Empty Voting: Private Solutions to a Private Problem EMPTY VOTING: PRIVATE SOLUTIONS TO A PRIVATE PROBLEM Michael Lee* I. Introduction ................................................................ 884 II. Applications of Existing Law ..................................... 885 A . D eephaven ............................................................ 886 B. Vote-Buying Doctrine .......................................... 887 1. A pplicability .................................................... 887 2. Schreiber ......................................................... 888 3. IXC and H ewlett ............................................. 890 C. Possible Applications of Other Legal Authority. 893 1. In re PN B ........................................................ 893 2. Secret Voting Trusts ....................................... 895 D . Sum m ary .............................................................. 897 III. Empty Voting Concerns that Still Remain ............... 897 A. Detection Problems .............................................. 898 B. Concerns Posed by Inaccurate Voting ................ 899 C. The Question of Fairness .................................... 901 D. Prevalence of Conflicted Voting Decisions ......... 904 IV. Possible Solutions to the Remaining Problems ........ 906 A. Increase Monitoring ............................................ 906 B. Increase Voting Accuracy .................................... 908 C. Implement Alternative Voting Structures ......... 909 V . C onclusion .................................................................. 911 * J.D. Candidate 2008, Columbia University School of Law; B.A. Economics 2005, Dartmouth College. The author wishes to thank Professor Harvey Goldschmid and Professor Merritt Fox for their patience and advice and the staff of the Columbia Business Law Review for their editing assistance. COLUMBIA BUSINESS LAW REVIEW I. INTRODUCTION "Empty voting"1 describes the process through which a shareholder separates and sells off the economic interest in a share, retaining only the voting rights. Empty voting can currently be accomplished through a variety of techniques. These techniques include using accompanying short sales to offset the long position, using options to hedge away some or all of the economic exposure, entering into swap agreements that trade the economic fluctuations of the shares for other cash flows, or through "record date capture," in which the empty voter temporarily borrows shares in order to vote them.2 In all of these instances, the empty voter retains little or no economic stake in the corporation, which distorts the empty voter's incentives to vote in the corporation's best interests. In extreme cases, such as when the empty voter is economically short, the empty voter may stand to personally gain from the corporation's loss. By separating decision- making power from the economic investment in a corporation, empty voting may undermine the assumption that shareholder votes will accurately reflect the economic interests of the corporation-an assumption that has largely justified the law's deferential stance towards shareholder voting outcomes. This apparent disconnect between the effects of empty voting and judicial perceptions of corporate democracy has prompted many commentators to suggest that existing legal doctrine is poorly equipped to respond to the issues raised by empty voting.' ' This term was originally coined by Hu and Black. Henry T. C. Hu & Bernard Black, The New Vote Buying: Empty Voting and Hidden (Morphable) Ownership, 79 S. CAL. L. REV. 811 (2006). 2 Id. at 827, 832. 3 See id. at 818; Marc Weingarten & Morgan 0. Mirvis, Vote-Buying Raises Questions Under Anti-Fraud Rules, 1565 PLICORP 407, 409 (2006); Jonathan J. Katz, Barbarians at the Ballot Box: The Use of Hedging to Acquire Low Cost Corporate Influence and Its Effect on Shareholder Apathy, 28 CARDozo L. REV. 1483 (2006); see also Shaun Martin & Frank Partnoy, Encumbered Shares, 2005 U. ILL. L. REV. 775 (2005) (arguing that the one-share/one-vote rule is flawed for economically and legally encumbered shares). But see Bruce H. Kobayashi & Larry Ribstein, [Vol. 2007 This Note attempts to re-examine the concerns raised by empty voting. Ultimately, this Note takes the position that while empty voting poses real concerns, it may not be as problematic as some commentators have suggested. Although some issues do remain, many of the more extreme dangers may be adequately addressed under existing law. While acknowledging that some public sector regulation would be beneficial, this Note argues that many of the unaddressed problems posed by empty voting are ones that private institutions have both the ability and the incentive to solve. Based on these assessments, this Note suggests some possible private sector solutions. Part II describes the current legal setting and how it deals with the problem of empty voting. Part III discusses the concerns raised by empty voting that still remain. Part IV discusses possible solutions to these remaining problems. Part V concludes. II. APPLICATIONS OF EXISTING LAW The sections that follow examine some of the existing Delaware case law.4 By determining which problem situations the law already addresses, the unaddressed fact patterns can be isolated and their associated problems identified with greater accuracy. Section A addresses Deephaven v. UnitedGlobalCom5 , which most directly discusses the role of courts with respect to empty voting. Section B discusses the analogous vote-buying doctrine, addressing the dual questions of whether and how vote- buying doctrine applies to empty voting. Section C examines possible interpretations or extensions of other legal authority. Section D summarizes. Outsider Trading as an Incentive Device, 40 U.C. DAvIs L. REv. 21, 24 (2006) (arguing against broad regulation of outsider trading). 4 Because this Note focuses on Delaware law, it may not fully account for jurisdictional variation. ' Deephaven Risk Arb. Trading Ltd. v. UnitedGlobalCom, Inc., No. Civ.A. 379-N, 2005 WL 1713067 (Del. Ch. 2005). No. 3:8831 EMPTY VOTING COLUMBIA BUSINESS LAW REVIEW A. Deephaven Deephaven v. UnitedGlobalCom most directly addressed the question of shareholder rights with respect to economically empty shareholders, as well as the role of courts in upholding or curtailing those rights. In Deephaven, a net short shareholder sought inspection rights under Del. Code Ann. tit. 8, § 220 (2003). The defendant corporation argued that shareholder inspection rights should not extend to shareholders who held net short positions in the corporation. The court held for the shareholder, analogizing the short and long positions as a loan and a bank deposit, respectively, and determined that for purposes of § 220, the offsetting "loan" of the short position did not curtail the rights or ownership of the long position "deposit."6 Initially, this holding would appear to grant free license for empty voters to vote as they wish, but the court went on to explicitly exclude voting contexts. In determining that § 220 rights should apply to net short shareholders, the court reasoned that "unlike in other situations such as voting, the § 220 analysis includes its own safeguard against plaintiffs with economic incentives that are not aligned with other stockholders: the proper purpose analysis."' This language clearly indicates the court's willingness to intercede in situations lacking safeguards against distorted economic incentives, and