[*pg 1405]

A THEORY OF LEGAL STRATEGY

LYNN M. LOPUCKI

WALTER O. WEYRAUCH

ABSTRACT

INTRODUCTION

I. THE STRATEGIC PERSPECTIVE DISTINGUISHED

A. The Texaco-Pennzoil Case as Illustration

B. Legal Strategy in Law-and-Economics Theory

II. THE FOUNDATIONS OF LEGAL STRATEGY

A. Definitions of Law

B. The Relationship Between Written Law, Social Norms, and Expectations Regarding Outcomes

C. The Residual Role of Legal Doctrine

III. THE VARIETIES OF LEGAL STRATEGY

A. Strategies Requiring Willing Acceptance by Judges

B. Strategies That Constrain Judges

C. Strategies That Transcend Judges

IV. IMPLICATIONS

A. The Effect of Superior Lawyering on Legal Outcomes

B. The Strategic Transformation of Law

C. Explaining Local Legal Cultures

D. Why the "Haves" Come Out Ahead

EVALUATION AND CONCLUSIONS

FOOTNOTES

ABSTRACT

By the conventional view, case outcomes are largely the product of courts' application of law to facts. Even when courts do not generate outcomes in this manner, prevailing legal theory casts them as the arbiters of those outcomes. In a competing "strategic" view, lawyers and parties construct legal outcomes in what amounts to a contest of skill. Though the latter view better explains the process, no theory has yet been propounded as to how lawyers can replace judges as arbiters. This article propounds such a theory. It classifies legal strategies into three types: those that require willing acceptance by judges, those that constrain the actions of judges, and those that entirely deprive judges of control.

Strategies that depend upon the persuasion of judges are explained through a conception of law in which cases and statutes are almost wholly indeterminate and strategists infuse meaning into these empty rules in the process of argumentation. Such meaning derives [*pg 1406] from social norms, patterns of outcomes, local practices and understandings, informal rules of factual inference, systems imperatives, community expectations, and so-called "public policies." Constraint strategies operate through case selection, record-making, legal planning, or media pressure. Strategists deprive judges of control by forum shopping, by preventing cases from reaching decision, or by causing them to be decided on issues other than the merits. The theory presented explains how superior lawyering can determine outcomes, why local legal cultures exist, how resources confer advantages in litigation, and one of the means by which law evolves.

[*pg 1407]

If the judges say a contract with your buyer that he will not resell below a certain price will be illegal, and not enforceable, if they are likely to fine you or send you to jail for making such a contract, but you still want your goods resold throughout the country at a single price -- what can you do? That is a problem for invention, for ingenuity; the problem of inventing a method of action which will keep you free of difficulty and will produce the results you want in spite, if you please, of what the judges in a case of dispute may be expected to do.

Karl N. Llewellyn1

INTRODUCTION

In the conventional view of the legal process, courts determine facts and then apply law to those facts to generate outcomes.2 In the strict, formalist version of the conventional view, the law consists principally of rules in which the outcomes of cases are already im- [*pg 1408] plicit.3 In the more popular version, the law is a mix of fixed rules and flexible standards that sometimes permit courts to inquire into purpose and to exercise judgment and discretion.4 In either version, the conventional view holds written law to be an important determinant of legal outcomes. To the extent that the conventional view acknowledges written law to be ambiguous or indeterminate, it assumes, as the original Realists did, that the effect is to lodge power in the judges.5

By any version of the conventional view, the role of lawyers is relatively minor. They gather facts, conduct research in the appropriate legal materials, and then attempt to persuade the courts to reach outcomes that favor their clients.6 In most disputes, no litigation will be necessary; the written law will be sufficiently clear that the lawyers will agree on what the outcome of litigation would be. In the vast majority of the disputes that are litigated, the views of judges and juries will determine outcomes. Only in a small minority could lawyers determine outcomes, and then only by persuading the judges or juries.

[*pg 1409]

Oddly, this conventional view coexists with another that sees the legal process as highly manipulable through legal strategy.7 Star litigators -- or "dream teams" of them -- can regularly win judgments in cases that have no merit,8 prevent meritorious cases from ever reaching trial,9 turn victims into wrongdoers,10 and make the system set the guilty free.11 Though some might dispute the particular examples of manipulation that we give, the belief that strategies often determine legal outcomes is widespread and generally consonant with the reality of legal practice. Lawyers devote substantial time and energy to the [*pg 1410] development of legal strategies and regard them as capable of determining outcomes across a wide spectrum of cases. Even though the best strategies are case-specific and must of necessity remain secret, entire publications are devoted to those that are general and that lawyers are willing to divulge.12

This Article argues that the strategic view captures the reality of the legal process while the conventional view misses it. The unpleasant implications that follow from the strategic view13 explain in part why the conventional view has continued to dominate.14 But perhaps an even greater impediment to recognition of the importance of legal strategy has been the lack of a coherent theory to explain how lawyers15 can overcome both law and judicial discretion to generate the pattern of legal outcomes.

Legal strategy is curiously absent from the realm of legal theory.16 Extensive accounts of the adversary process do not even mention it.17 The law-and-society literature has explored differential case out- [*pg 1411] comes extensively,18 but aside from the exploration of case-selection strategy initiated by Galanter,19 has not attempted to explain how legal strategy could generate them.

Strategy plays a central role in three methodologies employed in law and economics: economic modeling, game theory, and gaming. But the strategy explored by law-and-economics scholars is fundamentally different from that practiced by lawyers. The economic actor either yields to the rule of law and seeks maximum advantage under it, or violates the rule and accepts the consequences.20 The legal strategist, by contrast, often seeks to defeat the rule of law in a manner that avoids the penalty as well. To make the same point another way, the economic model treats the lawmaker -- like the game designer -- as omnipotent.21 In that model, players cannot challenge the rules; they can only seek advantage under them. Yet a central thrust of legal strategy is to control legal outcomes despite the contrary intentions of legislators or judges.22

This Article seeks to explain the relationship between law, judges, and legal strategy, and in so doing, to offer a theory that explains what lawyers do when they strategize. The principal task, as we see it, is to generalize from what seems to be an infinite number of imaginative, clever, fact-specific maneuvers to a theory of what strategy is and how strategy can defeat both the rules of law and the judges who interpret those rules. As part of that task, we attempt to distinguish the strategic perspective from competing perspectives, to identify the materials from which lawyers construct their strategies,23 [*pg 1412] to catalogue the variety of legal strategies they construct,24 and to explore the implications of our theory.25

Our theory can be summarized as follows: "Law" has direct effect through the rendition and enforcement of judgments in actual cases and indirect effect through the anticipation of such rendition and enforcement in hypothetical cases. Each such case is a complex undertaking that may require hundreds of strategic decisions by the parties and generate an indefinite number of actual or potential legal issues and extra-legal problems. The "merits" of the case, as conventionally conceived, may be only one among them.26 Each of those decisions, issues, and problems is potentially outcome-determinative. The odds that any one will determine the outcome are small. But cumulatively, the odds that some combination of these decisions, issues, and problems will determine the outcome are large. The legal strategist manipulates those odds in a game of skill, expanding and developing the array of decisions, issues, and problems in a manner calculated to confuse and ultimately overwhelm the opponent.27 Even if the "merits" should ever reach a decisionmaker, it will be a decisionmaker identified by the game, and the "merits" will reach that decisionmaker in a form determined by the game.

We assume that courts are generally hostile to legal strategy and will seek to nullify it when they can. In contrast to prominent scholars who consider formalism a precondition to strategy,28 we assume a thoroughly realist concept of law that leaves to judges a broad range of freedom in their decisionmaking. Still, we conclude that, within the wide range of what is culturally acceptable in legal outcomes, legal strategies are the primary determinants of who will decide cases, un- [*pg 1413] der what constraints, and with what consequences. Written law is just one of several kinds of normative material the system uses, and it is not even the most important.

Part I of this Article uses the Texaco-Pennzoil case29 to illustrate how the strategic perspective we advocate differs from the conventional perspective and then distinguishes our strategic perspective from that employed in law and economics. Part II of this Article begins by defining the terms we use to refer to various concepts of law and strategy. Part II then describes the fluid nature of legal doctrine that renders it displaceable by social norms and the prejudices of the decisionmaker and explains how the strategist exploits those norms and prejudices. Lastly, Part II notes the residual role of legal doctrine in a world dominated by strategy. Building on the theory propounded in Part II, Part III employs three main categories to catalogue the wide variety of activities referred to as "legal strategy": strategies that seek to persuade the judge, strategies that seek to constrain the judge, and strategies that seek to strip the judge -- and the law -- of all power over case outcome. Part IV identifies the principal implications of the theory we propose. Our theory explains why strategy can so dramatically affect legal outcomes, how strategy drives changes in law, the source and mechanisms of the set of variations in law commonly referred to as "legal culture," and why ingenuity and resources together are so effective in determining legal outcomes. Part V concludes that strategy, not law, is the principal determinant of legal outcomes, that awareness of strategy is likely to increase with growing public access to data on legal outcomes, and that legal systems should be redesigned consciously to minimize strategic opportunities.

At the same time, the theory we present does not depend upon a high level of indeterminacy in legal rules. Merely assuming that there is indeterminacy on some issues provides the flexibility necessary to render the ultimate outcomes of nearly all cases indeterminate. Small amounts of indeterminacy on each of the numerous issues that constitute a case combine to generate large amounts of potential variance in outcomes.30

[*pg 1414]

I. THE STRATEGIC PERSPECTIVE DISTINGUISHED

The first step in understanding the strategic perspective we promote is to distinguish it from both the conventional perspective and the strategic perspective sometimes employed in law and economics.

A. The Texaco-Pennzoil Case as Illustration

To frame the conflict between the strategic and conventional views of the legal process, consider their application to the struggle between Texaco and Pennzoil over Getty Oil. Recall that, in the conventional view, legal problems are resolved by the application of law to the facts of cases. That view recognizes that legal procedures are necessary to make the application, and reluctantly acknowledges that such procedures can be outcome-determinative. But it assumes that judges control outcomes and that they decide the vast majority of cases on their "merits."31

A conventional account of the Texaco-Pennzoil case would proceed something like this: Pennzoil negotiated an "agreement in principle" to purchase Getty Oil. That agreement in principle was approved by Getty's board, but while the remaining terms of the written agreement were being negotiated and drafted, Texaco persuaded Getty to sell to Texaco rather than Pennzoil.32 The legal theory of the Houston, Texas, action that resulted in the $10.53 billion judgment33 was tortious interference with contract. After the parties presented the facts to the jury, the court charged the jury to decide whether Getty and Pennzoil "intended to bind themselves" to an agreement, whether Texaco interfered with that agreement, and if so, the amount of Pennzoil's damages.34 The jury concluded that Getty and Pennzoil did intend to bind themselves to an agreement, that Texaco interfered with that agreement, and that Pennzoil was damaged in the amount of $7.53 billion dollars.35 The jury assessed an additional $3 billion in pu- [*pg 1415] nitive damages.36 On appeal, the court reduced the punitive damages to $1 billion but upheld the remainder of the judgment.37 The judgment supposedly bankrupted Texaco, which was then able to survive only by settling with Pennzoil for a cash payment of $3 billion.38 The bankruptcy and $3 billion payment were the consequences of the application of law to the facts of the case.

The strategic view provides a markedly different account of the Texaco-Pennzoil litigation. In the strategic view, legal outcomes are the product of complex strategic interaction designed to control outcomes by manipulating the environments in which the outcomes are determined. Strategically-minded lawyers devise and execute plans for reaching their goals. The lawyers' roles are, in the strategic view, primary; they decide when, by whom, and under what constraints legal issues are resolved. The judges' roles are secondary; they decide only what they are asked to decide, and within the constraints imposed by the strategists.39 Finally, the strategic view is phenomenological; it seeks to explain outcomes by examination of the strategies employed.

