《法律与金融》课程教学资源（阅读文献）A Political Theory of Corporate Finance

A POLITICAL THEORY OF AMERICAN CORPORATE FINANCE* Mark I.Roe** INTRODUCTION Why is the public corporation-with its fragmented shareholders on the stock excha the d omin ant form of ente Since Berie and Means,the conventional corporate law story begins with technology dictating large enterprises with capital needs so great that even a few wealthy individuals cannot provide enough.These enterprises consequently must draw capital from many dispersed shareholders.Shareholders diversify their own holdings,further fragm nting ownership. This ation of a huge enterprise ,concentrated management,and dispersed,diversified stock holders shifts corporate control from shareholders to managers.Man- gerscan pursue their own agenda,at times to the detriment of the survived because it best bala needs.In a Darwinian evolution,the large public firm mitigated the managerial agency problems with a board of directors of outsiders,with a managerial headquarters of strategic planners overseeing the operat- isions,and with ncentive Co. Hostile takeovers,proxy contess and the threat managers. Pomentedd adapted.They solved enough of the governance problems created by the large unwieldy structures needed to meet the huge capital needs of technology.In the conventional story,the large public f modern technolo ient respo onse to the economics of orga Iargue here that the public corporation is as much a political adap- tation as an economic or technological necessity.The size and technol- Tmereeented w nted ownership of the large public corporation;the most prominent alternative concen- t©l99l.Mark I.Roe Professor of Law,Columbia Law School.I benefited from presentations at Lucketer.Chicngo Columbia.Harvard:McCill.and OcI stein, Coffee, Andy Rutten,Susan Tobias,and John Wiley. 1.A.Berle G.Means,The Modern Corporation and Private Property (rev.ed. 1967)(original edition published in 1932). 10

A POLITICAL THEORY OF AMERICAN CORPORATE FINANCE* MarkJ. Roe** INTRODUCTION Why is the public corporation-with its fragmented shareholders buying and selling on the stock exchange-the dominant form of enter￾prise in the United States? Since Berle and Means, the conventional corporate law story begins with technology dictating large enterprises with capital needs so great that even a few wealthy individuals cannot provide enough. These enterprises consequently must draw capital from many dispersed shareholders. Shareholders diversify their own holdings, further fragmenting ownership. This combination of a huge enterprise, concentrated management, and dispersed, diversified stock￾holders shifts corporate control from shareholders to managers. Man￾agers can pursue their own agenda, at times to the detriment of the enterprise.' In the classic story, the large public firm survived because it best balanced the problems of managerial control, risk sharing, and capital needs. In a Darwinian evolution, the large public firm mitigated the managerial agency problems with a board of directors of outsiders, with a managerial headquarters of strategic planners overseeing the operat￾ing divisions, and with managerial incentive compensation. Hostile takeovers, proxy contests, and the threat of each further disciplined managers. Fragmented ownership survived because public firms adapted. They solved enough of the governance problems created by the large unwieldy structures needed to meet the huge capital needs of modern technology. In the conventional story, the large public firm evolved as the efficient response to the economics of organization. I argue here that the public corporation is as much a political adap￾tation as an economic or technological necessity. The size and technol￾ogy story fails to completely explain the corporate patterns we observe. There are organizational alternatives to the fragmented ownership of the large public corporation; the most prominent alternative is concen- * © 1991, MarkJ. Roe. ** Professor of Law, Columbia Law School. I benefited from presentations at Berkeley, Chicago, Columbia, Harvard, McGill, and UCLA, and from comments from Lucian Bebchuk, Bernie Black, Victor Brudney, Marvin Chirelstein, Jack Coffee, Alfred Conard, Jeffrey Gordon, Joe Grundfest, Cliff Holderness, Howell Jackson, Michael Jensen, Lou Lowenstein, Henry Manne, Robert Mundheim, Roberta Romano, Andy Rutten, Susan Tobias, and John Wiley. 1. A. Berle & G. Means, The Modern Corporation and Private Property (rev. ed. 1967) (original edition published in 1932)

1991] A POLITICAL THEORY OF THE CORPORATION trated institutional ownership,a result prevalent in other countrie But American law and politics deliberately diminished the power of fi- nancial institutions to hold the large equity blocks that would foster serious oversight of managers,making the modern American corpora- tion adapt to the political terrain.The modern corporation's origin lies in technology,economics,and politics Shareholder control of managers arises when the owner holds a large block of stock.Individuals rarely have enough money to buy big blocks.Institutional investors do.But law creates barriers to the insti- tutions'taking big blocks.Banks,the institution with the most money can own stock. Mutual funds generally can ot own cor ntrol blo cks c stock.Insurance companies can put only a fragment of their invest- ment portfolio into the stock of any one company.Pension funds own stock,but they also face restrictions.More importantly,corporate man- agers control private pension funds,not the other way around. And we have just exhausted the major financial institutions in America;none can readily and without legal restraint control an indus- trial company.That is the first step of my argument:law prohibits or raises the cost of institutional influence in industrial companies. The second step is to examine the politics of corporate financial structure.Many legal restraints had public-spirited backers;some rules would be those that wise regulators,unburdened by politics,would reach.But many important rules do not fit into this public-spirited mold.Examining financial regulation through the lens of the new icchoice literature reve als a complex and new repeatedly foreclosing alternatives to the Berle-Means corporation. Opinion polls show Americans mistrust large financial institutions with ulated d have aly h d of Wall Street con- trolling industry. Politicians responded to that distrust by enacting rules restricting private accumulations of power by financial institu- tions.Various interest groups also benefited from fragmentation;Con- gress and the administrative agencies also responded to them. During the SEC's formative years,its chairman said Ishould and will bel restricted to under Insofar as manag ent [and]for ustrial policies the banker will be superseded.The financial n er which he has exercised in the past over such processes will pass into other hands.? An inquiry into the interplay of political ideology and financial in stitutions follows.We shall examine the ideas of opinion leaders and ing us to speculate on a political expl nation for corporat e str cture 2.W.O.Douglas,Democracy and Fi rly every important Wall Street vestment banker speec n193%