in voting contexts in particular. What is not clear is whether this language indicates a carve-out whose boundaries must be delineated in the future, or whether the court was distinguishing the facts of Deephaven from prior case law. Although either interpretation would indicate that the voting rights of empty voters may be curtailed, it would seem appropriate to assume the latter if there existed a body of law that fit the description given by the Deephaven court: namely, one in which judicial safeguards were imposed in voting contexts where a voter's economic incentives were not aligned with those of the other stockholders. Existing vote- 6 Id. at *6. ' Id. (emphasis added). [Vol. 2007 EMPTY VOTING buying doctrine, which will be discussed in the following section, fits this bill. B. Vote-Buying Doctrine This section will take a look at the analogous vote-buying cases. Subsection One will address the question of applicability. Subsection Two will look at the decision of Schreiber v. Carney,8 the leading Delaware vote-buying case. Subsection Three will examine two of its progeny, IXC9 and Hewlett v. Hewlett Packard Co. 10, to determine the implications of present vote-buying doctrine. 1. Applicability The vote-buying cases provide examples of judicially- imposed safeguards in situations where distortion of economic incentives jeopardizes the shareholder franchise. Attempts to apply vote-buying doctrine, however, encounter the technical question of whether the new empty voting can be deemed to fit within the ambit of the classic vote-buying cases. The court has defined vote-buying as "any transaction by which a party directs a shareholder's vote for consideration personal to that shareholder."1 Empty voting does not fit neatly within that definition, since the empty voter is typically selling a separated economic stake rather than buying a separated vote.12 At least one commentator has remarked that due to this technical distinction, the vote- buying cases are inapposite. 3 ' Schreiber v. Carney, 447 A.2d 17 (Del.Ch. 1982). 9 In re IXC Commc'ns, Inc., No. C.A. 17324, 1999 WL 1009174 (Del. Ch. 1999). 10 Hewlett v. Hewlett-Packard Co., No. CIV.A. 19513-NC, 2002 WL 549137 (Del. Ch. 2002). 1' IXC Commc'ns, 1999 WL 1009174, at *8. 12 See Hu & Black, supra note 1, at 862. See Charles M. Nathan, "Empty Voting" and Other Fault Lines Undermining Shareholder Democracy: The New Hunting Ground for Hedge Funds, 1571 PLI/CoRP 425, 437 (2006). No. 3:883] However, the underlying policy rationale expressed by the courts in the vote-buying cases is equally applicable to the new "empty voting" situations. As one court explained: "[g]enerally speaking, courts closely scrutinize vote-buying because a shareholder who divorces property interest from voting interest[] fails to serve the 'community of interest' among all shareholders, since the 'bought' shareholder votes may not reflect rational, economic self-interest arguably common to all shareholders." 4 Under this rationale, there is no principled distinction between the classic vote-buyer and the new empty voter. The only distinction is that the empty voter has found in one case, the willing seller of a severed vote, and in the other, the willing buyer of a severed economic interest. If the applicability of vote-buying doctrine turned on who had the unsevered share to begin with, it could easily be avoided by arranging every vote-buying agreement as a two step process. The Hewlett court rejected a similarly technical distinction between explicit written agreements and unwritten understandings, noting that when determining whether vote-buying doctrine is applicable, the court's focus should be the "[plrotection of unsuspecting shareholders who are at risk of being defrauded or disenfranchised .. 2. Schreiber The earliest vote-buying cases, exemplified by Macht v. Merchants Mortgage & Credit Co., generally held that vote- buying was illegal per se as a matter of public policy.' 6 The rationale behind these decisions was that "each stockholder should be entitled to rely upon the independent judgment of his fellow stockholders."17 This duty to exercise independent judgment, however, was rejected by the court in Ringling Bros. Barnum & Bailey Combined Shows v. Ringling. 14IXC Commc'ns, 1999 WL 1009174 at *8. '5 Hewlett, 2002 WL 549137, at *4. 16 Macht v. Merchants Mortgage & Credit Co., 194 A. 19, 22 (Del. Ch. 1937). 17 Schreiber v. Carney, 447 A.2d 17, 24 (Del. Ch. 1982). COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 Upholding a voting agreement, the Ringling court noted that: Generally speaking, a shareholder may exercise wide liberality of judgment in the matter of voting, and it is not objectionable that his motives may be for personal profit, or determined by whims or caprice, so long as he violates no duty owed his fellow shareholders. The ownership of voting stock imposes no legal duty to vote at all.18 This decision was later codified by statute in 8 Del. C. § 218, which permitted voting agreements between shareholders.19 Schreiber v. Carney went on to hold that vote-buying was not illegal per se absent a purpose to defraud or disenfranchise the other shareholders.2 ° In Schreiber, the vote-buying in question was a loan agreement between Texas International, the defendant corporation, and Jet Capital, a significant shareholder who sought to veto a favorable merger. The loan agreement itself was conditional on the approval of the disinterested shareholders. The court upheld the loan agreement, rejecting Macht's per se rule because there was no intent to defraud or disenfranchise other shareholders.21 18 Ringling Bros. Barnum & Bailey Combined Shows v. Ringling, 53 A.2d 441, 447 (Del. 1947) (citation omitted). 19 Some have suggested a distinction between voting agreements and vote-buying agreements in that the first are generally vote-pooling arrangements while the latter are agreements which purchase votes for personal consideration. The line-drawing here would ostensibly occur between votes exchanged for votes in the same corporation (logrolling, so to speak) versus votes exchanged for anything else. See, e.g., EDWARD P. WELCH ET AL., FOLK ON THE DELAWARE GENERAL CORPORATION LAW § 151.3 (5th ed. 2007) ("[S]tockholders may lawfully contract with each other to vote in such manner as a majority of them may deem best, provided the voting agreement does not violate duties owed to other stockholders ... [blut votes may not be purchased."). As we will see, however, this is not the direction that the courts have taken. 20 Schreiber v. Carney, 447 A.2d at 25-26. 21 Id. EMPTY VOTINGNo. 3:8831 Some confusion remains with respect to the degree to which the Schreiber court rejected the Macht rule. Much of the commentary on the current vote-buying doctrine and even some subsequent court opinions seem to have interpreted Schreiber as establishing a rule that requires plaintiffs to show that shareholders were defrauded or disenfranchised in order to void a vote-buying agreement. In a peculiar turn of phrase, the Schreiber opinion noted that "voting agreements ... should not be considered to be illegal per se unless the object or purpose is to defraud or in some way disenfranchise the other stockholders,"22 indicating a limited per se rule that raises questions as to whether vote- buying agreements that do not necessarily defraud or disenfranchise other shareholders could still be illegal. The court clarified this statement by noting that "[t]his is not to say, however, that vote-buying accomplished for some laudible [sic] purpose is automatically free from challenge. Because vote-buying is so easily susceptible of abuse it must be viewed as a voidable transaction subject to a test for intrinsic fairness."23 Read correctly, the Schreiber court seems to have replaced the earlier per se rule with a heightened scrutiny rule in which vote-buying agreements with sufficient safeguards were not automatically void, but still voidable. 