A strategic account of Texaco-Pennzoil would proceed something like this: Pennzoil initially brought three legal actions seeking to block Texaco's deal with Getty Oil. First, it sued Texaco and Getty in the Delaware Chancery Court seeking an injunction against the deal.40 Second, while that lawsuit was pending, Pennzoil filed an antitrust action against Texaco in Tulsa, Oklahoma.41 The Tulsa filing was remarkable because neither Texaco nor Pennzoil was incorporated or headquartered in Oklahoma. Commentators speculated that Pennzoil chose Tulsa because the atmosphere there had been poisoned against takeovers by Occidental Petroleum's recently completed takeover of Tulsa-based Cities Service Company,42 or because of the political and economic connections of Pennzoil CEO Hugh Liedtke's family to [*pg 1416] Tulsa.43 (At the same time, Pennzoil surreptitiously financed another antitrust action against Texaco -- this one by a small heating oil and gasoline distributor in a federal court in Providence, Rhode Island.44) Probably none of the parties thought that any of these cases would go to trial; they were "tactical cases" filed to generate leverage, what one commentator called "the inevitable postscript to a high-stakes merger game."45 From a strategic perspective, Pennzoil was searching for a court receptive to its goal -- unraveling Texaco's deal with Getty. Each new case Pennzoil filed was another chance to find a court inclined to grant Pennzoil a remedy. The particular cause of action was of secondary importance. If one lawsuit seemed promising, Pennzoil could pursue it and dismiss or abandon the others.

Pennzoil's third legal action sought an injunction in the Houston, Texas, court; it was this action that eventually gave Pennzoil its $10.53 billion dollar verdict against Texaco. The Houston court was a strategic afterthought, selected in a conversation between Pennzoil CEO Hugh Liedtke and his friend Joe Jamail, a Texas personal injury lawyer, after Pennzoil had filed the same cause of action in Delaware.46 To file that cause of action in Texas, Pennzoil first had to dismiss the Delaware action. Pennzoil could dismiss as of right if it did so before Texaco filed an answer.47 If Pennzoil did not dismiss before Texaco filed an answer, Pennzoil could have dismissed only with leave of court, which probably would not have been granted.48 Had Texaco anticipated that Pennzoil would decide to switch courts, it could easily have filed that answer.49 Texaco's failure to file that answer became known later as the "10 billion dollar boo boo."50

[*pg 1417]

Pennzoil made the strategic decision to capitalize ruthlessly on Texaco's error. Deliberately ignoring a local custom that required Pennzoil to give Texaco notice of its intent to dismiss and an opportunity to prevent dismissal by filing the answer, Pennzoil dismissed without prior notice to Texaco.51 Fifteen minutes later, Pennzoil refiled in Texas.52

The strategic perspective recognizes the pivotal importance of Pennzoil's decision not to give notice. If Pennzoil had given notice, Texaco almost certainly would have filed an answer. If Texaco had filed an answer, the case would have remained in Delaware where Pennzoil probably could not have won53 and certainly could not have won more than about $800 million.54 The conventional perspective regards Pennzoil's decision as a matter of little or no importance to the outcome. Under the conventional perspective, the decision permitted Pennzoil to proceed in a different court, but that court would, under conflicts rules, be bound to apply the same law.55

The Texas case was assigned to Judge Anthony Farris.56 Two days after the assignment, Joe Jamail, then Pennzoil's lead counsel in the case, donated $10,000 to Judge Farris' reelection campaign.57 Texaco's motion to recuse on the basis of the contribution was denied.58 [*pg 1418] Jamail's generosity seemed to be rewarded with an almost continuous stream of rulings in Pennzoil's favor.59 Because such contributions do not meet the doctrinal standards of a bribe, conventional theory maintains the fiction that they have no impact on outcomes,60 but the actions of the parties show that they regarded the contribution as strategically important.61

Still considering Pennzoil's case weak as to liability at the close of evidence, Texaco chose not to dignify it by presenting evidence regarding damages.62 Texaco's strategy was carefully considered63 and is not without its defenders today.64 But its effect was to leave the jury with only Pennzoil's evidence regarding damages once they had found liability.65 Commentators generally regard the $7.53 billion compensatory portion of the verdict as grossly in excess of Pennzoil's actual damages66 and attribute it to what can be seen in hindsight as Texaco's strategic error.67 The conventional perspective, by contrast, would mask Texaco's strategic disaster with the homily that if each side in an adversary proceeding presents the evidence that is to its advantage, the truth will emerge.

[*pg 1419]

After the verdict, Texaco launched a media campaign seeking to convince the public that Texaco "could crumble under the weight of the history-making verdict unless [the Texaco trial judge] lifted the threat."68 The obvious purpose of the campaign was to affect the courts' future rulings, even though the conventional perspective recognizes no means by which that could occur.69

As the court was poised to enter a judgment that would place liens on property owned by Texaco, Texaco transferred one of the world's largest refineries and a chemical plant to subsidiaries that would not be liable for the judgment.70 The conventional analysis of such a transfer is that it is fraudulent and voidable if Texaco made it with the "actual intent to hinder, delay, or defraud" Pennzoil, or if Texaco made it in exchange for less than "reasonably equivalent value" while insolvent.71 In a creditor's action, the court would declare it void. The conventional analysis was, however, of little relevance because the mere transfer itself accomplished Texaco's objective: to keep record title clear of the judgment lien that would automatically take effect on entry of the judgment, and thereby prevent Pennzoil from gaining potentially outcome-determinative leverage. Pennzoil would have to take some additional legal action to nullify the transfer before its lien could attach.

Texaco's remaining property in the state continued to be subject to the threat of such liens. As a result, Texaco again faced a crisis on the day the Texas judge was to decide whether to impose liens on Texaco's remaining property within the state. Fearing that the judge might issue an adverse ruling from the bench, Texaco put in place an elaborate mechanism to enable it to obtain an automatic stay of such a ruling by filing for bankruptcy in White Plains, New York, during the few minutes between the judge's announcement of his ruling and Pennzoil's recordation of it.72 Imposition of the liens, which would [*pg 1420] have occurred immediately on recordation, might have severely disrupted Texaco's business, giving Pennzoil sufficient leverage to force Texaco to abandon its appeals.73

When the Houston, Texas, court entered judgment for Pennzoil (without imposing the liens), Texaco instead responded by filing a civil action in the United States district court in White Plains, New York. By that action, Texaco sought to remove authority over the case from the Texas state court and reverse the result.74 The strategic theory was that the Texas courts were prejudiced in favor of Pennzoil (the Texas company) and Joe Jamail (a personal friend of the judge). In contrast, the New York courts would presumably be prejudiced in favor of Texaco (the New York company), and the petition would be presented by a lawyer who knew the New York judge personally.75 Initially, the strategy worked. The New York judge reduced Texaco's appeal bond from $11 billion to $1 billion,76 opined that the Texas judgment was several times larger than what was appropriate,77 and suggested that if the Texas court did not set it aside, he might.78 Pennzoil's appeal of the New York ruling to the Second Circuit was accompanied by another intense lobbying campaign in both the state capitals and Washington.79 The Second Circuit affirmed the bond reduction.80

On April 6, 1987, the Supreme Court of the United States reversed, ruling that Texaco would have to comply with the Texas $11 [*pg 1421] billion bond requirement to pursue its appeal further.81 Although Texaco could have posted the bond82 or waited for a ruling on the bond from the Texas Court of Appeals, it chose instead to file bankruptcy in White Plains, New York.83 The immediate effect of the bankruptcy was to permit Texaco to continue pursuing its appeals without liquidating assets or posting the bond.84

Absent a bankruptcy filing, Texaco could have pursued its appeals only by posting an $11 billion bond; with the bankruptcy filing, Texaco was able to pursue its appeals without posting any bond at all. The conventional view justifies these starkly different entitlements with the theory that, first, bankrupt companies present different policy considerations that require different outcomes; and second, the courts control access to bankruptcy through the doctrine barring "bad faith filings."85 To put it another way, Texaco did not have a choice between posting the bond and filing bankruptcy; if Texaco was not bankrupt, it had to post the bond, but if it was bankrupt, it did not.

That perspective obscures the strategic nature of Texaco's decision. Texaco was not "bankrupt" in any common sense of the term. Its net worth was probably in the range of $22 billion to $26 billion -- easily enough to pay the $10 billion judgment.86 To justify the bankruptcy, Texaco launched a public relations campaign that one ob- [*pg 1422] server dubbed a "linguistic strategy."87 In that campaign, Texaco argued that it was financially impractical for it to post the bond, even though it was financially possible for it to do so.88 For reasons never made public, but which may also have been of a strategic nature, Pennzoil did not raise the bad faith filing issue in the bankruptcy case.89

The substantive law of bankruptcy offered Texaco no advantage against Pennzoil; because Texaco was solvent, bankruptcy law entitled Pennzoil to payment in full.90 The predominant and perhaps sole91 purpose of the bankruptcy appears to have been to create a delay that would enable Texaco to pursue its appeals.92 Bankruptcy's "automatic stay" prevented Pennzoil from enforcing its judgment while the case remained pending.93 If it remained pending long enough, Texaco might be able to finish its appeals, and in the meantime, Pennzoil would get no money.94

It was also possible that something benefiting Texaco might happen during the delay, and something almost did. The Securities Exchange Commission "gave Texaco a big boost" in its position by announcing that it would file a brief in the Texaco appeal arguing that [*pg 1423] Pennzoil had broken SEC rules in its bid for Getty.95 Ultimately, despite the SEC's support, the court ruled against Texaco.

The final resolution of the case -- settlement for a cash payment of $3 billion -- was precipitated by a procedural decision of the bankruptcy court. After the Texas Supreme Court upheld the verdict, the bankruptcy court modified Texaco's exclusive right to file a Chapter 11 plan. Such modification effectively permitted Pennzoil to settle the case with the creditors and equity committees if they could not settle it with Texaco's management.96 Faced with the possibility that the case would be settled without them, Texaco's managers settled quickly. Even so, the settlement was well below the expected value of the judgment in litigation under an economic analysis,97 suggesting that other, not-yet-revealed factors held down the amount.98

By condemning legal action taken solely for the purpose of delay -- a behavior the courts have no practical means of controlling -- the law conveniently escapes responsibility for Texaco's strategic success in delaying enforcement of Pennzoil's judgment. Either Texaco had some reason other than delay for filing the bankruptcy (which [*pg 1424] might be known only to Texaco)99 or Texaco's success was achieved unethically and illegally (in which case Pennzoil is at fault for not making appropriate objections).100 By contrast, the strategic perspective acknowledges that illegal strategies work in circumstances where the remedies available on paper do not.

The conventional view obscures the determinants of the outcome in Pennzoil versus Texaco by seeing the outcome as merely a question of the "proper" interpretation of ambiguous events -- was the original agreement in principle between Pennzoil and Getty a contract?101 The interpretation itself was an unverifiable process occurring in the minds of judges and jurors and hence was beyond effective examination or criticism.

By contrast, the strategic view demonstrates that the outcome in Pennzoil versus Texaco was highly serendipitous, and that each side had more than one opportunity to win by taking the right action at the right time.102 Texaco might have won by answering the Delaware complaint and objecting to voluntary dismissal, by presenting damage evidence to the Texas jury, or by reincorporating in New York so it could remove the Texas action to federal court. Pennzoil might have won by persuading the court to dismiss Texaco's bankruptcy as a bad faith filing103 -- leaving Texaco at its mercy -- or by demanding its right [*pg 1425] to absolute priority for the full amount of its judgment in the bankruptcy court.104

Settlement of the Texaco-Pennzoil case for about a third of the amount awarded by the jury is consistent with the conventional view that Pennzoil had a weak case on the "merits" -- the facts and the law. But the strategic analysis set forth here shows that to be mere coincidence. Pennzoil had overcome most of the conventional weaknesses of its case by winning the judgment and carrying it through the first two levels of appeal.105 Had it pursued the bankruptcy case, it might have taken ownership and control of Texaco, a $26 billion prize.106 The weaknesses in Pennzoil's case that remained and shaped the terms of settlement had little or nothing to do with the merits of its initial case as seen from the conventional perspective.

B. Legal Strategy in Law-and-Economics Theory

As noted in the Introduction, the kind of legal strategy we seek to explain -- the kind practiced by lawyers -- is curiously absent from current legal theory. Strategy does play a prominent role in three branches of economic theory: economic modeling, game theory, and gaming. Only in the third, little-used branch of the three applications, however, does the strategy employed at all resemble the type practiced by lawyers and explored in this Article.

Economic modeling is grounded in the conception of persons as "rational maximizer[s] . . . of 'self-interest.'"107 The choices these rational maximizers make are often strategic -- in the sense that they are plans for reaching goals -- and they are commonly referred to as [*pg 1426] strategies.108 But these models allow for only a single actor facing a choice among alternatives. Because the models lack interactivity -- strategist against strategist -- the strategies employed are simple and usually of limited interest in and of themselves.

