法学资料大全二（英文）：公司法的局限性

Columbia Law School The Center for Law and Economic Studies 435 West 116 th St. New York, NY 10027-7201 Working Paper No. 186 Updated Corporate Law’s Limits (Prior Title - The Quality of Corporate Law Argument and its Limits) Mark J. Roe Columbia University School of Law January 16, 2002 This paper can be downloaded without charge from the Social Science Research Network electronic library at: http://papers.ssrn.com/abstract=260582 An index to the working papers in the Columbia Law School Working Paper Series is located at: http://www.law.columbia.edu/lawec/ Corporate Law’s Limits Mark J. Roe A strong theory has emerged in recent years that the quality of corporate law primarily dete r mines whether securities markets arise, whether ownership separates from control, and whether the modern corporation can prospe r. The theory has been used convincingly to explain why we see weak corporate structures in transition and developing nations, but less convincingly to explain why concentrated ownership persists in continental Europe or why it became less important in the United States. The theory has its real-world correlates as international agencies focus on corporate governance for transition and developing economies, while several already-wealthy nations also focus on improving corporate la w. Surely, when an economically-weak society lacks regularity a gap that may be manifested by weak or poorly enforced corporate law that lack of regularity and that lack of economic strength precludes complex institutions like securities markets and diffusely- owned public firms. But in several nations in the wealthy west legal structures are quite good and, by measurement, shar e holders are well protected, but ownership has still not yet separated from control. Something else has impeded separation. W e can hypothesize what that something is by examining the calculus of owners and investors when they decide whether to diffuse ownership. Ownership cannot readily separate from control when man a gerial agency costs are especially high. And missing from current discourse is the basic concept that even American corporate law usually seen as high quality nowadays does not burrow into the firm to root out those managerial agency costs that arise from mediocre business decisions. Judicial doctrine and legal inquiry attack self- dealing, not bad business judgment. The business judgment rule, under which judges do not second- guess managerial mistake, puts the full panoply of agency costs such as over-expansion, over- investment, and reluctance to take on profitable but uncomfortable risks beyond direct legal inquiry. (This limit from the business judgment rule is not a “defect” in corporate law: a g gressive judicial attack on managerial error would replicate the costs of government management of business. Something other than direct legal attack has to control basic managerial agency costs, because judicial action here is far too costly.) The consequence is that even if corporate law as usually co n ceived is “perfect,” it eliminates self-dealing, not managerial mistake. But managers can lose for shareholders as much, or more, than they can steal from them, and law directly controls only the second cost not the first. If the risk of managerial error varies widely from nation-to-nation, or from firm-to-firm, ownership structure should vary equally widely, even if conventional corporate law tightly protected shareholders everywhere. There is also good re a son, and some new data, consistent with this analysis: by measurement several nations have fine enough corporate law; distant stockholders are protected from controlling stockholder and managerial thievery, but unco n trolled agency costs though seem to be especially high in those very nations. TABLE OF CONTENTS Introduction ................................ ................................ ................................ ................... 1 I. The Argument: Corporate Law as Propelling Diffuse Ownership .......................... 6 A. Protecting Minority Stockholders ................................ ................................ .. 7 B. The Attractions of a Technical Corporate Law Theory ................................ .. 9 II. Its Limits: Theory ................................ ................................ ................................ 10 A. Where Law Does not Reach : How Managerial Agency Costs Impede Separation ................................ ................................ ....................... 10 B. Improving Corporate Law without Increasing Separation ............................ 10 1. The model ................................ ................................ ........................... 11 2. An example ................................ ................................ ......................... 13 C. Corporate Law’s Limited Capacity to Affect Agency Costs ......................... 14 1. The business judgment rule ................................ ................................ 14 2. A gency costs: shirking and stealing ................................ ................... 16 D. Law’s Indirect Effect on Agency Costs ................................ ........................ 17 E. Even if Law Critically Affects Both ................................ ............................. 19 F. Precision in Defining Agency Costs and Private Ben efits ............................ 19 G. Ambiguity in the Legal Theory: Improving it Can Reduce Separation ......... 20 1. The offsetting effects ................................ ................................ ........... 20 2. Illustrating the countervailing movement ................................ .......... 21 H. The Tight Limits to the Purely Legal Theory ................................ ............... 27 III. Its Limits: Data ................................ ................................ ................................ .... 27 A. Measuring Quality ................................ ................................ ....................... 28 1. Corporate law: what counts? ................................ ............................ 28 2. Corporate law: the b ottom-line ................................ .......................... 30 B. Data: Nations with Good Corporate Law but Without Separation ............... 30 1. Market measures of the value of control ................................ ............ 30 2. Dual class common stock ................................ ................................ ... 33 3. The control block premium ................................ ................................ . 37 4. And not-so-rich nations? ................................ ................................ .... 41 5. Enforcing contracts ................................ ................................ ............ 41 C. What Beyo nd Law is Needed for Separation in the Wealthy West? ............. 42 1. Economic preconditions ................................ ................................ ..... 42 2. Political preconditions ................................ ................................ ....... 43 3. Social preconditions ................................ ................................ ........... 44 D. Data on Explanation s Beyond Law ................................ ............................. 44 Conclusion: The Quality of Corporate Law Argument and its Limits ........................... 47 Bibliography ................................ ................................ ................................ ................ 50 Tables and Graphs Table 1: When agency costs high, private benefits to contr oller irrelevant ................. ................................ 13 Table 2: Indeterminate effect of better corporate law in rich nation ........................... ................................ 26 Table 3: Voting premium and ownership separation ................................ .................. ................................ 34 Table 4: Control premium and ownership separation ................................ ................ ................................ 38 Table 5: Correlation matrix ................................ ................................ ....................... ................................ 46 Table 6: F-test: Law & politics ................................ ................................ .................. ................................ 47 Graph 1: Block premium vs. ownership dispersion ................................ .................... ................................ 39 Graph 2: Block premium vs. ownership dispersion (without Italy and Austria) ......... ................................ 40 Corporate Law’s Limits Mark J. Roe * Introduction The critical precondition to developing modern securities markets, and the economic and technological benefits that go with good stock markets, most recent analyses posit, is a foundation of solid corporate and securities laws that protect stockholders from the rampages of the dominant majority stockholders or controlling managers. Without such corporate law protections securities markets, it is said, will not arise. And if corporate law is good enough in technologically advanced nations, ownership will diffuse away from concentrated ownership into di s persed stock markets. This perspective contributes to understanding the fragility of capital markets in transition and third-world economies. But there is too much that is critical to separation that corporate law does not reach in the world’s richest, most advanced nations. And if those conditions—which depend on institutions other than corporate law—aren’t met, ownership will not diffuse. And in nations that do not meet those conditions, public policy makers would have little reason to invest much in developing good corporate law institutions, because they just would not be used. The conceptual problem is basic: Cu r rent academic thinking lumps together costly opportunism due to a controller’s self-dealing and costly managerial decision-making that inflicts losses on the owners. The first, self-dealing, corporate law seeks to control directly; the second, bad decision-making that damages shareholders, it does not. Other institutions must co n trol the latter and their strength varies from nation-to-nation. Owners tend to stay as blockholders if they expect managerial agency costs would be very high after full sep a ration. Corporate law does not even try to directly control the costs of * Berg Professor of Law, Harvard Law School. Thanks for comments go to Lucian Bebchuk, Victor Brudney, John Coates, Einer Elhauge, Howell Jackson, Ehud Kamar, Reinier Kraakman, Mitch Polinsky, and participants in workshops at Harvard Business School, the National Bureau of Economic Research, the Italian Securities Commission (CONSOB), the Sorbonne, and the Columbia, Harvard, Stanford, University of Southern California and Vanderbilt Law Schools. 2 CORPORATE LAW’S LIMITS misma nagement. Other institutions do. For these other institutions (product market competition, incentive compensation, takeovers, shareholder primacy norms, etc.), corporate law is at most a supporting prop, not the central institution. And, even if one thinks law has an equal role to play in both—in motivating managers as well as in deterring insider machinations—the two depend on different laws and different institutions. These would vary in strength, because of differing national histories, politics, and economic conditions; countries can, and do, better deal with one than the other, thereby affecting which organization—close or diffuse ownership—its institutions favor. Among the world’s richer nations, several by measurement have good minority stoc k holder protection, which the quality-of-corporate law theory would predict should have long ago facilitated separating ownership from control. But despite protective results that keep the ra m pages of the majority stockholders in check, ownership has not yet neatly separated from control. Our task is to assess the theoretical implications of why separation did not ha p pen, since these counter-examples tell us that deficient corporate law probably was not the basic impediment. The fact that ownership did not separate from control in a nation does not tell us whether it didn’t separate because blockholder rampages are uncontrolled or because managerial agency costs would be far too high if ownership separated. Each could have prevented separation. Or one alone could have, with the other not standing in the way. If underlying economic, social, or political conditions make managerial agency costs very high, and if those costs are best contained by a controlling shareholder, then concentrated ownership persists whatever the state of corporate law in checking blockholder mis-deeds. I speculate on what underlying economic, political, and social conditions could make managerial agency costs persistently high. I also speculate on how a shrin k ing of these agency costs, one plausibly now going on in continental Europe, could raise the demand to build legal institutions that facilitate separation. Many business features could keep agency costs higher in one nation than another: a weak product market is one; especially opaque businesses is another; an inability to use incentive compensation effectively because it would, say, disrupt relationships within the firm, is a third; a high level of social mistrust that impedes professionaliz a tion of management is a fourth. CORPORATE LAW’S LIMITS 3 Corporate law, when it’s effective, impedes insider machinations: it stops, or reduces, controlling shareholders from diverting value to themselves, and bars managers from pu t ting the firm into their own pockets. When, for example, controllers obtain very high private benefits from control, because they divert firm value into their own pockets, then distant shareholders mistrust the insiders, and are unwilling to buy. Ownership concentration should, all else equal, persist. Good corporate law (or substitutes like stock exchange rules, contract, media glare, or reputational intermediaries) can, by reducing this potential for thievery, facilitate separating ownership from control. But there is more to running a firm than controlling insider machinations. Managerial agency costs to distant shareholders come in two basic flavors: thievery and mismanagement. Law can reduce the first, but does very little directly to minimize the second. Not yet fully recognized in the current literature is that American law avoids dealing with the second. The business judgment rule has courts refusing to intervene when shareholders attack managerial mistake. Indeed, one might a r gue in only a modest over-statement that in modern American business history, there has been only one significant successful judicial attack on managers for mistake, that in Smith v. Van Gorkom , an attack the legislature promptly reversed. It’s business conditions, incentives, professionalism, capital structure, product and managerial labor market competition, and financial alignment with shareholders that directly impede managerial mistakes, not corporate law. Conventional, technical corporate law has little to say here. (True, law can create or destroy anything, so law isn’t irrelevant, but it is a second-order phenomenon: other institutions primarily control managerial mistake, law’s role here, if any, is either to support or drag on those primary institutional controls.) Even if one believes law to be central to competition, compensation, and so on—the institutions that reduce managerial agency costs—one must recognize that these laws differ from corporate law that controls insider machinations, and their efficaciousness could differ: one nation’s laws might control machinations well but managerial error poorly. Today’s corporate theory cannot explain why several wealthy European nations protect minority shareholders well, but nevertheless still have concentrated ownership. The most plausible theory is that