3. IXC and Hewlett At least one Delaware court has interpreted Schreiber to require plaintiffs to prove that the vote-buying agreement either defrauds or disenfranchises shareholders. In IXC Communications, the court refused to enjoin an upcoming shareholder vote regarding a merger, despite evidence of a vote-buying arrangement in which the acquirer paid a significant minority shareholder of the target to vote for the merger's approval. Citing Schreiber, the court stated that "[olnly when the vote-buying agreement defrauds or disenfranchises other shareholders can it be said that the 22 Id. 21 Id. at 26. COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 agreement is illegal and therefore void."24 The court went on to note that because the plaintiffs failed to allege that the vote-buying agreement defrauded other shareholders, it would only need to examine whether the other shareholders had been disenfranchised. Finding that the vote-buying arrangement had not "locked up" a majority of the votes needed for the merger to pass, such that an independent majority of the shareholders could still freely vote to defeat the merger, the court held that the other shareholders had not been disenfranchised and refused to grant injunctive relief.25 The court in Hewlett v. Hewlett Packard Co., in finding a colorable vote-buying claim, returned to the heightened scrutiny rule of Schreiber, but limited its application to scenarios in which the management had used corporate assets to purchase votes. Explaining the Schreiber decision, the court noted that "significantly... it was the management of the defendant corporation that was buying votes."26 The court explained that "Is]hareholders are free to do whatever they want with their votes, including selling them to the highest bidder."27 Management, however, would have to show that the "vote-buying activity does not have a deleterious effect on the corporate franchise."28 The court went on to hold that when management was buying votes, the plaintiff did not have to show that a majority of the shares had been locked up by the voting arrangement. The court reasoned that: Less than a majority of votes can be decisive in tipping the results of an election one way or another. If voiding the votes cast in accordance with a fraudulent vote-buying agreement with corporate management is sufficient to 24 In re IXC Commc'ns, Inc., No. C.A. 17324, 1999 WL 1009174, *8 (Del. Ch. 1999). 25 Id. at *8-9. There may be room to distinguish this case due to the nature of the relief sought. 26 Hewlett v. Hewlett-Packard Co., No. CIV.A. 19513-NC, 2002 WL 549137 at *4 (Del. Ch. 2002). 27 id. 28 Id. No. 3:8831 EMPTY VOTING change the result of a vote, I am again of the opinion that the defrauded or disenfranchised shareholders should not be prevented from bringing a vote-buying claim.2 9 IXC and Hewlett together adopt a standard for voiding voting agreements that resembles a majority or controlling shareholder test. As some commentators have already suggested, 0 empty voting is thus prohibited when there are fiduciary duty concerns. There are two important differences, however, between vote-buying doctrine and the majority or controlling shareholder test. The "majority"- equivalent prong of vote-buying doctrine may be more expansive than the analogous prong for imposing fiduciary duties, while the "controlling"-equivalent prong may be more limited. IXC potentially expands the "majority"-equivalent prong by indicating that a voting agreement that ties up less than a majority stake may be problematic if it is sufficient to make the voting outcome a foregone conclusion." Such a situation would disenfranchise the other shareholders, thereby making the agreement illegal. Although this is unlikely to come up often, it may be at issue when there is minority veto power over a proposal, or if a majority of the minority or similar requirement creates a situation in which less than a majority stake may in itself be outcome determinative. Hewlett potentially limits the "controlling"-equivalent prong in that it suggests the court will intercede only in cases where management is purchasing rather than selling votes.3 2 The opinion further suggests that vote-buying by 29 Id. at *5. 30 See Hu & Black, supra note 1, at 862; Weingarten & Mirvis, supra note 3, at 412. "' In re IXC Commc'ns, No. C.A. 17324, 1999 WL 1009174 at *8-9 (Del. Ch. 1999). 32 Hewlett, 2002 WL 549137, at *6. This is not as problematic as it seems, as vote-buying arrangements in which management is selling votes would be governed under Delaware statute, which imposes certain conditions on transactions by interested directors. DEL. CODE ANN. tit. 8, § 144 (2005). COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 management is only problematic if the votes are being purchased through corporate funds.3 Regarding the latter requirement, however, the court found that the vague promise of future business was the equivalent of purchasing with corporate funds,34 which suggests that the potential for unspoken quid pro quo arrangements may be sufficient to constitute "purchase" through corporate resources. Although IXC and Hewlett limit the situations in which a voting agreement would be found invalid, they directly address the two scenarios in which a voting agreement could be most abusive. The first scenario is one in which the arrangement actually disenfranchises the other shareholders by precluding the possibility of any other voting outcome. The second scenario is one in which the arrangement involves fiduciaries who have greater control over the corporate machinery. This would seem to leave unaddressed only those situations in which the tied-up voting stake is (a) insufficient in itself to preclude a voting outcome and (b) not brought about by fiduciaries who may have greater influence over the voting machinery. C. Possible Applications of Other Legal Authority This section examines possible interpretations or extensions of other legal authority that are less clearly applicable to empty voting contexts. The first subsection examines the implications of a broad interpretation of In re PNB .3 The second subsection examines the historical treatment of secret voting trusts and their potential applicability to empty voting. 1. In re PNB Though vote-buying doctrine effectively precludes the more obviously egregious abuses of empty voting, its focus on " Hewlett, 2002 WL 549137, at *6. 34 Id. at *5 n.12. 