To accommodate interactivity, law-and-economics scholars turn to a second methodology: game theory. Game theory posits the existence of two or more rational maximizers who compete with one another for desired outcomes.109 The theorist "solves" the game by determining the optimal moves for each player and proving those moves to be optimal.110 Economically-inclined scholars have used game theory extensively in studying the incentives that particular rule patterns create and the strategies that will best exploit them.111 As Rasmussen states:

Modern game theory has advanced law-and-economic analysis past the Coasian notion that parties, if they are well informed and transaction costs are low enough, will always reach the correct result. Any proposal for the creation of a default rule now has to be concerned with strategic obstacles that will impede a party from making the optimal choice.112

Game theory is effective in analyzing interactions involving two, or at most three, types of players in games where players can choose between only two, or at most three, alternatives.113 To analyze more [*pg 1427] complex interactions, a few economically-minded scholars turn to a third methodology: gaming.114 That is, rather than attempt to discover a comprehensive solution to the game, the researcher conducts plays of the game and evaluates the strategies employed by the players. Thus, for example, even though game theorists cannot prove "tit for tat" to be an optimal solution to a repeated prisoner's dilemma game, the strategy has defeated all others in empirical trials.115

Though gaming holds considerable promise for doing so,116 economic models have not yet managed to replicate the rich strategic interaction observable in legal practice. Almost invariably, economic models define in advance, with as little ambiguity as possible, both the goals of the strategists and the rules for interaction. That is, they adopt what we have referred to as the strictly formalist version of the conventional view.117 Usually, the economist's purpose is not to understand the process of strategizing, but to determine the pattern of outcomes that the strategic interaction will generate. Though surprises can occur within the restricted confines of these models, they are uncommon.

By contrast, lawyers conceive and execute their strategies in an environment in which the rules118 for interaction are unclear and constantly in flux. The challenge to the lawyers in Texaco was not to determine how best to respond to a rule pattern in certain conditions. It was to determine how best to respond in a context in which the rules were highly ambiguous and the conditions not entirely knowable. In practice, the strategist often succeeds by transcending what previously appeared to be the rules of the interaction. The emphasis is on crea- [*pg 1428] tivity, imagination, and flexibility. Surprises are routine.119 In addition, the environment in which the lawyers work is more complex than the environments of game theory and gaming because client goals are multifaceted, imperfectly known, and subject to change, in contrast to economic models in which the single goal is to maximize utility.120 Consequently, economic models fail to capture the crucial strategic element of legal practice.

II. THE FOUNDATIONS OF LEGAL STRATEGY

A strategy is a plan for action intended to accomplish some goal.121 It presumes some "field of play" -- people, data, or things that can be rearranged. The chess strategist arranges pieces by sequences of moves. The war strategist arranges troops, weapons, and propaganda. In either arena, the first task of the strategist is to understand the people, data, and things the strategist can manipulate directly.

The legal strategist works with decisionmakers, facts, legal cultures and law. The decisionmakers are judges, juries, arbitrators, administrators, boards, commissions, lawyers, and parties. The facts are events, both past and future. Those events morph into statements of fact, evidence, testimony, records, and finally the "facts" stated in court opinions. Legal cultures are sets of practices, perceptions, and expectations that differ from group to group and are often outcome-determinative.122 The "law" is perhaps the most difficult of the four elements to conceptualize for the purpose of a theory of legal strategy. Misunderstanding the nature of law has probably been the principal impediment to the integration of strategy into legal theory. If law were what some conventional theorists conceive it to be -- a set of written rules that specify comprehensively the appropriate outcomes [*pg 1429] for various fact patterns -- legal strategy could have far less impact on outcomes.

A. Definitions of Law

By what is probably the most common definition, "law" is a set of rules and standards promulgated by the state to govern conduct. In the American legal system, these rules and standards appear in statutes, court rules, court opinions, regulations, ordinances, and the like. The definition excludes social norms (the rules that spontaneously arise in groups of all sizes),123 oral legal traditions, and private contracts,124 even though the sanctions for violation of any of the three may be severe. This definition includes the "law on the books" whether or not it is in fact honored in the operation of the legal system. In the remainder of this Article, we will refer to the law thus described as "written law" or as "legal doctrine."125

Two leading Realists, Oliver Wendell Holmes, Jr. and Karl N. Llewellyn, used the word "law" to refer to what courts or other participants in the legal system will do in fact. As Holmes put it:

Take the fundamental question, What constitutes the law? You will find some text writers telling you that it is something different from what is decided by the courts of Massachusetts or England, that it is a system of reason, that it is a deduction from principles of ethics or admitted axioms or what not, which may or may not coincide with the decisions. But if we take the view of our friend the bad man we shall find that he does not care two straws for the axioms or deduc- [*pg 1430] tions, but that he does want to know what the Massachusetts or English courts are likely to do in fact. I am much of his mind. The prophecies of what the courts will do in fact, and nothing more pretentious, are what I mean by the law.126

Llewellyn expanded Holmes's definition to include the actions of officials, including lawyers.127 This conception of law ignores "rules of law" that are not enforced and ascribes to those that are enforced the same meaning given them by the officials. So, for example, if the rules of court provide that "if a party fails to respond to a request for admission within thirty days, the matter is deemed admitted," but the judges routinely extend the thirty-day period after it has expired, under the conception of law discussed here, failure to respond timely to a request for admission operates as an admission only if the court fails to excuse it. Using Pound's terminology, we will refer to this conception of law as the "law in action."128

Law, social norms, and physical constraints are, as means of social control, largely interchangeable.129 Hence, we would add to Holmes's and Llewellyn's definitions of law the effects of social norms and physical constraints to the extent they contribute to legal outcomes. The resulting conception of law, which we refer to as "delivered law,"130 attempts to link fact patterns with outcomes in a manner that is sensitive to "how much the [legal] system will cost, how long the system will take, what the system will require of the client along the way, and what the system will deliver in the end."131 Delivered law is the pattern of outcomes the legal system delivers. To continue with the preceding example, if judges routinely extend the thirty-day period for responding to requests for admission, then the [*pg 1431] delivered law is that responses to requests for admission are required only at such time as the court may fix.

As one of us has argued elsewhere, the lawyers, judges, and other officials who regularly interact in the processing of cases in a legal community forge and share mental models of the law that are both different from and simpler than the written law.132 They process routine cases according to the model, referencing the written law only when the model is challenged. For example, the law on the books provides complex, subjective tests for determining whether loans from insiders must be subordinated under a Chapter 11 plan. In the district in which one of us practiced, the delivered law was that debts owing to insiders could not be subordinated. At the same time, the delivered law of another district was that debt owing to insiders had to be subordinated. The lawyers and judges who processed cases in each district probably shared a mental model of the law that validated the law delivered in their district.133 That model was nowhere reduced to writing. We will refer to the law contained in these mental models as "the law in lawyers' heads."134

Finally, oral legal traditions that parallel but do not coincide with the other forms of law arise spontaneously in social groups.135 These traditions, which deal with virtually every subject touched by state-made law and many subjects that are not, are in large part the same phenomena discussed by sociologists as "social norms,"136 and sometimes discussed by economists as "spontaneous" or "private" ordering.137 Weyrauch and Bell138 and others,139 most notably Robert Ellick- [*pg 1432] son,140 have noted the ability of these informal rules to displace state-made law.

Some oral legal traditions, however, do not fit easily within the category of "social norms" because they appear, not in the form of rules that purport to govern behavior, but in the form of shared expectations about the outcomes which are appropriate under a given set of facts. For example, a particular community might expect that a youthful first offender should not do jail time for possession of a small amount of marijuana. We refer to the latter kinds of oral legal traditions as "expectations regarding outcome."

B. The Relationship Between Written Law, Social Norms, and Expectations Regarding Outcomes

Conventional legal theory regards the written law as specifying a set of rules that govern both social interaction and dispute resolution.141 For nearly any set of facts, the theory posits, written law specifies the appropriate outcome.142 The theorists who subscribe to this view acknowledge the existence of "gray areas" in which the appropriate outcomes are unclear. They differ regarding the size of these gray areas, but nearly all agree that written law specifies appropriate outcomes for a wide range of "easy" cases.143 In their view, social norms, prejudices, public policies, and social expectations are relevant only in the small minority of cases of first impression or cases in which litigants seek a change in the law to reflect modern conditions.144

[*pg 1433]

The theory we present here regards delivered law as the product of complex interactions among written laws, law in lawyers' heads, social norms, law in action, system imperatives, and expectations regarding outcomes. Legal strategists, who include lawyers, clients, judges, legislators, other officials, and sometimes persons who have not yet retained a lawyer or even become involved in a dispute, are the catalysts for these interactions. Through these strategic interactions, they construct the pattern of legal outcomes. Within broad cultural limits, the strategies these participants pursue and the quality with which they execute them -- not law or judges -- determine that pattern.145

Social norms and expectations regarding outcomes specify "just" or "fair" outcomes for virtually all human interaction, including interactions directly addressed by written law or the law in lawyers' heads.146 In regard to those interactions, the norms or expectations are often congruent with the written law and/or the law in lawyers' heads, but often they are not. One can easily think of examples of situations in which the written law requires one pattern of conduct, but social norms permit or require another. The possession and sale of illegal drugs is perfectly acceptable in particular subcultures. A testator has the legal right to disinherit a child, but any attempt to exercise that right clashes with a social norm that allows disinheritance only in the most extreme circumstances.147 The written law may applaud and protect the whistleblower at the same time that social norms render him or her unemployable.148

[*pg 1434]

These norms specify -- though with considerable imprecision -- the socio-legal entitlements of members of the group. Though the written law may seem to entitle the bank to call the loan "on demand,"149 the social norm may require that the bank act with some consideration for the borrower.150 Though the written law may regard the employer as entitled to discharge the employee "at will," the social norm may prohibit discharge for certain "unjust" reasons.151 Strategists who acquire the property of others in "perfectly legal" transactions may nevertheless be regarded as thieves.152

Conventional legal theory assumes wrongly that decisionmakers will apply written law to the exclusion of social norms,153 maintains falsely that expectations regarding outcomes are the direct product of written law,154 and does not even recognize the existence of the law in lawyers' heads.155 Were these assumptions accurate, legal strategists would have relatively few tools to employ. In fact, written law is sufficiently malleable that decisionmakers can interpret it to support virtually any position that finds support in social norms or expectations regarding outcomes.156 That is, whatever exists in a fact pattern that [*pg 1435] gives rise to rights or entitlements under social norms will find support in legal doctrine.157 As the language of written law departs from the entitlements of parties under social norms, that language becomes less effective.158

The view that written law drives legal outcomes is plausible only because written law (to the extent that it has any meaning at all) is usually in accord with social norms. The outcomes of cases in which the applicable norms differ from the written law demonstrate that the norms, not the written law, are the driving force.159 While written law is sufficiently flexible to support virtually any social norm, the social norms of a particular group are not sufficiently flexible to support virtually any written law.160

[*pg 1436]

Social norms differ from group to group and place to place. The geographical boundaries within which particular norms prevail are not congruent with the jurisdictional boundaries of the written law. Because the strategist can draw from such a wide variety of conflicting social norms and written legal doctrines, the strategic possibilities are virtually limitless.161 The strategist need only ascertain what in a given fact pattern gives rise to the client's conviction that the position is just. By generalizing on that quality of the case, the strategist can identify the governing social norm, which is probably already expressed in some existing legal doctrine. If it is not, the norm itself will be sufficient to establish the doctrine.

This does not mean, of course, that the social norm identified will appeal to anyone assigned to adjudicate the case. If the decisionmaker is from outside the group that forged the norm, the decisionmaker may not share it. But it does mean that if the decisionmaker shares the norm -- that is, believes after presentation of the case that the party should win -- then the doctrinal basis exists for a decision in the party's favor. Cases in which the decisionmaker sincerely162 opines that a particular party should have won, but the law required a contrary result, are virtually always cases in which the party's lawyer failed to pursue the most persuasive legal theory.163

[*pg 1437]

High-stakes litigation is fought principally on two issues. The first issue is who will decide the case.164 The second is who should win according to the values of, or the norms adhered to by, the decisionmaker. Parties can use legal doctrine to persuade decisionmakers that they should win, but legal doctrine is only one tool for doing so. It is seldom an effective one, because it inevitably carries within it support for the opposing position.165

The task of persuading the decisionmaker at the normative level is often an urgent one. Decisionmakers, particularly those who conceptualize law as a generally consistent set of rules that inform and bind their decisions, may frame the issue in accord with the first persuasive argument and be unable to give fair consideration to equally persuasive but conflicting arguments.166 Once the decisionmaker is persuaded at the normative level, the task of persuasion at the doctrinal level is easier and less urgent. The strategist need only provide the decisionmaker with legal doctrine that plausibly links the facts of the case to the strategist's desired conclusion. That other legal doctrines plausibly link the facts to other conclusions presents no real danger at this late stage. Presented with a bridge to the "right" conclusion, the decisionmaker is unlikely to adopt a doctrinal rationale that leads to a conclusion he or she believes is wrong.

