instance by the board. Merger agreements, sales of all assets, dissolution of the corporation and even amendments to its articles of incorporation must all be initiated and approved by the board prior to shareholder action. [FN36] As the board in the famous case Smith v. Van Gorkum [FN37] found to its discomfort, a board may not simply delegate these decisions to the shareholders. [FN38]

Similarly, tender offers in practice are subject to board decisions regardless of shareholder desires. Formally, to be sure, tender offers do not require board action, presumably because they are understood as individual shareholder actions with respect to individual holdings rather than as corporate actions. [FN39] However, with the emergence and judicial approval of the "poison pill" [FN40] and the " 'just say no' defense," [FN41] it is virtually impossible to consummate an uninvited tender offer without board acquiescence. The constituency statutes -- which effectively preclude judicial review of board resistance to hostile tender offers by allowing boards to invoke the interests of virtually any corporate participant to justify a board decision -- surely are mere icing on the poison pill cake. [FN42]

Furthermore, shareholders, unlike ordinary principals, may not issue binding instructions to their "agents." The general rule is that directors would be in breach of the law if they accepted such instructions. [FN43] Like Edmund Burke's statesman, [FN44] but unlike the Populist legislative model of direction, initiative and referendum, [FN45] directors are required to exercise independent judgment, not simply to follow their constituents' -- let alone principals' -- orders. [FN46] Indeed, in the precise opposite of the agency law rule, directors are not necessarily protected in their decisions by turning a matter over to shareholders for decision, and may not defend a shareholder suit on the ground that they were doing no more than they were instructed to do or that they pledged to do prior to their appointment. [FN47] Unlike agents, in short, directors are not viewed as mere emanations of their principals' personalities.

The fundamental agency right of termination is also missing. Agency law holds that an agent may always be terminated, regardless of prior agreements. An agent wrongfully terminated may have an action for damages, but never has a right to continue as the principal's agent. Partnership law follows the agency model, allowing for wrongful dissolution of the partnership at any time by any partner. [FN48] Corporate law, however, is quite different.

First, individual shareholders never have the right to dissolve the relationship between the shareholders as a group and the board: Ordinarily, the shareholders may never manage their own affairs. [FN49] Furthermore, the corporation may be dissolved only by vote of the board followed by vote of the shareholders. [FN50] A shareholder may, of course, transfer its interest to another person, but the shareholder/director relationship continues unchanged.

Second, shareholders have only limited rights -- even acting collectively -- to replace the directors. Unlike agents, directors are appointed for a term and may be terminated during the course of that term only by following the procedures set out in the statute; there is no equivalent to the agency concept of wrongful termination. Thus, a shareholder agreement to elect certain individuals directors is generally binding and enforceable. [FN51]

Furthermore, even in the absence of an agreement, shareholders may remove directors only by majority vote at a meeting duly called for that purpose. [FN52] Calling such a meeting requires fulfilling statutory requirements that often cannot be met on short notice. Moreover, if the corporation has a staggered board, the shareholders may be barred completely from removing directors during their term except on a showing of good cause; some of the older statutes required this showing of cause in all cases. [FN53] A stronger contrast to the agency rule cannot be imagined: While agents serve only at the pleasure of the principal, regardless of agreements to the contrary, directors serve for a fixed term and may not be removed except by prescribed procedures.

3. A Different Kind of Principal

We have seen, then, that shareholders do not have the kinds of disputes one would expect if they were a diverse group of Americans engaged in a struggle to make corporations in their images, and that as a matter of law, shareholders, even taken as a collectivity, lack the control over directors that characterizes an ordinary agency relationship. The facts are no surprise: Every reader of the Wall Street Journal knows that corporate elections are generally won by margins not seen in democratic politics. The law is no surprise, either: Virtually all my citations in Parts II.C.1 and II.C.2 above are taken from a leading casebook used to teach basic corporate law to second- year law students. [FN54]

One might conclude from this that the agency metaphor is simply wrong; that in fact directors are not agents of the shareholders and the shareholders are not the principals, or owners, of the firm. [FN55] Directors, after all, are explicitly authorized by statute in over half the states and by case law in Delaware [FN56] to consider the interests of corporate participants other than the shareholders. Thus, the law of directors defies even the remaining aspect of agency law, that the agent acts on behalf of the principal. [FN57] Taking these "constituency statutes" and the agency metaphor seriously, one might come to view the corporation as a coalition of bargaining groups with the shareholders as one among equals, [FN58] or as a quasi-state that has (presumptively wrongfully) limited the franchise to but one subsection of the governed, [FN59] or as a more amorphous kind of community. [FN60]

The persistence of the notion that the directors are agents for shareholders, in the face of well-known facts and law to the contrary, however, suggests that the metaphor should not be dismissed so easily. The fictional shareholder solves the puzzle and rescues the agency metaphor from otherwise hopeless obscurity.

The key to the puzzle, in my view, is that in a sense directors often do view themselves as acting on behalf of shareholders, and the shareholders do control the corporation, despite the law and appearances to the contrary. But the shareholders on behalf of which the directors act and the shareholders that control the corporation are not the owners of the shares. Rather, they are a kind of personification of the shares themselves, almost imaginary creatures driven by only one goal: to maximize the value of their shares.

I claim, then, that the agency picture of the corporation is right in this sense: The people who make corporate decisions -- directors and managers -- do so and are required to view themselves as doing so in an agency role. Their job, as they see it, is to put aside their own interests and views and act on behalf of someone else's views and interests. [FN61] In that strong sense, they are agents, regardless of whether the law gives that someone else the technical rights of a principal under agency law. But the someone whom the corporate agents represent, in whose behalf they must act, is not a full human being in all its complexity, much less a collection of half the citizens of the United States of America. Rather, it is the fictional simplification we call a shareholder.

4. Hypothetical Politics of Imagined Monads: Hobbes Meets the Fictional Shareholder

The fictional principal solves many of the puzzles of the agency metaphor. First, it explains the startling absence of intra-shareholder conflict and actual agency rules [FN62] in corporate law noted in the prior two sections. Second, it justifies an extraordinary level of deference to the professional managers of the corporation.

Fictional shareholders, unlike real ones, do not have strong conflicts in their attachments or ideologies. They are not Democrats and Republicans, religious and atheist, committed to New York or Iowa, tied to a job or a family or encumbered by the life stages of a real human being. They do not have a multiplicity of plans for a too-short life: They have one, to maximize the value of their shares. As a result, they are all the same (or almost all the same, as we shall see in a moment).

Now, timeless, ageless, familyless, unencumbered imaginary people with unified goals getting together and deciding what to do are a familiar image to students of Western political philosophy. That is a crude, one-sentence description of persons in the state of nature of the liberal political theory tradition of John Locke and Thomas Hobbes.

Liberal theory has long attempted to deal with the problem of contentious politics through the device of imagining such participants in politics. In the fictional shareholder we have, I believe, a clear application of this classic liberal methodology. The fictional shareholder does not need actual messy politics or, for that matter, the actual rights of a principal under agency law, for the same reason that Hobbes, Locke and John Rawls can base their political philosophies on imaginary agreements. In the simplified world of liberal theory -- and fictional shareholders -- expertise of the philosophers and the managers replaces actual politics. To see why this is so will require a short excursus into the arcana of liberal theory and back.

a. Liberal theory and the abolition of politics.