35 In re PNB Holding Co. S'holders Litig., No. Civ.A 28-N, 2006 WL 2403999 (Del. Ch. 2006). No. 3:883] EMPTY VOTING voting agreement contexts tends to limit its applicability to situations involving either a single empty voter or a group of empty voters that are acting together. Although Hewlett suggests that a vote-buying claim may be found even in cases where such collusion is not explicit, vote-buying doctrine does not address situations in which multiple empty voters, ostensibly acting independently, foreclose a voting outcome. While the concerns associated with such scenarios will be further explored in Part III, this subsection will briefly address possible judicial responses in light of the court's recent decision in In re PNB. In re PNB applied an intrinsic fairness test where a majority of the shareholders, including several directors, voted to reorganize in a manner that would leave as shareholders only those individuals who owned above a specified threshold of shares. The remaining shareholders would be bought out. The court held that the intrinsic fairness test was appropriate despite finding that the majority shareholders had acted independently in their own individual best interests, rather than as a "single, monolithic controller."36 According to the court, applying the intrinsic fairness test was appropriate because the majority shareholders had created a situation in which the interests of one group of shareholders were opposed to the interests of the remaining shareholders, creating what the court characterized as a "zero-sum game."37 A narrow reading of In re PNB would distinguish this case from most empty voting contexts based on the fact that some of the majority shareholders in In re PNB were directors who could influence the nature of the proposal. A broader interpretation, however, may potentially be applied to empty voting contexts involving multiple empty voters. This broader reading would consist of two factors. The first factor is the existence of a voting situation in which the individual interests of one group of voters are opposed to the 36 Id. at *9. "7 Id. at *1. COL UAMIA B USINESS LA W RE VIE W (Vol. 2007 interests of the remaining voters. 8 The second factor is that at least some members of the winning group of shareholders had unilaterally brought about the situation that caused the shareholders' interests to be opposed.3 9 Although the likeliness of finding the second factor seems otherwise limited when the majority shareholder group does not include the directors or management, this factor could potentially be extended to empty voting contexts. An empty voter, much like a director who dictates the terms of a proposal, has the ability to unilaterally create a situation in which shareholders' interests are opposed. An extension of In re PNB's rationale to empty voting contexts may provide a solution for situations involving multiple empty voters, but some caveats are in order. First, the case involves a very different factual context, and the rationale behind the court's decision here is not readily applicable to empty voting without a broad interpretation. Second, even if In re PNB were to be extended, it may be limited by the requirement that shareholder interests be actually opposed, which would likely restrict its applicability to those scenarios where the empty voters are net short rather than simply partially hedged. 2. Secret Voting Trusts Another potential application of existing law is to void empty voting on the grounds of secrecy. The requirements of Delaware's voting trust statute and explanatory dicta from some court opinions suggest intolerance towards clandestine empty voting. Although 8 Del. C. § 218 expressly permits voting agreements, it also requires that copies of the agreement be placed in the corporation's Delaware office where it may be inspected by the other shareholders.4" By way of explanation, the court in Oceanic Exploration Co. v. Grynberg noted that "the main purpose of a voting trust statute is 'to avoid secret, uncontrolled combinations of 8 See id. See id. 40 DEL. CODE ANN. tit. 8, § 218(a) (2007). No. 3:883] EMPTY VOTING COLUMBIA BUSINESS LA W REVIEW stockholders formed to acquire control of the corporation to the possible detriment of non-participating shareholders.' 41 The Schreiber court likewise characterized a secretive voting agreement from an earlier case as a "deceit obviously practiced upon the other stockholders."42 Both the statute and the language of the cases seem to indicate that voting agreements are appropriate only to the extent that they are made known to other shareholders. Although it is not clear whether the rationale behind a prohibition on secret voting arrangements is to help identify problematic situations or whether such secrecy in itself is viewed as a fraud on other shareholders, this suggests a predilection to void secretive arrangements that potentially endanger the shareholder franchise. Discussion of prohibitions on secrecy, however, has largely fallen by the wayside, and it is not clear whether such a prohibition would apply to empty voting arrangements. Although the policy rationale, if still valid, would seem likely to apply, there has been no clear exposition of that rationale. Though the Schreiber court characterized secret voting arrangements as an obvious deceit, it did so in conclusory terms without giving clear reasons as to why the other shareholders were entitled to knowledge of the agreement. If the true rationale is preventative, as was suggested by the Grynberg court, it seems that a disclosure requirement may not be as appropriate for empty voting as it would be for voting agreements, since the shorter duration and increased frequency of hedging transactions may mean that the administrative costs of requiring disclosure would outweigh any prophylactic value. Ultimately, although an argument may be made against the secret nature of most empty voting, the doctrine is neither sufficiently clear nor established to readily apply to empty voting scenarios. 41 Oceanic Exploration Co. v Grynberg 428 A.2d 1, 7 (Del. 1981) (quoting Lehrman v. Cohen, 222 A.2d 800, 807 (Del. 1966)). 42 Schreiber v. Carney, 447 A.2d 17, 24 (Del. Ch. 1982). [Vol. 2007 No. 3:883] EMPTY VOTING D. Summary An examination of existing legal doctrine suggests that the law precludes many of the more abusive empty voting practices. Although vote-buying doctrine was developed by the courts before the dangers of the new empty voting were known, the principles and rationale behind the vote-buying doctrine indicate that they should and will be applied to empty voting. Analysis of the vote-buying doctrine indicates that empty voting will be barred in situations in which the empty voting is itself sufficient to preclude the outcome of a vote or in situations where the empty voting is done by management in breach of fiduciary duty. The only situation left unaddressed is one in which a minority shareholder, who has no control over the affairs of the corporation, votes empty shares against the best interests of the other shareholders. In theory, such an action should be meaningless because the other shareholders, by voting in their own best interests, would render the empty voter's votes moot. In practice, however, it may still be possible for an empty voter to profit from such behavior at the corporation's expense. Although some of the potential dangers may be addressed through extensions of other legal doctrines, such as invalidating empty voting on the grounds of secrecy or subjecting certain opposed-interest shareholder votes to an intrinsic fairness test, these arguments require either broad interpretations or extensions by analogy, and the rest of this Note assumes that these potential applications are not currently applicable law. The following part will discuss these concerns in greater detail. III. EMPTY VOTING CONCERNS THAT STILL REMAIN The concerns that empty voting still raises are those that arise in the context of minority empty voters who cannot force the outcome of a vote. These concerns will be discussed in four sections. The