ownership hasn’t separated not because of weak corporate law, but because a) 4 CORPORATE LAW’S LIMITS managerial agency costs from dissipating shareholder value would be very high after full separation, and b) concentrated ownership reduces those costs to shareholders enough. I suggest why these costs to shareholders vary from nation-to-nation and firm-to-firm. Moreover, by shifting our focus from legally malleable private benefits to managerial agency costs, we can see why sub-standard corporate law persists in a few richer, well-developed nations. Low quality law might in some nations be a symptom of weak separation, not its base-line cause. If these kinds of managerial agency costs from dissipating shareholder value would be too high anyway for there to be much separation even if corporate law were pe r fect, then there’s little reason for the players (public policy-makers, investors, founding and family owners) to build good corporate law, because it wouldn’t be much used anyway. * * * A second theoretical limit afflicts the quality-of-corporate-law arg u ment. High quality corporate law could propel diffusion. But it can just as easily propel concentration. Its effect is indeterminate. High quality corporate law makes distant stockhol d ers comfortable with blockholders, because good law channels blockholders away from stealing from distant stockholders and into productive activity (such as overcoming shareholder free rider and informational problems or monitoring managers, for example). Channeling blockholders away from anti-stockholder action should mean in theory that improving the quality of corporate law could, all else equal, increase blockholding as easily as it could decrease it. Minority stockholders have less reason to fear the big blockholders when corporate law protects them. If the blockholders increase value, then we could see more of them develop, not fewer of them, as corporate law quality i m proved. * * * Good corporate law that stymies a grasping controller, or good substitutes like effective stock exchanges, effective reputational intermediaries and the like, is good for a nation to have. It reduces the costs of running a large enterprise. But it is insufficient to induce ownership separation. It’s thus at least possible that some reformers may be pinning their hopes too heavily on good corporate law institutions to propel development in third world and transition countries. CORPORATE LAW’S LIMITS 5 My logic here is that the current wisdom and theory tell us that when the core of corporate law is atrocious, and substitutes unavailable, complex firms cannot be stabilized. This is true, and the empirical contributions here are considerable. But the converse of the current wisdom is believed as well, although it is false: Bad law impedes separation, but when there’s no separation, law could be good with something else impeding that separation, not corporate law. Since several nations have protective corporate law but nevertheless have very little separation, we need some new theory to explain why. * * * A roadmap for this Article: I outline in Part I the quality of corporate law argument and why it is important. In Part II I show why when potential dissipatory managerial agency costs are perniciously high in a society, but containable by dominant stockholders, corporate law quality is irrelevant or tertiary: even if it’s good, ownership will not separate from control. (I distinguish two types of agency costs: those that shift value away from stockholders to controllers and those that dissipate shareholder value.) Conventional corporate law can contain managerial agency costs due to thievery, but does not directly contain managerial agency costs due to mismanagement. Concentrated ow n ership will persist in firms in high- agency-cost nations even if conventional corporate law quality is high as long as the owner can contain enough of these costs. In Part III I show why the data indicates that the quality of corporate law argument, although it explains transition economies nicely, is over-stated for several of the world’s richest nations: in too many of them basic shareholder protections seem adequate, stock can be and is sold, but ownership nevertheless doesn’t separate from control. Something else has made concentrated control persist. I speculate what that might have been. Lastly, I conclude. High quality, protective corporate law is a good institution for a society to have. It lowers the costs of building strong, large business enterprises. It can prevent, or minimize, controlling stockholder diversions, a necessary condition for separation. But among the world’s wealthier nations, it doesn’t primarily determine whether it’s worthwhile to build those enterprises. It’s a tool, not the foundation. 6 CORPORATE LAW’S LIMITS I. The Argument: Corporate Law as Propelling Diffuse Ownership Today’s most powerful and most widely-accepted academic explanation for why continental Europe lacks deep and rich securities markets is the purportedly weak role of corporate and securities law in protecting mino r ity stockholders, a weakness that is said to contrast with America’s strong protections of minority stockholders. A major European- wide research network, leading financial economists, and leading legal commentators have stated so. 1 One can imagine the Nobel Prize wi n ning Franco Modigliani shaking his head in disappointment when writing that nations with deficient legal regimes cannot get good stock markets and, hence, “the provision of funding shifts from dispersed risk capital [via the stock market] … to debt, and from [stock and bond] markets to institutions, i.e., towards intermediated credit.” 2 In a powerful set of important articles insightful economists showed that deep securities markets correlate with an index of basic shareholder legal protections. 3 These protections are important: “[P]rotection of shareholders … by the legal system is central to unde r standing the patterns of corporate finance in difference countries. Investor protection [is] crucial because, in many countries, expropriation of minority shareholders … by the co n trolling shareholders is extensive.” 4 Leading legal commentators have signed on to the law-driven theory. 5 At the same time, international agencies such as the IMF and the World Bank have admirably promoted corporate law reform, especially that which would protect minority stockholders. 6 The OECD has had major initiatives to improve corporate governance, both in the developing 1 . La Porta, Lopez-de-Silanes & Shleifer (1999); La Porta et al. (1998), at 1136- 37 and (1997), at 1138; Bebchuk (1999); Becht & R?ell, (1999); Carlin & Mayer (Oct. 1998), at 33; Coffee (1999). 2 . Modigliani & Perotti (1998), at 5; see also Modigliani & Perotti (1997). 3 . See La Porta et al. articles, cited supra note 2. 4 . La Porta, Lopez-de-Silanes, Shleifer & Vishny (2000), at 4.. 5. See Coffee (1999). 6 . The IMF’s journal, Finance & Development, tells us that “improved corp orate governance [institutions] are essential parts of economic reform programs under way in many countries.” Iskander, Meyerman, Gray & Hagan (1999). CORPORATE LAW’S LIMITS 7 and the developed world. 7 The “OECD and the World Bank agreed [in 1999] to cooperate in the promotion of improved corporate governance on a world-wide basis. Both institutions are committed to assisting governments in … improving the legal, institutional and regulatory framework for corporate governance in their countries. The move responds to mandates from finance ministers and central bank governors of the G-7 and the OECD countries.” 8 These are valuable initiatives. They will contribute to reaching their goals of more stable enterprises and better economic performance. But corporate law, and the reach of government policy-makers through corporate law reform, has limits. Here we demarcate those limits’ boundaries, beyond which corporate law ceases to be a primary institution. A. Protecting Minority Stockholders The basic law-driven story is straightforward: Imagine a nation whose law badly pr o tects minority stockholders against a blockholder extracting value from small minority stockholders. A potential buyer fears that the majority stockholder would later shift value to itself, away from the buyer. So fearing, the prospective minority stockholder does not pay pro rata value for the stock. If the discount is deep enough, the majority stockholder decides not to sell, concentrated ownership persists, and stock markets do not develop. 9 Or, approach the problem from the owner’s perspective. Posit large private benefits of control. The most obvious that law can affect are benefits that the controller can derive from diverting value from the firm to 7 . OECD (1999); Nestor (2001). 8 .OECD (1999a); Witherell (2000). To be clear here, the international authorities have not solely focused on corporate law, but have attended to governance practices as well. E.g., Nestor (2000); W ORLD B ANK (2000), at 20-24. 9 . If the public stock-buyers are non-naive, the selling blockholders are have contractual means to stymie raiders. Capped voting, mandatory bids, and poison pills can reduce or end the buying stockholders’ fears if the nation enforces contract satisfactorily, even if its corporate law is weak. Moreover, whether the common law ? as opposed to other non-law-based institutions ? did in fact well protect minority stoc k holders during the early development of the public firm is open to question. See Roe (2000), at 586 n.124 & 586-93 and (2001). But here we are making the quality- of-corporate-law argument, not yet evaluating it. 8 CORPORATE LAW’S LIMITS himself of herself. The owner might own 51% of the firm’s stock, but retain 75% of the firm’s value if the owner can over-pay hi m self or herself in salary, pad the company’s payroll with no-show relatives, use the firm’s funds to pay for private expenses, or divert value by having the 51%- controlled firm over-pay for goods and services obtained from a company 100%-owned by the controller. Strong fiduciary duties, strong doctrines attacking unfair interested-party transactions, effective disclosure laws that unveil these transactions, and a capable judiciary or other enforcement institution can reduce these kinds of private benefits of control. (Private benefits also arise from pride in running and controlling one’s own, or one’s family’s, e nterprise. About this, corporate law has little direct impact.) The owner considers whether to sell to diffuse stockholders. With no controller to divert value, the stock price could reflect the firm’s underlying value. But the rational buyers believe, so the theory runs, that the diffuse ownership structure would be unstable, that an outside raider would buy up 51% of the firm and divert value, and that the remaining m i nority stockholders would be hurt. Hence, they would not pay full pro rata value to the owner wishing to sell; and the owner wishing to sell would find that the sales price to be less than the value of the block if retained (or if sold intact). Hence, the block persists. 10 The controller refuses to leave control “up for grabs” because if it dips below 51% control, an outsider could grab control and reap the private benefits. 11 10 . See Bebchuk (1999); cf. La Porta et al. series, supra note 2. 11 . The literature correctly focuses on corporate law’s capacity to contain controlling stoc k holder thievery. Other features of corporate law ?agency theory in determining who can represent the corporation, corporate transactional flexibility, limited liability ? are down-played. Managerial capacity to mismange the corp o ration is also down-played, but it is quite relevant. It tends, improperly, to be ignored in the current literature. I will seek to introduce it here to modern analysis of the ownership separation decision. The theory has gaps. When the founder is selling out all of his or her stock, the potential stock buyers discount the purchase, the theory runs, because they anticipate a raider buying up the stock cheap and then diverting value to itself. But when raiders compete, the winning raider will bid the stock up to the value of the private ben e fits of control. Minority buyers would anticipate this competition and pay approximately full pro rata value for the stock when the founder sells. (The theory needs non- competitive—or secret—raiders.) CORPORATE LAW’S LIMITS 9 B. The Attractions of a Technical Corporate Law Theory One sees the appeal of the quality of corporate law argument. Technical instit u tions are to blame, for example, for Russia’s and the transition nations economic problems. The fixes, if they are technical, are within our grasp. Human beings can control and influence the results. Progress is possible if we just can get the technical institutions right. If it turns out that deeper features of society ?industrial organization and competition, politics, conditions of social regularity, or norms that support shareholder value ?are more fundamental, we would feel ill at-ease because these institutions are much harder for policy-makers to control. (To be clear here, I am not speaking simply of corporate law as just the “law-on-the-books,” but as “law-on-the-books,” including securities law, and the quality of regulators and judges, the efficiency, accuracy, and honesty of the regulators and the judiciary, the capacity of the stock exchanges to manage the most egregious diversions, and so on. 12 ) * * * And as self-contained academic theory, there is little to quarrel with in the quality-of-corporate law argument. It is sparse and appealing. Good corporate law lowers the costs of operating a large firm; it’s good for a nation to have it. But we need more to understand why ownership doesn’t separate from control even where core corporate law is good enough. Where managerial agency costs due to potential dissipation are substantial, concentrated ownership persists even if conventional corporate law quality is high . Given the facts that we shall develop in Part III ?there are too many wealthy, high quality co r porate law countries without much separation ?the quality-of-corporate-law theory needs to be further refined, or replaced. This we do next in Part II. 12 . The best compendium is in Black (2000). 