The possibility of an appeal may seem, at first, a significant barrier to the trial judge's imposition of personal values in determination of the case. However, with the exception of appeals by certain "domi- [*pg 1438] nant parties,"167 the losing party's right to appeal an adverse decision is likely to have little actual impact on case outcome. First, the judges' determinations of fact are for all practical purposes unappealable.168 A judge who wishes to avoid reversal on appeal can often do so simply by making a disingenuous ruling against the appellant on an issue of fact instead of on the issue of law that is actually in controversy. Second, practical considerations such as costs, delay, or an inadequate record prevent appeals of the large majority of judicial decisions.169 Third, many, if not most, appeals are disposed of without argument or actual consideration by the judges of the appellate court,170 making [*pg 1439] them little more than empty rituals.171 Finally, even if the decision is reversed on appeal, the case may be remanded to the judge who initially decided it "incorrectly." That judge may reach the same decision on remand, merely substituting a rationale acceptable to the appellate court or deciding the case on a different issue. Of course, if the appellate court does undertake to decide the merits of the case, its decision will not be the product of legal doctrine, but the product of the same factors discussed here, viewed from the appellate judges' own perspectives and based on the record presented to them.172

The basic strategy outlined here -- persuade the decisionmaker and provide a doctrinal bridge to the desired result -- is capable of producing virtually any result consistent with the decisionmaker's values. This does not mean, however, that the system is capable of doing so in every case. The kind of lawyering required for such strategy is skilled, time-consuming, and therefore expensive. For lawyers to do it too frequently may violate the local legal community's norms for case processing.173 In low-stakes cases, considerations of cost and [*pg 1440] case-processing resources are likely to dominate, producing outcomes in accord with the law in lawyers' heads.174

C. The Residual Role of Legal Doctrine

Written law is less influential in the legal system we have described than it is in common understanding. Though the large majority of decisions will appear to be in accord with law, they are in no meaningful sense the product of it.175 Legal strategy, playing on expectations regarding outcomes and social norms internalized by the decisionmaker, determines case outcomes.176 The rules of the written law are mere incantations177 that may or may not have some effect on the decisionmaker. Unsupported by norms, the rules become impotent technicalities, "scrivener's errors," or dead letters.178

Though our theory may at first seem to exclude written law from legal system content, it does not do so entirely. First, unskilled parties or their lawyers can render indeterminate law determinate through admissions that characterize the facts in such a way that particular adverse doctrine applies or through concessions that adverse doctrine applies. The legal process deliberately pressures lawyers and parties to make such admissions.179 Theoretically, one can deny every allegation and object to every piece of evidence. But in practice, a pattern of such denials and objections quickly discredits both parties' wit- [*pg 1441] nesses and their lawyers.180 The lawyers may be marginalized and ultimately ostracized.181 Lawyers must appear to be constructively engaged in a process that "narrows the issues" in preparation for the court's decision.182 Often, they come under intense pressure to admit facts or the applicability of law before they can entirely predict the effect of their admissions.183 Once a party has, against its own interests, admitted each of the facts in the antecedent of a rule184 in the language of that rule, the court may no longer have any alternative but to apply the rule. In such a case, legal doctrine may compel a particular outcome.

Second, participants in the legal process may see the rules of written law themselves as norms. That is, they may believe that a party should win because the rules of written law seem to direct that the party should win.185 This belief is an erroneous one; written law generally obtains normative content only when decisionmakers inject that content through interpretation in the particular case.186 Some decisionmakers' lack of sophistication in this regard, perhaps combined with poor lawyering on the losing sides, prevents those decisionmakers from seeing this. When decisionmakers feel genuinely compelled to a particular result because they believe it to be required by the rule, the result does, in a sense, follow from the rule.

That does not mean that the strategist should devote much effort to determining what result the rules compel on particular facts. The rules compel no particular result. Instead, the realization that decisionmakers may perceive themselves as compelled to accept a par- [*pg 1442] ticular rule recommends that the strategists attempt to discover how the process settles on a particular rule. The issue is explored further in Part III.

Third, even judges who believe, as we do, that law exists to support any conclusion they may choose to reach in the vast majority of cases will continue to be concerned with the doctrinal bridges by which they explain their decisions. Some outcomes will be more popular than others among the court's constituents. Particularly when the outcome will be unpopular within an important segment of that constituency, the ability of the judge to wrap the decision convincingly in legal doctrine can be important.187 The judge's willingness to make a particular decision may depend in part on the level of assistance counsel's briefs and arguments offer in the drafting of a persuasive opinion.188

Perhaps the most important role of legal doctrine today is in maintaining the respect of the public for the legal system and the decisions it renders. The public tolerates the current system because it erroneously believes that the system operates largely according to the conventional view.189 That is, that by and large, written laws bind judges to particular outcomes.

[*pg 1443]

But judicial opinions are, in essence, merely propaganda on behalf of the judiciary. Their purpose is to support the belief that legal doctrine plays a role in the system that it, in fact, does not. The difficulties of nations that lack a tradition of respect for written law -- such as Russia and China -- underline the importance of the exercise.190 Those difficulties should not, however, be considered justification for American legal scholars to maintain falsely that legal doctrine is the principal determinant of legal outcomes when it is not. Among other problems with that approach, steadily accumulating data on case outcomes will soon prove it wrong.191

III. THE VARIETIES OF LEGAL STRATEGY

This part describes the wide variety of legal strategies currently in use. Participants in the legal system have acknowledged most of these strategies individually but not their cumulative effect. We assert that the presence or absence of these strategies, and perhaps numerous others of which we are unaware, is the predominant determinant of case outcomes.

We present our taxonomy in three parts. First, most (but not all) legal strategies seek to enlist the judge. Other legal strategies seek to pressure the judge or limit the judge's alternatives, while still other legal strategies seek to deprive the judge of any say at all in the outcome of the case.

A. Strategies Requiring Willing Acceptance by Judges

Strategies that seek to persuade the judge, jury, or other decisionmaker to rule in the strategist's favor are both the most common and the most visible type. Some of these strategies of persuasion may have nothing to do with the particular case. For example, lawyers adopt the speech, mannerism, and dress of jurors or judges as a means of gaining their confidence.192 They attempt to undermine the credi- [*pg 1444] bility of adverse witnesses, even when they believe them truthful.193 They employ rhetorical techniques, such as theme, repetition, and innuendo.194 They pursue "linguistic strategies," employing words and phrases that seem concrete and evoke favorable reactions, while skirting the edges of falsifiability.195 They may attempt to construct logical arguments196 or they may simply throw out possibilities in the hope that the decisionmaker will seize upon one of them.197

Most strategy, however, relates to case-specific information: the evidence the lawyer will offer, the facts the lawyer will assert, or the rationale the lawyer will encourage the decisionmaker to adopt. The three combine to form an argument that the decisionmaker should reach a particular conclusion.

Ostensibly, the system approves only a single rationale for decisions:198 application of the written law to the facts proven requires the [*pg 1445] conclusion sought.199 Even when a matter is within the "discretion" of the court, a legal standard governing the exercise of that discretion will almost certainly already exist.200 Formally, the argument -- and the judge's explanation of the outcome -- must be that application of the standard for the exercise of that discretion requires the conclusion sought.201

Despite the relative ineffectiveness of written law, in practice it provides the sole basis for most legal argument. Nearly all written law is sufficiently general, however, that a sophisticated opponent can find substantial ambiguity in it as it would be applied in the case at hand. The basic technique for discovering such ambiguity is a simple one. The opponent examines each word of the proffered rule for ambiguity. If even a single word is ambiguous, application of the rule is ambiguous. Once that ambiguity is demonstrated, the opponent often asserts a rule from another area of law that, if applied, would lead to the opposite result in the case at hand.202 The proponent examines that new rule for substantial ambiguity and the process repeats itself.203

[*pg 1446]

If a party is unable to discover ambiguity in the rule proffered by its opponent, the party can turn to any of a number of meta-rules to invalidate the unambiguous rule: the rule is unconstitutional, the rule was not properly adopted, the rule is not authoritative in this jurisdiction, the (statutory) rule constitutes a scrivener's error, the rule should be changed because of changes in technology or society since it was adopted, etc.204 Most recently, the Supreme Court has been implicitly employing a test of "mandatory culpability" in the review of criminal convictions. That is, the defendant must not only have violated the written law; the defendant must also be "culpable."205 This use of meta-rules to nullify rules mimics a technique employed both in public debate and in computerized systems of analysis: when a pattern of thought is unsuccessful, shift to a higher or lower level of generality.206 If the parties are sufficiently skillful in manipulating the written law, the written law ultimately proves at least plausibly indeterminate.

Sophisticated lawyers argue the written law, but they anticipate its indeterminacy. Understanding that content must be injected into written law in every case, they ground their arguments in several nondoctrinal rationales. These include social norms, the law in action,

[*pg 1447]

Argument from Generalization of the argument Example of the argument

Social norms This is the fair thing to do. "An employer can't fire an employee based solely on false accusations."

Law in action or legal outcomes This is the result reached in like cases in the past. "The evidence shows that employees have never been given hearings in cases such as these."

Law in lawyers' heads This is what the law requires. "The law is clear that, unless otherwise specified, employment is 'at will' and the employee can be discharged for any reason or for no reason."

Informal rules of factual inference When fact A is present, fact B is present as well. "The defendant was late for an important meeting; so she was hurrying and therefore careless."207

System imperative Absent this result, the system cannot operate as intended. "If this employee is entitled to a hearing, then so are the other 500,000 employees discharged each year."

Community expectations regarding outcomes This result is in accord with commercial or community expectations. "Most employees would be shocked to learn that they could be fired on the basis of false accusations -- without being given the opportunity to present their side."

Public policy This interpretation is in accord with the lawmakers' intention. "The reason for the rule is that . . . ."

Figure 1

[*pg 1448]

the law in lawyers' heads, informal rules of factual inference, system imperatives, community expectations regarding outcomes, and public policy. Figure 1 contains illustrations of each kind of argument.

These arguments do not purport to be based in law. They purport only to buttress the maker's argument that the written law, correctly interpreted, requires the desired conclusion: "The law favors our side, and this argument is proof of its wisdom and rationality in doing so." In actuality, arguments such as these are outcome-determinative. If both lawyers are competent, both sides will have sound arguments from legal doctrine. The side that is most persuasive with their indirect arguments will almost always win. Through the process of interpretation, the sources in the first column of the table -- spontaneously generated "private" law -- are fused into and made the content of the rules of law promulgated by the state.208

The inclusion of "public policy" in the category of "spontaneously generated private law" requires explanation. While arguments from "public policy" are sometimes based on declarations of the legislature made simultaneously with the promulgation of a statute or declarations of a court made in the opinion that established the rule, most "public policies" are invented by lawyers or law professors based on examination of the rule itself. That is, they are mere speculations on what the intent behind such a rule might be. When accepted by a decisionmaker, however, they can fix a single meaning to a rule that on its face has several. Thus "public policy," like social norms, provides content to law.209 But public policy is a more flexible device because the policy, unlike the norm, can spring wholly from the strategist's imagination.

Before a judge who believes legal doctrine determinate, the most effective response to each of the arguments in the table is to point out that the argument is grounded in something other than legal doctrine. For example, in response to the argument from social norms, opposing counsel might reply that "my opponent argues that social norms prohibit an employer from firing an employee for something the employee did not do. That may well be, but the law is to the contrary." Similarly, the best response to an argument that relies on informal [*pg 1449] factual inference is probably to point out precisely the basis on which the inference is being drawn. "My opponent argues that just because a person was late for a meeting, you should infer that the person was careless."