A central issue for liberal theory is the majoritarian difficulty: put most crudely, that might does not make right. The mere fact that a majority has approved a law, without more, does not make a law just, and does not explain why the losers should view the law as legitimate, rather than as simple brute force to be obeyed in necessity and evaded where possible. In the particular American context, one of the oldest examples is the battle for religious freedom: We long ago agreed that mere majority vote does not justify an established church, let alone exiling or burning dissenters. An equally old and obvious example is slavery: The general consensus is that a majority vote to reinstate slavery or the apartheid system of Jim Crow would not be enough to justify it morally. Thus, few, if any, have ever tried to distinguish between the American version of apartheid and its South African variant on the ground that the former, but not the latter, was imposed by a majority on a minority. Such laws are unjust and illegitimate regardless of the procedures used to adopt them. [FN63]

The project of the Western liberal theories since at least Hobbes has been to explain why or when the law is legitimate -- when, that is, it has a claim beyond that of mere force, whether the force of arms or the force of numbers. [FN64] There are a number of approaches to this problem, but I will concentrate on only one -- the one I call "hypothetical politics." I deliberately ignore other solutions that seem less relevant to the problem of the fictional shareholder, and I discuss the hypothetical politics theory at a level of generality that is sure to offend the living authors and serious students of all concerned: My purpose here is not a philosophic examination but a broad picture of a general method that seems to have important implications for our understanding of corporate law.

The hypothetical politics story runs something like this. Real politics offers little or no guidance to the requirements of justice: Force too easily prevails. The majoritarian difficulty suggests that even if we could complete the John Hart Ely/Footnote Four [FN65] project -- that is, create a procedurally perfect majoritarian system -- we would still not have solved the problem of justice. Majority oppression of minorities is still oppression. Instead, justice seems to require some form of fair proportionality: taking turns, or equal distributions, or distributions appropriate to the nature of the thing in question, or something of that order. [FN66] Similarly, legitimacy seems to require some form of consent: If you have agreed (or should have agreed) to this law, you should not be heard to complain that it is coercive. [FN67] In contrast, majoritarian systems allow a unified majority to prevail, over and over again, without regard for fairness or minority consent. [FN68]

Consent, especially consent under fair conditions, is notoriously difficult to obtain. Those who premise justice on obtaining actual agreement tend to conclude that all existing societies are unjust and the possibility of creating a future just society is slim indeed, as do Robert P. Wolff, Robert Nozick and John Jacques Rousseau. Furthermore, their work can easily become a justification for creating agreements by force -- by killing or expelling those who disagree, as in the nationalist and revolutionary reinterpretations of Rousseau.

In contrast, if we could imagine an agreement that all rational people would agree to under fair conditions, some philosophers have argued that there would then be no need to reach an actual agreement. Real people might well refuse to agree -- but their refusal may be disregarded, since it must (by hypothesis) stem either from irrationality or from an unfair bargaining situation. [FN69]

Hobbes thus argued that all people, whatever else they want and whatever their goals in life, wish to stay alive; accordingly, under fair bargaining conditions they would all agree to create a government that will keep them alive. From this foundation, he constructed the Leviathan -- a massive defense of a rather unfree politics based on a hypothetical unanimous agreement. The power of his argument is that if we were persuaded that all rational people would, after reflection, agree to the society he describes, then such a political arrangement would be legitimate regardless of its history or origin: No investigation of real history, no questioning of real people, no actual debates or politics in this world are necessary to show the legitimacy or illegitimacy of the existing government. Hypothetical politics replaces the real form.

The preeminence of hypothetical politics -- the unanimous agreement into which all citizens would enter if, counterfactually, they were all fully rational, fully informed and in a fair starting position -- over actual messy, unfair, irrational ordinary politics is most striking in Hobbes since his hypothetical rational beings agree to a state that, to save their lives, bans anything resembling ordinary politics. But the model, in a less extreme form, underlies the work of philosophers ranging from Locke to Rawls to the scripted "dialogues" of the early Bruce Ackerman. [FN70] Indeed, I believe it is implicit in virtually every nontheological rights-based defense of constitutional law: The Constitution is meant to embody the hypothetical agreement that actual politics has no right to abrogate.

The exciting thing is that once the philosopher has identified a common goal -- life, or maximization of primary goods, or whatever -- the philosopher can then derive the agreement that such individuals would reach if they were in a fair starting point. No actual discussions with actual people are necessary: We can figure out what they would want by applied logic.

The hypothetical agreement reached has two broad elements, as a rule: certain basic rights that the parties would agree should be guaranteed in all circumstances, and a majority rule for the remainder of issues. This reduces or eliminates the majoritarian problem by shrinking the sphere of politics and eliminating most issues of uncompromisable principle: The difficult to resolve and truly controversial issues are removed from the majoritarian process. Freedom of conscience, freedom of movement, in Rawls' version a minimum standard of living -- each of the things about which one might be inclined to fight rather than switch are depoliticized, given over to the hypothetical agreement of the theory rather than the actual politics of a majoritarian democracy. [FN71]

Most remaining politics is not issues of principle but rather matters of administration, namely increasing the size of the pie (best left to experts and the market) and dividing it (best dealt with by general principles of fairness or the very rough equality of the spoils system in a pluralist democracy).

Thus, in the American application of this theory, the Constitution was meant to remove from majoritarianism each of the most divisive issues of the age: Religion and speech, each core areas of politics as understood by the ancients, were simply excluded from the sphere of governmental competence. Slavery, the most important issue of the day, was removed from majoritarianism: A mere majority vote would not have been enough to convince the losers to give up their position. What is left for majoritarianism is public works -- pork, in the modern deprecatory terminology. And for that, the constitutional solution is Publius' version of pluralism: splitting the governmental process into a multitude of different routes, with the goal of preventing anyone from creating a winner-take-all majority. Instead, to get anything done, ever-changing broad coalitions must be formed -- hopefully assuring that everyone gets a chance at the trough in something resembling a fair share. Indeed, the sharp decline in public respect for government in recent years may be due to the elective branches' increasing willingness to tackle ideological issues or issues with clear losers rather than sticking to issues where everyone (or at least all those represented in the bargaining process) can get some, like the interstate highway system. In times when the government commanded more respect, the elective branches tended to avoid principled issues, instead dumping them on our operationalized version of the hypothetical agreement, the courts, or hiding them in the invisibility of administrative agencies or local governmental bodies. [FN72]

b. The hypothetical agreements of corporate law: a unified goal reduces politics to administration.

In our corporate law, this liberal model of a hypothetical politics is taken to its fullest, Hobbesian, extreme. Fictional shareholders all want to maximize the value of their shares. They exist without context or history. Since the value of their shares is nothing more than the discounted value of the future income stream represented by the dividends, they are time indifferent. Since dividend streams are fully fungible, they are as uncommitted as persons in the state of nature. Since fictional shareholders function in a free market, they are individualist and self-interested. And like the persons in the state of nature or behind the veil of ignorance, they are fully equal and able to enter into a fair bargain.

Fictional shareholders, then, meet all the requirements of hypothetical politics. Here, as in the Hobbesian model, something very exciting happens: Once we agree that all the shareholders share this common goal, actual politics becomes an unnecessary distraction. We can calculate what rational and equal shareholders want by mere reason. Discussion is unnecessary; expertise can replace persuasion and voting.

