by Ian Ayres*
Preliminary Draft: August 5, 1996 (9:32a)
*William K. Townsend Professor, Yale Law School
PO Box 208215
Jennifer Brown provided helpful comments.
New Haven, CT 06520
203 432-7101
203 432-8260 (fax)
203 624-5654 (home)
ayres@mail.law.yale.edu (e-mail)
http://elsinore.cis.yale.edu/lawweb/faculty/ayrespub.htm
Table of Contents
Introduction 1
I. The Game-Theoretic Paradigm 2
II. The Contractarian Paradigm 6
Conclusion 11
Introduction
This excellent article argues that interjecting lawyers into the regulatory game played by firms and governmental agencies might foster socially-desirable cooperation. In particular, Professor Painter shows that regulators can manipulate the payoffs of attorneys to give them incentives to influence their clients to behave better. Even though Painter uses the adjectives "game-theoretic" and "contractarian" to organize his two major arguments, it might be clearer (and perhaps more parallel) to divide his two arguments into "reputational" (or non-legal) and "legal" effects on lawyers' cooperative payoffs.
In a nutshell, the "game-theoretic" section of the article emphasizes that lawyers' structural need to establish a reputation that will allow them to represent multiple clients restrains lawyers' temptation to defect from socially desirable types of cooperation. Painter applies this argument not just to firm lawyers (who would fear that defecting in one case will undermine their ability to effectively represent other clients before a particular agency), but also to agency lawyers. Even government attorneys who currently represent only one (agency) client may wish to establish a reputation for being neither a "pushover" nor a "jerk" -- so that they can develop better future job prospects. The core game-theoretic idea is that the structure of lawyers' payoffs -- because of the multi-client reputational concern -- might militate toward cooperation -- especially when "(i) firm lawyers and agency lawyers are repeat players who deal with each other on multiple occasions, (ii) noncooperative conduct by either is easy to detect and (iii) information about lawyers' reputations is readily available."
Painter is careful though not to be too Panglossian about these reputational effects. He acknowledges that when these three preconditions fail, lawyers may still aid their clients in defecting (or fail to sufficiently discourage client defection).
Because the reputational effects on lawyers' payoffs are not always sufficient, Painter also considers more direct ways that professional responsibility law could give lawyers better incentives to cooperate. Painter names this argument "contractarianism" because he envisions lawyers and agencies contracting for professional responsibility duties that are tailored to the specific characteristics (a.k.a. structure) of the industry.
Thus, while the title speaks of an "uneasy relationship between regulators and regulatory lawyers," there is no uneasy relationship between the game-theoretic and contractarian arguments: They are both stories about how lawyers' conditional payoffs (that is their payoffs for cooperation vs. defection) can induce cooperative behavior. I am deeply sympathetic with Painter's enterprise, but thrust, as I am, in the role of commentator, I will offer a few criticisms of his game-theoretic and of his contractarian analysis.
I. The Game-Theoretic Paradigm
The first half of the article extends and applies a prisoner's dilemma model of the "regulation game" originally formulated by John Scholz. John Braithwaite and I extended the model simply by thinking about how "capture" would effect the equilibrium of the game. We modified the prisoner-dilemma setup by assuming that when firms capture the "hearts and minds" of agencies, they cause the agency's payoffs to become more like the firm's payoffs. Like Painter, we were interested in how changed payoffs would affect cooperation. Our book was relatively ungrounded compared to Painter's analysis and many of the difficulties that Painter encounters are because he sets himself the harder task of relating a reductive game-theory model to a richer set of stylized facts.Ian Ayres, Playing Games with the Law, 42 Stan. L. Rev. 1291, 1294-95 (1990) ("Law review articles continue to be mindlessly mired in the game theory "technology" of the fifties. Countless articles rearticulate the Prisoner's Dilemma, but few even proceed to other bi-matrix games.")
Having said this, I am still concerned that Painter never explicitly includes lawyers in his game-theoretic models. After painstakingly setting out a game played by "firms" and "agencies," he never gets around to explicitly modeling (firm or agency) lawyers. The crucial move comes more than one-quarter of the way into the piece when Painter asserts that:
Because lawyers play their games at the same time as they advise clients on how to play the underlying firm/agency games, lawyer strategies influence client strategies and vice-versa.