Full articulation will damage these arguments and most legal strategies, whether the articulation is by the opponent or by the proponent-strategist. To avoid that articulation, the proponent-strategist should make its own argument sketchily and by innuendo -- expressing just enough of the argument to reach the decisionmaker, but not enough to alert the opponent to its nature or to destroy the argument by full articulation. Weyrauch and Bell give the following example:

[I]t is difficult for a divorced husband to obtain custody of his twelve-year-old daughter. There is an unarticulated cultural norm, no matter how factually inaccurate, that a single adult male cannot be trusted in an intimate living arrangement with a young female. It would be impossible, however, to articulate this highly prejudicial concern openly in court. Lawyers instead prefer veiled references to sexual concerns by arguing that the daughter is "maturing" and needs the guidance of the mother in "hygienic matters."210

Once such an argument has reached the decisionmaker, full articulation may be the best way to defuse it. The husband's lawyer points out that wife's counsel is insinuating, without supporting evidence, that the husband will sexually molest his daughter if the court gives him custody. Thus restated, the argument loses its impact.

Normative content can be infused into otherwise empty legal rules even before any case arises. That is, even though the ultimate interpretation of a rule remains open as a technical or legal matter, if a consensus forms in the relevant legal community that a particular interpretation is "correct," other interpretations may, as a practical matter, be foreclosed. For example, when a new law is enacted, "experts" may promote a particular interpretation at judicial education or continuing legal education programs.211 Even though the law is ambiguous on its face, once a consensus forms in the relevant legal community, individual judges may be under pressure to conform.

[*pg 1450]

B. Strategies That Constrain Judges

At least four types of legal strategy exist for pressuring judges into a favorable resolution of a case without entirely persuading them to it: case selection, making a record, legal planning, and media spin.

1. Case selection. In his landmark article, Why the "Haves" Come Out Ahead, Marc Galanter explained the many ways that "repeat players" -- who generally were business interests -- could tip the balance of law in their favor by selecting to litigate the cases most likely to make favorable law.212 He advocated the development of counterstrategies in which organizations of otherwise "single-shot" players would do the same thing.213

To illustrate case selection as a legal strategy, assume that the manufacturer of a product is interested in establishing that its product is not unreasonably dangerous as a matter of law. Plaintiffs have filed cases in which this issue might be raised in half a dozen state or federal courts spread throughout the United States. The manufacturer begins by evaluating the six cases on a single criterion: if the plaintiff raises the issue of whether the product is unreasonably dangerous as a matter of law, what is the likelihood that the court will decide it in favor of the manufacturer? Assume that the likelihood of decision favorable to the manufacturer is 60% in each of two courts and 10% in the four others.214 Assume also that the unfavorable courts are further along in processing the cases, so that their decisions are likely to be rendered first. Absent a case-selection strategy, the odds are nine-to-one that the first decision will go against the manufacturer. With that [*pg 1451] case argued as precedent,215 whether binding or not,216 the manufacturer's chances of winning the next three cases probably sink even lower. By the time the favorable courts reach the issue, the domino effect of precedent may already have put four decisions unfavorable to the manufacturer on the books. This "weight of authority," perhaps augmented by a "precedential cascade"217 effect against the manufacturer's position, may persuade the two remaining courts to abandon their inclination in favor of the manufacturer.

The manufacturer could improve the odds of the case law developing in its favor by assuring that the two courts favorable to the manufacturer218 are the first to render decisions. The manufacturer could employ at least four tactics for accomplishing that result. First, it could settle or delay the four cases pending in adverse courts so that the courts did not decide them, or at least, did not decide them first. Second, it could permit the cases to go forward to jury verdicts, but without raising the issue of whether the product was unreasonably dangerous as a matter of law. Third, it could strike a bargain with the plaintiffs in the unfavorable courts that would tie the outcomes of their cases to the outcomes in the two favorable courts and delay the former until the latter were decided. The inducement might be the elimination of much of the cost of litigation, or the guarantee of a minimum recovery regardless of the outcome of the case.219 Lastly, after trying and losing the first cases, the manufacturer could enter into [*pg 1452] settlement agreements conditioned upon the court not publishing an opinion or upon the court sealing the file.220 By assuring that the favorable courts decide first, the manufacturer puts favorable law on the books for use in arguing subsequent cases.

Though the strategy of case selection has been sufficiently effective to draw the attention of legal scholars, it remains relatively weak. Case selection is a means of generating written law favorable to one's position. However, litigation is unlikely to establish written law in any given case, and, as previously noted, even once established, written law is not very powerful.221

2. Making a record. Appellate courts decide cases on the record and the written and oral arguments of counsel.222 The record consists of selected portions of the pleadings and other documents filed in the case, transcripts of what transpired orally, and evidence offered. The appellate court may reverse the decision because the court below decided it wrongly or because the lower court did not follow appropriate procedures, as indicated by the record.

[*pg 1453]

When litigating a case, the lawyers generally have an incentive to "make a record" favorable to their own position. The record may be useful in overturning an adverse decision on appeal, or defending a favorable one.

Strategic opportunities abound. To maximize the possibility of reversal on appeal, the lawyer may make frequent objections, may attempt to incite the judge to intemperate action,223 may proffer evidence or testimony of questionable admissibility, or may argue positions ineffectually in the hope of leading the court into reversible error.

These strategies can become complex. For example, a lawyer who actually hopes to win the case may create the impression that he or she is just making a record for appeal. The court, knowing of this hostile intent, may be reluctant to rule against the lawyer on particular issues for fear that the ruling will provide the basis for reversal. The result may be that the lawyer gets the benefit of every doubt on the hundreds of legal issues that arise in a typical case.224 Merely bringing a court reporter to a hearing may have the same threatening effect.

Appellate courts assume that the judges below had legally acceptable reasons for their decisions unless the records affirmatively show that they did not.225 For that reason, a judge who does not wish to be reversed on appeal is best advised to give no more explanation than is legally required.226 That is what most judges do. They listen to arguments and sign an order stating the outcome of the case, but they [*pg 1454] either give no reasons or they give their reasons in a perfunctory manner.227

The task of a lawyer-strategist who seeks reversal is to get the judge to articulate an unacceptable reason for the decision. The lawyer might do that by directly asking the reasons for the ruling, or by accusing the judge of employing a particular line of reasoning, thereby provoking a response.228 Alternatively, the lawyer might make a series of motions or proffers that, when ruled upon, eliminate particular lines of reasoning as possible explanations for the overall decision.229

For present purposes, the key is to understand that these strategies are not meant to persuade the trial judge -- or in many instances the appellate judge either -- to the lawyer's position. The goal may be to intimidate the trial judge or to make sure that the case is decided on ancillary issues in the appellate court. Although reversal on ancillary issues is not automatically victory, it can get the lawyer a "second bite at the apple" -- another trial at another time, in different circumstances, and perhaps with a different trial judge -- and is also highly likely to lead to settlement.230

3. Legal planning. The term "legal planning" is used broadly to refer to action taken before contemplated litigation in an attempt to establish what will become the facts of that litigation.231 Professor Leo Katz provides the following examples of particularly strategic legal planning:

(3) A lawyer suggests to a client who owns a farm that she incorporate the farm and declare herself its employee in order to qualify for social security. . . . (5) A lawyer recommends to elderly clients that they distribute their assets to their children so that they qualify for governmental assistance more quickly. (6) A lawyer recommends that a client turn most of its employees into independent contractors [*pg 1455] to escape the burden of social security taxes or even simple tort liability.232

The planner may conceive of his or her purpose as avoiding litigation,233 but the plan typically accomplishes that only by assuring that, in the event of litigation, the planner's side would win.

The strategies Katz suggests consist of creating fact patterns that satisfy the antecedents of legal rules, and thus achieving the desired results. However, as we have seen, legal rules are not, in and of themselves, of much effect. For virtually any legal rule that would achieve the planner's objectives, there is another rule or doctrine that could be used to attack it. Corporate veils can be pierced; security interests can be subordinated; contracts can be rescinded, reformed or invalidated; and conveyances can be avoided. At a minimum, such planning is always vulnerable to attack as a "sham" or a "subterfuge."234

The sophisticated planner realizes that norms, customs, system imperatives, expectations regarding outcomes, and other informal bases for decisionmaking likely will be determinative. Accordingly, sophisticated planning will proceed largely on the basis of appearances. To illustrate the difference, consider the circumstances of a debtor who expects to file bankruptcy soon and wishes to keep as much property as possible. Mere attention to legal doctrine might suggest that the debtor convert all of the debtor's assets to exempt property, leaving it to the court to determine which of the conversions were made with fraudulent intent. But the strategist who took a broader view would realize that the greed and lack of consideration evidenced by that strategy would likely turn the court against the debtor. Bankruptcy lawyers express this truth in the familiar aphorism that "pigs get fat, hogs get slaughtered."235 Similarly, if the law exempts an automobile of a specified value, the doctrinalist might as- [*pg 1456] sume that the model and color were irrelevant.236 The sophisticated strategist, by contrast, would immediately recognize the difference between purchasing a staid blue Chevy Caprice and a flashy red Corvette -- even if the dollar amounts of the two purchases were identical.237

Legal planning is often an effort to persuade rather than constrain the court. Judges are to some degree politicians. Even those with lifetime appointments often seek higher offices or professional acclaim. To those ends, they seek low rates of reversal and reputations for sound, responsible decisionmaking. Despite the lack of any doctrinally-recognized difference between a Corvette and a Caprice, the former projects an image of extravagance and privilege that most observers would consider to be inappropriate for a person who is discharging debts in bankruptcy. The latter projects an image of practical necessity that fits more comfortably with bankruptcy discharge. A judge is therefore more likely to suffer criticism for permitting a debtor to retain a Corvette than a Caprice of the same dollar value.

4. Media spin. Because judges care what members of the profession and the public think of them, they are vulnerable to media spin regarding the cases that come before them. To release a defendant who was probably guilty of the crime charged because the police violated the defendant's rights is a great deal more difficult when the crime is a heinous one that has caught the press's attention. For that very reason, the prosecutor who wants to win the case has an incentive to make sure the crime catches the attention of the press. F. Lee Bailey, a highly successful trial lawyer in the 1960s and 1970s, was notorious for arguing his cases in the press, thereby "seeking to create a force outside the courts with the expectation and intention that the force so created would produce a desired result in a particular case [*pg 1457] within the courts."238 Since Bailey's heyday, the standards for what is acceptable have been liberalized.239

For the past two decades, manufacturers and insurers have been campaigning in the mass media against high jury verdicts and expanded remedies.240 Ironically, the campaign seems to have had more effect on judges than on jurors. Recent studies show astonishingly high rates of verdict reductions or reversals by both trial and appellate courts.241

C. Strategies That Transcend Judges

The most interesting types of strategies are those that seek a particular result "in spite, if you please, of what the judges in a case of dispute may be expected to do."242 These are legal strategies that seek to prevent the other side from obtaining an adjudication or that seek to control who the adjudicator will be.

1. Cost strategies. Legal remedies are available only through litigation.243 Litigation is expensive and can be made more so by the manner in which it is conducted. For example, by expanding the issues in litigation, a party can expand the scope of discovery or the [*pg 1458] opportunity to employ experts. Depositions may be taken in geographically remote areas, putting the opposing party to the expense of attending. Retention of experts by one party ordinarily necessitates the retention of other experts by the opposing party.244 As a result, it may cost millions of dollars in out-of-pocket costs for plaintiffs to prepare for trial a simple dispute over whether the dumping of toxic chemicals caused a cluster of children's leukemia.245 Occasionally, a party is able to win a case by exhausting the other side's resources, or to prevent the bringing of a case by eliminating the other side's resources,246 though the more common result is settlement of such a case at a discount.247

A manufacturer's or seller's announcement that it will defend all cases to final decisions by the courts, without consideration of settlement, can, in some situations, dramatically reduce the number of cases that are cost-effective to litigate.248 For example, the cigarette companies "decided they would defend every claim, no matter what the cost, through trial and any possible appeals" and held to that policy for thirty-five years.249 The strategy was phenomenally successful. At the end of those thirty-five years, few cases had come to trial, and the companies had not paid out a cent in tort awards.250

[*pg 1459]

2. Delay strategies. Litigation takes time, and defendants can expand the amount required by the manner in which they respond.251 An award of the same relief at a later time may be less valuable to the plaintiff or less costly to the defendant.252 For example, if the court delays the entry of an order barring a competitor's illegal practices, the competitor may succeed in driving the plaintiff out of business in the interim. Damages may not be an adequate substitute, because the damages actually incurred might be too speculative to recover253 or the judgment for those damages might not be collectible.254

Delay may reduce the expected value of a lawsuit by reducing the plaintiff's chances of winning. Witnesses may die or disappear. Evidence may be altered or destroyed. The pattern of judicial or community attitudes that made the case worth bringing may change. The damage inflicted on the plaintiff by the defendant's wrongful act may become less poignant with the passage of time -- perhaps through the death of the plaintiff. The plaintiff may become discouraged by the passage of time, or become less determined or less able to devote the necessary resources.