10 CORPORATE LAW’S LIMITS II. Its Limits: Theory A. Where Law Does Not Reach: How Managerial Agency Costs Impede Separation Managers would badly run some firms if their firms’ ownership separated from control. Effective corporate and s e curities laws constrain controllers’ stealing , but do much less to directly induce them to operate their firms well. A related-party transaction can be attacked or prevented where corporate law is good, but an unprofitable transaction law leaves untouched, with managers able to invoke corporate law’s business judgment rule to ward off direct legal scrutiny. Consider a society (or a firm) where managerial agency costs from dissipating shareholder value would be high if there’s separation, but low if there’s no separation, because a controlling shareholder can contain those costs. When high but containable by concentration, concentrated shareholding ought to persist even if corporate law fully protects minority stockholders from the ravages of dominant stockholders. Blockholders would weight their costs in maintaining control (in lost liquidity, lost diversification, etc.) versus their pote n tial loss in value from managerial agency costs. Control would persist even if corporate law were good. This is a basic but important point, and it is needed to explain the data that we look at in the next Part. B. Improving Corporate Law without Increasing Separation The arguments in the prior section ?that variance in managerial agency costs can drive ownership structure even if conventional corporate law is quite good can be stated formally in a simple model. High managerial agency costs precludes separation irrespective of the qua l ity of conventional corporate law. CORPORATE LAW’S LIMITS 11 1. The model . Let: A M = The managerial agency costs to shareholders from managers’ dissipating shareholder value, to the extent avoidable via concentrated owne r ship. C CS = The costs to the concentrated shareholder in holding a block and monitoring (that is, the costs in lost liquidity, lost diversification, expended energy, and, perhaps, error). When A M is high, ownership will persist in conce n trated form whether or not law successfully controls the private benefits that a controlling shareholder can siphon off from the firm. V = Value of the firm when ownership is concentrated. B CS = The private benefits of control, contai nable by corporate law. Consider the firm worth V when ownership is concentrated. Posit first that managerial agency costs are trivial even if the firm is fully public. As such, the private benefits of control, a characteristic legally malleable and reducible with protective corporate law, can determine whether ownership separates from control. Consider the controller who owns 50% of the firm’s stock. As such he obtains one-half of V, plus his net benefits of control. (In this simple first model, the value of the firm remains unchanged whether it has a controlling stockholder or is fully public.) He retains control when the following inequality is true: (1) V/2+B CS -C CS >V/2. The left side is the value to the controlling stockholder of the control block: half the firm’s cash flow plus the private benefits diverted from minority stockhol d ers, minus the costs of maintaining the block (in lost diversification and liquidity). The right side is the value he obtains from selling the block to the public. Equation (1) states that as long as the private benefits of control (e.g., in value shifted from minority stockholders) exceeds the costs of control, then concentrated ownership persists. B ecause corporate law can dramatically shrink the private 12 CORPORATE LAW’S LIMITS benefits, B CS , corporate law matters quite a bit in equation (1). 13 This is the conventional theory 14 that we shall next amend. We amend by introducing A M, managerial agency costs from dissipating shareholder value. If those managerial agency costs are non- trivial, then the controllers’ proceeds from selling into the stock market would be (V-A M )/2. Concentration persists if and only if (2) V/2+B CS -C CS >(V-A M )/2. Re-arranging: concentration persists if the net benefits of control (B CS - C CS ) are more than the controller’s costs of diffusion (A M /2): (3) B CS -C CS >-A M /2. Or, further re-arranging, concentration persists if: (4) B CS +A M /2> C CS . Quality-of-corporate-law theory predicts diffusion won’t occur when B CS >C CS , with corporate law the means of containing B CS . But we now can see the limits to the theory: True, without agency costs, A M , the level of private benefits, B CS , can determine whether ownership diffuses or concentrates. But where A M is high, diffusion won’t occur even if B CS is zero, because A M could take-over and drive the separation decision. B CS , the contro lling shareholder’s private benefits, are relatively unimportant if A M is very high. Only when A M →0 do legally malleable private benefits kick in as the critical determinant. 13 . Some private benefits are matters of taste, preferences for power, fam ily recognition in a family firm, etc. These are not readily containable by law; they might be better analyzed here as part of the costs of control, C CS , as mitigating the usual costs (lost diversification, liquidity, etc.) They also might vary from firm-to-firm and nation- to-nation. And the capacity to off-load illiquidity and non-diversification might vary similarly. Where risks can be hedged, owners should indulge themselves and keep control more readily than where they cannot. 14 . See Bebchuk (1999) . CORPORATE LAW’S LIMITS 13 2. An example. Agency costs as impeding separation can be exemplified. Firm has value V of 150 under concentrated ownership, 100 if it’s diffusely owned. Managers will not steal. Law is good enough here. But they will be loose with shareholders’ investment in the firm. Unlike a controlling shareholder, they will over-expand, react slowly to changing markets, and avoid the tough decisions. Hence A M =50. Consider two situations for private benefits to the controlling shareholder: in one it’s high, equal to one-third of the firm; in the other B CS is low, equal to zero. Table 1: When agency costs high, private benefits to controlling shareholder irrelevant; concentration persists even if law drives down private benefits of control V, firm’s value to shareholders Value to 50% owner Value to minority shareholders Notes Private benefits low: B CS =0; Managerial agency costs high: A M =50 Concentrated ownership 150 75 75 Diffuse ownership (A M =50); controller sells block for 50 100 50 50 Management loses A M ; concentration persists Private benefits high: B CS =1/6 of V Concentrated ownership 150 100 50 Diffuse ownership (A M =50); controller sells block for 33 100 33 33 Raider grabs 33 (and perhaps A M ) 15 ; hence, concentration persists Separation loses shareholders 50 in value. Because that lost value, A M, is high, at 50, shareholders would seek, and pay for (in C CS , for instance) a structure that would preserve those profits for themselves. If blockholding keeps A M low (which it does as the model defines A M : the 15 . Again, this conventional scenario has us assuming that the raider can attack by secretly accumulating the block or without having to pay stockholders the value it will acquire. If raiders compete, the price will be bid up to 100, or perhaps 150, and the standard corporate law story would fade. See supra note 9 . 14 CORPORATE LAW’S LIMITS agency costs that blockholding would avoid), and if the costs of blockholding, C CS , are low, blockholding will persist irrespective of whether private benefits, B CS, are 50, 25, or zero . True, if the costs of blockholding exceed A M , or if the efficiency benefits of diffuse ownership ove r whelm A M , then private benefits once- again become relevant. The most plausible scenario under which they become relevant is, again, when A M →0. But when A M , the containable managerial dissipation, is very high, then this managerial agency cost determines whether ownership can separate. C. Corporate Law’s Limited Capacity to Reduce Agency Costs One might reply that core corporate law when improved reduces both the controlling stockholder’s private benefits (B CS , by reducing the controller’s capacity to siphon off value) and managerial agency costs (A M , by reducing the managers’ capacity to siphon off benefits for themselves). 1. The business judgment rule. This criticism is both right and wrong, but mostly wrong. The reason it’s mostly wrong is simple. Managerial agency costs are the sum of managers’ thievery (unjustifiably high salaries, self-dealing transactions, etc.) and their mismanagement. Economic analyses typically lump these together and call them “agency costs.” But agency costs come from stealing and from shirking. It is correct to lump them t ogether in economic analyses as a cost to shareholders , because both costs are visited upon shareholders. 16 But it is incorrect to think that law affects each cost to shareholders equally well. The standard that corporate law applies to managerial decisions is, realistically, no l i ability at all for mistakes, absent fraud or conflict of interest. 17 But this is where the big costs to shareholders of having managerial agents lie, exactly where law falls into an abyss of silence. Conventional corporate law does little, or nothing, to directly reduce shir k ing, mistakes, and bad business decisions that squander shareholder 16 . Fama (1980) (agency costs come from “shirking, perquisites or incompetence”). 17 . Dooley & Veasey (1989), at 521 (Veasey is now the Delaware Supreme Court chief judge); Bishop (1968), at 1095 (managers without a conflict of interest always win); Rock & Wachter (2001), at 1664-68. CORPORATE LAW’S LIMITS 15 value. Elaborate do c trines shield directors and managers from legal action. The business judgment rule is, absent fraud or conflict of interest, nearly insurmountable in America, insulating directors and managers from the judge, and not subjecting them to scrutiny. Consider this statement from a well-respected Delaware chancellor: There is a theoretical exception to [the business judgment rule, protecting directors and managers from liability] that holds that some decisions may be so “egregious” that liability … may follow even in the absence of proof of conflict of interest or improper motiv a tion. The exception, however, has resulted in no awards of money judgments against corp o rate officers or directors in [Delaware] . … Thus, to allege that a corporation has suffered a loss … does not state a claim for relief against that fiduciary no matter how foolish the i n vestment …. 18 One does not exaggerate much by saying that American corporate law has produced only one major instance in which non-conflicted managers were held liable to pay for their mismanag e ment: Smith v. Van Gorkom , 19 a decision excoriated by managers and their lawyers, and promptly overturned by the state legislature. 20 Nor should we think, “Oh, this is just a gap in American law, one that could be filled if other legal institutions didn’t sufficiently control these managerial agency costs. If the other institutions failed, corporate law would, and could, jump in.” One wouldn’t want the judge in there, regularly second-guessing managers anymore than one would want the commissar or the bureaucrats in there directing the managers. Most American analysts would assume that it would be costly for judges to regularly second-guess managers’ non-conflicted business decisions. 21 18 . Gagliardi v. Trifoods Int’l, Inc., 683 A.2d 1049, 1052 (Del. Ch. 1996) (Allen, J.) (emphasis su p plied). 19 . 488 A.2d 858 (Del. S up. Ct. 1985). And not just Delaware and not just recently: “[I]t is only in a most unusual and extraordinary case that directors are held liable for negligence in the absence of fraud, or improper motive, or personal interest.” Bayer v. Beran, 49 N.Y.S.2d 2, 6 (N.Y. Sup. Ct. 1944). 20 . Del. Corp. Code § 102(b)(7). And thirty-eight legislatures, aghast at the possibility of directorial liability if their courts followed Delaware in Smith v. Van Gorkom , passed similar exemptive legislation. See Smith (1998), at 289 n.52. Corporate casebooks have to go back quite far to find other judicial attacks on managerial error, much less find managerial judicial defeats. Shlensky v. Wrigley, 237 N.E.2d 776 (Ill. App. 1968); Dodge v. Ford Motor Co. 170 N.W. 668 (Mich. 1919). 21 . See Joy v. North, 692 F.2d 880 (1982) (Winter, J.). 16 CORPORATE LAW’S LIMITS 2. Agency costs: shirking and stealing . Stated more formally: A M = A LD + A MM ), where total managerial agency costs are the sum of legally controllable diversions, A LD (stealing), and legally uncontrollable managerial error that dissipates value, A MM (shirking in the economic literature). So subst i tuting into (2) above, we obtain: (5) V/2+B CS -C CS >(V-A LD -A MM )/2. Good basic corporate law reduces B CS and A LD , a component of agency costs, but it doesn’t touch A MM . With perfect corporate law, B CS =0 and A LD = 0, which yields: (6) V/2-C CS >(V-A MM )/2. Or: (7) A MM /2>C CS . When (7) holds, ownership does not separate from control, even if a perfect corporate law reduces the private benefits of control to zero. And corporate law does not directly affect A MM . Good corporate law is insufficient to induce separation. * * * One might refine this analysis by adding a term to account for controlling shareholder error. One could, but the costs of these errors would be smaller than legally unco n trollable managerial error. True, similar legal doctrines (the business judgment rule) shield the controlling shareholder from lawsuits for a non-conflicted mistake. But because the controlling stockholder owns a big block of the company’s stock, it internalizes much of the cost of any mistake (unlike the managers). Because the controlling stockholders, in contrast to the managers, bear the costs of their error, they immediately internalize their errors, unlike managers who must be made indirectly to internalize them (via incentive compensation, labor market constraints and the like). Controlling CORPORATE LAW’S LIMITS 17 shareholders who make many mistakes can sell their firm. 22 Only if the compensating gains (such as power, prestige, and pride of ownership) exceed the costs of their errors (and the costs of carrying the block), do they keep that block of stock. Managers who make errors do not face such immediate constraints and incentives. Simply put: controlling stockholders impose the cost of their stealing on diffuse stockholders but absorb half of the cost of their own mismanagement. Co ntrolling managers impose the cost of their dissipation, as well as of their thievery, on diffuse stockholders. Law reduces stealing, not unconflicted mismanagement. The point is not that founders and blockholders never make mistakes, but that they bear most of the costs of their waste, while managers do not. D. Law’s Indirect Capacity to Affect Agency Costs We have thus far focused on the effects on separation of conventional corporate law, the law of fiduciary duties, of derivative suits, and of corporate waste. Conventional corporate law can reduce stealing and, where it or a substitute fails to, separation should not be wide. But even if it succeeds, managerial agency costs to shareholders could be high and, when high, ownership cannot readily separate. Institutions other than conventional corporate law raise, lower, and control managerial agency costs, reducing them via competitive markets, shareholder wealth maximization norms, incentive compensation, hostile takeovers, and corporate transparency. For these institutions, law is also relevant. But its relevance is indirect. True, law can pote n tially encompass everything in a society. Law could ban the institutions that indirectly reduce agency costs. Anything can be taxed, destroyed, and prohibited. That is, although ordinary but mistaken managerial business decisions and corporate transactions are immune from any direct judicial attack, other institutions in society affect these business decisions and transactions; and law can facilitate or ban these other institutions. But in each case, the other institution (the competitive product market, the 22 . That is, controlling stockholders might manage badly, because they maximize personal utility not firm value. Cf. Demsetz (1983). The usual analysis is that they internalize these costs and would usually sell if they couldn’t manage well. 18 CORPORATE LAW’S LIMITS incentive compensation, the pro-shareholder norm, the hostile takeover) is the primary control, with law just assisting or impeding that institution. But for insider thievery, basic corporate law is a primary deterrent. It’s the judge who bars the transfers, orders recovery of the diverted value, and punishes the wrong-doer. The judge intervenes directly. For the indirect constraints, the judge doesn’t attack directly and often never gets involved at all. Consider five major direct constraints on managerial agency costs: product market competition, shar eholder primacy norms, incentive compensation, hostile takeovers, and transparency. Strongly competitive markets, for example, can be prodded along by good antitrust law, or lost by bad antitrust law. 23 But the primary constraint is the product market, not law, and law only acts as a secondary, and perhaps sometimes inconsequential means of enhancing or demeaning product competition. And shareholder primacy norms, for example, can be facilitated, or demeaned by legal pronouncements. But the norm, not the pronouncement, is the direct means to affect managerial performance. Or incentive compensation can be spurred, or taxed. But once again, it’s not the tax rule that spurs managers directly, but the incentive compensation that pushes them. Hostile tak e overs can be allowed, or banned. And again, a market- player launches the takeover; judges and rules only modulate the offeror’s chances of success. 