1991] A POLITICAL THEORY OF THE CORPORATION 11 trated institutional ownership, a result prevalent in other countries. But American law and politics deliberately diminished the power of fi￾nancial institutions to hold the large equity blocks that would foster serious oversight of managers, making the modem American corpora￾tion adapt to the political terrain. The modern corporation's origin lies in technology, economics, and politics. Shareholder control of managers arises when the owner holds a large block of stock. Individuals rarely have enough money to buy big blocks. Institutional investors do. But law creates barriers to the insti￾tutions' taking big blocks. Banks, the institution with the most money, cannot own stock. Mutual funds generally cannot own control blocks of stock. Insurance companies can put only a fragment of their invest￾ment portfolio into the stock of any one company. Pension funds own stock, but they also face restrictions. More importantly, corporate man￾agers control private pension funds, not the other way around. And we have just exhausted the major financial institutions in America; none can readily and without legal restraint control an indus￾trial company. That is the first step of my argument: law prohibits or raises the cost of institutional influence in industrial companies. The second step is to examine the politics of corporate financial structure. Many legal restraints had public-spirited backers; some rules would be those that wise regulators, unburdened by politics, would reach. But many important rules do not fit into this public-spirited mold. Examining financial regulation through the lens of the new pub￾lic choice literature reveals a complex and new political story, of law repeatedly foreclosing alternatives to the Berle-Means corporation. Opinion polls show Americans mistrust large financial institutions with accumulated power and have always been wary of Wall Street con￾trolling industry. Politicians responded to that distrust by enacting rules restricting private accumulations of power by financial institu￾tions. Various interest groups also benefited from fragmentation; Con￾gress and the administrative agencies also responded to them. During the SEC's formative years, its chairman said: [T]he banker... [should and will be] restricted to ... under￾writing or selling. Insofar as management [and] formulation of industrial policies . . . the banker will be superseded. The financial power which he has exercised in the past over such processes will pass into other hands.2 An inquiry into the interplay of political ideology and financial in￾stitutions follows. We shall examine the ideas of opinion leaders and political actors, and the content of major political investigations, lead￾ing us to speculate on a political explanation for corporate structure: 2. W.O. Douglas, Democracy and Finance 32, 40-41 (1940) (Douglas stunned his audience of nearly every important Wall Street investment banker when he gave this speech in 1937)

COLUMBIA LAW REVIEW [Vol.91:10 Main Street America did not want a powerful Wall Street.Laws dis- couraging and prohibiting control resulted. I.BERLE AND MEANS A.The Classic Story Berle and Means'vision is central to corporate law scholarship. Their story is straightforward:"[T]he central mass of the twentieth- century American economic revolution [is a]. massive collectiviza- tion of cline of Pa oted to pro oduction with [an]accom pa nying de- d ecision-making and and a]mass dissociation of wealth from active management.This restructuring turns corporate law on its head:stockholders,the owners,become powerless.The stockholder's vote"is of diminishing importance as the number of shareholders in each corporation increases. diminishing in fact to ne gible 0 ance as the orpora ome giants.As the om rmely me Aoehela number of stockhol shareholders as a "passive"investment while managers control the corporation. The problem is not solely the separation of shareholders from managers or,as Berle and Means put it,the "massive dissociation of wealth from active management."The problem is atomization.Most public companies are held by thousands of shareholders,each with only a small stake.As a result,an active shareholder cannot capture all of the ng nagers. The costs of monitoring are ecomes ed,studies th te prise,or sits on the board of directors and thereby takes the risk of enhanced liability.But since monitoring gains would be divided among all shareholders,most fragmented shareholders rationally forego involvement. This disincentive to monitor has costs enterprises are run imper fectly as managers pursue their own agenda.Managers want high sala- ries,nice offices,and,if the firm is large enough,corporate jets. Occasionally they take corporate opportunities for themselves.Much ial ursue operating policies thati ealth. T ney want to grow ond an effi cient size for the prestige,power,and salary tha t comes with running a larger organization.Too large,the enterprise slows down;it runs less efficiently.5 25时o a shareholder with ldveto empire-building acquisiti ons 3.A.Berle G.Means,supra note 1,at xxv;see id.at 4-7. 4.Id.at xix. 5.See id.at 112-14;O.Williamso

COLUMBIA LAW REVIEW Main Street America did not want a powerful Wall Street. Laws dis￾couraging and prohibiting control resulted. I. BERLE AND MEANS A. The Classic Story Berle and Means' vision is central to corporate law scholarship. Their story is straightforward: "[T]he central mass of the twentieth￾century American economic revolution [is a] ... massive collectiviza￾tion of property devoted to production, with [an] accompanying de￾cline of individual decision-making and control[, and a] massive dissociation of wealth from active management."'3 This restructuring turns corporate law on its head: stockholders, the owners, become powerless. The stockholder's vote "is of diminishing importance as the number of shareholders in each corporation increases-diminishing in fact to negligible importance as the corporations become giants. As the number of stockholders increases, the capacity of each to express opin￾ions is extremely limited." 4 As a result, corporate wealth is held by shareholders as a "passive" investment while managers control the corporation. The problem is not solely the separation of shareholders from managers or, as Berle and Means put it, the "massive dissociation of wealth from active management." The problem is atomization. Most public companies are held by thousands of shareholders, each with only a small stake. As a result, an active shareholder cannot capture all of the gains from monitoring managers. The costs of monitoring are borne by the shareholder who becomes involved, studies the enter￾prise, or sits on the board of directors and thereby takes the risk of enhanced liability. But since monitoring gains would be divided among all shareholders, most fragmented shareholders rationally forego involvement. This disincentive to monitor has costs: enterprises are run imper￾fectly as managers pursue their own agenda. Managers want high sala￾ries, nice offices, and, if the firm is large enough, corporate jets. Occasionally they take corporate opportunities for themselves. Much more perniciously, many managers pursue operating policies that di￾minish social wealth. They want the enterprise to grow beyond an effi￾cient size for the prestige, power, and salary that comes with running a larger organization. Too large, the enterprise slows down; it runs less efficiently.5 A large shareholder might make a difference: a shareholder with 25% of the company's stock could veto empire-building acquisitions, 3. A. Berle & G. Means, supra note 1, at xxv; see id. at 4-7. 4. Id. at xix. 5. See id. at 112-14; 0. Williamson, The Economics of Discretionary Behavior: Managerial Objectives in a Theory of the Firm 34-35, 79 (1964). [Vol. 91:10