Delay may also reduce the plaintiff's chances of collecting any money judgment that may be rendered; that is, a defendant that is not judgment-proof at the commencement of litigation may become judgment-proof by its conclusion.255 The defendant may even be able [*pg 1460] to delay until the case becomes moot.256 When cases settle, the amounts of the settlements are likely be discounted for the time, expense, and uncertainty that would have been involved in completing the litigation.

The fact that delay affects case outcomes is widely recognized. What is not widely recognized is the capacity of legal strategy to manipulate the extent of delay.257 Strategies that can change the extent of delay can change both the parties' chances of winning and the amounts by which their cases will be discounted in early settlements.

Various rules of law may seem to prohibit strategies that seek delay, but, on close examination, they prove ambiguous. Thus, for example, the ABA Model Rules of Professional Conduct impose on lawyers an obligation to "make reasonable efforts to expedite litigation" but only "consistent with the interests of the client."258 The comment to that rule provides that "[d]elay should not be indulged merely for the convenience of the advocates, or for the purpose of frustrating an opposing party's attempt to obtain rightful redress or repose."259 The term "merely" suggests that deliberate delay is acceptable strategy when those motives coincide with loftier ones, even if the loftier ones are of little weight in the decision. Rule 11 of the Federal Rules of Civil Procedure refers to "unnecessary delay" as an "improper purpose,"260 suggesting that delay is a proper purpose when the delay is "necessary." The effect of these ambiguities is to provide [*pg 1461] the basis for an array of strategies that lawyers can use to pursue, or counter, delaying strategies.

3. Extralegal strategies. Much strategic effort is devoted to extralegal means of deterring those entitled to legal remedies from suing or from continuing suits already filed. These efforts range from common courtesy -- an airline's making relatives comfortable and providing them with information in the days immediately following a fatal crash -- to physical violence. Between those extremes are a variety of pressures that parties to a litigable dispute can apply against their opponents. One can embarrass an opponent by the kinds of questions asked in trial or in deposition, or by the nature of the matters inquired into in discovery. One can make an opponent's life unpleasant by conducting litigation in an uncivil manner. One can threaten an opponent -- perhaps explicitly, but more likely implicitly -- with the loss of business relationships, unwanted publicity, the loss of a job, criminal prosecution,261 deportation,262 or embarrassment in matters having no direct relationship to the litigation. Or one can go entirely outside the law, fabricating evidence, bribing judges,263 or murdering witnesses.264

The effect of these extralegal strategies on patterns of litigation is difficult to overestimate. Businesses tend to sue their trading partners [*pg 1462] only when the business relationship expires.265 A recent empirical study found that only one in fifty patients with a valid claim for malpractice against a medical provider actually made the claim.266 Cases against professional killers or other violent felons are notoriously difficult to bring because witnesses understandably fear for their lives.267

4. Contractual strategies. Through a variety of contracts, legal strategists can install their allies as decisionmakers, bar the bringing of otherwise meritorious cases, or shift risks to persons who do not realize their magnitude. Examples of the first are contracts that require arbitration of customer claims against industry by arbitrators whose primary loyalty is likely to be to the industry.268 Examples of the second are contracts in which customers waive their right to sue the businesses they patronize. An example of the third is an agreement in which the customer indemnifies the company against claims by third parties.269

[*pg 1463]

In most of these circumstances, the strategy is both contemplated and authorized by legal doctrine. That is, policymakers have made a deliberate choice to permit customers to waive their rights. But legal strategy is capable of extending these contracts and waivers beyond the limits contemplated or authorized. For example, Professor Katherine Van Wezel Stone documents how legal strategists for industry have transformed the Federal Arbitration Act from a guarantor of a level playing field to a tool for the oppression of employees or customers.270

5. Forum shopping. Given what we have already said about the ineffectiveness of legal doctrine, the fact that different decisionmakers reach different results on the same legal doctrine should not be surprising.271 To take advantage of those differences, legal strategists have devised numerous ways of controlling who will decide a dispute. In some circumstances, legal doctrine deliberately [*pg 1464] affords the party filing the case a choice among courts272 or even among judges of a single court.273 The most prominent example is the right of a plaintiff to sue a defendant in any jurisdiction in which the defendant does business.274 Most plaintiffs so entitled use their "venue privilege" to bring their actions in the jurisdiction most convenient for them. But some choose the jurisdiction for the likelihood that its courts will rule in their favor.275

The doctrines that initially afford plaintiffs choices among courts may also give opposing parties means of upsetting the plaintiffs' choices, leading to an interaction of strategies. For example, if the plaintiff files a case against an out-of-state defendant, the defendant may have the right to remove the case to the federal court.276 Plaintiffs who anticipate the possibility of removal may be able to prevent it by joining additional defendants,277 by suing only on state law causes of [*pg 1465] action,278 by requesting particular relief,279 by suing in the defendant's home state,280 or by otherwise destroying diversity.281 Defendants who [*pg 1466] fear adjudication by a state or federal trial court can sometimes avoid it by filing for bankruptcy and having the initial adjudication made by the bankruptcy court as a determination of the amount of the plaintiff's claim.282

With respect to some types of actions, the party who would otherwise be the defendant can file first, thereby seizing the venue privilege for itself. These actions include divorces,283 bankruptcies,284 and virtually any matter that can be the subject of an action for declaratory relief.285 Parties with bargaining leverage may attempt to control venue through a contract provision imposed at the time of the initial transaction.286

The system's response to these kinds of strategies is to pass or create laws, such as the doctrine of forum non conveniens,287 that authorize judges to transfer cases. But these laws fall far short of providing an antidote to forum shopping for legal outcomes. First, these laws do not even theoretically contemplate a single, most appropriate venue for each case. Second, in at least some contexts, practical and [*pg 1467] political pressures on judges have caused a general failure in the venue correction process.288 Internationally, no correction mechanism may exist at all.289 Empirical studies suggest that the struggle over venue is often outcome-determinative.290

Forum shopping among the judges of a single court is generally prohibited, and violations are usually treated as serious violations of legal ethics.291 The predictable differences in legal outcomes across judges are, however, so great that legal strategists continue to find ways of picking their judges. The simplest means is to file the case in a court that has only a single judge,292 or in a court that will advise filers in advance what judge will be assigned.293 In a system that rotates the assignment of judges in a predictable order, the strategist has merely to observe which judge was assigned to the previous case or cases to know which judge will be assigned to the next case filed.294 In a system that assigns cases randomly, the strategist typically looks for situations in which the judge it seeks to avoid is removed from the draw temporarily because of vacation, illness, imbalance in already assigned cases, or some other reason. Often, the chief judge of the court [*pg 1468] has the authority to bypass the random assignment system, opening the way for political corruption.295 Strategists may file several identical cases, and then dismiss all but the one assigned to the desired judge.296 In more sophisticated versions of the same strategy, the strategist files only a single case in the court, observes the assignment, and then decides whether to go forward or dismiss. If the strategist chooses dismissal, the strategist may later file in the same court or file in a different court. Other variations are possible.297

Strategists can also select courts -- and thereby select more favorable panels of judges -- by changing the nature of the relief they request. For example, wealthier taxpayers can avoid the notoriously pro-government judges of the tax court298 by paying the tax and then suing for a refund in the United States district court. In the Texaco-Pennzoil example with which this Article began, Pennzoil tested the waters in Delaware by filing for injunctive relief, and then changed judges by dismissing that action and filing for money damages in Texas, before having New York judges imposed on them by Texaco's filings in New York federal and bankruptcy courts.299

6. Settlement strategies. The conventional view considers the settlement of cases a means of reaching approximately the same result litigation would reach, but at less expense. The concept is [*pg 1469] crystallized in Professors Mnookin and Kornhauser's memorable phrase "bargaining in the shadow of the law."300

Settlement is, however, more than just an amicable resolution of a dispute. It is a complex, often adversarial process with its own norms, procedures, strategies, and range of possible outcomes. Settlement strategies can include the arguing of facts that would be inadmissible at trial or of legal theories that would have little chance of winning in court, the making of threats that would be considered extortionate in any other context, and the substituting of what the parties really want for the remedies the legal system offers.

Settlement can generate outcomes that lie entirely outside the range of possible outcomes from litigation. The operative leverages may result from the court's restrictions on continuing the litigation301 or from the consequences of continuing aside from the ultimate decision.302 For example, the litigant who can credibly threaten to embarrass the opposing party in the course of the litigation -- perhaps by discovering facts that then become part of the public record -- may be able to settle for more money than the litigant could win in court.303 Similarly, the litigant may threaten an opponent with disclosure of business strategies or as-yet unpublicized vulnerabilities. An adverse result in litigation may have consequences that extend far beyond the specific matter in issue. Loss of a civil suit may lead to a criminal indictment. The filing of a civil action may bring the plaintiff's or the defendant's conduct to the attention of law enforcement agencies.304 [*pg 1470] One plaintiff's success in an obscure lawsuit may cause an avalanche of such suits by others. Settlement of the initial case prior to filing may have enhanced value because it could prevent such occurrences.

Commentators generally consider the settlement process harmless because they assume that parties are free to refuse a settlement that gives them less than they think they would win at trial.305 The assumption is incorrect. Some judges attempt to "encourage" settlement of their cases through implicit threats of retaliation against recalcitrant parties, and thus make it virtually impossible for those parties to continue to litigate.306 Though the written law may give the parties the right to refuse to discuss settlement, the party who actually refuses may suffer retaliation from the court.307 Once the party is forced into negotiations by this threat, the party will encounter other norms, such as those requiring the party to divulge aspects of its case that would otherwise be confidential.

In multiparty litigation, settlement norms may be enforced by strategic alliances among parties. For example, strategic alliances are generally regarded as important determinants of recoveries -- by liti- [*pg 1471] gation or negotiation -- in large bankruptcy reorganization cases.308 The party who fails to accede to settlement norms -- such as the norm that everybody at the table gets something, regardless of legal entitlement -- may find itself excluded from key alliances.309 The power of these alliances derives in part from the fact that if the case comes before the court with all parties in agreement except one, that one is likely to be overruled.310

IV. IMPLICATIONS

The theory of legal strategy we present here is capable of explaining several phenomena for which current explanations are inadequate or nonexistent.

A. The Effect of Superior Lawyering on Legal Outcomes

By the conventional view -- made manifest both in bar examinations and judicial opinions -- law is an extensive set of instructions to courts on how to decide cases.311 Those instructions are necessarily incomplete in some respects and ambiguous in others. But the instructions are sufficient to render most cases "easy."312 The judge's task in the remaining cases is to select the best result by resolving the ambi- [*pg 1472] guity in light of social conditions and other factors. Were this traditional view correct, lawyers could, in most cases, look up the law before litigation and predict those cases' outcomes. Only rarely could the quality of lawyering change those outcomes.

Observations of the legal system in operation suggest that the quality of lawyering is highly correlated with success. The theory we have presented explains why. Legal outcomes are the products of complex human interactions in which the lawyer can draw not just on written law, but on social norms and prejudices, the law in action, the law in lawyers' heads, informal rules of factual inference, apparent system imperatives, community expectations regarding legal outcomes, and virtually anything else that might persuade the decisionmaker.313 Strategy can constrain judges and other decisionmakers -- or replace them altogether. No a priori limits on this process exist, in or outside of the law. The strategic possibilities are almost limitless, which accounts for the importance of creativity and imagination on the part of the strategist.314 The skilled strategist knows that one can no more predict the outcome of a case from the facts and the law than one can predict the outcome of a game of chess from the positions of the pieces and the rules of the game. In either case, one needs to know who is playing.

B. The Strategic Transformation of Law

In the conventional view, the common law evolves according to the policy choices or "insight" of judges.315 The alternative view -- that litigants drive changes in the law -- has been argued principally by Professors Galanter and Macaulay in the law-and-society literature316 and principally by Professors Rubin and Bailey in the law-and-economics literature.317 In Galanter and Macaulay's theories, the [*pg 1473] mechanisms of change are essentially case selection.318 Although Rubin and Bailey have compiled impressive empirical evidence in support of their position,319 they have not specified mechanisms by which litigants can substitute their own views for the views of judges.320

Our theory of legal strategy suggests an additional, and we think, more powerful, mechanism by which litigants can change the law. The mechanism has essentially three stages. In the first, the litigant succeeds in bringing about a result that is contrary to expectations and conventional understanding of the written law.