Under the model of hypothetical politics, then, the wills of the people who own shares are no longer particularly important. Since management and the board, a reviewing court or any other expert can determine what the hypothetical shareholder interests in fact are, it is not necessary to ask shareholders what they think. Indeed, in light of the well-known problems of bounded rationality, it probably would be foolish to do so. [FN73]

So the fictional shareholder model explains why corporate law reverses the usual agency law and contract law presumptions. If shareholders all share the same basic goal, namely maximization of the present discounted value of future dividends, then we can construct the agreement they would have reached under more ideal conditions. It follows that freedom and autonomy for the principal and contracting party are completely unnecessary. Paradoxically, the principal can be empowered by being barred from deciding the key issues. Indeed, even if an overwhelming number of shareholders support a given action, the board, a reviewing court or other expert may be better positioned to determine that shareholder interests are different from their expressed views. [FN74]

Following the liberal model, actual contracts and actual agency agreements can be replaced by imaginary ones. All the important issues are removed from the political sphere, determined instead by the hypothetical agreement of the fictional shareholders in a state of nature. Thus, shareholders may not, by majority vote, transform a corporation into an eleemosynary institution. [FN75] Indeed, they may not even decide that some shares of common stock will receive more dividends than other shares in the same class. Stepping away from the fictional shareholder's fixation on mammon, shareholders may no more determine whether the firm is to be pagan or Christian in its outlook than the voters in a liberal democracy may make the state Catholic or Baptist. Indeed, shareholders may not even determine that the purpose of the corporation is to produce some particular product: If, for example, the Coca-Cola Corporation decided to stop producing Coca-Cola (or to change its formula), shareholders would have no vote on this decision. [FN76]

This lack of power, strange on the agency view, is natural on the hypothetical politics view. For fictional shareholders, whatever else the people behind them may want, all want to maximize the value of their shares. And as follows from the basic teachings of Adam Smith regarding the division of labor, rational share-value-maximizers would agree to delegate management of the company to professionals. Maximizing the value of the shares is a job for technical experts; there is no reason to think that the average (or indeed any) shareholder is particularly good at it.

It follows, then, that the separation of ownership and control is not a vaguely illegitimate deprivation of the rightful prerogatives of ownership, but rather a supremely sensible application of the division of labor. Companies need professional managers; the shareholding system allows competent managers to be chosen without regard for whether they also have the wealth to be shareholders.

But see the Hobbesian transition: If the corporation were run by and for real people, it would be a hotbed of political controversy. Real people argue about goals -- so they must be represented by politicians, not technicians. The fictional shareholders, however, all agree on the goal; they have no need or use for politics. If the real people disagree with the fictional representation, the real people may simply be disregarded as not real shareholders. [FN77]

Corporate democracy, in short, is not a Greek forum in which men seek to see and be seen or struggle to define a way of life. The goal is set before "politics" begins; the function of shareholder votes is not to set the goal but simply to police the experts. Indeed, given that shareholders have only one goal, expert managers acting in good faith are more likely to be able to do what shareholders want than shareholders themselves are. In this model, the only purpose to giving the shareholders any actual control at all is to ensure that their experts, the managers, are in fact acting in good faith. And in this model, it is entirely natural that we would only rarely see actual political action by shareholders: They have no disputes and they can generally rely on the market to take care of the occasional corruption problem.

We have come then to a solution to the shareholder puzzle. The fictional shareholder reduces politics to administration. Fictional shareholders do not have the conflicting goals that drive real politics and that are the reason why real principals must have ultimate control over their agents. Rather, like the Hobbesian monads in the state of nature, they know that whatever their goals in the real world, as shareholders they want more stock value rather than less. Given this common interest, they can be imagined to have agreed to let the philosophers (or in this case, the MBAs) decide how best to reach the common goal. Having entered into this original agreement, they can, like Hobbesian subjects, drop out of politics entirely. [FN78]

Note the importance of the particular goal: Maximizing value of shares, unlike, say, the merits of Time Culture or Classic Coke or American Cans, is something on which all shareholders can be deemed to agree. Thus, no vote is necessary on the ultimate goal; anyone who proposes an alternative goal can be assumed to be acting in bad faith. [FN79] But no vote is needed on particular methods of reaching the goal, either -- the decision whether to produce cans or insurance, sweet or less sweet cola, or turgid or less turgid magazine prose, appears as merely an issue of implementation, best left to the experts and the full-timers.

So, if a corporation is run in the interests of fictional shareholders, it need not have actual owners of shares controlling it, contrary to the implications of the agency model if one assumed that the principals were real human beings. Nor need it have the actual political struggles that the diversity of the Americans who own shares would suggest is inevitable. Instead, experts considering how best to serve the single interest of the fictional shareholder can run the company all by themselves. Indeed, most likely, professional managers will run it far better than the real owners of shares would, for unlike the managers, the people behind the shares are neither trained experts nor full-time professionals.

c. Politics as policing corruption rather than setting goals.

Only one function for politics remains within this paradigm; the same one that ultimately led Hobbes' successors to replace his authoritarian king with the thin majoritarianism I have discussed above. That is the problem of corruption. The logic of the division of labor counsels in favor of giving the maximum power and discretion possible to the expert managers who will run the company in the theoretically derived interests of shareholders. But, as we know, power corrupts. Thus, the key remaining function of shareholder democracy is to eliminate corrupt or self-serving managers -- in the modern jargon, to reduce agency costs.

Finally, the fictional shareholder model explains the strikingly primitive understanding of agency problems in the corporate law agency cost literature. Corporate law agency theory concerns itself almost exclusively with corruption costs -- the problem of agents who deliberately refuse to do the job they are hired to do, or who ignore their duties, intentionally putting their own interests ahead of their principals'.

Compare this thin view of the difficulties of the role of the professional agent to, for example, the elaborate discussions of how best to represent another that arise entirely within good faith models of professionalism in other fields: lawyers and doctors struggling to understand how to pursue their clients' interests and goals in a world where those interests and goals may be nonexistent, underdeveloped or incoherent. [FN80] These issues drop out of corporate law because the fictional shareholder -- unlike the human clients of doctors and other professionals -- is seen as having only a single, consistent and clear goal.

Corporate law's limited understanding of agency issues flows directly from the hypothetical politics view: The simplicity of the fictional shareholder eliminates any need to talk about more sophisticated failures of agency and representation. Rather, the directors and managers are seen as mere administrators in an uncomplicated professional role of implementing a policy decision made prior to politics. If they would only keep their promises, corporate law would be dead. [FN81]

d. The solution to the agency puzzle.

In this Part, then, we have seen that the claims of agency theory would be absurd if the principal were the American citizenry, in its full committedness and diversity. On the other hand, if shareholders are viewed as simple beings interested in nothing but share value maximization, the existing intellectual legal structure makes a great deal of sense. Given the clarity of the fictional shareholder's single goal, it is highly appropriate to delegate to professionals enormous discretion in managing the daily affairs of the firm. Under the fictional view, it is reasonable to restrict shareholder votes: Predictably, most shareholder interventions will be irrational, uninformed or in bad faith and doomed to be overwhelmingly rejected by a rational majority -- surely fictional shareholders would agree that such exercises are a waste of time and money. Similarly, most shareholder litigation will be, as is commonly claimed, no more than a form of blackmail, in which a share owner (or an entrepreneurial lawyer purporting to act on behalf of such a person) presents a meritless claim and then seeks to negotiate a settlement based on the expense to the firm of winning in court.

The fictional shareholder/hypothetical politics analysis also clarifies the business judgment rule, which often bars courts from examining whether directors have in fact fulfilled their duties to their supposed principals. Courts need concern themselves only with possible corruption, not problems of representation, shareholder self-governance or the primacy of the principal.

The key question remains, however: Do real owners of shares take the same one-sided view of corporate goals as does the fiction? If not, our replacement of real debate with the thin politics of corporate law must be as troubling as Hobbes' Leviathan.

III. THE IMAGES OF THE FICTIONAL SHAREHOLDER

In this Part, I will discuss some of the ways the fictional shareholder appears in the law and the cases and then describe some of its basic characteristics. The picture of the fictional shareholder, however, is confused by two complicating factors.

First, corporate law and corporate culture today reflect two different notions of the shareholder. In both, the shareholder has no interests other than the shares it owns. Under the traditional view, the shareholder owns nothing but the shares of the company in question. This I will call the "corporate law" fictional shareholder.