19911 A POLITICAL THEORY OF THE CORPORATION 13 question managerial performance,and in the extreme instance replace the managers.Since it could capture a quarter of the gains,the large shareholder would often have the incentive to act,an incentive frag- mented shareholders lack. B.Modern Complaints About the Berle-Means Corporation odewitersto sharchoders, who they say va sh ly,to the deriment of nation. They claim that managers would take the long-view but are stymied by Wall Street's short-run goals;companies shun long-term in- vestment,and industry underinvests in research and development.6 The lon ng/short co ntroversy posits a market failure.After all,insti- tutions should know how to d long-term value to present value While no one has demonstrated that the long/short phenomenon exists, consider the effect of fragmentation.Fragmentation diminishes the value to a single shareholder of assessing long-run soft information not reflected in the hard numhers of current financial statements.To as- sess long-term research nd devel s often equ res st consultants with specialized expertis f there are sca economies in evaluation,then a fragmented shareholder may rationally decline to in- cur the cost.?Theoretically,the investor would invest in research and buy up undervalued stocks.That is,the size of shareholding would in- crease to reflect any scale economies in information.But.as I shall document in Part I,institutions have ceilings on their stock positions. Consider the difficulty in conveying soft long-term information to a fragmented marketplace;dispersed investors cannot cheaply distin- Wr g With Wall St ,5-6.9,56-63,76 1881 e,e.gAb What's Ha 1yAug.1980,at67,68-70,Lp in the of Fina Corporatism,136 U.Pa.L.Rev.1,7-9,23-25(1987);Dobrzynski,More Than Ever,It's Ma gement fo the Short Term,E Wk. 4.1986,at 82;Drucker,A Crisis of Wall S ept.5 986 colRid at32 6, of the de of managers [who push]c orate ma agement to focus on short-term results."). 5ed byTenology AreLot onaly Wall analyst ng 0nge 8.Furthermore,managers may feel constrained to follow average shareholder opin ion,for fear that the average will sell out and prompt a change in control.An isolated institution with a har of the enterprise will ac ingly be una protect managers fom gation of the go priect.SceT.Hoshi,A.Kashy p&D.Scharfstein Co orate Structure,Liquidity,and Investment:Evidence from Japanese Panel Dat 9-15,19-250M firm af Working Paper No..Sep (da liation with banks in Japan mitigates some markets

1991] A POLITICAL THEORY OF THE CORPORATION 13 question managerial performance, and in the extreme instance replace the managers. Since it could capture a quarter of the gains, the large shareholder would often have the incentive to act, an incentive frag￾mented shareholders lack. B. Modern Complaints About the Berle-Means Corporation Modem writers blame corporate mismanagement on shareholders, who they say value short-run profits excessively, to the detriment of the nation. They claim that managers would take the long-view but are stymied by Wall Street's short-run goals; companies shun long-term in￾vestment, and industry underinvests in research and development. 6 The long/short controversy posits a market failure. After all, insti￾tutions should know how to discount long-term value to present value. While no one has demonstrated that the long/short phenomenon exists, consider the effect of fragmentation. Fragmentation diminishes the value to a single shareholder of assessing long-run soft information not reflected in the hard numbers of current financial statements. To as￾sess long-term research and development plans often requires staff or consultants with specialized expertise. If there are scale economies in evaluation, then a fragmented shareholder may rationally decline to in￾cur the cost. 7 Theoretically, the investor would invest in research and buy up undervalued stocks. That is, the size of shareholding would in￾crease to reflect any scale economies in information. But, as I shall document in Part II, institutions have ceilings on their stock positions. 8 Consider the difficulty in conveying soft long-term information to a fragmented marketplace; dispersed investors cannot cheaply distin- 6. See, e.g., L. Lowenstein, What's Wrong With Wall Street 1, 5-6, 9, 56-63, 76 (1988); Hayes & Abernathy, Managing Our Way to Economic Decline, Harv. Bus. Rev., July-Aug. 1980, at 67, 68-70; Lipton, Corporate Governance in the Age of Finance Corporatism, 136 U. Pa. L. Rev. 1, 7-9, 23-25 (1987); Dobrzynski, More Than Ever, It's Management for the Short Term, Bus. Wk., Nov. 24, 1986, at 82; Drucker, A Crisis of Capitalism, Wall St. J., Sept. 30, 1986, at 32, col. 3; Altman & Brown, A Competitive Liability, Ridding Wall Street of a Short-Term Bias, N.Y. Times, June 1, 1986, § 3, at 3, col. 1 (former Assistant Secretary of the Treasury decries "a new cadre of professional money managers . . . [who push] corporate management to focus on short-term results."). 7. Cf. Wilke, Among Those Baffled by Technology Are Lots of Stock Analysts, Wall St. J., Sept. 28, 1989, at Al, col. 6 (stock analysts have trouble evaluating long-term research). 8. Furthermore, managers may feel constrained to follow average shareholder opin￾ion, for fear that the average will sell out and prompt a change in control. An isolated institution with a small share of the enterprise will accordingly be unable to "protect" managers from other shareholders; both managers and the fragmented institution see￾ing an undervalued long-term opportunity will fear that other shareholders will force termination of the good long-term project. See T. Hoshi, A. Kashyap & D. Scharfstein, Corporate Structure, Liquidity, and Investment: Evidence from Japanese Panel Data 9-15, 19-25 (MIT Working Paper No. 2071-88, Sept. 1988) (data suggests industrial firm affiliation with banks in Japan mitigates some informational defects of securities markets)