24 Transparency can be mandated, or muddled. Law obviously can be, or is, in play, albeit indirectly. Although law affects these institutions, law’s effects here differ from the effects of conventional corporate law. First off, they do not directly invoke the core explanation for good corporate law, namely that it grows out of co m mon law and the judge’s capacity to control interested-party, conflict-of-interest transa ctions that divert corporate value into the 23 . Strong capital markets can also constrain managers, but I do not add capital markets here because they are closely connected with good corporate law. 24 . Although law affects takeovers, even perfec tly pro-shareholder takeover law will not drive managerial agency costs to zero. Critics of American corporate law point to British rules and enforcement structure as being as good for shareholders as possible. E.g., Bebchuk & Ferrell (1999), at 1193; DeMott (1983); Miller (1998), at 58. An indicator of the size of impediments to takeover is the size of the premium offerors pay to overcome them. In the United States, the premium in a hostile bid averages 50%; in Britain it’s noticeably lower at 40%, see Thomson Financial (database) (accessed Oct. 2001), but if British corporate law is the realistic ideal, there’s more (40%) impeding a perfectly smooth hostile market than law alone. CORPORATE LAW’S LIMITS 19 controller’s bank account. Most importantly, law here doesn’t attack the cost to shareholders directly (as law does when the judge punishes a controlling shareholder who diverts value to herself). Law’s role is not to attack directly but to enhance or impede the private institution that would reduce the dissipation. More generally, these institutions (excepting perhaps basic transparency) have the potential to be politically charged, and in other nations one or the other or all four are politically charged. E. Even if Law Critically Affects Both One might reject the proposition that law is secondary and indirect here, only enhancing or demeaning the institutions that more directly control agency costs. One might believe that law is so central to these other institutions that control managerial agency costs (competition, compensation, and so on) that one would be driven to believe that law is still central to whether public firms can exist and whether ownership can separate from control. Bu even so, the structure of my argument persists: Different institutions and different laws affect managerial agency costs than the institutions and laws that affect insider machinations. The two sets are not identical. They barely overlap. If one society does better with one set that with the other, the degree of diffusion is deeply affected. Corporate law might minimize insider transactions, but the other laws might fail to reduce managerial agency costs. F. Precision in Defining Agency Costs and Private Benefits B CS we’ve defined as the private benefits that a controller can grab from the firm by diverting value away from the firm’s stockholders. A typical such diversion would be for the controller to fully own a private entity that sells product to the firm at inflated prices. Dissipating shareholder wealth isn’t the actor’s goal; shifting that wealth to him or her is. But to be precise here as well, dissipation would be a secondary result. To make the transfer from the firm into her own pocket, the controller might have to distort the firm’s operations. But her primary goal is to divert, not to dissipate. 20 CORPORATE LAW’S LIMITS A M we’ve defined as the value that managers can dissipate in the firm, due to mistaken management. They might over-invest, under-invest, or mis-invest. They might over-pay suppliers or fail to adopt profitable technologies. They might react too slowly to changing market conditions. But to be precise here, diversion could still be a secondary result here too. The dissipation occurs because the managers could work a little harder, or a little longer, or take on the tough decisions. Their action is a kind of self-dealing, in that they benefit from the easier life. But the primary effect of A M is to dissipate value; the diversion managers get by working a little less hard is secondary to the dissipation of shareholder value. G. Ambiguity in the Legal Theory: Improving it Can Reduce Separation There is more. Thus far I have accepted the conventional wisdom that strengthening corporate law facilitates separation across-the-board, in rich, developed nations as well as in transition and developing ones. But even if managerial agency costs are constant (at any initial level of ownership concentration), a theory of separation based on corporate law is softer here than the dominant literature has it. Improving corporate law in the world’s richest nations is in theory as likely to increase blockholding as to decrease it. 1. The offsetting effects . Recall the core corporate law argument from Part I. If minority stockholders are u n protected by corporate law, they will not buy or will only buy at a discount. With private benefits of control high, the controller must hold onto control, because those benefits ca n not be sold (other than by selling the block intact), lest someone else be able to grab those benefits of control. In such settings, distant investors invest reluctantly, fewer firms go public, and those that do will retain a concentrated owner. For some firms in some nations when those firms do go public, they set up roadblocks to a raider entering, via poison pills, capped voting, and mandatory bid rules. (Managers ofte n times seek such rules to entrench themselves, but such rules keep nasty raiders out as well.) Consider the firm that is public, with pills, caps, and other charter terms that keep raiders out. Consider two nations, A and B, with A protecting minority stockholders imperfectly and less well than B. (Nation CORPORATE LAW’S LIMITS 21 A is moving through law-reform from imperfect but not atrocious minority protection to better mino r ity protection.) The minority protection argument tells us that minorities would feel more comfortable in “protective” nations, such as B, or the reformed A, than in non- protective nations, such as the old A. Hence, the “better” corporate law nations could end up with more blockholders in those firms that go public. More controllers would be willing to go public, because investors, feeling well protected, would pay full pro rata value for the minority stock that they would buy. More dispersed stockholders would be willing to accept a blockholder, because better law would lower the blockholder’s capacity to rip them off. Blockholders provide critical good services to the firm and one powerful bad service: the good ones are monitoring managers, 25 facilitating information flow from inside the firm to capital owners, 26 and making implicit deals with stakeholders when soft deals are efficient; 27 their one big bad activity is their stealing from the minority stockholders. But if a nation’s laws limit their potential to do bad without diminishing their ability to do good, then one could expect that nation’s firms to get more blockholders, not fewer. 2. Illustrating the countervailing movement. Consider the impact of corporate law improvement on three categories of large-firm ownership: 1. Diffu se 2. Public, but with a dominant stockholder, and 3. Privately-held. The quality-of-corporate law thesis assumes that law reform would, monotonically, increase diffusion. Without satisfactory corporate law protections, diffuse ownership would be unstable, with a raider able to capture control and siphon off private benefits. Large owners (in categories 2 or 3) therefore cede control only reluctantly. If a nation 25 . Shleifer & Vishny (1986), at 465 (“our analysis indicates that [by monitoring managers] large shareholders raise expected profits and the more so the greater their percentage of ownership”); Bertrand & Mullainathan (2000) (managers in firms with blockholding stockholders have less performance-based pay than managers in firms with blockholders). 26 . Roe (1994), at 260-61 (1994) (large owners can mitigate); Stein (1989) (diffuse ownership creates informational inefficiencies). 27 . Cf. Shleifer & Summers (1988). 22 CORPORATE LAW’S LIMITS improves its corporate law, large owners’ reluctance to cede control diminishes. But now consider a nation that has plausible but flawed corporate law. Some firms are fully public firms in category 1, but most large firms stay in category 2 or 3. Corporate law has gaps but it’s good enough for some public firms, and there are a few. The diffuse firms use contract to keep out future blockholders, i.e., capped voting that stops blockholders from taking control. Some firms are sufficiently valuable when fully public that they can absorb the costs of these barriers to controller entry. But many firms stay in category 2 or 3, because the dominant owner cannot cheaply enough construct structures that minimize any future grab for the private benefits of control. (Think France, Belgium, or, as usually viewed, Germany.) Corporate law (or a substitute, or its enforcement) improves. 28 Current discourse focuses on the motivations of the owners in the middle category—those in category 2 that fear a loss of private benefits. No longer so fearing, they relinquish control. Ownership diffuses. This is surely one logical effect of improving corporate law. But it’s not the only effect. Consider first those firms that are already public. They have devices that minimize the intrusion of blockholders, to impede a raider from entering to siphon off private benefits. But with private benefits less available, the firm could drop its barriers to entry, as the siphoning—possible when corporate law was weak—is less important when improved corporate law reduces the private benefits. A blockholder can enter if it can add value. 29 This might be especially so if the reason the firm moved from category 2 to category 3 was because of reduced firm value because of the existence of private benefits. That is, if shareholders and dominant stockholders were always wary, always wrestling for position, and sometimes engineering value-decreasing transactions (by the blockholder to siphon off value, or by the blockholder to “warranty” to outsiders that it wouldn’t 28 . Path dependent drags will impede legal change, and if such c hange occurs, path dependence will slow change. Roe (1996); Bebchuk & Roe (1999). Here we simply analyze that even without path dependent impediments, improving law propels the firms in several directions, not just the one conventionally assumed. 29 . Not all will, of course. Path dependence, id., and positional advantage will deter many. The point is that the pressures here from improving corporate law don’t all point toward greater diffusion. CORPORATE LAW’S LIMITS 23 siphon off value and that hence the outsiders should pay full “pro rata” value) then the firm might have moved to diffuse ownership as the less costly alternative. With this drag on value—blockholder vs. diffuse stockholder infighting and bonding—removed, the firm could stabilize with a blockholder. (We could use the inequalities in the earlier model: improving corporate law reduces C CS by eliminating the costs of conflict, the resulting sub-optimal operating strategies, etc. Law by lowering those costs of concentration, would make concentration more viable for more firms.) Other movement toward public firms with dominant blockholders could occur. Many firms might be in Category 3—privately-held— because the owners refuse to sell out a discount. They do not sell for several reasons: a) firm value might decline as they are forced to take on value-minimizing transactions to keep the private benefits flowing to themselves, b) the outsiders and insiders might not agree on the expected future private benefits, so the right price might not be obtained. But with corporate law improved, the blockholder could sell, and investors could buy, confident that future rip-offs would be minimal. With future rip-offs minimal, value-decreasing transactions could not occur (or would occur less frequently and less severely). Hence, more initial public offers would be made. And with future rip-offs minimal, the sellers and investors could focus on fundamental value, instead of the future division of the pie, in pricing the stock in the initial public offer and thereafter. To illustrate, consider a nation with 30 large firms. Ten are diffusely- held, ten are public but with blockholders, ten are fully private. In the commonly used indices of diffusion, this nation would have a .5 (i.e., of the twenty largest public firms, ten have blockholders). Corporate law improves. As suggested by current theory and intuition then, five of the blocks diffuse in the public firms with blocks. Ownership for these five fully separates. Owners can at last sell out at full value, because they don’t have to worry about a future raider grabbing control and the concomitant private benefits. If that were the only move, then the standard index of diffusion would jump to .75. But consider the ten firms that were fully public. Five now can afford blockholders because the dissipation of value from infighting would decline, and the blockholder would add value. If this were the second 24 CORPORATE LAW’S LIMITS move, then the index would, when smoke cleared and all the separation adjustments were made, end up right back where it began, at .5. But that’s not all. Of the ten firms that were fully private (category 3), five owners decide that they can take the firm public and, under the newly- improved corporate law regime, investors readily buy up the stock. When this third set of transactions is completed, that nation would have five ten fully-public firms and fifteen public but block-held firms. Five firms would remain completely private. The index of diffusion would have dropped from .5 to .4. 30 Improving corporate law would thereby reduce the density of separation in that nation’s public firms. 31 And some fully public firms might go private, because they know that one barrier to cashing in on any improvement in the firm value—the difficulty of taking the firm public later—has been removed. (Public to private to public ownership again does happen in the United States.) To summarize: One effect of improving corporate law is that public firms with blockholders could more easily transit to fully diffuse firms. This is just what current theory, intuition, and policy-making would predict. But that’s only one of the effects; overall, improving corporate law has ambiguous effects on diffusion. This may not just be idle theory: “The mean percentage of common stock held by a [U.S.] firm’s officers and directors as a group rose from 13 percent in 1935 to 21 percent in 1995. Median holdings doubled from 7 30 . Because the standard index only looks at the 20 largest firms, we assume here that the size distribution is random. Hence, the top twenty would include 4 of the previously public firms and 4 of the old public but concentrated firms that went diffuse. The public but diffuse category would include 4 diffuse firms that took on a now more- trusted blockholder, 4 that stayed with their old structure, and 4 previously private firms that decided that could now handle an IPO but would stay with block ownership. 