The Possibility of Inefficient Corporate Contracts, 60 Cincinnati Law Review 387 (1991)
Return to Professor Ayres' Homepage
Return to Professor Ayres' Resume

As in other areas of law, the application of game theory to corporations began with the use of the prisoners dilemma. Several scholars noted that the use of two-tier tender offers might place target shareholders with the dilemma of either tendering their shares at a price below their evaluation or facing a the back end freeze out purchase at an even lower amount. The inroads of the "new learning" in game theory have, until very recently, been rarely applied to corporate law issues. This "new learning" developed by economists in the 1970's and 1980's consists largely of theoretical breakthroughs in how to model dynamic games in which "players" have private information.

For those uninitiated in this new learning, let me suggest three ways to tell when you are reading a game-theory piece that is by-product of this new learning. First, games will often be depicted with game trees instead of pay-off matrices. Second, the game will specify that there are two (or multiple) types of a particular class of player. And third, there will often be asymmetric information. If you seen an article with any one of these attributes, odds are that the author is toiling in this new vineyard.

Not surprisingly, these three shibboleths are related. The use of game trees (what game theorists call extensive form representations) much more adeptly captures informational assumptions and their impact on how the game is played. And the assumption that there are two types of a particular class of player is often used to generate asymmetric information when other players in the game cannot tell initially what type they are playing with.

The explicit introduction of multiple types of players also represents an important advance in the economic analysis of contractual bodies of law, such as corporations, because if there are heterogeneous types on one side of the contract, then efficient contracting will often necessitate heterogeneity in the contractual equilibrium. This is a significant departure from what I would call "first generation" law and economic analysis of contracts, which often sought to discover the single legal rule that would maximize the gains from trade for the contracting parties. This is also often the mode of corporate and economic analysis: many articles propose a single legal rule that promotes the efficiency or gains of trade for all corporate contracts. But this "first generation" mode of analysis, in an important sense, proves too much: if the authors have really discovered a rule that enhances efficiency for all contracting types, then one should be agnostic about whether to make the rule mandatory or merely a default that parties can contract around. Since there is a strong consensus that many corporate and contractual rules should be defaults that parties may contract around, the new game theory models which generate contractual heterogeneity are much better suited to analyze the three primordial questions of contractual freedom:
  1. should a particular rule be immutable?
  2. if not immutable, what is the most appropriate default?, and
  3. if a default, what are the necessary and sufficient conditions for contracting around the rule?

In this piece, I would like to continue a discussion, or more precisely revise an answer that I gave to Dan Fischel a couple of years ago at Chicago's Law and Economics Workshop. Rob Gertner and I were presenting an article about how to choose efficient contractual defaults. Among other things, Rob and I argued that contractual parties with private information might strategically refuse to bargain for socially efficient rules in order to protect the returns to their private information. A part of our article suggested that efficiency-minded lawmakers might at times want to promulgate "penalty" defaults that induce the private contracting parties to contract around the undesirable defaults and thereby reveal their private information. Although this contract article proposed a few applications in the corporate context, Fischel asked at the seminar whether our model had any implications in the corporate context. His interest in the issue is not surprising. Easterbrook and Fischel have forcefully argued that corporate default rules should "duplicate the terms the parties would have selected . . . if they had contracted explicitly." If the possibility of inefficient strategic bargaining in private contracts could be generalized to the corporate context, these new game-theoretic models might limit or qualify this hypothetical contracting norm. Although at the time of that seminar, I told Dan that I thought there were few possibilities for the inefficiencies of strategic bargaining in the corporate context, I am here today to recant in a small way and give a slightly different answer. In particular, my thesis is that strategic interactions may lead to inefficient corporate contracting (a) even in a world where there are numerous shareholder/investors competing to make investments and (b) even when it is costless to contract around a given default.

At first blush, one might argue that game theoretic approaches to modeling the corporate governance contracts of large publicly traded businesses are inappropriate because the cost of contracting around any default rule will be negligible in proportion to the potential effect on corporate value. If either side can costlessly contract around a given default rule, traditional law and economics scholarship would posit that parties would be able to find and include the contractual provisions that maximize the gains from trade.

Secondly, the potential for strategic inefficiency might be limited by the sheer number of players that ex ante are potentially on the various sides of the corporate contract. Numerosity, after all, undermines strategic interactions in the standard price theory models. This piece argues that neither costless contracting nor the ex ante numerosity of potential contracting parties sufficiently ensures efficient contracting. To support this thesis, the article will analyze a model based on a seminal article by Rothschild and Stiglitz (and recently updated by Aghion and Hermalin) which demonstrates how asymmetric information and adverse selection can generate inefficient corporate debt contracts even though contracting is costless and there is vigorous competition among numerous lenders. I will then suggest how the results of this corporate debt model could be applied to the inefficient selection of more traditional legal rules. In particular, I'll discuss how attempts at "signaling" can lead to inefficient separation in the equilibrium contractual terms regardless of the initial default choice. Fischel and Easterbrook are vindicated in so far as they argue that default choice does not change the equilibrium -- but they are wrong in arguing that the equilibrium will necessarily be efficient.

I also briefly discuss types of government intervention that may be successful in improving a strategic problem. Having done all of this, in the final portion I will try to undercut this simple signaling model by discussing how signaling theory does not explain major aspects of current corporate contracting in this age of enabling statutes. In the end, Easterbrook and Fischel's hypothetical contracting norm remains a powerful tool for analyzing default choice. My thesis is not that corporate contracts are inefficient, but that we can't rely on small transaction costs or numerosity to ensure efficiency. To the extent that we retain a belief in the efficiency of the corporate nexus of contracts, this article stimulates us to look further for the structural forces that foster efficiency. In a world in which major financial decisions are best analyzed as signals in models of asymmetric information, the possibility for inefficient "excessive" signaling through contractual provisions should remain an important area for further research. Because signaling theories so dominate the portion of the corporate contract relating to financial structure, it would be premature to wholly discount the possibility that inefficient signaling has infected the portion of the corporate contract relating to governance structure.


Return to top of the Page