14 COLUMBIA LAW REVIEW [Vol.91:10 guish egoistic empire-building from a high net present value project. And managers are wary of revealing proprietary information to scat- tered shareholders:9 without good information,even highly paid,tech- nically com petent analysts c nnot evaluate long-term research and opment In contra cial analysts with gene ric skills can eas ily evaluate short-term financial data. Complaints also are heard that shareholders do not monitor man- agers.Managers build empires and pursue bad strategies without reholder in lenc e of theh tervention until matters are so out-of-hand that the vio- or the of the leve raged buyout results.Persistent shareholder involvement could lead to intervention before bloodshed. the Bne perniclous monitoring and intormnation stores depend on oration.Imploring an owner of $10 million of the stock。 a $10 billi be sha reholder is us less. The sharehol lder can capture only 1/1,000 of the corporat ' gain.The shareholder should rationally decline to invest $100,000 of his or her time and wealth,even if that $100,000 would yield a $100 million gain for the corporation But a 25%shareholder could invest millic ons of dollars in monitor ing and in costly evaluation of soft research and development informa- tion.Beneficial monitoring does not even depend on the monitors being better than managers at most tasks,as I discuss in Part IV.The 25%shareholder could in fact generally be worse,but could still help if it were art en ough to resist tion until corporate results w poor,and had some skill in assessing wheth the poor re were due to poor management. Moreover.management would more willingly reveal proprietary information to the large long-term shareholder,who has the incentive to maintain secrecy. The large sha reholder would protect secrets and protect managers from outsi ders who would guess truly pro fita ble long-run investments.If an owner could take 25%stock positions in a few firms,it might find it worthwhile to assemble a staff with the expertise to monitor effectively.Finally,managerial motivation to pur- wealth maxim managers knew tha change 95 was already if the owner was inactive day-to-day;managers would realize tha deviation from long-run profitability could activate the large-block holder,or ac tivate an outsider who would find it easy to buy the control block.Fi- aeconomcague that large block sharcholding improves corporate operai Markets and Hierarchies:Analysis and Antitrust Implica 10.See Holderness &Sheehan,The Role of Majority Shareholders in Publicly Held 24(1989).But see I emsetz

COLUMBIA LAW REVIEW guish egoistic empire-building from a high net present value project. And managers are wary of revealing proprietary information to scat￾tered shareholders; 9 without good information, even highly paid, tech￾nically competent analysts cannot evaluate long-term research and development. In contrast, financial analysts with generic skills can eas￾ily evaluate short-term financial data. Complaints also are heard that shareholders do not monitor man￾agers. Managers build empires and pursue bad strategies without shareholder intervention until matters are so out-of-hand that the vio￾lence of the hostile takeover or the instability of the leveraged buyout results. Persistent shareholder involvement could lead to intervention before bloodshed. These pernicious monitoring and information stories depend on the Berle-Means corporation. Imploring an owner of $10 million of the stock of a $10 billion industrial firm to be an active shareholder is use￾less. The shareholder can capture only 1/1,000 of the corporation's gain. The shareholder should rationally decline to invest $100,000 of his or her time and wealth, even if that $100,000 would yield a $100 million gainfor the corporation. But a 25% shareholder could invest millions of dollars in monitor￾ing and in costly evaluation of soft research and development informa￾tion. Beneficial monitoring does not even depend on the monitors being better than managers at most tasks, as I discuss in Part IV. The 25% shareholder could in fact generally be worse, but could still help if it were smart enough to resist intervention until corporate results were poor, and had some skill in assessing whether the poor results were due to poor management. Moreover, management would more willingly reveal proprietary information to the large long-term shareholder, who has the incentive to maintain secrecy. The large shareholder would protect secrets and protect managers from outsiders who would second guess truly profita￾ble long-run investments. If an owner could take 25% stock positions in a few firms, it might find it worthwhile to assemble a staff with the expertise to monitor effectively. Finally, managerial motivation to pur￾sue wealth-maximizing operating policies could change for the better if managers knew that a 25% block was already assembled, even if the owner was inactive day-to-day; managers would realize that deviation from long-run profitability could activate the large-block holder, or ac￾tivate an outsider who would find it easy to buy the control block. Fi￾nancial economists argue that large block shareholding improves corporate operations. 10 9. See 0. Williamson, Markets and Hierarchies: Analysis and Antitrust Implica￾tions 35-37, 97-98 (1975). 10. See Holderess & Sheehan, The Role of Majority Shareholders in Publicly Held Corporations, 20J. Fin. Econ. 317, 333-45 (1988); Wruck, Equity Ownership Concen￾tration and Firm Value, 23 J. Fin. Econ. 3, 14-16, 23-24 (1989). But see Demsetz & [Vol. 9 1: 10

1991]A POLITICAL THEORY OF THE CORPORATION 15 s11 Few i are position.Although banks,insurance companies.mutual funds and pension funds have enough money,they do not take large positions. yfinancial institutions do not because would fail to be believe they But e rom the past tantalizing prospects General Motor the largest American industrial corpora tion-today has no controlling shareholder.But once it had one.Du Pont owned 25%of GM until the courts ordered antitrust divestiture. In the 1920s,an ineptly managed GM neared bankruptcy.Its manage ment was reinvigorated by neither a p stile takeo mor a leverage oxy fight,nor a h t in react on to th e prospect ofa takeover,but by the intervention of Detroit.rcorganized the company and installd w anaerSm larly,the J.P.Morgan investment bank monitored many of the country's railroads when reorganizing them at the turn-of-the-century.13 Certainly con entrated control by financial institutions is imagina ble:Japanese and German corporate ownership is quite concentrated; their financial institutions are more actively involved in their companies than are financial institutions in the United States.14 Daimler-Benz,the attomotve concem that is the largest Wrenan industrial company,has a 28%shareholder. Deutsche Ba nk ma erial inf ng re the company withou ban Dameseniot management without the replaced of a hostile takeover.15 The result in Germany- -like GM during the du Pont intervention-should be contrasted with the recent result in the United States when financial institutions(in this instance,public pe the selecti n of the next rebuffed them i6 If Ame ncan financia institu tions had a stake in GM approaching that controlled by Deutsche Bank in Daimler-Benz.that rebuff would have been difficult. Lehn,The Struc Econ,he structure of Corporate Ownership:Causes and Consequences,98 J.Pol. com:M.Ei laaiopanisnhnememaiharchodnn T1、 g4559 (1976(6 udies show that 10%blocks.while n ore thar one-third of the various s les of large firms). 12.A.Chandler S.Salsbury,Pierre S.du Pont and the Making of the Modern on V Ca 0(1971) 371-72987:RC80 onal Bankers 1854-1913,at 44 00 14.See Aoki,Toward an Economic Model of the Japanese Firm,28 J.Econ.Litera- ture 1,14(1990);Ingersoll,The Banker Behind the Shakeup at Daimler-Benz,Bus.Wk., Juy21986 36 r Muscle,N.Y.Times,Feb.11,1990,,at 13, col.2

