| Part I |
| Fiduciary Duties and Corporate Governance |
1 | Controlling Corporate Self-Dealing |
| Convergence or Path-Dependency? |
| Zohar Goshen |
The problem of corporate self-dealing is a manifestation of the fundamental âagency problemâ pervading corporate law (Jensen and Meckling 1976). The self-dealing problem may be found in a wide variety of corporate transactions such as those between a corporation and the party that controls it, a subsidiary, or a director or officer of the corporation.1 Legal systems in different countries address the self-dealing problem in different ways. For instance, in Delaware, transactions involving a conflict of interest are subject to the âentire fairnessâ test,2 while in the United Kingdom transactions involving conflict of interest require the approval of the disinterested shareholders (â a majority of the minorityâ).3
Which approach is preferable? Convergence theories imply that there exists âone efficient corporate law,â while path-dependency theories suggest that the efficient law might vary depending on the specific characteristics of a given country. Is there, indeed, âone efficient solutionâ to self-dealing? Answering this question requires an analysis of the economic effects of the diverging solutions provided in different countries. The primary aim of this essay is to provide such an analysis through the application of the liability rule/property rule distinction (Calabresi and Melamed 1972) to the different solutions provided for corporate self-dealing.
A property rule validates consensual transactions. A transaction can be performed only with the consent of the parties, at a price that is a function of their subjective evaluation. This category includes systems that require the approval of the disinterested âmajority of the minorityâ to validate self-dealing. A liability rule, on the other hand, validates nonconsensual transactions. A transaction can be imposed on an unwilling party subject to objectively determined adequate compensation. This category includes systems that use the âfairness testâ as a measure to validate self-dealing.
Characterizing the solutions to the self-dealing problem as either property-rule or liability-rule protections provides a powerful framework to examine the parameters influencing the choice between them. The choice between a liability rule and a property rule is affected by adjudication and negotiation costs. The level of these costs, in turn, is influenced by factors such as the effectiveness of the judicial system and the efficiency of market mechanisms. Consequently, the efficiency of a given rule is dependent on the economic, legal and social conditions in a given country, ruling out the possibility of âone efficient solution.â Indeed, the efficient solution to corporate self-dealing is path-dependent.
Section I describes the voting process as a means of extracting group consensus and presents the problem of conflicts of interest in voting. Section II analyzes the possible solutions to the problem. In Section III, the liability/property rule distinction is applied to the fairness test and the majority of the minority rule. Section IV describes the different factors affecting the choice of an optimal solution. Section V examines the relative weight of each factor empirically, as these come into play in Delaware and the United Kingdom.
I. On Voting and Self-Dealing
Voting is most commonly accepted as the best method for extracting group consensus from among the disparate subjective assessments of individuals within a group. The voting mechanism is based on the assumption that the majority opinion expresses the âgroup preference,â that is, the optimal choice for the group as a whole (Nitzan and Procaccia 1986). In voting for or against a transaction, each member of the group subjectively assesses the merits of the proposed deal and expresses her particular informational perspective. The voting process aggregates the subjective assessments of individual group members into a single coherent stance that expresses the groupâs consent. The majority view is presumed to be the best indicator of an efficient transaction (Nitzan and Paroush 1982; Bebchuk 1988).
Voting, however, is by no means foolproof.4 It does provide an effective means of formulating the groupâs stance from among the various individual positions of its members, but only if each memberâs vote is based on an honest appraisal of her best interests as a member of the group (âsincere votingâ) (Sen 1973). Whenever voters take into account how other members of the group will vote (âstrategic votingâ) (Goshen 1997) or vote according to a personal interest conflicting with the interest of the group (âconflict-of-interest votingâ) (Goshen 2003), the voting procedure ceases to function as an indicator of efficient transactions. Here I focus on solutions to the problem of conflicts of interest.
The conflict-of-interest problem manifests itself in circumstances where some voters in the group have interests as members of the group which conflict with their interests external to the group, resulting in balloting which does not necessarily express the âgroup preference.â The problem is common where a transaction is proposed between the group and one of its members. Then, the basis for the voterâs decision will no doubt focus on self-interest and her personal stake in the outcome, and not on the transactionâs value for the group as a whole. A conflict-of-interest situation may therefore neutralize the voting mechanismâs ability to determine group preference (Arrow 1963).
Nonetheless, it does not necessarily follow that all transactions bearing an element of conflicting interests are bad (inefficient) transactions. In certain situations, a transaction with an interested shareholder may be the best option available to the group. Indeed, a self-dealer may have a competitive edge in the market, or even an advantage stemming from her proximity to the group, which ensures that a deal with her is in the groupâs best interests.
The fact that self-dealing may be either good (efficient) or bad (inefficient) is at the root of the conflict-of-interest problem. It requires a system that can screen self-dealing and provide a mechanism that maximizes the execution of efficient deals and minimizes the execution of inefficient ones. The main corporate law mechanisms designed to accomplish these goals are reviewed below.