first section addresses the difficulties of monitoring and detecting problematic empty voting, which may hamper the law's ability to deal with problem scenarios No. 3:883] EMPTY VOTING COLUMBIA BUSINESS LA W REVIEW that it already seeks to prevent. The second section discusses the concern of "inaccurate" voting by other shareholders, which may provide opportunities for empty voters with minority stakes to bring about voting outcomes that are detrimental to the corporation. The third section examines fairness concerns based on the possibility that the transactions utilized by empty voters may improperly price the value of votes. The fourth section addresses the possibility that conflicted voting decisions may be more prevalent than is generally assumed, which would increase opportunities for empty voting strategies to profit at the expense of the corporation. A. Detection Problems The first concern is that existing law is not properly equipped to detect problematic empty voting scenarios. Although Securities and Exchange Commission ("SEC") rules require disclosure of related transactions at the ten percent threshold,43 practices such as record date capture may fall under the radar. Perhaps the most significant concern in this area is that posed by consciously parallel behavior, sometimes called "wolf pack" behavior," by multiple empty voters. If, in fact, multiple empty voters have conspired and acted together in order to force a voting outcome, this would seem to fall under "actual disenfranchisement" of other shareholders under the holding of IXC.4 5 However, a legal rule that presumed concerted action simply based on voting results would likely allow any and every voting decision to be challenged by those who disagreed with the outcome. Proving that such a conspiracy existed, on the other hand, could be a formidable task.4" To some extent, the dangers of 15 U.S.C.A. § 78p (2002); see also 17 C.F.R. s. 240.16a-1 et seq. Stephen Fraidin & Daniel S. Hoverman, Hedge Fund Activism, 1571 PLIICoRP. 401, 407 (2006). 4" In re IXC Commc'ns, No. C.A. 17324, 1999 WL 1009174, at *8 (Del. Ch. 1999). 41 See, e.g., Hallwood Realty Partners, L.P. v. Gotham Partners, L.P., 286 F.3d 613 (2d Cir. 2002) (upholding lower court's decision to dismiss where there was circumstantial but no direct evidence of collusion). [Vol. 2007 such consciously parallel action may be mitigated by the collective action problems that the parallel actors face. However, some have suggested the existence of various signaling devices47 that may help to overcome these collective action problems. Although a disclosure requirement may be helpful in identifying and detecting these problem scenarios, the administrative burden of enforcing such disclosure requirements may exceed the benefits that increased detection brings. B. Concerns Posed by Inaccurate Voting The next category of concerns involves voting accuracy, which may result in opportunities for a less than decisive empty voting position to be outcome determinative. The first concern in this category is that many shareholders cannot accurately determine whether a transaction is in the corporation's best interests. The second concern is that many shareholders are rationally apathetic and either choose not to vote in an informed manner or choose not to vote at all. In either of these scenarios, voting results may deviate from those that would reflect the best interests of the corporation. Empty voting in these circumstances will have the effect of exacerbating these failures and increasing the likelihood of voting outcomes that are detrimental to the corporation as a whole. The first concern is primarily one of information. There are two potential reasons that a shareholder may inaccurately assess the desirability of any proposal. The first is that the shareholder lacks the level of expertise necessary to completely understand the proposal. To some extent, this may be mitigated by better information flow from those who do have the requisite knowledge, such as directors, portfolio advisers, or other fiduciaries, provided that they are representing this information correctly. Conflicts of interests, perverse incentives, or simple error may result in inaccurate portrayals, which would increase the chances that the shareholder will vote incorrectly. The "7 See Fraidin & Hoverman, supra note 45, at 406-07. EMPTY VOTINGNo. 3:8831 COLUMBIA BUSINESS LA W RE VIEW second potential reason is that due to suspicion, distrust, or a perceived lack of reliability, the shareholder ignores or rejects the validity of the information that these experts provide. In summary, voting "inaccuracy" can be largely mitigated through better and more trustworthy information flow, which would in turn decrease the probability that a minority empty voting stake would be outcome- determinative. The second concern is somewhat more of a problem. Assuming that the process of voting entails some non-zero cost for the voter, it may be rational for a voter not to vote even in cases where one voting outcome is clearly preferable to the other. First, assume a shareholder would gain X if a proposal passes, and Y if the proposal fails, and that X > Y (the shareholder is better off if the proposal succeeds). If the process of voting costs the shareholder c, each outcome will result in c less for the shareholder: X-c if the proposal passes, and Y-c if the proposal fails. Assume that the initial probability that the proposal will pass is p, but that by voting for the proposal, the shareholder increases the probability that the proposal will pass by t. If the shareholder does not vote, his expected value is p(X) + (1- p)(Y). If the shareholder does vote, his expected value is (p+t)(X-c) + (1-p-t)(Y-c). Thus, the only situation in which the shareholder rationally should vote is one in which by voting, he increases the probability that the proposal will pass by at least c / (X - Y). This result is complicated by the fact that some initial cost, such as reading the proxy statement, is necessary for the shareholder to determine X and Y in the first place. In some cases, the shareholder may have an incentive not to commit even this initial cost, if X - Y is perceived to be on average small compared to this initial cost, or when his voting influence is minimal. Due to the presence of rationally apathetic voters, a minority empty voter may have the practical ability to force a voting outcome. The Delaware statute, which requires a majority of outstanding rather than voting shares to approve certain extraordinary [Vol. 2007 EMPTY VOTING transactions,48 is a mixed blessing. Although it makes it more difficult for an empty voter to force the approval of an unfavorable extraordinary transaction, it gives the empty voter greater ability to block a favorable one. At the same time, by diminishing each individual shareholder's probabilistic influence over voting outcomes, they reduce those shareholders' incentives to vote. To some extent, this factor may be mitigated by decreasing the costs of voting, such as by making proxy statements clearer and easier to read, but this may come at the expense of increased accuracy of information. An increased presence of minority bloc holders or vote pooling arrangements may also help to alleviate this concern, both by increasing effective voting influence and decreasing relative voting costs. By not addressing concerns raised by voting inaccuracy, existing legal doctrine appears to permit empty voters to take gambles on the