In the second stage, others similarly situated copy the successful strategy. This copying is problematic. Successful legal strategy is not always recognizable as such; indeed, it is generally more effective when it goes unnoticed. But the moves that execute the strategy are usually disclosed in public hearings or on public records. Careful observers can piece them together. Even strategies never publicly disclosed or admitted are nearly always in some manner revealed to sufficiently observant opponents.

At this second stage, the decisionmakers or other defenders of the status quo may attempt to neutralize the strategy. If they succeed, the strategy dies. They may fail, however, for a variety of reasons. The legal system relies on certain basic principles that are widely known and virtually impossible to change.321 Those principles cannot be abandoned simply to deny the strategist's victory. Practical diffi- [*pg 1474] culties in making a necessary distinction may prevent the system from making an exception directed against the strategist.

Consider the example of a strategist who creates a judgment-proof business to engage in hazardous activity. To avoid exposing assets to the liability likely to be generated, the strategist rents both plant and equipment at arms length, sells its products only for cash, and distributes earnings to shareholders frequently. In an action by an injured plaintiff, a court might like to defeat the strategy by subjecting the plant and equipment to later judgments against the strategist. The court cannot, however, hold leased plant and equipment generally subject to the claims of tort creditors, because such a basic change in principles would disrupt the plant and equipment leasing industries. Nor can the court make an exception applicable only to strategists who deliberately judgment-proof their businesses. All uses of limited liability are deliberate, and every case in which the exception could be at issue is one in which the debtor ultimately had insufficient assets to pay the tort creditors. The court must allow the strategist to win, because if it does not, the court finds itself on a slippery slope that leads to massive social change.322 The ultimate problem is the inability of the courts to make certain kinds of decisions with sufficient reliability. Unreliable rules or standards generate uncertainty that ultimately creates anxiety and interferes with planning. The legal system must draw lines where it can, which is not always where it thinks best. If the decisionmaker cannot defeat the strategy at this second stage, the strategy changes the pattern of case outcomes.

At the third stage, the legal system recognizes the triumph of the strategy by changing the written law to make it consistent with the case outcomes.323 The system does so to eliminate the expense and [*pg 1475] embarrassment to the system of the strategic activity and the disparity in outcomes between strategists and nonstrategists.

Another example may help in understanding this process of transformation. Nineteenth-century law recognized no method by which a corporate debtor could continue to operate its business without paying its debts as they became due -- a process currently referred to as "bankruptcy reorganization."324 Strategists representing corporate debtors forced the development of such a process.325 As Professor Skeel described it:

In a pattern that reorganization lawyers perfected through time, the [debtor] railroad would arrange for a friendly creditor (generally an out-of-state creditor, to create federal diversity jurisdiction) to file a creditor's bill asking for the appointment of a receiver. Rather than preparing to liquidate assets, as a creditor's bill contemplated, the receivers, who generally included members of the [debtor's] management, worked out the terms of a reorganization. At the same time, the [debtor's] investment bankers formed bondholder "protective committees" and attempted to persuade the bondholders to deposit their securities with the committee, which would commit the bondholders to the terms of the eventual reorganization. Once everything was in place, the bonds and other security interests were foreclosed and the [debtor's] assets were "sold" in a foreclosure sale. In reality, the "sale" simply effected a reorganization of the railroad's capital structure.326

The foreclosure sale theoretically was an exposure of the company to the market -- anyone could have bought it. But in reality, the process was closed. Only the purchaser organized by the insiders could bid because the insiders had unique knowledge of the business, some of the relationships necessary to continue the business may have been personal relationships of the insiders, and the insiders manipulated the timing of the proceedings to their own advantage. Sale by the receiver for less than the amount owing to the creditors left the creditors with less than the full amounts owing to them, yet permitted the shareholder-insiders to remain in ownership and control in their [*pg 1476] role as purchasers at the foreclosure sale. The "reorganization" thus accomplished was a result superior to piecemeal liquidation,327 but the courts undoubtedly would have preferred to have the reorganization without insider dominance. The strategists had, however, eliminated that possibility by being the only ones in a position to bid. Even though a good case might have been made that the insiders were usurping corporate opportunities in breach of their fiduciary duties, the courts had no practical alternative to approval of the proposed sale to the insiders.

After a period in which these equity receiverships enabled insiders to remain in control through a process that reduced the claims of creditors, Congress enacted a bankruptcy reorganization law designed to recognize and regulate the reorganization process.328 An early version, Chapter X of the Bankruptcy Act, attempted to eliminate insider domination by requiring the appointment of a trustee in every case.329 Strategists for the insiders responded by filing their cases under Chapter XI of the Bankruptcy Act -- in violation of the written law -- and negotiating.330 In the Bankruptcy Code adopted in 1978, Congress acknowledged its inability to contain the insiders' strategy331 [*pg 1477] and gave them broad reign.332 Bankruptcy law now formally recognizes a right in the owner-managers of a debtor corporation to remain in ownership and control through a reorganization proceeding specifically designed to reduce the corporation's debts.333

Additional examples of strategic transformation of law abound. During the late 1980s, aggressive debtors' lawyers changed the terms on which consumer debtors could adjust their debts under Chapter 13 of the Bankruptcy Code. Prior to the change, the conventional wisdom held debtors to a choice between the relief available under Chapter 7 (discharge of all debt, without the necessity to make payments to unsecured creditors) and the relief available under Chapter 13 (the ability to save the debtor's home from foreclosure, but only on the condition of paying all of the debtor's disposable income to unsecured creditors for a period of three years). Almost a decade after adoption of the legislative scheme, strategists invented a method for circumventing this choice that became known as a "Chapter 20" -- a Chapter 7 case promptly followed by a second filing under Chapter 13.334 By discharging his or her debts in Chapter 7 before filing the Chapter 13 case, the debtor assured that the "price" of Chapter 13 relief -- payments to unsecured creditors -- would be zero. Apparently unwilling to read into the statute words that were not there, the Supreme Court permitted a Chapter 20 case to stand, thereby putting its imprimatur on the change the strategists had effected.335

Finally, attorney Lincoln Brooks has developed a strategy that gives business debtors what is generally viewed as the most important advantage of a bankruptcy filing -- the ability to stay and bind dissenting creditors -- without having to file bankruptcy.336 Prior to Brooks's strategy, the conventional wisdom held that "dissenting creditors cannot be bound in a workout, but they can be bound in [*pg 1478] Chapter 11."337 Brooks's strategy is to grant a security interest in all of the debtors' assets to the Credit Managers' Association as trustee for all of the unsecured creditors. If a dissenting creditor attempts to levy, Brooks opposes the levy on the novel ground that it violates the newly minted secured creditors' right to priority. Though the case is weak on the written law, it is normatively strong and appealing from a systems standpoint. Brooks can argue that a few dissenting creditors should not be able to overturn the efforts of the majority to negotiate a solution to the company's financial problems. That appeal, combined with favorable economics of litigation, has enabled Brooks to win about fifty consecutive cases. Other lawyers have adopted similar techniques, thereby changing the delivered law.338

In each of these instances, the decision to make a change in how things were done was not made by a judge or a legislature, but by a strategist working on behalf of a party. Perhaps sufficiently determined judges or legislatures could have reversed the outcomes -- if not in the cases where the strategies were first employed, at least in later cases. But even that is not clear. Legal strategy exploits vulnerabilities of the existing system. Comprehensive identification of these vulnerabilities is beyond the scope of this Article. Almost certainly among them, however, are (1) the inability of the system reliably to make desired distinctions, (2) the deliberate retention of "flexibility" in the legal system on the theory that it is necessary to accommodate change, (3) the lack of any social consensus as to the appropriate fora for the resolution of legal disputes, and (4) the diffusion of decisionmaking power among tens of thousands of judges in the United States alone. Vulnerabilities such as these give rise to the indeterminacies that legal strategists exploit.

C. Explaining Local Legal Cultures

Written law purports to be universal within geographical boundaries.339 Because the same law applies throughout each jurisdiction, the conventional view suggests that legal outcomes should be [*pg 1479] uniform throughout each as well. Empirical research, however, has repeatedly discovered large variations in legal outcomes from place to place within jurisdictions, a phenomenon referred to as "local legal culture."340 The mystery is why and how these systematic variations in legal outcome occur from one part of a jurisdiction to another when the law that supposedly generates them is the same.

The answers lie in the fact that several of the factors we have identified -- social norms, the law in lawyers' heads, and expectations regarding outcomes -- are not uniform within the neat boundaries of legal jurisdiction. They are forged by frequent interactions among members of groups. The locations of these groups are rarely coextensive with the city, state, or national boundaries that define the reach of legal doctrine. Because social norms, the law in lawyers' heads, and expectations regarding outcomes are oral traditions, they may have little credibility to outsiders and little effect outside the group. The law in lawyers' heads may be different from city to city, or city to rural area, yet the same throughout national or international groups whose members interact frequently. Social norms exist within virtually every group of people who interact repeatedly, from the citizens of a nation to the members of a family.341 They differ by religion, race, interests, geographical location, occupation, institution, and any other factor that brings groups together.

[*pg 1480]

Such differences explain, at least in part, the processes by which local legal cultures arise, change, and disappear. The cultures are constantly in flux because the underlying factors that give rise to them are in flux. Through legal strategy, changes in social norms and expectations are translated into changes in legal outcomes, even while the written law remains constant. These changes in outcomes vary from place to place within a jurisdiction and change abruptly over time for the simple reason that the factors producing them, including legal strategies, change. If local legal cultures were merely the products of general cultural differences from place to place -- the general practices, perceptions, and expectations of the people who live in a community -- they could not appear and disappear nearly as quickly as they do.342

D. Why the "Haves" Come Out Ahead

The theories we present here extend Professor Galanter's theory as to why the "haves" come out ahead, by explaining new means by which spending more resources343 on litigation can produce more favorable outcomes for the spenders. If law operated in accord with conventional legal theory, resources would affect results only in the small minority of cases in which determinative facts remained undiscovered or the result specified by law was unclear. Yet, resources consistently produce good or acceptable results in what had seemed to be the most hopeless of cases.

Resources matter because they unleash strategy, and strategy is capable of altering legal outcomes across a wide range of possibilities. Adverse legal doctrine defeats only those who believe it can. For nonbelievers, the strategic application of resources can construct out- [*pg 1481] comes to order, within cultural limits. The more resources the party can spend, the better the outcomes the party is likely to get.

That does not mean that the party with greater resources will prevail -- or even have an advantage -- in every case. The resources that a party can profitably spend on a case are limited by the value of the case to that party. It follows that if a case involves only money -- as many commercial disputes do -- neither party should be willing to spend more than the amount in issue. Of course, if one of the parties is not sufficiently liquid, that party may not be able to invest optimally in the case. In addition, a given level of expenditure by opposing sides in a given case may do more for one side than the other, as where one side is a frequent litigator and can reap economies of scale.

If the dispute is over a nonmonetary right -- the validity of a patent or the right to control a corporation -- the win may be worth more to party A than to party B. The party with more to gain can probably cost-justify larger expenditures than the opposing party, giving the former an advantage in the litigation. This clash of strategists might have some tendency toward efficiency because the party that places a higher value on a favorable outcome is more likely to get it.344 But for strategy to be "efficient" in the economic sense, its tendency toward efficiency would have to be greater than the tendency toward efficiency of the other systems that might be adopted. That seems unlikely, because strategy is of greatest use in cases that the legal system would otherwise regard as nonmeritorious. The types of cases in which Professor Galanter initially propounded his theory of why the "haves" come out ahead -- cases in which the "haves" are repeat players but their opponents are not -- fall within this class.

Probably few legal disputes are simply over money. Among those that are, settlement usually produces a better result for both sides. The cases that go to litigation tend to be either cases in which emotional investments are high and parties are interested not only in what [*pg 1482] benefit they can win but also in what harm they can inflict on others, or cases the parties believe will establish important precedents. The responsiveness of the legal process to infusions of resources will tend to give the "haves" advantages over the "have nots" in these cases.345

As discussed above, innuendo -- the oblique allusion to matters that must remain unarticulated -- is a principal skill of lawyer-strategists. Some lawyers will be better at it than others. That may occur because they have superior linguistic skills generally, or a better understanding of what can be left unsaid with the particular decisionmaker. This difference in skill level inures to the benefit of those who can afford the best advocates. Though that is generally the wealthier side, because of the contingency fee system it is far from invariably so.