II. Possible Solutions to Self-Dealing
A. Prohibition on Self-Dealing
One extreme solution is the outright prohibition of self-dealing. This solution endorses a fundamentally negative view of any transaction tainted by a conflict of interest. Historically, courts adopted a solution consistent with this view: any deal born of conflict-of-interest voting was voidable and could be repudiated by the corporation, regardless of its terms or its desirability to the corporation (see Marsh 1966). Indeed, this approach âsolvesâ the problem; self-dealing will rarely occur under a regime prohibiting it. If self-dealing is considered pernicious, an outright prohibition is a simple solution: such a rule is easy to apply, obviates the need to grapple with evaluations, and prevents most bad (inefficient) deals. If, however, the initial position is that a significant number of transactions are efficient despite the presence of a conflict of interest, an outright prohibition will exact too heavy a price: the loss of too many efficient transactions. An outright prohibition is irreconcilable with the goal of increasing the performance of efficient transactions, and is therefore too extreme to serve as a general solution to the problem.
B. The Fairness Test
Another approach allows the self-dealer to vote but requires that the minority receive fair compensation. When a minority claims the compensation is unfair, a courtâs scrutiny will be needed to objectively determine the fairness of the deal. Suppose, for example, that a controlling shareholder has sold the company an asset for $100. A claim that the transaction violates the fairness test will obligate the court to determine whether the asset is indeed worth $100.
A solution based on the fairness standard, in effect, allows the person with a conflict of interest to effect a âtakingââto impose the transaction on the minority, subject to the right to challenge the transaction as unfair. The minority, however, is not assured the best attainable deal. The fairness protection is no more than a guarantee that the transaction will be fair, and that the minority will gain some portion of the profit reaped by the transaction on terms similar to those that might be expected of a transaction between willing buyers and sellers.
This solution requires routine recourse to the courts for an objective assessment of the transaction. When called upon to do so, the court must base its decision on value assessments made by professionals. However, a determination of the âobjectiveâ value of an asset is not an exact science: evaluations are influenced by subjective assumptions. Moreover, evaluations are subject to tendentiousness; specifically, a rendered opinion is liable to be slanted to meet the demands of the party who has commissioned it. This will often mean favoring the party interested in seeing the transaction performed (Bebchuk and Kahan 1989). The court will be compelled to decide between the inevitably differing opinions tendered by the opposing parties in order to determine the âcorrectâ fair value. Yet, despite the drawbacks of a reliance on courts and professional evaluations, there is no other means of determining the objective value of an asset.
C. The Majority of the Minority Rule
Another solution to the problem is to use the voting mechanism as a means of determining the groupâs consent. Those with a conflict of interest are excluded from the vote, resulting in the decision being made by âa majority of the minority.â The vote of a self-dealer contributes no pertinent information as to the benefits of the transaction to the group as a whole. Indeed, only those members of the group with no ulterior interests are relevant if the vote is to express the âgroup preference.â
The ban on conflict-of-interest voting will prevent a self-dealer from imposing a transaction on an unwilling minority. That is, the minorityâs consent is required. Since such an approach is based upon the subjective assessments of the participants in the ballot, it is not necessary to bring the transaction before the courts for an objective evaluation. If the remaining participants in the ballot (the minority) form a large group, it will be reasonable to assume that the vote does, in fact, reflect the group preference. Placing decisionmaking in the hands of the minority, however, is liable to prevent the attainment of efficient transactions in certain situations. When the minority is composed of a small group, the threat of strategic voting will arise. Since the interested majority obviously will support the transaction, the minority, or some of its members, can attempt to hold out for a larger share of the transactionâs expected profit. So long as the extorted sum leaves the majority with some amount of profit, the transaction may still be performed. But if the minority, or some of its members, hold out for too great a share, an efficient transaction may be lost. Likewise, even a âreasonableâ hold-out will preclude a transaction if the majority refuses, for strategic reasons of guarding its reputation, to accede to the minorityâs demands.
D. Nonintervention
An approach at the other extreme of possible solutions is to avoid any intervention and leave the issue to market forces. It might be argued that nonintervention will provide corporations with the opportunity to generate appropriate tailor-made protections against self-dealing. Investors, for their part, will pay an appropriate price for a security with a defense against self-dealing, and little or nothing for one without such protection (see Easter-brook and Fischel 1982). In an efficient market, and absent transaction costs, the prices of securities will reflect the value of the different defenses they carry. In such a market, each corporation can accord its securities with the protection most appropriate for its needs, and each investor can choose the type of protection she requires. Nonintervention, therefore, would prove to be an effective solution in a perfect market (Fischel 1982).
However, where transaction costs are present and the market is not efficient enough to accurately price the different securities, nonintervention will fail. Different protections or their absence will not be priced accurately due to free riding and asymmetric information. Consequently, a dynamic leading to diminishing standards known as âthe market for lemonsâ (see Akerlof 1970; Leland 1979) will lead firms to arrive at a common and inefficient point: no defense. The conclusion that nonintervention is inefficient in imperfect markets is supported by empirical findings.5 For this reason, imposing a mandatory protection will be more efficient, leading to information cost savings and reduced transaction costs (see Gordon 1989; Romano 1989; Klausner 1995).
III. Characteristics of the Solutions
The preferable solutions that emerge out of the foregoing discussion are the âfairness testâ and the majority of the minority rule. The ensuing discussion, accordingly, will concentrate on these solutions and their basic tenets.
A. Property Rules Versus Liability Rules
The protections against self-dealing can be classified ...