Recently, the corporate law view of the shareholder as a permanent investor in a single company has come under sharp challenge from an image of the shareholder as holder of a diversified portfolio; that is, as an investor holding shares of many companies. This second view I will call the "portfolio investor" shareholder. [FN82]

The corporate law and portfolio investor images of the shareholder share certain important characteristics. Each is a version of economic man, driven only by profit maximization, with no connections to time, place or community -- in effect a type of immortal, unencumbered monad. But because they differ in the shares they hold, they differ in how they profit-maximize, with dramatic differences in how a corporation run in their interest should be run. [FN83]

Second, the two images of the shareholder differ in the way they influence the corporation. The first one, the corporate law shareholder, is largely a metaphor -- a powerful organizing theme that influences how human actors in the corporate world, particularly directors, managers and judges, go about doing their jobs. Like the Pope, the corporate law shareholder has no army divisions, but enormous power. [FN84] This shareholder is an abstraction from the full human beings who ultimately own shares; in the process of abstraction, it becomes radically different from any human.

Directors acting in the interests of the corporate law shareholder are under its control in a powerful, but limited, sense. The corporate law shareholder is an image of the law and corporate culture, not a human being. Directors who avoid its strictures may be censured by their consciences, and in extreme cases by the courts. But as a rule, there is no one who forces them (in the "how many divisions does he have" sense of force) to act as the corporate law and its shareholder wish. [FN85] Indeed, there is no one who wishes as the corporate law shareholder wishes. It is a construct, driving our behavior much as Locke's men in the state of nature influence the behavior of humans and the history of states without ever having been flesh and blood, living or breathing. Corporate managers and directors manage the corporation in the interests of the corporate law shareholder because they believe it is right to do so, not because they have a gun pointed to their heads.

The second image, in contrast, has Stalin's divisions -- perhaps more divisions than ideological legitimacy. The portfolio shareholder is an accurate description of the way our large institutional shareholders act. But those shareholders are themselves corporate entities, driven to act not by the decisions of thoughtful human beings acting in their own interests and according to their own beliefs, but rather by the legal and moral compulsion to serve their own fictional corporate law shareholders.

Thus, the two fictional shareholders control the corporation in two quite different ways. Unfortunately, this makes the story rather complicated -- I hope my readers will bear with me as I try to separate the interconnected strands.

A. The Corporate Law Shareholder

The corporate law fictional shareholder is a person with no interests other than its shareholdings in the particular corporation at issue, and no will other than the desire to maximize the value of that shareholding. [FN86] It is, then, no more than a personification of a share of the particular corporation.

This, I take it, is why shareholders are often described as widows and orphans. Widows and orphans, presumably, are stereotyped as purely passive investors who accept the deceased father's choice of stock and hold on to it indefinitely. [FN87] Widows and orphans in Iowa presumably have no connection to the company's factories in Flint, or its headquarters in New York, or its union members nationally, or its consumers abroad, or its charitable contributions in Pittsburgh. Like the fictional shareholders, these stereotyped widows and orphans are entirely dependent on the performance of the stock and have no other relationships with the firm.

Perhaps because the corporate law shareholder is so similar in its interests and will to a reified share, lawyers and academics writing in corporate law tend to blur the distinction between shareholders and shares. For example, advocates of so-called shareholder democracy generally support what would better be called share democracy, with a voting rule of one share (not one shareholder)/one vote. This corporate governance rule suggests that shares, not shareholders, are the bearers of corporate rights or, as it were, the citizens of the corporate polity. In this Article, I use the traditional terminology, referring to shareholders rather than shares, because of the ideological importance of the terminology: It makes share rights seem like human rights, and share interests seem like human interests. [FN88]

The corporate law shareholder as share is deeply embedded in corporate law. An example from a well-known case in a sophisticated court should make clear how powerful the image is.

In Revlon v. MacAndrews & Forbes Holdings, Inc., [FN89] the Delaware Supreme Court took the most extreme pro-shareholder position it has taken in recent years: It held that once a board has decided to put a company up for sale, its sole fiduciary obligation is to the shareholders and its sole duty is to maximize the price they will receive for their shares. The case is striking, in part, because it abandons the usual Delaware rule that directors have a duty to the corporation and its shareholders -- this case, more than almost any other, accepts the position of the agency cost theorists that an obligation to the corporation can only mean an obligation to the shareholders.

But even more extraordinary is that the facts of the case make clear that the Revlon board in fact was acting in the interests of at least some owners of Revlon shares when it took the "anti-shareholder" action the Delaware court condemned. Revlon had been the target of a hostile takeover attempt. Its board ran through a number of devices in combating it. One of its maneuvers was a coercive exchange offer of up to ten million shares of its stock for newly issued notes. The exchange offer was structured so as to guarantee a large tender: thirty-three million shares (about eighty-seven percent of the total) were tendered and about one third of those were accepted pro rata. Thus, it appears that virtually all Revlon shareholders became noteholders as well. [FN90] Subsequent defensive measures led to a sharp drop in the value of the notes. Apparently fearing a lawsuit from disappointed noteholders, Revlon's board then structured a sale of the company that was designed to shore up the value of the notes, at some reduction in the consideration given to shares. [FN91] This the Delaware court said was forbidden: Once the company was to be sold, the "directors' role changed . . . to auctioneers charged with getting the best price for the stockholders at a sale of the company . . . . [C]oncern for the non-stockholder interests is inappropriate when an auction . . . is in progress." [FN92]

The key fact to note is that the history of the transactions makes clear that, by and large, the noteholders and the shareholders were the same people. The Delaware court does not require any investigation into the degree or consequences of this overlap: whether the stock or notes had traded, or whether investors who held both notes and shares were injured -- on average or at any extreme -- on balance by the benefit given to the notes at the expense of the shares. It does no investigation into whether even in the aggregate the notes gained more or less than the shares lost.

In short, despite its language, which talks of the interest of the stock holders, the court is utterly uninterested in the implications for the financial well-being of the owners of the shares. It does not take note of the fact, recited in its own opinion, that the owners of the shares and the noteholders are, to a large and undetermined extent, the same people and that consequently a benefit to noteholders is a benefit to the holders of shares (if not necessarily a net benefit). Similarly, it does not consider the possibility that the shareholders also might be employees with a far greater interest in assuring that Revlon chose a course with institutional continuity (even if in the name of a new legal entity), less likelihood of downsizing or a greater ability to resist demands to dedicate a large part of revenues to financial interests rather than employees. Rather, its focus is purely on the price of the shares themselves: "Stockholders," as the court uses the term, refers not to the people who own the shares, but to an imaginary creature with no interest but the shares themselves, no plans or desires but maximizing the value of those shares.

The Revlon case is unusual in the explicitness with which it rejects any consideration of the interests of the real, as opposed to the fictional, shareholders, but it is not unusual in its narrow focus. In a world that has long since adopted the principle of diversification, it will often be the case that bondholders and shareholders are the same people. [FN93] Similarly, it will commonly be the case that shareholders are also neighbors, employees or customers -- yet no case that I am aware of ever suggests that shareholder interests might include those other interests simply because the people are the same.

The cases and articles assume that it is unproblematically beneficial for "shareholders" if corporate profits or the share price goes up, even though it is easy enough to imagine situations where that is not true. If the real owners of the shares, rather than the fiction, were at issue, corporate directors -- and courts adjudicating breach of duty cases -- could not assess competitive activities without first examining the actual interests of the actual people behind the shares at some determinate point in time to see if they were in fact benefitted or injured by the action. This is never done; rather, inquiry is invariably restricted to the fictional shareholder.