probability that some voters will vote incorrectly. The problem here can be characterized as resulting from two factors: (a) the opportunistic empty voters and (b) the inaccurate voting that creates that opportunity. Although it may feel morally more satisfying to condemn the opportunistic empty voters, the problem may also be addressed by eliminating voting inaccuracy. Even a reduction, rather than a thorough elimination, would drive up the cost of the gamble that empty voters are taking, which may substantially reduce its occurrence. C. The Question of Fairness The question of intrinsic fairness is in essence a question of whether some investors should have access to leveraged voting rights. Through a combination of transactions, empty voters are able to obtain voting influence with less economic "8 See DEL. CODE ANN. tit. 8, § 241 (2007). Compare REVISED MODEL Bus. CORP. ACT § 7.27 (2003). Delaware statute and the Model Act also differ in that Delaware requires majority approval of those represented at a meeting for ordinary matters, while the Model Act requires majority approval of those who vote. The latter approach increases individual shareholders' probabilistic influence on voting outcomes and thus increases their incentives to vote. No. 3:883] COLUMBIA BUSINESS LAW REVIEW exposure than if they owned those shares without those transactions. Under this assumption of fairness, the difference between a purely empty voter and a partially hedged voter is merely a difference in degree. To take an extreme example, for a partially hedged voter who is exposed to only one percent of the normal fluctuations of an equivalent unhedged voting position, that minimal exposure is little more than an added transaction cost. If this principle of fair treatment is to be applied consistently, then, it should also disfavor any hedging done by long shareholders as well as classes of stock with increased voting rights. It is important to recognize that leveraging voting is not actually free. By hedging the economic stake attributable to the voting rights, empty voters are giving up upside profit potential as well as downside risk. This is in effect little different from weighted classes of voting stock. A transaction that retained the upside potential, such as a put, would come with its own cost, and even "zero-cost" collars are in effect trading increased downside risk or more limited upside potential in return for the decreased cost.49 If empty voters are paying adequately for the vote leveraging that they acquire, the claim that such vote leveraging is intrinsically unfair becomes much less tenable. Thus, an argument that vote leveraging through hedging is intrinsically unfair ultimately requires two assumptions. The first assumption is that the cost of this vote leveraging is too low. Empty voters, in effect, are obtaining votes for less money than they are worth. The second assumption is that this method of vote leveraging is not equally accessible for all investors, such that those who can access the vote market in this way gain an advantage over those who cannot. These assumptions have some validity. Generally, most options, derivatives, or futures pricing do not calculate in a " Admittedly, the existence of leveraged voting positions may translate into a discount of upside potential relative to downside risk for equivalent fundamentals. See Hu & Black, supra note 1, at 859-60. [Vol. 2007 separate vote value.5" Further, the market for OTC swap transactions, a popular form of hedging, is accessible only to certain enumerated "appropriate persons," such as investment companies, broker-dealers, or corporate entities with total assets exceeding $5,000,000.51 To some extent, these assumptions can further be supported by examining the relative value of votes. As Hu & Black noted, the value of a vote can be calculated as the probability that the vote will be outcome-determinative.52 In majority vote contexts for large public companies where the voting decisions of the "ocean" of other voters are randomly distributed, this probability can be approximated by its "oceanic" Shapley value.53 Although a vote's relative value will change depending on the size and existence of other bloc voters, the marginal value of an additional vote will generally increase as the voter's voting influence increases, up until the point where the voter has attained a decisive voting bloc. 54 The "oceanic" game suggests that a shareholder with a large voting influence will generally value votes more than small shareholders.55 This suggests that there will be an access differential in fact if not in form, because purchasing votes will generally be more worthwhile to voting blocs. In summary, the fairness argument is really a complaint about the market for votes. As opposed to most other "goods," the marginal value of a vote tends to increase as one gains more votes, rather than the reverse. Furthermore, if we assume that voting influence itself has a value, this value '0 The Black-Scholes formulation, for example, is primarily a measure of volatility over a period of time. See Fischer Black & Myron Scholes, The Pricing of Options and Corporate Liabilities, 81 J. POL. ECON. 637 (1973). 5' 7 U.S.C.A. § 6(c) (2000). 512 Hu & Black, supra note 1, at 853. 5 See id. See J.W. Milnor & L.S. Shapley, Values of Large Games II: Oceanic Games, 3 MATHEMATICS OF OPERATIONS RES. 290, 300-06 (1978). " See id. It should be noted that this is not always the case. The relative value of one voting bloc's votes is affected by other voting blocs. In a situation involving two large, roughly equivalent and opposed voting blocs, neither of which have majority control, the relative vote value of the "ocean" considerably increases. EMPTY VOTINGNo. 3:8831 should increase in situations where shareholders' voting incentives are more likely to be opposed. In an ideal setting, in which the interests of all shareholders are perfectly aligned with those of the corporation, the value of voting influence should theoretically be worthless, since relative voting influence should make no difference to a voter's payoffs. The true concern with regard to empty voting in particular, then, is that by allowing votes to be severed from the economic stake, empty voting exacerbates the access and pricing problems of the "vote market." D. Prevalence of Conflicted Voting Decisions The final concern is that shareholder conflicts of interest may be both more prevalent and comparatively greater than is generally assumed, causing any given shareholder's voting decision to substantially deviate from that which would maximize the welfare of the corporation as a whole. Under this model, each shareholder's payoff for any given proposal would consist of two parts. The first part would be the pro- rata economic gains or losses for the corporation - the "aligned" interest. The second part would include gains or losses personal to the shareholder - the "idiosyncratic" interest, which is tied more closely to voting influence than to the economic shares. 