1991] A POLITICAL THEORY OF THE CORPORATION 15 Few of the largest public firms are controlled by a holder of a sub￾stantial block of shares. " Few individuals have the wealth to take that large a position. Although banks, insurance companies, mutual funds, and pension funds have enough money, they do not take large positions. Possibly financial institutions do not because they believe they would fail to be effective. But examples from the past offer tantalizing prospects. General Motors-the largest American industrial corpora￾tion-today has no controlling shareholder. But once it had one. Du Pont owned 25% of GM until the courts ordered antitrust divestiture. In the 1920s, an ineptly managed GM neared bankruptcy. Its manage￾ment was reinvigorated by neither a proxy fight, nor a hostile takeover, nor a leveraged buyout in reaction to the prospect of a takeover, but by the intervention of its large shareholder. Pierre du Pont moved to Detroit, reorganized the company, and installed new managers.' 2 Simi￾larly, theJ.P. Morgan investment bank monitored many of the country's railroads when reorganizing them at the turn-of-the-century.' 3 Certainly concentrated control by financial institutions is imagina￾ble: Japanese and German corporate ownership is quite concentrated; their financial institutions are more actively involved in their companies than are financial institutions in the United States.' 4 Daimler-Benz, the automotive concern that is the largest German industrial company, has a 28%o shareholder, Deutsche Bank. When managerial infighting re￾cently left the company without clear direction, the bank replaced Daimler-Benz' senior management without the organizational violence of a hostile takeover.' 5 The result in Germany-like GM during the du Pont intervention-should be contrasted with the recent result in the United States when financial institutions (in this instance, public pen￾sion funds) sought to influence the selection of the next chairman of GM; GM management rebuffed them.16 If American financial institu￾tions had a stake in GM approaching that controlled by Deutsche Bank in Daimler-Benz, that rebuff would have been difficult. Lehn, The Structure of Corporate Ownership: Causes and Consequences, 93 J. Pol. Econ. 1155, 1173-76 (1985) (comparing companies with fragmented shareholders to companies with concentrated stockholders). 11. M. Eisenberg, The Structure of the Corporation: A Legal Analysis 45-52 (1976) (studies show that 10%o blocks, while not unheard of, were found in no more than one-third of the various samples of large firms). 12. A. Chandler & S. Salsbury, Pierre S. du Pont and the Making of the Modem Corporation 457-560 (1971). 13. See V. Carosso, The Morgans: Private International Bankers 1854-1913, at 371-72 (1987); R. Chernow, The House of Morgan 44 (1990). 14. See Aoki, Toward an Economic Model of theJapanese Firm, 28J. Econ. Litera￾ture 1, 14 (1990); Ingersoll, The Banker Behind the Shakeup at Daimler-Benz, Bus. Wk., July 27, 1987, at 36. 15. Ingersoll, supra note 14, at 36. 16. Lorsch, Funds Should Flex Their Muscle, N.Y. Times, Feb. 11, 1990, § 3, at 13, col. 2

16 COLUMBIA LAW REVIEW [Vol.91:10 Natural selection sugge adaptive mechanisms to mitigate the agency costs.Survival of the fit- test organizational structures makes the modern corporation as effi- cient as it can be.17 The political perspective raises doubts,since as I argue next in anizational form At best,th d States rn corporation is cient as it can be"only within powerful political constraints. 1 do not here argue that institutional control would have been bet- ter for the United States.Serious disadvantages would arise as well.8 Rather than prescription,I want first to understand d how politics dete mined corporate structure. II.THE FINANCIAL INSTITUTIONS:WHO HAS THE MONEY AND WHAT CAN THEY DO WITH ITP pes of financial institutions dominate: mutual fun ension funds ercial banks, They have as sets respectively,of $2 trillion,54 billion,$18 trillion,and trillion. Clearly these financial institutions have enough assets to influence orporati ons.But rules anti-n ng contro. blocks.Demonstrating this requires regulatory detail,but the rules can be summarized quickly for those who do not want the detail:Banks and bank holding companies were repeatedly prohibited from owning con- trol blocks of stock or from affiliation with investment banks that did. Insurance companies r q uite so e pr hibi ed fro nds mt eploymors hanrt o o rules still restrict their abili contro concentrated position;buying more than 5%of a company triggers on- er&Friedland,The Literature of Economics:The Case of Berle 19 18.See infra Part IV (conflicts of inter st,banking instability,politically intolerable accumulations of power,"capture"by management).And do not think that there were tional monitoring that ha e ended.Most institutions never coul k position.Fo the fe that cou d,some damaged shar lders by See L.Brandeis.Other P -And Ho oh Ban Part III. 19.Board of Govern s of the Federal Reserve System,Flow of Funds Accounts: gures for year-end 1989) ot sa 0 of the as in financial intermediaries.See Board of Governors of the Federal Reserve System,Bal- ance Sheets for the U.S.Economy 1945-1989,at 42,line 4 (Apr.1990).Mutual funds nd money market Ban also had 679 bilhon m Fed ina33（1989 tutions Examinatio】