31 . In addition, some diffusely-held firms might go privat e. They didn’t do so before because of reasons that improved corporate law would eliminate: a) the firm’s charter barred going private because when corporate law was weak going private was too dangerous in its potential to rip-off public stockholders. And b) the engineers of the going private would have been reluctant to take their firm private, because they would have had more difficulty doing a later IPO when it became warranted. Exit barriers (i.e., via going public later) are entry barriers (i.e., to taking the firm private now). But with corporate law improved, these two considerations fade in importance. Bottom line: Improving corporate law could decrease the incidence of diffusion in public firms from .5 to .25. CORPORATE LAW’S LIMITS 25 percent to 14 percent.” 32 As corporate law improved in the United States in the 20 th century through better securities laws and enforcement, for example, from passable to very good, blockholding increased. * * * 32 . Holderness (2001), at 2. 26 CORPORATE LAW’S LIMITS T able 2. Indeterminate effect of better corporate law in rich nation. Type of firm ownership Diffuse public Blockholder but public Fully private Index of concentration of public firms Time 1. Country begins with serviceable but not excellent corporate law 1. Initial ownership distribution 10 10 10 .5 (10/20) Time 2. Corporate law improves 2a. 5 blockholders sell out 15 5 10 .75 (15/20) 2b. Caps, pills removed; 5 public firms get blocks 10 10 10 .5 (10/20) 2c. 5 fully private go public but blockholder remains 10 15 5 .4 (10/25) 2d. 5 fully public go fully private in LBO’s 5 15 10 .25 (5/20) More starkly: concentrated blockholders have two major roles inside the firm: they steal from stockholders and they monitor managers. Minority stockholders would see a trade-off and reduce the price they’d pay accordingly: pay more to the extent monitoring raises firm value, but less to the extent the blockholder steals from the minority. But if law limits the negative possibility ?less, or no, shareholder stealing, because law protects the minority stockholders ?then that good law should make stockholders more comfortable, not less comfortable , with blockholders. Private owners would feel more comfortable in selling some stock because they would be able to get full, pro rata value for it, rather than the discount that buyers would insist upon in bad-law regimes. Improving corporate law should, in such settings, all else equal, increase , not decrease, the incidence of blockhol d ing. This offsetting effect from improving corporate law doesn’t tell us that improving it is bad. (Getting more public firms for a nation and separating ownership from control is not inherently good.) Ownership choice and shareholder welfare expand by improving corporate law. So improving it is worthwhile. But one could not measure the increased quality of corporate law by measuring the change in the number of public firms and the density of CORPORATE LAW’S LIMITS 27 separation over time. Corporate law might improve, and its very improvement might diminish the density of separation . H. The Tight Limits to the Purely Legal Theory Thus, using poor minority protection from a controller’s diversion of value to explain why blocks persist in Russia is sufficient and convincing, but for Western Europe is fuzzy, or maybe wrong. If blocks persist, one cannot a priori know whether they persist because minority stockholders fear the contro l ler, or because they fear the managers , who might dissipate shareholder value if the controlling stockholder disappears. Even if better corporate law usually increases diffusion in rich nations with adequate but not outstanding corporate law (a proposition open to theoretical challenge 33 ), concentration might be due to high managerial agency costs and have little to due with core corporate law’s constraints on insider machinations. If distant shareholders fear unrestrained managers, the controller cannot sell stock at a high enough price and thus she keeps control to monitor managers or to run the firm. III. Its Limits: Data If we could measure the quality of corporate law, then we could see whether owne rship is concentrated where corporate law protects shareholders and diffuse where it does not. True, if diffusion correlated with high-quality law, the primacy of the law-as-cause thesis would not be proven: when ownership is made diffuse for some other reason (say, technology or, say, politics) then the diffuse owners may demand legal protections. Co r porate law might follow, not lead, market development. But if, among nations with satisfactory corporate law, ownership is still concentrated in several, we would need more than just the legal theory to explain the result. 33 . See section G, pp. 20 - 27 , supra. 28 CORPORATE LAW’S LIMITS A. Measuring Quality 1. Corporate law: what counts? Judging how well corporate law protects minority stockholders across nations by examining their corporate law is hard. Not only must one judge which laws are critical (how did Britain succeed without a derivative suit, the very institution that plaintiff- oriented counsel in the U.S. would cite as a sine qua non? and one that France, seen as a weak corporate law nation by American analysts, allows? 34 ), but interaction effects can make a rule that is a loophole in one nation into a roadblock in another. Or a protection might be missing, but an even stronger substitute might be present. Moreover, the rules-on-the- books could be identical in two nations but if the quality of enforcement (because of a corrupt, incompetent, or inefficient judiciary or reg u latory system) might make the bottom line protections differ greatly. Or practices not r e quired by a nation’s corporate law could protect shareholders: a legal index might look bad, but the reality could be the opposite if contract, corporate charter terms, or business practices counter-act a deficient corporate law. Undaunted by lawyers’ skepticism that one can qualitatively assess corporate law directly, finance-oriented students of corporate governance built legal indices for many nations. They have accomplished a major undertaking, one that should embarrass many (of us) corporate law professors who have not even attempted what the financial economists have completed. They have argued convincingly that corporate law institutions are weak in many third world and transition nations, that these weaknesses cripple securities markets. 35 These stu d ies have also been interpreted, less convincingly, as showing weak corporate law to be the primary culprit for the weak securities markets on the European continent. Not only do corporate players in France, Germany, and Scandinavia think 34 . Art. L225-252 C. Com.; Art. 200, Decree No. 67-236 of March 23, 1967, J.O. March 24, 1967, p. 2843 (France); Stengel (1998) (Germany). Germany has the derivative suit in theory (in that the company must bring suit when 10% of the stockholders seek that it do so), but in practice it just isn’t used. Id. 35 . La Porta et al. series, supra note 2, and the followers. Economically less developed countries have added reasons why they haven’t developed securities markets. Good securities and corporate rules might come with wealth, and not the other way around. CORPORATE LAW’S LIMITS 29 their corporate law is fine, but they sometimes proclaim its superiority in some dimensions over the Amer i can variety. 36 The indexers consciously do not seek to measure the bottom-line quality of each nation’s corporate law but uses a handful of proxies: Possibly the index focuses on rules that aren’t right at the core of shareholder protections, but rather on proxies for a total set of institutions that protect shareholders, a set for which there might be more direct measures. Improvement is possible. 37 Organized qualitative analysis cha l lenges legal academics’ preference for nuance and discussion. 38 Anecdotes abound, and one can find anecdotes of fleeced minority stockholders on both sides of the Atla n tic. 39 One can list differing rules, but it’s difficult to know a) which rules are substitutes and, hence, which countries truly have gaps in protection, b) which rules really count, c) the extent to which players follow announced rules, and d) whether the rules in focus are the kind that securities market 36 . Tunc (1982) (French law bans dangerous transactions that American judges weigh, balance, and sometimes approve); cf. Agnblad, Bergl?f, H?gfeldt & Svancar (2000) (not even anecdotes of insider machinations in Sweden). 37 . Wall Street lawyers might have reservations about heavily using preemptive rights, cumulative voting, and the minimum percentage needed to call a special shareholder meeting—items not likely to be near the top of most American lawyers’ lists of Delaware corporate law’s most important legal protections—and of partly abandoning Delaware law for the index. (The index uses Delaware corporate law except on the minimum percentage needed to demand a meeting.) Delaware allows firms to decide the issue by specifying a low percentage in their charter, a right that, I understand, firms rarely use. Sticking with Delaware here would have made Delaware corporate law protection look mediocre, when it’s probably pretty good. The point isn’t that Delaware is bad—the index probably hits the right bottom line—but that developing an accurate index is hard. For a critique more skeptical than mine of the index, one that is probably more in line with legal academic views than mine, see Detlev Vagts, Comparative Corporate Law—The New Wave , in F ESTSCHRIFT FOR J EAN -N ICOLAS D RUEY ( edited by Rainer Schweitzer and Urs Gasser) (forthcoming, 2002). Vagts argues that the coding judgments for the German index are incorrect. For instance, although German stockholders are viewed as unable to vote by mail, most send their instructions in to their bank (by mail) and the bank then votes the stockholders. Hence, German corporate law is “better” than the index suggests. 38 . E.g., Tunc (1982). 39 . Cf. Norris (2000); Stewart (1992), at 119-27 (machinations of Victor Posner in NVF, DWG, Pennsylvania Engineering, APL, Royal Crown and Sharon Steel). 30 CORPORATE LAW’S LIMITS players demand up front as necessary to build securities markets, or whether the rules are just the polish on financial markets that comes once deep securities markets exist for other reasons. Some rules are window- dressing, some really bind. Which is which? 2. Corporate law: The bottom line. Can we measure the bottom-line, overall quality of corporate law? If we knew the nation-by-nation average premium for control and could compare it to the value of the traded stock, one would have a bottom-line number for the value of control in a firm. In nations where the premium is high, we’d surmise corporate law or its enforcement is inferior; in nations where that premium over the price available to diffuse stockholders is low, we’d surmise corporate law is superior. Consider a firm worth $100 million, with a 51% blockholder who values that block at $60 million and minority stock that trades for an aggregate value of $40 million. If we can observe those numbers, we have roughly measured the value of control: the controller plausibly pays the 10% premium (measured as a percentage of total firm value) because he or she can divert 10% of the firm’s value from minority stockholders into his or her own pocket. If one could measure this diffe r ence across nations, then one would have a “bottom-line” number indicating the value of control. If the quality of corporate law were the, or a, principal determinant of separation, then nations with high gaps between the value of control and the value of the minority stock would have more concentrated ownership than nations where that gap is weak. 40 B. Data: Nations with Good Corporate Law but Without Separation 1. Market measures of the value of control . We have data on the value of a control block. Researchers have looked at the premium paid for 40 . The measurement of the private benefits of control and, hence, of law’s ability to keep those ben e fits low will be imperfect. Some of the premium could come from the cost of assembling a block. Some of the premium may come from the controller’s power to decide, say, when to sell, although the sale would be made at a fair price for all. If the transaction costs are high, then a pre-assembled block should command a premium because it side-steps the transaction costs. Hence, high premia may over-state law’s weakness because some fraction of the premium come from unrelated transaction costs, not from uncontrolled private benefits. CORPORATE LAW’S LIMITS 31 a voting block over the pre-trading price. In the United States, it was found to be 4% of the firm’s value. 41 For Italy parallel research suggests a premium of 25 to 30% or more, 42 a premium consistent with the quality-of- corporate-law theory (since ownership is concentrated there and corporate law apparently poor). But in Germany, the control block premium was recently, and surprisingly, found to be about 5% of the firm’s value, 43 inconsistent with the corporate law theory, because German ownership is quite concentrated. 44 To be sure here, the data could under-state the private benefits: ben e fits might have already been taken before the sale and, hence, the sales price wouldn’t r e flect them. And firms for which blocks are sold could be those with low private benefits, while those where diversion is high don’t trade. But even the reduced fact remains that for those blocks sold, the future private benefits are expected to be about equal to those expected in American block trades. So, the block premium in Germany is about 5% of firm value; that in the United States is 4%. We should pause at this finding for Germany. That number casts doubt on the pure-form of the law-driven theory, because Germany, the world’s third largest national economy, has very concentrated ow n ership. If control blocks trade at such a low premium, perhaps something else induces concentration to persist . An explanation for Germany is that German codetermination ?by which labor gets half of the seats in boardrooms of large firms ?fits snugly with concentrated shar e holding as a counter-balance in large, especially large smoke-stack, industries. 45 That 5% premium is less than the decline 41 . Barclay & Holderness (1989). 42 . Nicodano & Sembenelli (2000). See also Zingales (1994) (premium for voting stock). The Italian number comes from the voting premium for dual-class co m mon stock. More about that below. 43 . Franks & Mayer (2000), at 24. 44 . The premium is the difference between block price and the trading value of the diffusely held stock. American blocks traded at a 20% premium over he price of diffuse stock, for blocks of (typically) one-fifth of the firm’s stock. If the pr e mium represents what the controller can grab for itself, the blockholder would be able to grab 4% of the firm’s value (one-fifth of 20%). German blocks are larger, with many equal to half of the company’s stock. A premium of 10% for half of the company, the typical numbers, indicates the controller could grab 5% of the firm’s value for itself. 45 . Roe (1999). 32 CORPORATE LAW’S LIMITS in shareholder value measured when Germany e n hanced its co- determination statute in 1976 and increased employee representation in the boardroom from one-third to one-half. 46 The low German 5% control premium also shows why constructing an index for corporate law quality is so hard. To divert big value, the controlling shareholder typ i cally needs a big transaction ?a buyout, a merger, a related party sale of good or services. And to get a big transaction through a firm, one needs board approval and, hence, a co m pliant board. Because the majority stockholder in the United States typically a p points the entire board, it’s plausible to expect it to control a compliant board. But in Germany the blockholder can never control the full board, because German law mandates that labor get half of it, and the practice is that banks holding their brokerage customers’ proxies get some board seats. Other German corporate law features might be weak, thereby generating an appearance of weak corp o rate law protections in a cross- country index. But even so, the German blockholder may be stymied in pushing a related-party transaction through because he or she cannot control the full German board. Interaction effects impede putting our finger on one or two key features as indicative of whether technical corporate law is overall good or bad. Two other researchers use a differing methodology, but come up with similar data. They investigate the effects of ownership concentration in 100 of Germany’s public firms. They conclude that moving a German firm from diffuse to concentrated ownership would double the value of the firm’s shares. 