While this claim has superficial appeal, the article would be much stronger if the mechanisms of influence were more tightly specified. If we stick with the dichotomous choice, defect or cooperate, it is far from clear how we should model lawyer's influence on client's behavior. Should we assume that a firm lawyer's threat of whistle blowing can force firms to cooperate (when the firm would prefer to defect)? Focusing on the influence mechanism might have yielded valuable insights. For example, it might be that lawyers have a harder time forcing firms to defect (than forcing firms to cooperate) -- so that lawyer influence tends to rachet firms toward cooperation. Or given Painter's detailed knowledge of the interaction, he might have been able to break out of the simple prisoner's dilemma dichotomy and tell a richer story of the ways lawyer strategies influence client strategies.
One trivial way that lawyers might have been included in the game would be simply to relabel the axes of the matrix to make the prisoner's dilemma game between the firm lawyers and the agency lawyers. The twist would be to argue that the lawyers' reputational interest gives them a lower discount rate (than their clients) and thus might lead them to act more cooperatively. But I'm not sure that the explicit game-theory model is doing enough work here to justify its inclusion. The basic reputational argument is that "because lawyers representing agencies and lawyers representing firms play the 'regulatory game' with each other on a frequent basis, cooperative play between them can have substantial payoffs for both." But this thesis was well understood in certain contexts without formal game theory. Painter is persuasive when he argues: "because lawyers play their own games cutting across client representations at the same time as they participate in formulation of client strategies, firm lawyer/agency lawyer games influence the outcome of the underlying firm/agency games" But the persuasiveness does not come from the explicit game-theory.
Now in fairness, Painter does try to build on the non-intuitive result of the Ayres/Braithwaite capture model -- that is, that there are multiple forms of capture, including a welfare-enhancing form. Painter argues that agency lawyers have "a lower susceptibility to inefficient and zero-sum capture" but are "prone to efficient capture." Making this distinction provides some justification for trotting out the big technological guns of a game-theory model. But Painter's ultimate distinction turns on a rather strong assumption about what Painter calls a "reputational paradigm" which posits that agency lawyers "invest substantial human capital in building a reputation for thoroughness, integrity and zealous representation of their [agency] clients." While Painter helpfully lists several aggressive government attorneys who have flourished in private practice, I can think of several non-zealous lawyers who have found think-tank sinecures waiting for them at the end of their government service. At the end of the day, while Painter persuades that the repeated interaction between firm and agency lawyers may foster cooperation, I'm not sure that the explicit game theory served to illuminate the phenomenon.
II. The Contractarian Paradigm
The second half of the paper is a tremendously ambitious attempt (i) to describe the ways that agencies contract with firm lawyers for tailored professional responsibility duties and (ii) to envision how a much more explicit contractual regime might be implemented. In the end, Painter argues for professional duties that have three core attributes:
1. Tailored (instead of untailored) duties. The duties should be tailored to the needs of the particular representation context. Agency-specific, tailored duties can better serve the public interest than untailored (a.k.a. "one-size-fits-all" or "off-the-rack") duties.
2. Default (instead of immutable) duties. The parties should have a mechanism to contract around presumptive or default duties to allow even more tailoring.
3. Rules (instead of standards). Clarity is a primary value in creating enforceable ethical duties.
Combining these attributes, Painter argues for tailored (agency-specific) but precise (rule-like) default duties. While Painter makes a powerful case for agency-specific duties, he might have done more to defend his arguments for defaults (instead of immutable) duties and for rules (instead of standards). As a theoretical matter, these three dichotomous choices give rise to eight possible permutations. One way to frame my criticism is that I am not sure that Painter has sufficiently considered all of the individual permutations.
I am particularly worried that Painter's analysis at times conflates the tailoring and default dimensions -- so as to over look the possibility of tailored immutable rules. For example, Painter characterizes the SEC's holding in In Re Carter and Johnson as "infer[ing] a 'quasi contractual' understanding from its relationship with an entire group of lawyers, those who practice security law." As Painter explains, the opinion imposes heightened ethical duties on lawyers who take on "significant responsibilities" with respect to SEC filings. These heightened duties are at times described as being quasi-contractual but at other times more like "an implied-in-fact" contract based upon "implied understandings between regulators and lawyers." Beyond the doctrinal difference that quasi-contracts are implied-in-law instead of implied-in-fact, Painter's analysis fails to capture the core dimensions of the decision -- that it established a tailored, immutable standard. Painter is persuasive in arguing that the immutable standard did "not provide lawyers with clear guidance for future conduct," but he is not persuasive that tailored immutable rules would not have been more appropriate. And In Re Carter and Johnson shows that tailoring might be conditioned not merely on agency-specific representation but on the types of representation ("significant responsibilities") at issue. It mischaracterizes this line of cases to think of the heightened duty as the byproduct of some type of "quasi-contractual understanding." Rather, it is an immutable (albeit fuzzy) duty that is unilaterally imposed by the lawmaker.