Think, for example, of a diversified investor that owns shares of several of the large airline companies. Aggressive competition by one company ("Aggressive Air Inc.") might well increase that company's market share and its profit while leading to great injury to other companies in the industry. Clearly, this is beneficial for the fictional unified shareholder of Aggressive Air: Providing that the market concludes that Aggressive Air's long-term profit potential is improved by its behavior, Aggressive's stock price should increase and its shareholders -- thought of in the corporate law fictional way -- will be better off. But for an investor holding shares of both Aggressive Air and its corporate victims, the issue should be one of aggregate welfare: Did the competition help or hurt industry-wide (or at least portfolio-wide) profit potential? If Aggressive is doing well entirely at the expense of its competitors, this more diversified investor would not benefit from its activities.

Similarly, the famous case of Dodge v. Ford [FN94] established that it is a potential breach of duty to shareholders, for example, to pay employees more than the minimum the market demands, or to employ more of them than it might, or to charge consumers less than what the market might bear, or even to reinvest proceeds in expanding the existing operation rather than pay out a high dividend. In each of these instances there is an arguable injury to the fictional shareholders. [FN95]

But while many cases and authors have sought to loosen the Dodge v. Ford rule or to allow corporations to show some degree of social responsibility, [FN96] I know of no instance where a court has considered the possibility that it might be a breach of duty to shareholders to underpay employees, overcharge customers or pay out profits as dividends instead of reinvesting them. Nor is there any law suggesting that a company decision to abandon a traditional site of a manufacturing plant or headquarters might be against the interests of shareholders: Proponents of restrictions on such changes always invoke other interests.

In each of these cases, the interest of the fictional shareholder is clear and explains the law. But the interests of actual people behind the shares are far from clear. Most indirect holders of AT&T stock, for example, no doubt have a bigger financial stake in the company in their role as customer than in their role as shareholder. Similarly, many owners of the stock of the old Reynolds Tobacco were far more invested in the town and the surrounding tobacco-based economy than in the shares they owned. [FN97] Quite commonly, employees of a firm own stock in it as well, yet find their identity as employees far more important than their identity as shareholders.

These real people who own shares might be very interested in the corporation taking actions that would injure the value of their shares, if there were a compensating benefit in another relationship to the company. Dodge v. Ford would look quite different if the court had considered the case from the perspective of a shareholder who held three shares of Ford stock but was looking for employment in one of its expanding divisions and planning to buy a Model T each year for the foreseeable future.

The fictionality of the shareholder also explains other puzzling aspects of corporate law, including the odd concept of long-term shareholder value in takeovers -- where by the time any important takeover is consummated, the real owners of the shares are virtually all either professional short-term investors (arbitragers), insiders or hopeful insiders with other more salient identities than "shareholder," or simply dead. The only player that looks even remotely like the imaginary long-term shareholder is, paradoxically, entrenched management. Undiversified and well fortified with share options and golden parachutes, top managers increasingly have an economic interest closely paralleling that of the fictional shareholder. [FN98] (It follows that the view of corporate law as centrally about the conflict between shareholders and management may be near obsolescence -- management has become something approaching a fictional shareholder, and the next story will be about something else.)

The fictional shareholder also explains a key difference between securities class actions and, for example, a normal class action over a tort. In other areas of the law, after a class has been certified and class-wide claims have been adjudicated, each member of the class must establish that he or she was in fact injured. Thus, a member of a class of asbestos exposure victims who died in an automobile accident long before the appearance of any asbestosis symptoms would be denied recovery. In sharp contrast, in securities actions, the individual claimant need only establish that he, she or it owned, purchased or sold the security at the relevant time. No court, to my knowledge, has ever suggested that a class member's damages relating to one security might be diminished by a showing that another security held by the same class member (or some other relationship with the firm) increased in value. Yet in many competitive situations it should be the case that an injury to one company (and its securities) will be a benefit to its competitors (and their securities), so that a diversified investor would have little or no net injury from the original violation. [FN99]

We have seen, then, that if the focus of interest were the actual people who own the shares, rather than the fictional unified shareholder with no other interests, boards and courts would have to make detailed factual inquiries before they proceeded to analyze the interests of shareholders. From the absence of such inquiries, if nothing else, we can see the dominance of the fictional shareholder in our corporate law. [FN100]

B. The Portfolio Investor

The first fictional shareholder is a very odd beast. It is a profit maximizer, in the manner of economic man, with no obligations or responsibilities to constrain its attempts to maximize the value of its stockholdings. On the other hand, it defies the first rule of rational investing as understood by modern financial theory: It is undiversified and a permanent investor in the company. [FN101]

Since the financial market is dominated by investors that are diversified, this oddity has resulted in the second major image of the fictional shareholder: not a share but a portfolio.

Roughly half the shares of publicly traded American corporations are held by institutions. Far more than half the stock market activity is institutional trading. Institutional shareholders, therefore, have become an enormously important part of the financial markets and the corporate structure. An entire academic enterprise has grown up to describe (and prescribe) the behavior of the institutional investors, and both law and ordinary discourse have sometimes adjusted to reflect the dramatic differences between the institutional investors and the traditional corporate law fictional shareholder.

Institutional investors come in several varieties. Some, such as bank trust funds or some insurance companies, are straightforward business corporations, or corporation-like entities such as mutual funds, mutual insurance companies or mutual savings banks. Others are ERISA-regulated pension funds, often attached either to business corporations or unions, and sometimes managed by one of the other listed entities. Still others are endowment funds for not- for-profit organizations. These differences make generalization risky, but I believe that certain broad characteristics are shared by virtually all the institutional investors and are key to understanding the image of the portfolio investor in the law.

1. The Fictional Shareholders of the Institutional Investors

The key point about institutional investors is that they invest on behalf of the interests of their own fictional corporate law shareholders; that is, they act as if they were agents for an undiversified investor with no interests beyond the portfolio held by the institution, without offering any mechanism by which their own investors could indicate any contrary interests or desires.

Institutions that hold shares vote those shares as an institution; they do not pass the vote through to their own shareholders. [FN102] In managing their shareholdings, the institutional investors act as if they were controlled by corporate law fictional shareholders. That is, they attempt to maximize the return to an imaginary being who is solely invested in the institution in question. Why they do so varies from organizational type to organizational type.

a. Mutual funds and insurance companies.

Mutual funds, for example, are even less controlled by shareholder voice than are standard Berle and Means corporations: Shareholder votes are meaningless as a control mechanism. On the other hand, market pressures force mutual funds to act as if they were completely controlled by the fictional shareholder. Mutual funds are subject to constant rating and comparison. Any fund that increases the value of its shares -- that is, acts in the interests of its fictional shareholders -- is effectively rewarded by a rapid and vast flow of money, as its shares are purchased by investors seeking high returns. Conversely, funds that do not perform well shrink, though less rapidly. For these entities, far more than for large publicly held corporations, the corporate law shareholder is a real and powerful presence.

Note, however, that the shareholder remains fictional. Investors have and continue to have many values and conflicting goals, even as they move their money around in search of the highest possible return. Those other goals and values, however, drop out of the message as it is transmitted to the mutual fund. From the perspective of the mutual fund, its shareholders are not full, conflicted and situated human beings but rather merely capital flows, attracted to the latest quarterly results as bugs are photo-tropically attracted to a light. Whatever complexity the owners of the fund's shares may have as human beings, as shareholders they are simple, one-sided and monolithic.

Other portfolio investors in competitive industries face more or less the same market pressures. Insurance companies must maximize their investment returns to offer competitive insurance rates. Those that do survive; the others fail. Even mutually owned insurance companies, without clear shareholders and with virtually complete management autonomy (in the Berle and Means sense) will be compelled, if the market is competitive enough, to act as if they were maximizing returns for a fictional shareholder.

b. Pension funds.