47 Increases in ownership concentration typically benefit the diffuse stockholders. That’s a story that fits quite poorly with a poor- corporate-law theory, 48 but quite nicely with a managerial agency cost theory. 46 . FitzRoy & Kraft (1993); Gorton & Schmid (2000); Schmid & Seger (1998); but see Baums & Frick (1999); Frick, Speckbacher & Wentges (1999) . 47 . Edwards & Weichenrieder (1999). 48 . It could fit with the poor corporate law theory if a) corporate law was poor, but b) big blockholders desisted from transferring value to themselves, while c) small blockholders insisted on massively transferring such value. Plausible, yes (big blockholders incur more deadweight costs if the transfers demean total firm value), but this confluence would seem implausible as accounting for all, or even most of, the doubling of value to minority stockholders of blockholding. CORPORATE LAW’S LIMITS 33 All of the new German data on control block premium presents a counter-example, and a very big one, to the law-driven theory. Counter- examples are important, but perhaps there is some German-specific factor, not replicated elsewhere, that could make the theory ge n erally true, but just inapplicable in Germany. To check, we turn to other data. 2. Dual class common stock. Corporate law’s effectiveness can be roughly measured otherwise. Some firms issue dual class common stock. Class A stock votes, class B stock doesn’t, but both have the same dividend rights. (Variations abound.) A controller cannot reap ben e fits by controlling the class B stock, but can by controlling class A stock. Both are formally entitled to the same cash coming out from the company. If the value of class A stock is higher than that of B’s, we have a measure of the value that the controller can surreptitiously divert from outside shareholders to herself. Good law should keep that value low. If we could measure the differences across nations, we’d have an indicator measuring the quality of corporate law in controlling the ravages of a dominant stockholder, as one Nobel Prize winner sought to do a few years ago. 49 More specifically: Posit a $100 million company issues 500,000 shares of class A stock, which vote, and 500,000 shares of class B stock, which doesn’t. If the diversionary benefit of control is near zero, then the difference A stock and B stock should trade at the same price. If law is weak and control allows the controller to divert $10 million in value from the minority stoc k holders into his own pocket, then A stock should trade for about $60 million in the aggregate, the class B stock for $40 million. The quality-of-corporate-law theory would then predict that, in gross, as the private value of control went up, ownership concentration would increase. And vice versa: as the value of control decreased, contro l lers would loosen their grip and ownership would diffuse. Unpublished voting premium data has recently become available. Table I shows the voting premium in the world’s richer nations. Italy’s and France’s voting premium is high, America’s low—a difference consistent with the legal theory, as Italy is said to have poor protections 49 . Modgiliani & Perotti (1997 or 1998), at 525, who sought to prove that the dual class premium varied with the quality of a nation’s security market. They had a smaller, and less up-to-date sample of seven nation’s voting premia. The current dual class data better measures the premia and the implied quality of legal protections, but as we shall see shortly it still doesn’t predict the degree to which ownership separates from control. 34 CORPORATE LAW’S LIMITS and has concentrated owne r ship, and the United States the converse. But this new data increases the tension for the legal theory; Germany is a weak corporate law nation in the finance economists’ indices, but the dual class numbers here again show it protects non-voting stockholders rather well, vindicating defenders of the quality of German corporate law. And not just Germany: Four Scandinavian nations all have very concentrated ownership but protect minority stockholders well. Table 3: Voting premium and ownership separation 50 Country Voting premium Portion of large firms that are widely-held Australia 0.23 0.65 Canada 0.03 0.60 Denmark 0.01 0.40 Finland 0.00 0.35 France 0.28 0.60 Germany 0.10 0.50 Italy 0.29 0.20 Norway 0.06 0.25 Sweden 0.01 0.25 Switzerland 0.05 0.60 United Kingdom 0.10 1.00 United States 0.02 0.80 Source: Voting premium data comes from Nenova (2000); the ownership concentration data comes from La Porta et al., supra note 2. The percentage of widely-held firms for a nation is the percentage of the nation’s twenty largest firms that have a 20% or larger blockholder. This dual class premium data casts doubt on whether a uni-variable model is enough to explain the richer nations’ degree of ownership separation. True, further co n firmation, with data collected by other researchers, should be added. And the number of observations ?a dozen 50 . The voting premium measures the percentage excess of the trading price of voting stock over non-voting stock. Technical cautions for this data are raised and discussed infra at notes 51 and 56 and accompanying text (comparing price of minority of voting class to price of nonvoting stock). CORPORATE LAW’S LIMITS 35 or so of the richer nations ? isn’t enough to allow crosschecks and controls to weight competing factors. And the dual class data as measuring the value of control is soft. If the contro ller has a majority of the class A voting stock, then the researchers are observing the tra d ing value of the minority stockholders on the class A level, and comparing that data to the trading value of the non- voting class B stock. But the minority class A stockholder is not a controller, it just has a chance of sometime joining a control block. 51 Thus while this is the best data set available, it is not perfect data. (Data sets never are.) We can take comfort in that ancillary i n formation comports with the numbers. The American premium is low, and U.S. corporate and securities law is usually seen as highly protective. The Swedish premium is low as well, and Swedish researchers assert that there are not even anecdotal instances of contro l lers shifting value to themselves. Two Scandinavian researchers tell us that “the value of control does not derive from the possibility to expropriate the fringe of minority shar e holders … [but] has to be motivated by some other economic motives.” 52 Other Swedish researchers report that: Outside shareholders do not refrain [from] investing on the Stockholm Stock Exchange since 55% of the Swedish population own shares … and 33% of outstanding shares are owned by foreign investors … . [T]he ratio of the stock market capitalization held by m i nority shareholders in relation to GDP … is 0.51 for Sweden compared to 0.58 for the U.S. … [I]t is not likely that weak investor protection has hampered financial ma r ket development in Sweden …. 53 The other Scandinavian nations have similar reputations, and they also have low premiums. Moreover, the leading blockholding Swedish investor typically uses dual class stock, but not in a way that locks up control: the Wallenberg family holding company doesn’t take majority control but more typically ends up with 5% of the cash flow and 25% (not a m a jority) 51 . Statistical analyses could try to control this problem, by observing the value of the vote when control is incomplete, i.e., what’s the premium for the minority of stock A when there’s a 40% shareholder. The incomplete control means that the other stock’s vote could be decisive for a controlling coalition. See Nenova (2000); Zingales (1994). Such calculations are inherently imprecise. 52 . Bergstr?m & Rydqvist (1990) (emphasis supplied). 53 . Holmén & H?gfeldt (1999), at 39 (emphasis supplied). 36 CORPORATE LAW’S LIMITS of the votes, 54 leaving potential control in the other 75% of the votes. And although the German premium is low compared to the usual prejudice, it comports with the control block premium recently measured by two leading financial r e searchers. 55 The voting premiums in the world’s poorer nations, which one could believe to be in the process of developing securities markets, are high. 56 * * * To repeat a proviso: I hardly mean that this data tells us that high- quality corp o rate and securities law is irrelevant. Rather, it’s that some rich nations have high quality corp o rate and securities law but ownership still remains close and securities ownership does not diffuse. The point in this Part is not that good corporate law is irrelevant in the world’s richer nations ? it keeps the costs of running a big enterprise low ?but when it’s already pretty good, subtle gradations in its quality do not determine 54 . Leser & Rocco (2000). 55 . Franks & Mayer (2000), at 24. They attribute concentrated ownership to the high private benefits of control, although the benefit they found, 10%, is on the low- side world-wide. Other researchers recently found German bankers able to extract little in private benefits of control. Gorton & Schmid (2000), at 70. The density of dual class usage could indicate which propellant—private benefits or managerial agency costs—is central. If control is usually had via dual class stock (or pyramids) that might indicate that extraction of private benefits is primary. But this would be so only if the controller pulls his or her wealth out of the firm. If managerial agency costs are high, control with some financial commitment is important for firm value: if wealth constraints mean the controller cannot commit most of the family’s wealth and still influence managers, then dual class could be privately efficient for managerial agency cost reasons. 56 . And, as long as the “real” premium varies in tandem from nation-to-nation then, even if the data under-states the premium, country-by-country comparisons can be made. We can take some further comfort with the data in that a well-known winner of the Nobel Prize thought it the best available way to measure legal quality. He was, however, working with what has become stale data. Modgiliani & Perotti (1997 or 1998), 524-25. Data measuring the value of the vote can be corrected to reflect that the minority voting stock repr e sents the probably of joining a control group and not the direct value of control. The value of control rises or falls with the of the control block: in a country where 51% of the voting stock can control ever y thing, the value of minority voting stock should approach that of non-voting stock; but if there are two voting blocks of 40%, the minority voting stock’s value should reflect the value of control. Sophisticated tests can approximate the correction. See Zingales (1994). CORPORATE LAW’S LIMITS 37 whether ownership diffuses . Something else is more important, with the leading altern ative being that high managerial potential to dissipate precludes separation even if corporate law is good. 3. Control block premium. Other new data on corporate law quality is also suggestive, and not helpful to a pure corporate law thesis. The premium in a block sale could reveal the quality of the governing corporate law. Consider a controller who owns 50% of a $7,500 company with 100 shares, and sells her 50 shares not at $75 per share, but at $100 per share, when the dispersed stock trades at $50 per share. The $50 per share premium could be seen as measuring the private benefits of control, benefits that better corporate law would force the controller to share pro rata with all of the firm’s stockholders. The controller could plausibly siphon off $50/share x 50 shares in value, or $2,500 of the firm’s total value of $7,500. The premium is 33% of the company’s value. If instead the controller’s 50 shares sold for $80 per share while dispersed stock sold for $70 per share, then we could calculate that the market was expecting that the controller could siphon off $250 (from 50 shares x $10/share), or 3.3% of the firm’s value. As always, other considerations could be in play, but we would surmise that the first scenario is one of weak corporate law, the second one of (relatively) stronger corporate law. (We’d have to be careful that we weren’t reading data in which only the control sale was regulated, while rampant shifts occurred elsewhere: If the relevant corporate law forced a mandatory bid for dispersed stock upon a control shift, the premium might be low, but incumbent controllers could still be otherwise shifting value. So the better data would measure the results before mandatory bids became common; or would measure control sales not subject to the mandatory bid, i.e., sales of less than the typical 30% trigger.) Two financial economists have just accumulated such data, in an important undertaking. Here, in the last two columns, is what they have for the world’s richer nations: 38 CORPORATE LAW’S LIMITS Table 4. Control premia and ownership separation. Country (1) Widely-held at 20% for Med Corps (2) Mean premium as a fraction of total equity (3) Mean premium from col. 3, "corrected" for industry effects Australia 0.30 0.02 0.04 Austria 0.00 0.38 0.34 Canada 0.60 0.01 -0.04 Denmark 0.30 0.08 0.03 Finland 0.20 0.10 -0.01 France 0.00 0.02 0.04 Germany 0.10 0.10 0.02 Italy 0.00 0.37 0.30 Japan 0.30 -0.04 -0.04 Netherlands 0.10 0.02 0.02 Norway 0.20 0.01 0.04 Sweden 0.10 0.06 0.03 Switzerland 0.50 0.06 -0.06 United Kingdom 0.60 0.02 0.04 United States 0.90 0.02 0.04 med20 Block premium/ equity value Source: Alexander Dyck & Luigi Zingales, Why are Private Benefits of Control so Large in Certain Countries and What Effect Does this Have on their Financial Development (working paper, 2001). Even a cursory examination shows no persistent pattern. If the sample only had, say, Austria, Italy, the U.S., and the U.K., we would get a nice pattern for the corporate law thesis: Austria and Italy have concentrated ownership and high sale-of-control premiums; the U.S. and the U.K. have low premiums and low ownership concentration. For these four nations, there’s an excellent fit for a quality-of-corporate-law thesis, suggesting it has some importance. But the other wealthy nations also have low sale-of-control premiums, despite that most of them have concentrated ownership. Take a look at the Scandinavian nations again: low premium for control, high concentration. And take a look at Germany and Switzerland: fairly low premia, but concentrated ownership for Germany and middling- CORPORATE LAW’S LIMITS 39 concentration for Switzerland. And in this sample, both the Netherlands and France, countries with concentrated ownership, don’t look bad in protecting minority stockholders. Overall, the “sign” of the relationship is as corporate law theory would predict, but the significance is low, and the portion of the variation explained (12%) is quite low. (The last column adjusts the premium for industry type: some industries are more likely to “naturally” protect dispersed shareholders, other industries are riskier for them. The adjustments refine the index but do not radically change the results: Austria and Italy still have the largest premiums; the U.S. and the U.K. are still at the protective end of the spectrum. Some of the middle range nations change rank: Germany and Finland look more protective; most other changes meander.) The following shows the relationships graphically and the statistical results. (Similar results came from seeing the extent to which the block premium predicted stock market capitalization: Pretty good results overall, but the results are “driven” by Italy and Austria.) Graph 1. Block premium vs. ownership dispersion . U.S. Jap. U.K. Can. Swi. Fra. Austral. Nor. Neth. Ger.Swe. Fin. Den. Italy Aus. -0.20 0.00 0.20 0.40 0.60 0.80 1.00 -0.1 0 0.1 0.2 0.3 0.4 0.5 Block premium/Equity value Widely-held at 20% for Med Corps Technical data: med20 v. control premium Regression y = -1.02x + .36 Adj R-Sq -.12 t-stat -1.71* * Not signi ficant. P value = .11. 