It is initially hard to conceive of agencies engaging in ex ante negotiations over ethical duties, but Painter imaginatively begins to consider what such a world would look like. He might have spoken more about the mechanisms for contracting around default ethical duties. For example, the article doesn't say whether the corporate clients as well as individual attorneys (and the attorneys' partners) would be necessary parties to the ethical contract. It also doesn't address whether government attorneys would be subject to contractible ethical duties. Given the importance of government attorneys in inducing cooperation in the first half the article, it is somewhat incongruous that these attorneys play such a minor role in the second half of the article.
But Painter does consider an interesting menu of "default rules for lawyers to choose among." The agency here would not individually bargain ex ante but would give firm lawyers a unilateral option to choose among a restricted list of ethical duties. I could imagine allowing a lawyer to disclaim certain ethical duties and thus signal the SEC that it cannot rely on the attorney as whistle blower.
Painter goes most off the mark, however, in his tentative analysis of default choice. Early on, he makes the bold and interesting claim that duties should be contractible (defaults) because default duties are more likely to be cast as clear rules (instead of fuzzy standards). His argument that defaults should be rule-like instead of standard-like is at bottom a conventional (but not therefore necessarily wrong) argument about clarity:
[I]mmutable rules can be very unclear when drafted in circumstances where different constituencies that participate in the rule-making process cannot agree on what a lawyer's responsibilities should be. The result may be convoluted language that reflects search for compromise rather than clarity. . . . By contrast, lawyers and regulators have less vested interest in default rules and are likely to value clarity in such rules over content, knowing that they can probably contract around a rule they don't like. . . .This clarity alone is a substantial benefit to those lawyers, perhaps a majority of lawyers, who are not as concerned about the rules as they are concerned about knowing what the rules are.
I want to pause to praise this quotation as a persuasive and new predictive argument about the likely content of immutable and default rules. However, while Painter persuades that immutable duties are likely to be unhelpful standards, his implicit conclusion that default standards would be less helpful than default rules does not follow.
First, Painter does not consider that default standards might be valuable because they might induce more explicit contractual tailoring. Painter acknowledges that "penalty default" duties might be appropriate "to encourage lawyers to negotiate their own tailor-made rules with the agency," but he does not see that by this very logic the uncertainty caused by a standard-like duty might induce lawyers to contract. Painter repeatedly argues that the rules that lawyers and agencies agree to substitute for default rules are likely to be clear and easy to follow, and given the high value that Painter places on tailoring, this penalty default justification for default standards deserves more consideration.
Second, Michael Klausner and I have each offered non-penalty justifications for default standards in the corporate context. We argue that contracting parties may have a much harder time contracting for standard-like duties than contracting for rule-like duties. The utility of standards relies much more on building a thick common-law precedent of interpretation and it may be harder to develop such precedent if private parties independently choose a variety of fuzzy provisions in order to contract around a precise default rule. Painter has not convinced me that some substantial minority of contractors (attorneys/firms/agencies) would not jointly prefer to have attorneys bound by some open-ended standards that attorneys must act "reasonably," or "in good faith," or not "overreach". While each of these standards is imprecise, it may allow attorneys to commit to their clients and the agencies that they will not in some sense "defect." Indeed, returning to the "game-theoretic" analysis, if we reintroduce the repeated reputational contact of the agencies, firm lawyers need not be nearly as scared by exposing themselves to open-ended liability of a standard since the reputational constraints might deter egregious agencies opportunism in its interpretation. Even if only 10% of the contracting parties jointly prefer to be bound by a standard, it might be best to choose the standard as a default (so that a thick interpretive precedent could most easily form) and let the remaining 90% explicitly contract for the clarity that they prefer.
Conclusion
This fine article reminds us that discussions of regulatory enforcement often forget the prominent role that both agency and firms lawyers can play in inducing cooperative reliance. Attention to the structural/reputational influence and more direct legal influence on lawyer's incentives can yield better policy. In game-theoretic terms, Painter's attention to lawyer's "payoffs" thus has an important payoff of its own. The article's foray into game-theoretic and contractual analysis has few professional responsibility antecedents. Painter, however, has made out more than a prima facie case for further application of these tools to the difficult problems of legal ethics.