COLUMBIA LAW REVIEW Natural selection suggests that if the separation of ownership from control created agency problems, the organization that survived found adaptive mechanisms to mitigate the agency costs. Survival of the fit￾test organizational structures makes the modem corporation as effi￾cient as it can be. 17 The political perspective raises doubts, since as I argue next, important organizational forms are prohibited or made costly in the United States. At best, the modem corporation is "as effi￾cient as it can be" only within powerful political constraints. I do not here argue that institutional control would have been bet￾ter for the United States. Serious disadvantages would arise as well.' 8 Rather than prescription, I want first to understand how politics deter￾mined corporate structure. II. THE FINANCIAL INSTITUTIONS: WHO HAS THE MONEY AND WHAT CAN THEY Do wiTH IT? Four types of financial institutions dominate: commercial banks, mutual funds, insurance companies, and pension funds. They have as￾sets, respectively, of $3.2 trillion, $548 billion, $1.8 trillion, and $1.9 trillion. 19 Clearly these financial institutions have enough assets to influence large corporations. But portfolio rules, anti-networking rules, and other fragmenting rules disable them from systematically taking control blocks. Demonstrating this requires regulatory detail, but the rules can be summarized quickly for those who do not want the detail: Banks and bank holding companies were repeatedly prohibited from owning con￾trol blocks of stock or from affiliation with investment banks that did. Insurance companies were for quite some time prohibited from owning any stock, and portfolio rules still restrict their ability to take control. Mutual funds cannot deploy more than a fraction of their portfolio in a concentrated position; buying more than 5% of a company triggers on- 17. See, e.g., Stigler & Friedland, The Literature of Economics: The Case of Berle and Means, 26J.L. & Econ. 237, 258-59 (1983). 18. See infra Part IV (conflicts of interest, banking instability, politically intolerable accumulations of power, "capture" by management). And do not think that there were halcyon days of institutional monitoring that have ended. Most institutions never could take a large stock position. For the few that could, some damaged shareholders by "si￾phoning" off value, or helped shareholders only by organizing monopolies and cartels. See L. Brandeis, Other People's Money-And How the Bankers Use It 33 (1914); infra Part III. 19. Board of Governors of the Federal Reserve System, Flow of Funds Accounts: Financial Assets and Liabilities (Mar. 1990) (preliminary figures for year-end 1989). Bank assets are only those of FDIC-insured commercial banks, not savings associations, which have another $1.7 trillion. These institutions account for nearly 90o of the assets in financial intermediaries. See Board of Governors of the Federal Reserve System, Bal￾ance Sheets for the U.S. Economy 1945-1989, at 42, line 4 (Apr. 1990). Mutual funds exclude closed-end and money market mutual funds. Banks also had $679 billion in discretionary trust funds at year-end 1988. Federal Financial Institutions Examination Council, Trust Assets of Financial Institutions: 1988, at 33 (1989). [Vol. 91:10

1991] A POLITICAL THEORY OF THE CORPORATION 17 erous rules.Pension funds are less restricted,but they are fragmented; rules make it difficult for them to operate jointly to assert control.Pri- vate pension funds are under management control:they are not con- structed for a palace revolution in which they would assert control ov their manager al bosses Do not be deceived by the detail that follows:A pattern is there 三 that emerges with clarity in Part III when we look at legislative history, ough to cont rol op erating compan destroyed the most prominent alternative to the Berle-Means corpora- tion:concentrated institutional ownership. A.Banks and Commerce Banks have half of the financial assets available for investment e the mo ial ownership is simple: s cannot own stoc 1.The Historical Separation of Banks from Commerce. Some early American banking charters expected banks to be in commerce.The Manhattan Bank was also a waterworks;some manufacturers had cor- achaters that included the right to open a bank.But other cha opied English ch which sepa rated banks By the mid-nin omThe important National Bank Act of 1863 gave national banks eteenth century,separation was the only limited powers.Control of an industrial company was out of the question.When controversy arose over whether banks could own stocks,the Supreme Court resolved that question against the banks: the pow t wn stock s no t listed, ngly it wa 2.The Glass-Ste granted.21 gall Act.- To avoid the direct prohibition on stock dealing,commercial banks dealt in securities through affiliates. In 1933,Congress believed that the failure of these stock affiliates dam- aged banks.and the resulting bank failures caused.not refected.the Dep ession The ng Glass-Ste eagall Act of 1933 pr ohibited bank tes f owning and dealing in curties,there everng com mercial banks from investment banks. Today,prohibitions on com- mercial banks'underwriting and affiliation with companies dealing in securities are breaking down.2s But the prohibition on bank ownership 20.B.Hammond.Banks and Politics in America from the Revolution to the Civil war149-55(1957). k v.Ker nedy,167 U.S.362,366-67(1897);National Bank Ac of Fet th s5⊙，12U.s.C.s3351988). Banking Banking Act of (Glass-Stcagal)6,12U..C.4(scventh)() 23.Securities Indus.Ass'n v.Board of Governors of the Fed.Reserve Sys.,468 U.S 207,214-21 (1984)(upholding FRB's authorization of Bank of America to acquire