6 These "idiosyncratic" interests could range from the worrisome to the benign. The more abusive interests are addressed by law, such as the incentive for a controlling or majority shareholder to "freeze out" the minority shareholders for its own gain.57 Less often addressed are non-monetary incentives, such as incentives by management to increase the size of the corporate empire under their control even in cases where it would not be beneficial to the ' Others have described these in terms of "security" and "private" benefits. See Sanford Grossman & Oliver Hart, One Share-One Vote and the Market for Corporate Control, 20 J FIN. ECON. 175, 177 (1988). "' See, e.g., Kahn v. Lynch Commnc'n Sys., Inc., 638 A.2d 1110 (Del. 1994) (applying "entire fairness" review in a cash-out merger arranged by a 43.3% shareholder). COL UMBIA B USINESS LA W RE VIE W [Vol. 2007 corporation. Another idiosyncratic interest may be a founder's desire to pursue a vision of the company even in cases where it would not be profit maximizing to the corporation as a whole. Additional idiosyncratic interests may include an upstream manufacturer-shareholder's desire to prevent potential competition or an employee- shareholder's desire to improve working conditions. The Schreiber case involved a significant minority shareholder with an idiosyncratic interest in the form of tax liability.5 8 Although these idiosyncratic interests, if relatively small and occasional, would not be greatly problematic, it is not clear that this would be the case. All other things being equal, the voting shares of a corporation and the accompanying decision-making influence would be more valuable to individuals who have a significant and sustained idiosyncratic interest in the corporation's decisions. This list would likely include founders, management, employees, competitors, creditors, insurers, and upstream or downstream businesses. Because individuals or entities with sustained idiosyncratic interests in the corporation's business activities would gain greater value from the corporation's voting shares, it would be expected that these shares would tend to wind up in their hands. Moreover, there would be a tendency for these shareholders to be bloc holders. In a world where there is a tendency for the corporation's shares to drift to parties with substantial and recurring idiosyncratic interests, there is less guarantee that shareholders as a whole will consistently vote in the corporation's best interests. As a result, there may be greater opportunities for an empty voter to have a decisive influence on voting outcomes. In addition, there may be incentives for idiosyncratic bloc holders to exacerbate the problem by emptying or partially emptying their own votes. Finally, the decreased costs that empty voting offers may allow new idiosyncratic voters to enter the votes market, thus making the problem even worse. "s See Schreiber v. Carney, 447 A.2d 17, 24 (Del. Ch. 1982). No. 3:883] EMPTY VOTING Similar to the fairness concern discussed in Section C, the concerns here can be most accurately characterized as issues relating to the "vote market." The market depiction here, however, assumes a market that already distorts incentives away from the ideal in which shareholder interests are generally aligned. This distorted market creates openings for empty voting to be profitable, which in turn magnifies and exacerbates the existing distortion, which subsequently may lead to more empty voting in a cycle that feeds on itself. IV. POSSIBLE SOLUTIONS TO THE REMAINING PROBLEMS Although public sector regulation and expansion of existing law may be greatly beneficial, many of the problems discussed in Part III are those that private-sector institutions, notably the corporations themselves, would have strong incentives to solve should the concerns become great enough. Many of the solutions include actions which regulatory agencies such as the SEC have been trying to promote. These private-sector responses may include the adoption of disclosure requirements that help ferret out problematic empty voting, measures to reduce voting "error," and adjustments in voting structures that could mitigate distortions in the market for votes. These possible responses will be discussed in the following sections. A. Increase Monitoring The first possible response is to increase the likelihood of detecting abusive empty voting practices. One method may be to disclose the voting decisions of the top twenty percentage shareholders following every meeting, whether they voted to approve, reject, or did not vote. This may make it easier for individuals to find irregular voting patterns or anomalies that suggest empty or conflicted voting and may help to deter it from happening in the first place. Widespread adoption may also decrease monitoring and enforcement costs by increasing the likelihood of detecting both conflicting positions across corporations and patterns of COL UMIA B USINESS LA W RE VIE W [Vol. 2007 EMPTY VOTING behavior that suggest collusion. Two caveats, however, are in order. The first is that this practice may, to some extent, result in frivolous litigation as individual monitors "detect" problem patterns that are not actually problematic. The second is that in some limited scenarios, it may actually increase the possibility of secret voting agreements by providing readily verifiable evidence of how shareholders have voted. This second caveat assumes that the inability to verify whether a shareholder has indeed voted according to the terms of a clandestine agreement makes it more difficult to enter into these agreements in the first place. Another method would be to amend the charter to require the disclosure of leveraged voting positions, making the ability to vote contingent on this disclosure being current.5 9 Although this would have the adverse effect of increasing voting costs for hedged shareholders, its value in revealing empty or partially empty positions may be sufficient to offset this effect, particularly if the method of disclosure were relatively simple. A proposal by Fraidin and Hoverman simply requires voters to certify the economic ownership of the shares that they are voting.60 As mentioned earlier in Part III, however, one of the important benefits of increased monitoring would be the ability to ferret out collusive or parallel behavior. In the absence of widespread adoption, these particular benefits would be limited. To the extent that the increased costs of these measures outweigh the more limited benefits in the absence of widespread adoption, individual corporations may not have an incentive to adopt measures that increase monitoring. To the extent that this is the case, a universal regulatory approach, such as the one suggested by Hu and Black,6" may be more effective. " Because such a change would potentially affect voting rights, an amendment to the certificate of incorporation would be required. See DEL. CODE ANN. tit. 8, § 212(a) (2007). o Fraidin & Hoverman, supra note 45, at 416. Fraidin & Hoverman additionally note that because such an amendment would affect the proxy machinery, SEC approval may be required. 