Pension funds generally are local monopolists and need not, at least in the short run, concern themselves with the pressures of competitive markets. Usually, the beneficiaries are more or less captive, since they can change pension funds only by changing jobs, and it is rare that pension decisions drive job decisions. Thus, the fund and its managers need not worry about flight of customers in the way that insurance companies and mutual funds do. However, ERISA compels pension funds, as a matter of law, to act for the sole benefit of a person much like the fictional shareholder: A plan fiduciary must act "solely in the interest of the participants and beneficiaries and for the exclusive purpose of providing benefits to [them] and defraying [administrative expenses]." [FN103] Thus, ERISA plans are required by law to reduce the plan participants to a fictional creature who is interested only in benefits from the plan, regardless of any other considerations. Since fiduciaries are required to act in the interest of this variant on the fictional shareholder, there is little left for them to do but assess how to maximize those benefits.

Specifically, ERISA plans have been barred from voting their shares in ways that arguably might benefit the individuals who are the plan's ultimate beneficiaries, but would be bad for fictional undiversified investors in the plan. For example, investing plan assets to rescue an employer is generally barred. This rule makes sense from a fictional perspective: Such an investment almost certainly would not be in the interest of an imaginary person whose sole connection to the company is a desire to maximize returns from the pension, since it reduces diversification in a particularly unsound way, tying pensioners' fortunes even closer to the employer's. [FN104] In contrast, from the perspective of the employees who make up the bulk of the fund holders, this result is by no means clear. Thus, employees might find it in their interest to increase the odds of preserving or improving their current salaries by, for example, using shareholder voting rights to cause the employer's business to be run more in the interests of employees and less in the interests of fictional shareholders. Even if this benefit came at substantial expense to their future pensions, they might well conclude that obtaining it is worth giving up a good deal of value in a pension many years down the road. [FN105]

In contrast to the problem of how best to maximize the welfare of a fictionalized fund beneficiary with only one interest, the tradeoff for full people who are both employees and pension beneficiaries (and perhaps also long- term residents of a local economy with few alternative employment opportunities) cannot be resolved a priori. Instead, the proper resolution depends on the risk and consumption preferences, alternative job opportunities, age and family circumstances, fears and hopes, political and religious attitudes of each individual employee/pension fund holder/citizen. Pensioners who have already retired and moved to Florida or Arizona and have no remaining connection with the employer's activities, for example, will have dramatically different views regarding proper fund investment from a young highly mobile employee or from a long-time resident of the company-dominated town with extensive diversified savings. In contrast to the Hobbesian expert calculation by a disinterested fiduciary contemplated in the fictional model, this type of conflict between different and important goals can only be resolved by genuine political debate and struggle.

ERISA, by mandating that the pension fiduciaries consider only the interests of the fictional beneficiaries, cuts off this potential political debate. [FN106] In effect, it requires the trustees to assume that all the present and future pension beneficiaries have no connections or commitments to the investments made by the fund.

c. Not-for-profit endowments.

Private endowment funds are the most difficult of the institutional investors to understand as driven by a fictional investor. Most of the time, no doubt, a university endowment fund will act as if the university were a fictional shareholder with only one interest, namely maximizing the risk-adjusted present value of the future returns on

its endowment. But great universities do not hesitate to assert other interests as well: Yale, for example, used endowment funds for urban renewal projects in its immediate vicinity that probably could not have been justified on grounds of purely economic returns to a fictional shareholder/beneficiary of the endowment. Perhaps its more controversial use of endowment principal to cover current operating expenses for a period of time in the last decade can also be understood as the university acting as a real owner rather than a fictional shareholder of the endowment fund. Nonetheless, it seems safe to assume that these interventions remain exceptional. Most of the time, endowment funds are managed by professional managers in precisely the same way that other institutional portfolios are managed: that is, in the interest of a fictional shareholder with no interests or goals other than the performance of this investment. [FN107]

2. Serving the Fictional Shareholder, Portfolio Style

Paradoxically, the institutional investors serve their own fictional shareholders by acting in a way significantly different from the way that the corporate law shareholder is presumed to act. While the corporate law shareholder is assumed to be a permanent investor who is completely dependent on the fate of this particular investment, the institutional investors diversify. In order to reduce the risk of any particular investment failing to perform as hoped, they spread their investments among many varying companies and many varying types of investment: not merely stocks, but also bonds, other forms of debt instruments, real estate and various derivative instruments.

The institutional investor, thus, is a portfolio investor. It is concerned not with maximizing the value of a particular security, as the fictional shareholder is, but rather with maximizing the value of an entire portfolio. In other respects, it is much like the fictional shareholder. Because of its obligation -- whether legal, market or moral -- to act in the interests of its own fictional shareholders, the portfolio investor is as one- sided as the other fiction. It, too, will neither balance economics against other values nor consider economic values that do not appear in its portfolio. And, again like the corporate law shareholder, one portfolio shareholder is much like another.

In some ways, the portfolio investor is even more inhuman than the corporate law fiction. Portfolio investors have a very different attitude towards risk, time, change and place than do most human beings. If the fictional shareholder was a stripped-down human, deprived of many of the characteristics that make us who we are, the portfolio investor is a souped-up version, more like a god than a human. In significant ways, its interests are premised on its very much not- human character as an infinite and immortal being, living at the average, indifferent to local variation or individuality. Each of these claims requires some argument, and I will address them sequentially in the following sections.

3. Unsystematic Risk Indifference

Corporate finance has centered on the portfolio investor's approach to reduction of risk. As the now familiar argument goes, imagine that a corporation has a possible investment with a small chance of an extremely high return and a large chance of no significant return at all -- for example, drilling an oil well, developing a new drug or investing in a biotechnology startup. In each case, the attempt will probably fail, but if it succeeds, the payoff will be huge.

To make the example more determinate, assume the investment has a 20% chance of yielding a return of $1,000, and an 80% chance of yielding a return of $2. To simplify further, assume the payoff is immediate (so as to eliminate any problem of the time value of money). A risk-neutral investor would value this investment by simply taking the probability-weighted average of the possible results: 20% x 1,000 + 80% x 2 = $201.60. That is, if the investment costs less than $201.60, it is worthwhile.

However, for most investors, the average is not a very accurate measure of the value of the investment taken in isolation. The result will be either $2 or $1,000, not $201.60, and the difference can be critical if, for example, one has an obligation to pay $100 in the near future. When I discuss the difficulties of relying on averages with my corporate finance classes, I often ask the students to consider packing for a trip to Utah's national parks. Spring temperatures in the desert could reach 100 degrees at midday, and then drop to freezing at night. The weighted average temperature might be a balmy 72 degrees -- but dressing for that average would be a mistake. Human beings normally need to take into account the dispersion as well as the average.

More technically, few investors are risk-neutral; most investors would refuse to make this investment unless the weighted average payoff were substantially higher than the cost of the investment. Many gamblers, on the other hand, would be willing to pay far more than the expected value, particularly when the probability of payoff is extremely low and the size of the payoff extremely high: That is the psychological fact that makes lotteries, casinos and retail trade in the commodities future markets possible. [FN108] Thus, the value of this investment to any given investor will depend on the investor's particular attitude toward risk, need for predictable cash flows and the like.

When the investor is diversified, however, a miraculous change takes place: The value of the investment becomes fixed, altogether independent of risk attitudes, and the same for every investor. Imagine an investor who is able simultaneously to invest in 10,000 instances of the above investment, perhaps as a 1/10,000th investor in each. For each oil well, or drug development, the odds remain the same: a high probability of a low payoff ($2), a low probability of a high payoff ($1,000) and no probability whatsoever of the average payoff ($201.60). But if the different instances are independent of each other, that is, if the success or failure of one has no influence on the others, [FN109] the investor faces very different odds. It is overwhelmingly likely that this diversified investor will, in fact, receive the weighted average return or something quite similar to it. The diversified investor can go to Bryce Canyon dressed only for 72 degrees.