1991] A POLITICAL THEORY OF THE CORPORATION 17 erous rules. Pension funds are less restricted, but they are fragmented; rules make it difficult for them to operate jointly to assert control. Pri￾vate pension funds are under management control; they are not con￾structed for a palace revolution in which they would assert control over their managerial bosses. Do not be deceived by the detail that follows: A pattern is there that emerges with clarity in Part III when we look at legislative history, popular ideology, the power of interest groups, and the views of opin￾ion leaders. Politics never allowed financial institutions to become powerful enough to control operating companies. American politics destroyed the most prominent alternative to the Berle-Means corpora￾tion: concentrated institutional ownership. A. Banks and Commerce Banks have half of the financial assets available for investment. They are still where the money is. But legal analysis of bank power in industrial ownership is simple: Banks cannot own stock. 1. The Historical Separation of Banks from Commerce. - Some early American banking charters expected banks to be in commerce. The Manhattan Bank was also a waterworks; some manufacturers had cor￾porate charters that included the right to open a bank. But other char￾tering authorities just copied English charters, which separated banks from commerce. By the mid-nineteenth century, separation was the norm. 20 The important National Bank Act of 1863 gave national banks only limited powers. Control of an industrial company was out of the question. When controversy arose over whether banks could own stocks, the Supreme Court resolved that question against the banks: the power to own stock was not listed, accordingly it was not granted. 2 1 2. The Glass-SteagallAct. - To avoid the direct prohibition on stock dealing, commercial banks dealt in securities through affiliates. In 1933, Congress believed that the failure of these stock affiliates dam￾aged banks, and the resulting bank failures caused, not reflected, the Depression. The resulting Glass-Steagall Act of 1933 prohibited bank affiliates from owning and dealing in securities, thereby severing com￾mercial banks from investment banks. 22 Today, prohibitions on com￾mercial banks' underwriting and affiliation with companies dealing in securities are breaking down.23 But the prohibition on bank ownership 20. B. Hammond, Banks and Politics in America from the Revolution to the Civil War 149-55 (1957). 21. California Bank v. Kennedy, 167 U.S. 362, 366-67 (1897); National Bank Act of Feb. 25, 1863, § 11, 12 U.S.C. § 24 (seventh) (1988). State member banks of the Federal Reserve System were later similarly restricted. Banking Act of 1933 (Glass￾Steagall), § 5(c), 12 U.S.C. § 335 (1988). 22. Banking Act of 1933 (Glass-Steagall), § 16, 12 U.S.C. § 24 (seventh) (1988). 23. Securities Indus. Ass'n v. Board of Governors of the Fed. Reserve Sys., 468 U.S. 207, 214-21 (1984) (upholding FRB's authorization of Bank of America to acquire

18 COLUMBIA LAW REVIEW [Vol.91:10 of equity remains stable. Bank trust departments are commercial banks'only remaining di- rect link to equity.Rules fragment trust fund investments:no more than 10%of a bank's trust f nds may be invested in the stock of any corpora ration.The regulation first came down a few years after Report warned against the growing power of bank trust de partments.Other trustee laws foster a hyper-fragmentation of the bank trust portfolio,well beyond that which financial economists say is needed for diversification.26 3.The Bank Holding Compa Legislation -Many bank owne wished to operate from several loc s or to chain several banks to- o6二 reincorporated themselves as holding companies,each owning several separately incorporated banks.The holding company had other regu latory advantage s:it was not subject to the same ngent regulatio as were the ubs diaries the 082 own businesses and stock,including control block company could In response, Congress enacted the Bank Holding Company Act of 1956,restricting a holding company's activities to those closely related to banking.A bank holding company cannot own more than 5%of the voting stock of any nonbanking company and cannot otherwisec ndustrial firm.28 Holdi con rol an ng compa only if the additional stock is nonvoting. Schwab.a securities dealer): o IP Mo &Co..11988-1989 Transfer Binder Fed.Banking ep.(CCD)(an.19RB approves application aR28 to establish under writing affiliate);Bankers Trust New York Corp.,75 89)(FRB approves application of commercial banks oc22CF.R.s9.18( 9 25.1 Staff of House Subcomm.on Domestic Finance,Comm.on Banking and Cur rency,90th Cong.,2d Sess.,Commercial Banks and Their Trust Activities:Emerging Repoitee on the American Economy 14 Comm.Print 1968)Chereinater Patnan rd Investor Challenges and(Samset Bicksler (forthcoming). d Tucker,Interst te Banking and the Feder Historical Per. Act of 1956,4(c (1988).Shares acquired in a fiduciar not included in the5%limit 29.Id.;P.Heller,Federal Bank Holding Company Law $4.03[2][b](1990).Au i companics couldm recn issu ompany Manual -11--14 (Feb.1990) xchange

COLUMBIA LAW REVIEW of equity remains stable. Bank trust departments are commercial banks' only remaining di￾rect link to equity. Rules fragment trust fund investments: no more than 10% of a bank's trust funds may be invested in the stock of any single corporation. 24 The regulation first came down a few years after the Patman Report warned against the growing power of bank trust de￾partments. 25 Other trustee laws foster a hyper-fragmentation of the bank trust portfolio, well beyond that which financial economists say is needed for diversification. 26 3. The Bank Holding Company Legislation. - Many bank owners wished to operate from several locations or to chain several banks to￾gether. However, many states prohibit banks from branching and pro￾hibit out-of-state banks from operating locally. In the 1950s, banks reincorporated themselves as holding companies, each owning several separately incorporated banks. The holding company had other regu￾latory advantages: it was not subject to the same stringent regulation of activities as were the bank subsidiaries-the holding company could own businesses and stock, including control blocks.27 In response, Congress enacted the Bank Holding Company Act of 1956, restricting a holding company's activities to those closely related to banking. A bank holding company cannot own more than 5% of the voting stock of any nonbanking company and cannot otherwise control an industrial firm. 28 Holding companies can own more than 5% (up to 25%o), but only if the additional stock is nonvoting.29 Schwab, a securities dealer); see also J.P. Morgan & Co., [1988-1989 Transfer Binder] Fed. Banking L. Rep. (CCH) 87,554 (Jan. 19, 1989) (FRB approves application of commercial bank to establish underwriting affiliate); Bankers Trust New York Corp., 75 Fed. Reserve Bull. 829 (Oct. 30, 1989) (FRB approves application of commercial banks to engage in some brokerage activities). 24. 12 C.F.R. § 9.18(b)(9)(ii) (1990). 25. 1 Staff of House Subcomm. on Domestic Finance, Comm. on Banking and Cur￾rency, 90th Cong., 2d Sess., Commercial Banks and Their Trust Activities: Emerging Influence on the American Economy 1-4 (Comm. Print 1968) [hereinafter Patman Report]. 26. See Roe, Institutional Fiduciaries in the Corporate Boardroom, in Institutional Investors: Challenges and Responsibilities (A. Samsetz & J. Bicksler eds. 1991) (forthcoming). 27. Clair & Tucker, Interstate Banking and the Federal Reserve: A Historical Per￾spective, Fed. Reserve Bank of Dallas Econ. Rev., Nov. 1989, at 1, 6-12. 28. Bank Holding Company Act of 1956, § 4(c)(4)-(5), 12 U.S.C. § 1843(c)(5)-(6) (1988). Shares acquired in a fiduciary capacity are not included in the 5% limitation. Id. 29. Id.; P. Heller, Federal Bank Holding Company Law § 4.03[2][b] (1990). Au￾thority to own nonvoting stock of public companies was until recently largely illusory. Public companies could not issue nonvoting stock for most of the post-1956 era because of one-share, one-vote rules. New York Stock Exchange, New York Stock Exchange Listed Company Manual 3-11-3-14 (Feb. 1990). [Vol. 9 1: 10