40 CORPORATE LAW’S LIMITS Graph 2. Block premium vs. ownership dispersion (without Italy and Austria. Can. U.K. U.S. Swi. Den. Fin. Ger.Swe. Nor. Neth. Fra. Austral.Jap. 0.00 0.20 0.40 0.60 0.80 1.00 -0.05 0 0.05 0.1 0.15 Block premium/Equity value Widely-held at 20% for Med Corps Technical data: med20 v. control premium (without Italy and Austria) Regression y = -1.58x + ,38 Adj R-Sq -.02 t-stat -0.84* * Not significant. P value = .42. If control premia were our only data, and our only way to measure the quality of corporate law, we’d be driven to conclude that that corporate law quality only weakly explained the variation in ownership dispersion and the strength of securities markets in the world’s richest nations. Something else is as, or more, important. 4. And the not-so-rich nations? One might observe that many poorer nations have decrepit corporate law institutions. This is true, and possibly weak corporate law is keeping them back, but the coincidence of bad law and a bad economy does not tell us enough. To learn that, say, Afghanistan, has poor corporate law doesn’t tell us whether its weak economy is primarily due to its weak corporate law or to its other weak institutions. If the other institutions, particularly the other property rights institutions are decrepit, these may be the critical debilities preventing Afghanistan from developing the wealth and sufficiently complex private institutions that get it ready to start needing public firms and ownership diffusion. Only then , when it gets that far, will we be able to tell whether weak corporate law holds it back. The omitted variable might be weak CORPORATE LAW’S LIMITS 41 property rights institutions generally, with weak corporate law institutions just being a visible, and perhaps minor, surface manifestation of the deeper weakness. 5. Enforcing contracts. Bad law sufficiently explains weak securities markets where law is so weak that basic contract s cannot be enforced ?as they cannot be in co n temporary Russia, many transition economies, and significant parts of the less developed world ? thereby rendering complex corporate institutions impossible. 57 This is important because a) the quality of contract and corporate law ought to correlate and b) much that is useful in corporate law can be built out of good contract law, either directly by public authorities or indirectly by private parties. Many of the same nations that by measurement have good corporate law also have good contract law. All the Scandinavian nations, Germany, and several other cont i nental European countries enforce contract as well as the United States does. 58 This casts more doubt on whether the quality of corporate law thesis explains enough of why ownership separ a tes or doesn’t in the world’s richer nations. Contract law seems good, and corporate law, which also seems good, is in many dimensions a special form of contract law. Nations that can build one should be able to build the other. And the rudiments of a corporate law can, in primitive form, be built out of contract. Studies of business climate are consistent: continental Europe and the Anglo-Saxon basic business institutions are generally seen as equally business-friendly, but the cont i nental European labor markets are seen as much less business-friendly (and that unfriendliness can raise managerial agency costs, as we shall see in the next section). 59 Nor is it logically correct to assume that where corporate rules are weakly enforced, that weakness is the primary cause for weak stock markets in nations that have already built satisfactory contract and property 57 . Black & Kraakman (1996) (Russia); Sachs & Pistor (1997), at 3. 58 . O’Driscoll, Holmes & Kirkpatrick (2001), at 18 (Denmark, Finland, Germany, Norway, Sweden, and the United States protect private property and contract strongly and have a largely efficient legal systems). The index, a crude one, purports to measure both property rights and “the ability of individuals and businesses to enforce contracts.” Id. at 57. Cf . Levine (1999), at 14-15, 20 (risk that government will not respect a contract it has signed: low for the United States, but lower for France, Germany, and Scandinavia). 59 . Sachs & Warner (1998), at 24. 42 CORPORATE LAW’S LIMITS institutions. Were the demand for diffuse ownership sufficiently strong in such nations, investors and firms could try to build the institutions needed for good securities markets. If societies that successfully built other complex business and legal inst i tutions, especially those that effectively enforce commercial contracts, did not try to build these corporate law institutions, then a deeper reason might explain why they did not try. Thus, one could synthesize the legal and managerial agency costs theories into a two-step argument: When corporate law, contract law, and court systems are decrepit, public firms will not emerge, because the system fails to protect minority stockholders. This describes many third- world and transition nations. But when either contract or basic corporate law becomes satisfactory, as it is in several western European nations and the United States, then whether a nation builds on what it has (by writing complex contracts, by further impro v ing corporate law, or by developing the ancillary institutions such as stock exchanges or effective intermediaries), becomes a question of whether the underlying potential for low managerial agency costs to sharehol d ers makes it profitable for the players to do so. C. What Beyond Law is Needed for Separation in the Wealthy West? I have argued here that corporate law could be fine and ownership might still not separate from control. Corporate law does not readily control the important managerial agency costs of dissipating shareholder value. For firms for which these costs to shareholders are high, ownership will not readily separate from control. For nations where these costs are systematically high, separation is more rare than where these costs are low. It is more rare even if corporate law quality is high. Co r porate law is insufficient to induce separation. Other conditions have to be met. Here for the sake of completeness I briefly outline the other conditions. 1. Economic preconditions. Economic and technological conditions must yield a d e mand for public firms with lots of capital. If the economy is too poor to have such a demand (many nations still are in this category) or if the reigning technologies do not demand large economies of scale, then public firms won’t be sought. Moreover, the distrib u tion of wealth CORPORATE LAW’S LIMITS 43 and income must be flat relative to the demand for large firms. 60 Strongly competitive product markets keep managerial agency costs lower than weakly competitive product markets. In the latter, managers have much slack; in competitive markets, they don’t. 2. Political preconditions. Some modern societies are rich, have technological demands for large firms, but their politics stymies separating ownership from control. In strong social democracies, politics drives a wedge between shareholders on the one side, and managers and employees on the other side. Politics there presses firms to expand, to avoid down- sizing, and to avoid disrupting employment conditions. These are just the kind of goals that unconstrained managers were said to have in the United States, just the kind of things that the arsenal of agency-cost-reducing tools is designed to handle. And the tools that make managers tolerably loyal to shareholders in the United States ?transparent accounting, incentive compensation, hostile takeover, and strong shareholder primacy nor m s ?are denigrated in the strong social democracies. Social policies there may raise the well-being of most people, but they would do so without much ownership separation in large firms. In contrast, a more conservative nation typically wouldn’t drive a wedge between shareholders on the one side, and employees and managers on the other side. It would facilitate, or at least allow, shareholders and managers to ally themselves. When they are loosely allied, ownership and control can separate. In technical terms, managerial agency costs if unremi t tingly high can induce concentration to persist, rendering corporate law quality secondary. In societies of the first type, concentrated shareholding is capital’s next best means to control managers, and it persisted even after the other economic conditions for separation were met. Moreover, it has persisted even in such nations that have good quality corporate law . It budged in recent years only as these nations’ social democratic polit i cal parties shifted rightward. 60 . Thus the United States has today a skewed distrib ution of wealth and income, but it has a very high demand of large-scale firms. In the nineteenth century the distribution was flatter, and, with the railroads creating a single huge market, the demand for large firms with widely gathered capital was even higher. If technology flattens and shrinks firms, making them more “pocket-sized” then wealthy people can co n trol them more easily than if the optimal scale is very large. 44 CORPORATE LAW’S LIMITS 3. Social preconditions. Some societies are so in turmoil that private institutions ca n not be built. Reputations are not worth developing, because no one is sure to be able to use the reputation once built. Private-ordering via, say, a stock exchange won’t work, because investors lack confidence in the exchange and fear who might capture it. But once a society has sufficient regularity so that reputations, private instit u tions, and, if need be, corporate law can be built, then if political and economic conditions are otherwise ripe, large enterprises can arise and ownership can separate from control. If there’s suff i cient social and political regularity, corporate law or substitutes can arise to do the job. D. Data on Explanations Beyond Law We have seen agency-cost- based theoretical limits to the legal theory. Can we measure these limits, however crudely? If we can find a proxy that measures one of these other conditions, then we could see whether the two explain the degree of separation equally well, or whether adding the political variable substantially strengthens the explanatory power of the tests. Two institutions affect managerial agency costs and vary from nation- to-nation. One is conventional, one not. Competitive product markets are conventionally said to reduce managerial agency costs. And, less conventionally, politics affects managerial agency costs, in that strongly social democratic nations historically pressed managers to side with employees when managers made operating decisions that would either favor employees or shareholders but couldn’t favor both. (The idea here is not that, say, business schools and leadership skills in these countries are weak—although perhaps schools there do not strongly inculcate shareholder primacy—but that managers in such countries are pressed to run the firm other than purely in shareholder interests. Hence, A M , broadly interpreted, is higher in nations where such political pressures are higher.) The quality of corporate law helps to predict the degree of ownership separation, as Table 3 shows. But the other social and political variables—measures of the political pressure on managers— also predict separation well, or better. When we add political variables to the corporate-law-driven “model,” we get much stronger predictive power than we do with law alone . Law alone doesn’t do as well as law and politics. CORPORATE LAW’S LIMITS 45 The “F-test” measures whether a factor significantly adds power in explaining the results. Take the legal theory: a qualitative index of corporate law quality, one used frequently now in the finance literature, predicts ownership separation plausibly. (The voting premium data described on p. 34 , however, doesn’t predict separation as nicely as an index of corporate law rule; see Table 5, suggesting that the common index might be imperfect. But let’s stay with the standard index.) Using several measures of ownership separation and stock market strength (large firms, medium firms, 10% as the blockholder cut-off, 20% as the cut-off), we see in line 1 of Table 6 that the commonly-used qualitative corporate law index predicts the depth of separation, with several results statistically significant. But when we add political variables, we get a significantly higher explanatory power than with law alone, as the next two lines of Table 5 show, than the legal test alone. A plausible correlation becomes much stronger, and the F-statistics (shown via asterisks) reveal that half of the results of added explanatory power from politics are quite significant statistically. Even if law is important, politics is independently quite important too. Roughly this suggests that controlling insider thievery—the type of costs of public firms that law can reach—gets us half-way to making public firms viable. But if the political environment impedes managerial- shareholder alliances, the second type of managerial agency costs would rise, and ownership could not easily separate from control. * * * Table 6 correlates a set of political, ownership, legal, and competitive variables. All three political variables predict ownership separation and of the depth of a nation’s securities market. The two legal variables also predict separation and stock market depth, as does the measure of monopoly strength. 