Diversification, then, under the right circumstances, allows the investor to live at the average without regard to dispersion. In the corporate context, the consequence is that (taxes, information, agency and other transaction costs held equal) the diversified investor will be indifferent between (a) an investment in a single large oil company that drills many oil wells in many different places, or (b) multiple small investments in many separate small corporations, each of which drills a single oil well. While the large company is likely to be a fairly safe, stable, 72-degree investment, and the small companies are likely to be highly risky freeze or boil investments, the aggregate is the same: In each case the diversified investor lives at the average, receiving the 72-degree weighted average payoff rather than the extremes.

This means that corporate expenditures to achieve diversification or protection from unsystematic risks at the corporate level add no value for the diversified shareholder. Corporate level diversification may well be good for the corporation, considered as an entity -- the large diversified oil well drilling company is likely to survive for many years, while most of the small undiversified wildcatters will be out of business when their first drilling attempt fails. Some of the corporation's employees may also prefer it: Risk- averse employees may prefer to work for a stable employer who will move them (administratively) from one well to another, rather than to rely on a market to move them from one corporation to another.

But unless the diversified corporation has a cost advantage over the aggregate of the small ones -- that is, unless administrative allocation of resources is cheaper than market allocation -- diversified portfolio shareholders will be indifferent to corporate-level risk reduction. And if it costs money for the corporation to reduce its risk -- if, for example, a market is able to move resources from one project to the next more cheaply than administrative action -- the portfolio shareholder's interest is better served by leaving (corporate-level) risk high.

Current political ideology often assumes that the market will have a cost advantage over administration. However, this is a difficult empirical issue which must be determined in each instance by examination of the facts, not a priori reasoning. For example, if oil well drilling (or whatever) is most efficiently done by a team of workers who have extensive experience working with each other, the diversified corporation may have a dramatic cost advantage over the market of undiversified ones. On the other hand, if administration adds extensive overhead, the market may have the cost advantage.

In many areas of production, neither the market nor administration appears to have a decisive social cost advantage. History, or contingent factors of organization, or ability to externalize costs of doing business onto other sectors of society, makes the difference. For example, in many sectors of the economy, when a corporation goes bankrupt, its employees are likely to lose their pensions or medical insurance (which, for those with pre-existing conditions, may be irreplaceable). Thus, in these sectors, diversified employers can offer a substantially more attractive package to employees, which may lead to a more motivated workforce and a significant cost advantage over less diversified employers. Alternatively, firms that learn how to use unmotivated workforces (or to motivate with sticks instead of carrots) and firms in industries with a highly mobile and very young workforce, such as the early days of Silicon Valley computer programming, will have a substantial cost advantage due to their externalization of significant costs of doing business: Unlike the diversified firm, they do not have to absorb much of the cost of pensions or adequate medical insurance. [FN110] In yet another variation, industries, such as trucking, with transferable benefits packages, often offered through centralized unions, may leave small firms able to offer big firm employment packages without adding administrative costs.

Similarly, small undiversified companies may be more profitable for portfolio shareholders in industries where this allows externalization of tort or pollution control costs, [FN111] while large diversified ones may be more profitable where externalization is by lobbying rather than threats of bankruptcy or avoidance of regulation.

In some industries, both market and administrative diversification coexist for extended periods of time. Thus, Lloyds of London, an insurance market with diversification at the investor level, competes successfully with conventional insurance companies that diversify at the corporate level, without either organizational form dominating the other. [FN112]

For our purposes, however, the key point is that portfolio shareholders care about internalized cost, not social costs or unsystematic risk. Because they can live at the average, they -- regardless of their personal or institutional risk preferences -- will be benefited by corporate attempts to reduce fluctuating income only to the extent that the corporation can diversify more cheaply than its shareholders can. Even the proverbial risk-averse widow and orphan are better off investing in multiple risky single oil drillers rather than a safe diversified oil driller, if the former have an internalized cost advantage and externalized costs (and noneconomic values) are ignored.

For an undiversified fictional shareholder, of course, the reverse would often be true: The greater predictability of the large company would be worth a great deal, and might well more than offset any cost advantage associated with riskier organizational forms. For the undiversified shareholder, the $2 and $1,000 figures are the salient ones, not the $201.60. [FN113]

Critically, the portfolio shareholder's interest is different not only from the corporate law shareholder's fictional interest but also from the social interest and the interests of the human beings who stand behind the portfolio shareholder's fictional shareholders. Often the cost advantage of one form of organization may be a simple result of externalization of costs to outsiders -- including the people behind the portfolio investor.

This might be the case, for instance, if a market imposes much of the cost of moving resources from an unsuccessful oil well to the next attempt on workers (by leaving them unemployed between jobs and making them pay the costs of searching for a new team) or on taxpayers (who finance the bankruptcy courts used for reallocation of resources, and pay for the unemployment or welfare benefits used to cover the costs of reallocating workers but do not pick up the salaries of headquarters officials who perform these functions in the diversified firm). The portfolio shareholder will not include these social costs in calculating the relative return of market and administrative organization: Those costs will not be reflected in the returns of the portfolio of wildcatters. This gives an unfortunate boost to the wildcatters relative to the diversified company, which must internalize much of the costs the wildcatters avoid. Unless the administrative mechanism has such a large cost advantage that it exceeds even the invisible costs of the market mechanism, the portfolio shareholder seeking to further the interests of its fictional shareholders will be forced to choose the wildcatters.

The people behind the fictional shareholders, however, are roughly representative of the general public. They, thus, bear the externalized costs ignored by the market calculation and the portfolio shareholder. In addition, some of those citizens are likely to have views about the morality or political expediency of the more stable administrative structure or the more dynamic market one, and those views may or may not correlate with their economic interests. [FN114] Both these economic calculations and the political ones drop out of the considerations of portfolio investors and the firms they invest in: The structure of the fictional shareholder's relationship to the portfolio investor guarantees that the portfolio will consistently support the market's externalization of costs onto workers or the general public unless the internalized costs in employee morale are stupendous.

In short, then, the portfolio shareholder has radically different notions of risk from both ordinary people and corporate law shareholders. Like the corporate law shareholder, the portfolio shareholder's fictional unity suppresses many of the difficult political and economic decisions that affect the real people behind them.

4. Wealth Transfer Indifference

The second main area where portfolio shareholders have fundamentally different interests from the corporate entity or undiversified shareholder is that they are indifferent regarding transfers of wealth from one security to another. For example, a corporate action that expropriates bondholders in favor of shareholders -- or that takes market share from a competitor without increasing the total profits derived by the two companies -- is of no value to the portfolio investor. The portfolio investor stands on both sides of such transactions and thus should view them as, at best, a waste of energy that could be put into something useful, such as increasing the total pie. [FN115]

In most instances, however, portfolio investors will view such transactions, including ones that are at the core of a competitive market system, as downright harmful. First of all, any transaction that takes from one investment and gives to another costs money -- money that goes to unsecuritized investment bankers [FN116] and lawyers rather than investors, or money that is simply lost when management attention shifts to financial gamesmanship rather than more productive enterprise. [FN117] Thus, unless there is some synergy or increase in efficiency involved, the portfolio investor loses out, because the benefit to the gaining security will be less than the loss to the losing one.