61 Hu & Black, supra note 1. No. 3:883] B. Increase Voting Accuracy Another possible response is to reduce both voting error and voting costs. One method to reduce error may be to increase the clarity and reliability of disclosures regarding proposals. Another possibility may be to enhance the perceived reliability of information conveyed by voluntarily accepting greater potential liability, such as by waiving certain defenses regarding the accuracy of specific statements. Although materially accurate disclosures are currently required under federal law, the presence of potential abuse by empty voters may have the ironic effect of increasing the incentives of significant shareholders to ensure that other similarly long shareholders vote in the corporation's best interest. It may be possible to reduce voting cost by tying certain dividend payments to whether or not a shareholder has voted. This measure would have the effect of offsetting voting costs-perhaps to the point where shareholders gain a net benefit from voting regardless of their ability to influence the outcome-thereby decreasing voter apathy. This measure, however, may be problematic for several reasons. The first is that it would have the effect of discriminating against shareholders who choose not to vote, thus creating disparate treatment of shareholders who own the same class of stock.62 Although Delaware courts generally view dividend payments as contractual obligations subject to any "fixed scheme"63 and have allowed some degree of disparate 62 Generally, in the absence of a clear agreement otherwise, all shares of stock should be treated equally. See EDWARD P. WELCH ET AL., supra note 19, § 151.4. ' Penington v. Commonwealth Hotel Constr. Corp., 155 A. 514, 517 (Del. 1931). DEL. CODE ANN. tit. 8, § 151(a) (2007) also provides that the rights . . . limitations or restrictions of any . . . class or series of stock may be made dependent upon facts ascertainable outside the certificate of incorporation . . . provided that the manner in which such facts shall operate upon the ... rights . . . limitations or restrictions of such class or series of stock is clearly and expressly set forth in COL UMIA B USINESS LA W RE VIE W [Vol. 2007 treatment within a class in the case of tiered' or tenured65 voting rights, adoption of such a measure would almost certainly require a charter amendment approved by the shareholders.66 The second concern is that although tying dividend payments to voting may increase incentives to vote, it would not necessarily increase incentives to vote in an informed manner. To the extent that those shareholders who are least likely to vote now are those who would also be the most likely to vote incorrectly if they did vote, implementing voting incentives may actually have an adverse effect. C. Implement Alternative Voting Structures Finally, alternative voting structures may be able to mitigate many of the concerns associated with inefficient vote markets. One possible method may be the weighted English structure discussed by Martin and Partnoy,7 which limits a shareholder's voting influence by requiring more economic shares for each voting right after certain thresholds. One example, upheld in Providence & Worcester Co. v. Baker,6" provided one vote per share for the first fifty shares, but only one vote per twenty shares for all shares above fifty. Although empty voters could theoretically still hedge against the greater economic stake required for increased voting influence under such a structure, this tiered voting structure would drive up empty voting costs considerably. A tiered structure would also counteract the the certificate of incorporation or in the resolution or resolutions providing for the issue of such stock. Providence & Worcester Co. v. Baker, 378 A.2d 121 (Del. 1977). 65 Williams v. Geier, Civ.A. No. 8456, 1987 WL 11285 (Del. Ch. 1987). See id. at *1. A possible alternative arrangement may be to issue to all shareholders a second class of shares with preferred dividend rights that convert to the common if the shareholder fails to vote, and to include a convertibility option on the common to the preferred which vests after a set history of voting. For many reasons, a mid-stream implementation is a much thornier task than adopting such a scheme at the outset. 6 Martin & Partnoy, supra note 3, at 782-83. 68 Providence & Worcester Co., 378 A.2d at 121. EMPTY VOTINGNo. 3:8831 anomalies of the vote market described in Part III by increasing the cost of additional votes as one gained greater voting influence. Although a mid-stream change in voting structure has generally been disfavored, Fraidin and Hoverman have suggested an alternative method for achieving a similar result. 9 Fraidin and Hoverman's proposal suggests declaring a dividend of non-voting stock, then applying a price protection provision that requires shareholders to own a specified number of non-voting shares in order to exercise their votes at certain thresholds. 0 Delaware statute permits the board of directors to issue new stock by resolution if this power is expressly granted by the certificate of incorporation,7 and Delaware courts have interpreted this statute to "permit[] the creation of stock having voting rights only, as well as stock having property rights only."72 Although the New York Stock Exchange ("NYSE" or "the Exchange") rules generally provide that the "[v]oting rights of existing shareholders ... cannot be disparately reduced or restricted through any corporate action or issuance [of stock] "7 a proposal substantially identical to the one described above was deemed permissible under the Exchange rules.74 69 Fraidin & Hoverman, supra note 45, at 416. 70 Id. 7' DEL. CODE ANN. tit. 8, § 151 (2007). Section 151(g) provides that such a resolution will have the effect of amending the certificate of incorporation. Under the doctrine of "independent legal significance," this would indicate that the board may issue new classes of shares without the need for a shareholder vote. See, e.g., Nat'l Educ. Corp. v. Bell & Howell Co., C.A. No. 7278, 1983 WL 18035 at *3 (Del. Ch. 1983) (denying injunctive relief to shareholders in case where directors bypassed the need for a supermajority vote by amending the certificate of incorporation through board action). 72 Lehrman v. Cohen, 222 A.2d 800, 807 (Del. 1966). 73 NYSE, Inc., Listed Company Manual, § 313 (2006). 71 See Para. 313.00 Interpretation No. 95-02, available at http://www.nyse.com/pdfs/voting-rights.pdf; see also Para. 313.00 Interpretation No. 96-02, available at http://www.nyse.com/pdfs/ voting-rights.pdf. (Vol. 2007COL UMBIA B USINESS LA W RE VIE W EMPTY VOTING Adopting an alternative voting structure may be able to offset many of the problems related to the votes market. Although this structure would not completely eliminate empty voting, it would reduce both the profit margin of problematic empty voting and the prevalence of vote capture by conflicted shareholders, both of which should reduce the frequency of empty voting. V. CONCLUSION Though empty voting practices pose some concern, many of the most abusive practices are addressed by existing legal doctrine. The problems that remain, such as difficulties in detection, voting inaccuracy, and anomalies within the votes market, have the beneficial effect of creating incentives for directors and majority shareholders to adopt measures that benefit minority shareholders. Rather than being a significant problem that can only be addressed by government action, the new empty voting may align incentives in a way that causes directors and incumbent boards to voluntarily pursue actions that previously were required by federal mandate. The threat of empty voting may induce directors to establish disclosure regimes to help detect abusive empty voting, convey information to shareholders in a clearer and more accurate manner, and adjust voting structures to better account for the eccentricities of the votes market. Although regulatory solutions to empty voting concerns may be preferable when responses by private actors are more expensive to implement, the alignment of the board's incentives in empty voting contexts and their relative ability to respond to the unaddressed problems suggest that the government's best response may be to do nothing. No. 3:883]