1991] A POLITICAL THEORY OF THE CORPORATION 19 B.Mutual Funds Mutual funds pool the investment funds of hundreds of investors, thereby enabling the investors both to diversify and to buy the invest- ment expertise of the fund's managers.If mutual funds could have taken la rge blocks,they mi ave nitoring worthwhile theory,they could have evolved into the missing monitoring link be- tween fragmented investors and large operating firms. However,Congress was suspicious of mutual funds'potential to control industrial companies.A 1934 Senate securities repor t distin guished two functions:investment and management control Only un- scrupulous financiers mixed investment with control: “The investment company [has]bec[o]me the instrumentality of financiers and industri- alists to facilitate acquisition of concentrated control of the wealth and industries of the "30 As a cons gres vent the of these [inves trust from their normal chan of diersified investment to the o avenues of cor of Congress might have "to completely divorce investment trusts from investment banking."32 Mutual funds should invest only passively;holding companies were organized for control and they w den gated,becag nflicts of inte ged outsider sha ers Congress directe the SEC to draft legislation. 1.The Investment Company Act of 1940.- -The sEC understood its marching orders from Congress.Its bill declared that "the national public inte [is]adversely affected. when investment compa- e .[have great [and]have exce in the na tional economy. C wanted mutual funds and employees off the boards of all portfolio companies,essentially a Glass- Steagall-type severance.35 Enactment ultimately entailed compromise with the industry,but the SEC achieved much of its severance goal. A mutual fund cannot advertise itself as diversified if the regulated part of its portfolio has more than 10%of the stock of any company, even if that stock is a small portion of the fund's portfolio.36 (Only 30.Co Stock Excha C D 73d c ort,in reference to its final chief counsell. 31.ld.at 333(emphasis added);see also Investment Trusts and Investment Com- 30 Betore ong3d Sess.6(1 mm.of the Senate Comm.on Banking and 40)[herei Act F (state nent of 6 (quoting eport). 33.Public Utility 30,at3 gs Co Act of193530,15U.S.C.s79a-4(1988 conce ment and pr 35.1d.at216-20. 36.Investment Company Act of 1940,$5(b),15 U.S.C.80a-5(b)(1988)[herein- after 1940 Act

1991] A POLITICAL THEORY OF THE CORPORATION 19 B. Mutual Funds Mutual funds pool the investment funds of hundreds of investors, thereby enabling the investors both to diversify and to buy the invest￾ment expertise of the fund's managers. If mutual funds could have taken large blocks, they might have found monitoring worthwhile. In theory, they could have evolved into the missing monitoring link be￾tween fragmented investors and large operating firms. However, Congress was suspicious of mutual funds' potential to control industrial companies. A 1934 Senate securities report distin￾guished two functions: investment and management control. Only un￾scrupulous financiers mixed investment with control: "The investment company [has] bec[o]me the instrumentality of financiers and industri￾alists to facilitate acquisition of concentrated control of the wealth and industries of the country."30 As a consequence, Congress must "pre￾vent the diversion of these [investment] trusts from their normal chan￾nels of diversified investment to the abnormal avenues of control of industry." 3 ' Congress might have "to completely divorce investment trusts from investment banking."'3 2 Mutual funds should invest only passively; holding companies were organized for control and they were denigrated, because conflicts of interest damaged outsider sharehold￾ers. Congress directed the SEC to draft legislation.33 1. The Investment Company Act of 1940. - The SEC understood its marching orders from Congress. Its bill declared that "the national public interest ... [is] adversely affected ... when investment compa￾nies ... [have] great size ... [and] have excessive influence in the na￾tional economy." 34 The SEC wanted mutual funds' directors and employees off the boards of all portfolio companies, essentially a Glass￾Steagall-type severance.3 5 Enactment ultimately entailed compromise with the industry, but the SEC achieved much of its severance goal. A mutual fund cannot advertise itself as diversified if the regulated part of its portfolio has more than 10% of the stock of any company, even if that stock is a small portion of the fund's portfolio.36 (Only 30. Comm. on Banking and Currency, Stock Exchange Practices, S. Rep. No. 1455, 73d Cong., 2d Sess. 333-34 (1934) [hereinafter Pecora Report, in reference to its final chief counsel]. 31. Id. at 333 (emphasis added); see also Investment Trusts and Investment Com￾panies, Hearings on S. 3580 Before a Subcomm. of the Senate Comm. on Banking and Currency, 76th Cong., 3d Sess. 36 (1940) [hereinafter 1940 Act Hearings] (statement of Robert E. Healy, Comm'r, Sec. and Exchange Comm.) (quoting Pecora Report). 32. Pecora Report, supra note 30, at 393. 33. Public Utility Holdings Company Act of 1935, § 30, 15 U.S.C. § 79a-4 (1988). 34. 1940 Act Hearings, supra note 31, at 2. The statement of purpose also showed concern for efficient investment management and protection of investors. Id. 35. Id. at 216-20. 36. Investment Company Act of 1940, § 5(b), 15 U.S.C. § 80a-5(b) (1988) [herein￾after 1940 Act]