Second, the possibility that corporations will opportunistically raid one security for the benefit of another should lead to a "market for lemons" problem, [FN118] to the ultimate detriment of all market participants. If bond investors are concerned about the possibility that bond issuers will increase their risk by such transactions, they will attempt to increase the bond return to reflect that risk. But since the risk is entirely in control of the issuer, rational bond buyers should assume the worst -- thus forcing all issuers to pay higher interest rates regardless of whether they actually do plan to engage in such transactions. In such a market, corporations that do not engage in raids are at a competitive disadvantage (since they are paying for a right to raid but not using it); presumably, then, both raiding and the premium charged for the risk of a raid will increase over time in a downward spiral. [FN119] If such raids could be eliminated, everyone would benefit -- issuers from lower interest rates and investors (and the economy as a whole) from refocusing of managerial energy and other resources towards more productive activity.

Similarly, competition for market share often hurts portfolio investors who hold both competitors: When competition drives down profits, as in the airline or the personal computer industries, the gains to the winning stock may be less than the losses to the losing one. [FN120]

More generally, often when the corporation spends money (or executive attention) in order to reduce its risk from volatile markets, or to shift risks or resources from one security to another -- whether in this corporation or outside it -- or, in many instances, to take market share from (securitized) competitors, it will not be acting in the interests of the fictional shareholder of the portfolio investor, since that shareholder may well be a cheaper risk reducer than the corporation or may be on both sides of the transaction.

The portfolio shareholder, then, is largely indifferent or hostile to changes in the division of the economic pie between different security holders. Gains that come at the expense of corporate competitors, suppliers or customers or other types of publicly traded securities in the company are of no value to a portfolio holder who holds both the loser and the winner. To this extent, the portfolio investor is analogous to the utilitarian who, in Rawls' description, does not take the distinction between persons seriously. [FN121] All that matters is the aggregate change in wealth of securities holders, not the distribution among them.

Critically, however, portfolio shareholders are far from indifferent to divisions of the pie between securities holders, on the one hand, and entities in which they do not hold an interest, on the other, such as employees, retirees, noncorporate consumers or the public health. For portfolio shareholders -- who are thought of as pure securities holders with no other interests -- gains at the expense of employees, retirees or (unsecuritized) customers are valuable even if they result in substantial leakage or shrinkage of the total social pie. It is the wealth of securities holders alone that interests the portfolio investor.

Additionally, even competition in its purest form may be contrary to the interests of the portfolio shareholder. As one director put it:

For example, if a firm takes on a new research project, then the pension fund managers will sell your shares, and your share price will decline even though the project may be an essential move to put you in a stronger competitive position for the future, and may also be in the shareholders' best interests. [FN122]

The director's puzzlement at shareholders punishing the company for taking actions that are in the shareholders' interest stems from a confusion of the two different images of the shareholder. The research project necessary to preserve the corporation's competitive position very well may not be in the portfolio investor's interest. First, as noted above, competitive advantages are often zero sum games for diversified shareholders, when one company's gain is another's loss. Again, the classic example is airline price wars.

Second, even research that allows the company to provide a better product (without increasing the size of the market) may be contrary to portfolio investor interests. After the expenditure and the innovation (and its match by competitors), total profits in the industry may be no higher despite a higher capital investment. In these cases, the diversified portfolio shareholder is rational, from a self-interested point of view, to oppose the "research project." Note that the portfolio shareholder rationally may oppose the research even if the director is right that the investment is necessary to preserve the company's competitive position: An investor who also owns shares of the competitor may not care if this company withers and dies.

In short, while the company may benefit, fictional corporate law undiversified shareholders may benefit and consumers will certainly benefit from this type of competition, the fictional shareholder of the portfolio shareholder will not. What it gains on the one hand it loses on the other. So, the director's problem is in part the result of a shift in the shareholder about which he is thinking: The pension fund managers -- representing here the diversified portfolio shareholder -- have a fundamentally different interest from the "widow in Iowa" undiversified corporate law shareholder in whose best interest the project is.

The other source of the director's puzzle has to do with the will/interest distinction. Many research projects may be in the long-term interests of even diversified portfolio shareholders: If an innovation increases the size of the market or allows a company to exploit a market imperfection to charge higher prices, diversified portfolio shareholders -- the pension funds -- benefit from the price rise in this stock without an offsetting drop in some other security. However, agency and information cost problems may result in the expressed will of diversified shareholders diverging from their long-term interests. They may drive down the stock price even when the research project is in their long-term interest when, for example, institutions rely on quarterly results of their investment managers as a cheap proxy for more accurate assessment measures and investment managers are unable to distinguish, at a reasonable cost, research projects that will benefit diversified shareholders from ones that will not. [FN123]

5. Immortality

Portfolio investors are time indifferent. They are simultaneously the shortest of short-term investors and the longest of long-term investors, with oddly paradoxical results from the perspective of the corporations in which they invest. [FN124] A diversified investor is time indifferent because the market allows the sale of any liquid investment at any time for the present discounted value of its future income stream. Thus, any long-term investment can instantly be converted into a short-term one, and vice versa. Stocks, which from the perspective of the issuing corporation are permanent investments (the corporation is under no obligation ever to return the shareholder's investment) are, by this strange alchemy, the quintessential short-term investment from the portfolio investor's perspective.

A portfolio investor is in a deep sense always a short-term investor. This shareholder has no interest in any stock or any company except to the extent that the security represents an interest in a future income stream. At all times, the portfolio investor seeks to hold the portfolio with the best risk/reward ratio -- it has no commitment to any product or institution behind that ratio. Thus, if the ratio changes, as it does with each additional bit of information about the corporation's prospects or each fluctuation in security price, the investor is always prepared to switch to another security offering a better ratio. Acting rationally, this shareholder has no preference for existing holdings, no commitment and no friction: Like the sinner in the Weavers' song, it is always prepared to slide down into the Promised Land. [FN125]

In another sense, however, the portfolio investor is long-term. Just as stocks are permanent investments in the corporation regardless of the constant changes in the identity of the shareholders, so too many portfolio investors are permanently invested, regardless of shifts in the identity of the stocks held. Retirement funds, mutual funds, insurance companies and the like, unlike widows and orphans, do not grow up, go to college, buy houses, move to retirement homes or die: They exist in a world of statistical probability in which each year is the same as the last. They are time indifferent in the strong sense that they have no life cycle with its accompanying seasons to spend and seasons to save. [FN126]

6. The Inhumanity of the Portfolio

Portfolio investors, then, are deeply inhuman. Humans are never diversified and never time indifferent: A job here is quite different from an (otherwise similar) job in five years or 2000 miles away. Humans live in particular communities, enmeshed in particular relationships, connected to particular landscapes, people and products. They need money now, or later, but are rarely indifferent between the two.

For humans, fungibility is the rare exception, not the norm. For humans it always matters whether the local company (or your employer, or your customer) wins or loses the fight for competitive advantage; the aggregate changes most important to the portfolio investor are likely to be far less salient and perhaps even imperceptible.

Perhaps most significantly, portfolio investors view humans as pure objects of exploitation. For the portfolio investor a transfer from a nonsecuritized entity (such as most humans) to a securitized entity is a pure gain, even if total wealth diminishes as a result; for humans, at least in the aggregate, that is never true. Indeed, since the majority of citizens hold relatively little securitized wealth, a transfer of wealth to securities normally would cause more citizens harm than benefit even if the transfer were -- in the aggregate -- beneficial.

This disparity of interests has concrete results in important conflicts: For example, if corporations were able to reduce their employees' share of the corporate pie in favor of security holders, the portfolio investor would support this action even if the reduction to employees were far larger than the benefit to securities holders and even if most of the people behind the fictional shareholders found their identities as employee far more salient than that as shareholder.

Consider, for example, a hypothetical situation in which corporations suddenly began to abrogate implicit contracts of lifetime employment. [FN127] Conceivably, downsizing and reduced real wages might lead to demoralized employees spending more energy planning their future careers outside the company than inside