46 CORPORATE LAW’S LIMITS Table 5. Correlation matrix Political indicators Legal indicators Competitive indicator Dispersion and strength of stock market indicators Political place Employ- ment pro- tection Gini La Porta Law Voting premium Monopoly mark-up Stock/ mktcap/GDP Widely- held at 20% Political place 1.00 Employment protection -0.41 1.00 Gini after-tax 0.49 -0.53 1.00 La Porta Law 0.39 -0.95 0.65 1.00 Voting premium -0.26 0.50 0.28 -0.39 1.00 Monopoly mark-up -0.57* 0.35 -0.42 -0.41 0.39 1.00 Stock mkt. capitalization/GDP 0.55* -0.56* 0.64* 0.70* -0.23 -0.80** 1.00 Widely-held at 20% for Medium Firms 0.67* 0.84** 0.67* 0.87** -0.46* -0.57* 0.73 1.00 * Significant at .05 level. ** Significant at .01 level. (Not all significant correlations are highlighted.) Sources for data: Political place comes from Cusack (1997); employment protection from OECD (1994); GINI from Deininger & Squire (1996); La Porta law and widely-held at 20% from La Porta (1999); voting premium from Nenova (2000); monopoly markup from Martins, Scarpetta & Pilat (1996); and stock-market capitalization from OECD (1998). The correlation matrix shows politics persistently correlating with dispersion: the more conservative the nation, the more dispersed is ownership. The matrix also shows the good quality corporate law correlates with dispersion: the higher the quality of corporate law, the more dispersed is ownership. Sorting out which factor is the base-line cause is hard. But analysis can illuminate some issues. We might want to know if the two potential causal institutions—corporate law quality and politics—are co-linear: Is corporate law good everywhere that politics is conservative or middle-of- the-road? And do left nations uniformly have weak corporate law protections for minority shareholders? As it turns out, adding political measures to the legal explanation strongly increases explanatory power. CORPORATE LAW’S LIMITS 47 Table 6 first correlates an index of corporate law quality with ownership separation, showing that it explains much separation. But explanatory power jumps dramatically when we add political measures.. Table 6. F-test: Law & politics 20% Large 20% Medium 10% Large 10% Medium Stock mkt. Cap /GDP La Porta Law Adj. R-Squared .38 .43 .35 .12 .11 La Porta + Political Place Adj. R-Squ ared .65*** .64*** .62*** .28* .40*** La Porta + Gini Adj. R-Squared .73*** .57** .79*** .38** .43*** * Significant at.05 level (i.e., adding politics significantly adds to law in explaining separation). ** Significant at .01 level. *** Significant at the .001 level. Conclusion: The Quality of Corporate Law and its Limits I have here neither denied the value of strong corporate law that protects distant stockholders, nor denigrated its usefulness in building efficacious business enterprises, nor sought to refute its academic utility in explaining some key aspects of corporate differences around the world. It is valuable in protecting distant shareholders, as it is often the lowest costs means to protect them. It is useful in thereby building big firms. If it doesn’t exist, the society needs substitute institutions. And it is helpful in explaining corporate structures in the world’s developing and transition economies, many of which cannot establish good corporate rules of the game. I have instead sought to map out the limits to the quality-of-corporate- law argument. High quality corporate law is insufficient to induce ownership to separate from control in the world’s richest, most economically-advanced nations. Technolog i cally-advanced nations in the wealthy West can have the potential for fine corporate law in theory, and several have it in practice, but ownership will not separate from control if managerial agency costs are high. And managerial agency costs, unlike insider self-dealing, are not closely connected with corporate law. Indeed corporate law’s business judgment rule has corporate law avoid dealing with managerial agency costs. 48 CORPORATE LAW’S LIMITS Today’s reigning academic theory leaves too many unanswered questions. Why doesn’t strong, pro-minority shareholder corporate law lead to more blockholders, not fewer, because distant minority stockholders would have less to fear of controllers’ trampling as law improved? Why do some rich nations lack even a single anecdote of over- reaching behavior from controllers’ ne v ertheless lack strong separation? Why are there so many rich nations with, by measurement and anecdote, low private benefits of control, high quality corporate institutions, and much minority stock, yet without ownership separation? The quality of conventional corporate law does not fully explain why and when owne r ship concentration persists, because corporate law does not even try to directly control managerial agency costs from dissipating a firm’s value . The American business judgment rule keeps courts and law out of basic business decisions and that is where managers can lose, or make, the really big money for shareholders. Non-legal institutions control these costs. In nations where those other institutions, such as product competition or incentive compensation, fail or do less well, managerial dissipation would be higher and ownership cannot as easily separate from control as it can where dissipation is lower. Corporate law quality can be high, private benefits of control low, but if managerial agency costs from dissipation are high, separation will not proceed. Even if we believed law to be critical to building these other institutions, the analysis would persist because different laws support the agency cost controlling institutions (antitrust and product market competition; tax law and incentive compensation, etc.). A nation doesn’t control insider machinations and motivate managers equally well; and to the extent it does one better than the other, concentration and diffusion are affected: The diffusion decision is based on the sum of private benefits of control and managerial agency costs. Even if traditional corporate law drives private benefits to zero, concentration should persist if managerial agency costs are high. Data is consistent. Several nations have, by measurement, good corporate law, but not much diffusion and hardly any separation. These nations also have a high potential for managerial agency costs: relatively weaker product market competition and relatively stronger political pressures on managers to disfavor shareholders CORPORATE LAW’S LIMITS 49 The quality of a nation’s corporate law cannot be the only explanation for why diffuse Berle-Means firms grow and dominate. Perhaps, for some countries at some times, it’s not even the principal one. 50 CORPORATE LAW’S LIMITS Bibliography Agnblad, Jonas, Erik Bergl?f, Peter H?gfeldt & Helena Svancar. "Ownership and Control in Sweden---Strong Owners, Weak Minorities, and Social Control." Stockholm: Stockholm School of Economics, Feb. 10, 2000. Barclay, Michael J. & Clifford G. Holderness. "Private Benefits from Control of Public Corporations." 25 Journal of Financial Economics 371 (1989). _______ & Dennis P. Sheehan. "The Block Pricing Puzzle." Rochester, N.Y.: Simon School of Business, University of Rochester, March 2001. (Working Paper No. FR 01-05.) Baums, Theodor & Bernd Frick. "The Market Value of the Codetermined Firm." In Employees and Corporate Governance, edited by Margaret M. Blair & Mark J. Roe, 206. Washington, D.C.: Brookings Institution Press, 1999. Bayer v. Beran , 49 N.Y.S.2d 2 (N.Y. Sup. Ct. 1944). Bebchuk, Lucian Arye. "A Rent-Protection Theory of Corporate Ownership and Control." Cambridge, Mass.: Harvard Law and Economics Working Paper, 1999. (Discussion Paper No. 260.) ________ & Allen Ferrell. “Federalism and Corporate Law : The Race to Protect Managers from Takeovers.” 99 Columbia Law Review 1168 (1999). _______, Reinier Kraakman & George Triantis. "Stock Pyramids, Cross- Ownership, and Dual Class Equity: The Creation and Agency Costs of Separating Control from Cash Flow Rights." In Concentrated Corporate Ownership, edited by Randall K. Morck, 295. Chicago: University of Chicago Press, 2000. ________ & Mark J. Roe. “A Theory of Path Dependence in Corporate Ownership and Governance.” 52 Stanford Law Review 127 (1999). Becht , Marco & Ailsa R?ell . "Blockholdings in Europe: An International Comparison." 43 European Economic Review 1049 (1999). Bergstr?m, Clas & Kristian Rydqvist . "The Determinants of Corporate Ownership : An Empirical Study on Swedish Data." 14 Journal of Banking and Finance 237 (1990). Bertrand, Marianne & Sendhil Mullainathan. "Agents with and without Principals." Princeton, N.J.: Industrial Relations Section, Princeton University, February 2000. (Working Paper No. 430.) CORPORATE LAW’S LIMITS 51 Bishop, Joseph W., Jr. "Sitting Ducks and Decoy Ducks: New Trends in the Indemnification of Corporate Directors and Officers." 77 Yale Law Journal 1078 (1968). Black, Bernard S. "The Core Institutions That Support Strong Securities Markets." 55 Business Lawyer 1565 (2000). _______ & Reinier Kraakman. "A Self-Enforcing Model of Corporate Law." 109 Harvard Law Review 1911 (1996). Carlin, Wendy & Colin Mayer. "Finance, Investment and Growth." London, Oxford: University College and Said Business School, Oxford, October 18, 1998. Coffee, John C. "The Future as History: The Prospects for Global Convergence in Corporate Governance and Its Implications." 93 Northwestern University Law Review 641 (1999). Cusack, Thomas R. “Partisan Politics and Public Finance : Changes in Public Spending in the Industrialized Democracies, 1955-1989.” 91 Public Choice 375 (1997). Deininger , Klaus & Lyn Squire. “A New Data Set Measuring Income Inequality. 10 World Bank Economic Review 565 (1996). Delaware Code Annotated ( Michie ). DeMott, Deborah A., “Current Issues in Tender Offer Regulation: Lessons from the British.” 58 NYU L. Rev. 945 (1983). Demsetz, Harold. "The Structure of Ownership and the Theory of the Firm." 26 Journal of Law and Economics 375 (1983). Dooley, Michael P. & E. Norman Veasey . "The Role of the Board in Derivative Litigation: Delaware Law and the Current ALI Proposals Compared." 44 Business Lawyer 503 (1989). Dyck , Alexander & Luigi Zingales. Why are Private Benefits of Control so Large in Certain Countries and What Effect does this Have on their Financial Development (working paper, 2001). Edwards, Jeremy S. & Alfons J.Weichenrieder, Ownership Concentration and Share Valuation: Evidence from Germany (working paper, 1999). Fama, Eugene F. "Agency Problems and the Theory of the Firm." 88 Journal of Political Economy 288 (1980). 52 CORPORATE LAW’S LIMITS FitzRoy, Felix R. & Kornelius Kraft. "Economic Effects of Codetermination." 95 Scandinavian Journal of Economics 365 (1993). France. Art. 200. Decree No. 67-236 of March 23, 1967. Journal Officiel de la République Fran?aise, March 24, 1967, 2858. France. Art. L225-252 Code de Commerce. (Art. 245. Law No. 66-537 of July 24, 1966 Journal Officiel de la République Fran?aise, July 26, 1966, 6420.) Franks, Julian & Colin Mayer. "Ownership and Control of German Corporations." Review of Financial Studies (forthcoming). Frick, Bernd, Gerhard Speckbacker & Paul Wentges. "Arbeitnehmermitbestimmung und moderne Theorie der Unternehmung." 69 Zeitschrift für Betriebswirtschaft 745 (1999). Gagliardi v. Trifoods Int'l , 683 A.2d 1049 (Del. Ch. 1996). Gorton, Gary & Frank A. Schmid. "Class Struggle Inside the Firm: A Study of German Codetermination." Cambridge, Mass.: National Bureau of Economic Research, October 2000. (Working Paper No. 7945.) _______. "Universal banking and the performance of German firms." 58 Journal of Financial Economics 29 (2000). Holderness, Clifford G., “A Survey of Blockholders and Corporate Control.” FRBNY E CON . P OL . R EV ., ____ (2001) (forthcoming). Holmén, Martin & Peter H?gfeldt. "Corporate Control and Security Design in Initial Public Offerings." Stockholm: Stockholm School of Economics, Dec. 15, 1999. Iskander, Magdi, Gerald Meyerman, Dale F. Gray & Sean Hagan, Corporate Restructuring and Governance in East Asia, 36 Finance & Development: A Quarterly magazine of the IMF, No. 1, at 1 (March 1999). Joy v. North , 692 F.2d 880 (2d Cir. 1982). La Porta, Rafael, Florencio Lopez-de-Silanes & Andrei Shleifer. “Corporate Ownership Around the World, 54 Journal of Finance 471 (1999). _______ & Robert Vishny. "Investor protection and corporate governance." 58 Journal of Financial Economics 3 (2000). _______. "Law and Finance." 106 Journal of Political Economy 1113 (1998). CORPORATE LAW’S LIMITS 53 _______. "Legal Determinants of External Finance." 52 Journal of Finance 1131 (1997). Leser, Eric & Anne-Marie Rocco. " Les Wallenberg veulent faire cohabiter capitalisme familial et mondialisation ." Le Monde (Paris), November 24, 2000, 23. Levine, Ross. "Law, Finance, and Economic Growth." 8 Journal of Financial Intermediation 8 (1999). Martins, Joaquim Oliveira, Stefano Scarpetta & Dirk Pilat. “Mark-up Pricing Ratios in Manufacturing Industries: Estimates for 14 OECD Countries.” Paris: OECD Econ. Dep’t Working Paper No. 162, 1996. Miller, Geoffrey. “Political Structure and Corporate Governance : Some Points of Contrast Between the United States and England.” 1998 Columbia Business Law Review 51 (1998). Modigliani, Franco & Enrico Perotti. "Protection of Minority Interest and the Development of Security Markets." 18 Managerial and Decision Economics 519 (1997). _______. "Security Versus Bank Finance: The Importance of a Proper Enforcement of Legal Rules." Cambridge, Mass.: MIT Sloan School of Management, 1998. Nenova, Tatiana. "The Value of Corporate Votes and Control Benefits: A Cross- Country Analysis." Cambridge, Mass.: Harvard University, Sept. 21, 2000. Nestor, Stilpon. Corporate Governance Trends in the OECD Area: Where Do We Go From Here? Paris: OECD. Working paper. 2000. [ check date] _______. International Efforts to Improve Corporate Governance: Why and How? Paris: OECD. Working paper, 2001. Nicodano , Giovanna & Alessandro Sembenelli . “Private Benefits, Block Transaction Premiums and Ownership Structure.” Turin, Italy: University of Turin. Working paper, Jan. 2000. Norris, Floyd. "Perelman's Plan: Take Profits While Public Owners Suffer." The New York Times, November 24, 2000, C1. O'Driscoll, Gerald P., Kim R. Holmes & Melanie Kirkpatrick. 2001 Index of Economic Freedom. Washington, D.C.: The Heritage Foundation, 2001. OECD. Ad hoc Task Force on Corporate Governance, Principles of Corporate Governance. Paris: OECD. 1999. 54 CORPORATE LAW’S LIMITS _______. “Corporate governance: Getting it right together—and in Russia,” The OECD Observer. Paris: OECD. Summer 1999a, 5. _______. The OECD Jobs Study: Evidence and Explanations—Part II—The Adjustment Potential of the Labour Market. Paris: OECD. 1994. _______. Financial Market Trends. Paris: OECD. Feb. 1998. Rock, Edward B. & Michael L. Wachter. "Islands of Conscious Power: Law, Norms, and the Self-Governing Corporation." 149 University of Pennsylvania Law Review 1619 (2001). Roe, Mark J. "Rents and Their Corporate Consequences." 53 Stanford Law Review 1463 (2001). _______. "The Quality of Corporate Law Argument and Its Limits: History's Gaps." Working paper, 2001. ________. “Political Preconditions to Separating Ownership from Corporate Control.” 53 Stanford Law Review 539 (2000). ________. “Chaos and Evolution in Law and Economics.” 109 Harvard Law Review 641 (1996). ________. Strong Managers, Weak Owners: The Political Roots of American Corporate Finance. Princeton, N.J.: Princeton University Press, 1994. Sachs, Jeffrey D. & Andrew M. Warner. "Executive Summary." In The Global Competitiveness Report 1998, 24. Geneva: World Economic Forum, 1998. _______ & Katherina Pistor. "Introduction: Progress, Pitfalls, Scenarios, and Lost Opportunities." In The Rule of Law and Economic Reform in Russia, edited by Jeffrey D. Sachs & Katharina Pistor, 1. Boulder, Colo.: Westview Press, 1997. Schmid, Frank A. & Frank Seger. " Arbeitnehmermitbestimmung, Allokation von Entscheidungsrechten und Shareholder Value." 5 Zeitschrift für Betriebswirtschat 453 (1998). Shleifer, Andrei & Lawrence H. Summers. "Breach of trust in hostile takeovers." In Corporate Takeovers: Causes and Consequences, edited by Alan J. Auerbach, 33. Chicago: University of Chicago Press, 1988. _______ & Robert W. Vishny. "Large Shareholders and Corporate Control." 94 Journal of Political Economy 461 (1986). Shlensky v. Wrigley , 237 N.E.2d 776 (Ill. App. 1968). CORPORATE LAW’S LIMITS 55 Smith v. Van Gorkom , 488 A.2d 858 (Del. Sup. Ct. 1985). Smith, D. Gordon. "The Shareholder Primacy Norm." 23 Journal of Corporate Law 277 (1998). Stein, Jeremy C. "Efficient Capital Markets, Inefficient Firms: A Model of Myopic Corporate Behavior." 104 Quarterly Journal of Economics 655 (1989). Stengel, Arndt. "Directors' Powers and Shareholders: A Comparison of Systems." 1998:2 International Company and Commercial Law Review 49 (1998). Stewart, James B. Den of Thieves. New York: Simon & Schuster, 1991. Tunc, André. "A French Lawyer Looks at American Corporation Law and Securities Regulation." 130 University of Pennsylvania Law Review 757 (1982). Vagts, Detlev. Comparative Corporate Law—The New Wave. In Festschrift for Jean-Nicolas Druey, edited by Rainer Schweitzer and Urs Gasser. Zurich: Schulthess Juristische Medien . Forthcoming, 2002). Witherell, William. Corporate governance: A basic foundation for the global economy. The OECD Observer. Paris: OECD. Summer 2000. 24. World Bank, Corporate Governance: A Framework for Implementation. Washington, D.C.: World Bank. 2000. [Confirm date and publication information.] Zingales, Luigi. "The Value of the Voting Right: A Study of the Milan Stock Exchange Experience." 7 Review of Financial Studies 125 (1994).