EVALUATION AND CONCLUSIONS

In the conventional model of the legal process, the lawyers present the facts of a case and the governing law to the judge. The judge applies the law to the facts to reach a decision. If the decision is wrong, the losing party can appeal to a higher court that will set it right. Settlements can, and usually do, occur along the way, but these settlements are merely estimates or approximations of the results the courts would otherwise reach. No room exists within that model for legal strategy.

The alternative model we have presented is a more open-ended one. When the stakes are sufficiently large, lawyers assume the role of strategists. They compete in the selection of decisionmakers from an unlimited array of judges, arbitrators, private parties in settlement, or other persons. They employ a wide variety of strategies to reach or prevent adjudication. In the small minority of cases in which adjudication occurs, the strategists argue from a similarly wide variety of materials, which include social norms, law in action, law in lawyers' heads, informal rules of factual inference, system imperatives, community expectations, public policy, and written law. Because adjudicated outcomes depend so much on the predispositions of the adjudicators and the strategies employed in presentations, those outcomes differ widely from judge to judge, case to case, and place to place. [*pg 1483] Appeals are rare, and reversals even rarer. Strategy, not law, is the principal determinant of outcomes.

Once the dominant role of legal strategy is accepted, some undoubtedly will defend it on the ground that the vulnerability of the system to strategy gives the system an added measure of stability. That is, the system bends to accommodate power. We argue, to the contrary, that legal strategy renders legal outcomes unpredictable, making the system less stable. To the extent that strategies determine the outcomes of cases, the merits of cases as traditionally perceived do not. The injustice causes disrespect for the legal system, which ultimately undermines its legitimacy.

Because legal strategies are attempts to manipulate the outcomes of cases irrespective of their supposed merits under written law, strategies are widely viewed as unethical. Lawyers are reluctant to publicize the strategies they pursue partly for that reason. But an even more important reason is that most lawyers understand that strategies work best when unnoticed. Articulation can, and usually does, render them ineffective. These factors combine to drive legal strategy largely underground.

Some of the legal scholars who share our view of the ubiquity and importance of legal strategy view the problem of legal strategy as insoluble. Strategy, they say, permeates not only law, but life.346 No matter what system is in place, strategists will attack it and transform it into something other than what was intended.347

Though we generally agree with this view, we think it ignores a key aspect of law-related systems:348 they differ in the degree of their vulnerability to strategy. Better systems generate less strategic activity. That strategic activity is more likely to be of the kinds anticipated or even intended by the systems' designers. Strategies can still gain advantage in the better systems, but they can gain less. Accordingly, the first means we propose for reducing the impact of strategy on le- [*pg 1484] gal outcomes is to redesign the legal system to provide fewer incentives for system-unintended strategy. 349

Articulation offers a second means for reducing the impact of strategy on legal outcomes. The articulation need not be by participants in the case. Legal scholars can do it as well. Articulation will not be easy. Because the strategist conceals the strategy, the articulation is necessarily speculative, and therefore dangerous to the articulator. Inevitably, some articulations will miss the mark. Moreover, articulation operates largely by embarrassing the strategists, and, if it comes from a third party, such as the press or an empirical researcher, may also embarrass the judges and opposing lawyers who have been the strategists' unwitting victims. Thus none of the participants are typically willing to divulge information regarding the nature, use, or effects of strategy.

Articulation alone cannot solve the problem of legal strategy. Although its application tends to neutralize the articulated strategy, the effect is uncertain and usually incomplete. The court that receives a case may be embarrassed by the articulation of a forum-shopping strategy but not always enough to send the case back. The neutralization of one strategy still leaves others effective.

In the coming age of information, the task of articulating legal strategy may become easier. Computerization of the legal process -- in the courts, in lawyers' offices, and in the offices of business and governmental clients -- is documenting the pattern of cases and outcomes. Though such data generally does not capture the motives of participants and therefore does not document the strategies themselves, it does capture the footprint of those strategies in the form of disparities in legal outcomes from place to place, race to race, judge to judge, and, perhaps soon, lawyer to lawyer.350

[*pg 1485]

This development parallels one that has already had a profound effect on the medical services delivery system. Data generated for a variety of other purposes has been analyzed to show that medical procedures employed and outcomes obtained differ sharply from place to place and provider to provider.351 The effect has been to focus attention on the causes of those differences.352

Legal scholars, working with the huge masses of electronic data that have been accumulating in the legal system in recent years, may soon be able to focus similar attention on legal strategy. The legal system already has been profoundly affected by scholars' documentation of the relationship between defendants' race and the imposition of the death penalty,353 the rampant forum shopping in major business bankruptcies,354 the nullification of securities class action reform by the simple expedient of moving the litigation to another set of courts,355 and the lack of relationship between formal circuit rules governing the rules for publication of opinions and actual publication practices.356

[*pg 1486]

Data capable of documenting the effects of legal strategy on legal outcomes already exists. Those who have it -- principally the courts -- are not likely to surrender it easily in usable form.357 But as the data is inevitably released, the existence and importance of legal strategy in the generation of legal outcomes will become increasingly difficult to deny. When that occurs, scholars, judges, and other participants will be forced to acknowledge strategy's dominant role in the legal process.

FOOTNOTES

唵 Stephen C. O'Connell Chair and Distinguished Professor, University of Florida Levin College of Law (weyrauch@law.ufl.edu). We thank Jane Aiken, Maureen Bell, Debbie Carmichael, Frances Foster, Lawrence Friedman, Marc Galanter, Mitu Gulati, Jeffrey Harrison, Peter Joy, Ken Klee, Ron Levin, Steve Legomsky, Michael Lopez, Stewart Schwab, Gary Schwartz, Elizabeth Warren, Bill Whitford, and participants in the Washington University Law School Workshop for comments on earlier drafts; Eric Zolt for assistance with respect to income tax strategy; and Jeff Raven for assistance with research.

1. THE BRAMBLE BUSH 14 (1960). Llewellyn's reference was to statutes banning resale price maintenance contracts. In those contracts, retailers agreed not to sell the manufacturer's product for a price less than that specified in the contract. See id.

2. The conventional view is typically described in terms such as these:

The common-law decisional process starts with the finding of facts in a dispute by a fact-finder, be it a jury, judge, or administrative agency. Once the facts are ascertained, the court compares them with fact patterns from previous cases and decides if there is sufficient similarity to warrant applying the rule of an earlier case to the facts of the present one.

RUGGERO J. ALDISERT, LOGIC FOR LAWYERS: A GUIDE TO CLEAR LEGAL THINKING 2-4 (1992).

The breadth of acceptance of the conventional view is illustrated in Marc Galanter's classic article on why the "haves" come out ahead. Although Professor Galanter ultimately argues for a partially strategic explanation of case outcomes, he begins by starkly assuming the conventional view as a frame of reference:

This society has a legal system in which a wide range of disputes and conflicts are settled by court-like agencies which purport to apply pre-existing general norms impartially (that is, unaffected by the identity of the parties). . . . The rules applied by the courts are in part worked out in the process of adjudication (courts devise interstitial rules, combine diverse rules, and apply old rules to new situations). There is a living tradition of such rule-work and a system of communication such that the outcomes in some of the adjudicated cases affect the outcome in classes of future adjudicated cases.

Marc Galanter, Why the "Haves" Come Out Ahead: Speculations on the Limits of Legal Change, 9 LAW & SOC'Y REV. 95, 96 (1974). That the primary objective of a player in Galanter's model is to win rules of law favorable to its side suggests their centrality.

3. See, e.g., Frederick Schauer, Easy Cases, 58 S. CAL. L. REV. 399, 426-38 (1985) (arguing that the textual language and precedent of the law limit its permissible interpretations). Bankman overstates the conventional view when he argues:

[A]cademics believe that the vast majority of issues can and should be decided without recourse to legislative history or intent. Legislative intent or purpose becomes relevant, however, in those situations in which the seemingly "plain language" of a statute or rule admits more than one meaning, conflicts with the plain language of other statutes or a body of well-accepted doctrine, or produces a nonsensical result.

Joseph Bankman, The Proposed Partnership Antiabuse Rule: Appropriate Response to Serious Problem, 64 TAX NOTES 270, 271 (1994). But a substantial number of academics do hold these beliefs, and probably most judges believe that they act in accord with them.

4. See, e.g., Mark D. Rosen, Nonformalistic Law in Time and Space, 66 U. CHI. L. REV. 622, 623 (1999) ("'Standards' are laws that describe a triggering event in abstract terms that refer to the ultimate policy or goal animating the law."). Nearly all academics subscribe to one or the other version of the conventional view. See, e.g., Richard H. Pildes, Forms of Formalism, 66 U. CHI. L. REV. 607, 620 (1999) ("To be sure, as Hart and Sacks recognized, there is much at stake in our age over whether the central products of our modern legal system-statutes-are read literally or purposively.").

5. See, e.g., Felix S. Cohen, Transcendental Nonsense and the Functional Approach, 35 COLUM. L. REV. 809, 842-47 (1935) (presenting a "theory of legal decisions" that assumes social forces are given effect through the judge).

6. But see Thomas Michael McDonnell, Playing Beyond the Rules: A Realist and Rhetoric-Based Approach to Researching the Law and Solving Legal Problems, 67 UMKC L. REV. 285, 290-300 (1998) (rejecting formalism and proposing in consequence that law schools teach students to research judges, lawyers, and other participants in litigation).

7. See, e.g., Leo Katz, Form and Substance in Law and Morality, 66 U. CHI. L. REV. 566, 566 (1999) (presenting six "familiar examples" of controversial legal strategy); id. at 595 (concluding that "[l]awyers routinely exploit law's formality").

As employed in this Article, "strategy" has been defined as "the art of devising or employing plans or stratagems toward a goal." MERRIAM-WEBSTER'S NEW COLLEGIATE DICTIONARY 1162 (10th ed. 1996). A "stratagem" is "a cleverly contrived trick or scheme for gaining an end." Id.; see also 2 HAROLD D. LASSWELL & MYRES S. MCDOUGAL, JURISPRUDENCE FOR A FREE SOCIETY 1046 (1992) ("A strategy is a sequence of practices in which base values are utilized to influence outcomes and effects.").

8. For example, Dow Corning is settling the breast implant cases for over $2 billion even though the scientific evidence seems clear that silicon cannot cause the injuries claimed. See Gina Kolata, A Case of Justice, or a Total Travesty? How the Battle over Breast Implants Took Dow Corning to Chapter 11, N.Y. TIMES, June 13, 1995, at D1 (stating that there is no scientific evidence that implants cause serious disease). Securities class action lawyers have continued to win cases despite legislation intended to thwart them. See, e.g., America's Entrepreneurs Advocate Securities Litigation Reform, BUS. WIRE, June 10, 1997 ("Congress in 1995 overwhelmingly passed a sweeping measure that was intended to stop the filing of abusive securities class action lawsuits in federal court. But some lawyers who specialize in these types of suits are thwarting the 1995 law by filing them in state courts.").

9. See generally, e.g., JONATHAN HARR, A CIVIL ACTION (1995) (demonstrating that financial pressures on plaintiffs' attorneys made it virtually impossible for them to take a case to trial).

10. See, e.g., William T. Pizzi & Walter Perron, Crime Victims in German Courtrooms: A Comparative Perspective on American Problems, 32 STAN. J. INT'L L. 37, 46 (1996) (referring to the American legal strategy of "attacking the victim's character while keeping the defendant's prior record away from the jury").

11. Only an omniscient observer could attest that a guilty person had been set free. But skilled defense lawyers have freed defendants against whom the evidence could hardly have been more compelling. See, e.g., W. William Hodes, Lord Brougham, The Dream Team, and Jury Nullification of the Third Kind, 67 U. COLO. L. REV. 1075, 1077 (1996) (arguing that O.J. Simpson's lawyers did not act unethically in obtaining his acquittal even though the evidence suggested guilt); Jay Mathews, DeLorean Acquitted of All Eight Charges in Drug-Scheme Trial, WASH. POST, Aug. 17, 1984, at A1 (reporting that John Z. DeLorean was acquitted of all charges even though he was "caught on videotape discussing cocaine deals with government agents posing as drug dealers"); John Needham, The Lure of Fame, Fortune; Bruce Cutler Defended a Mob Don; Now a High-Profile Case Brings Him West, L.A. TIMES, Apr. 28, 1992, at E1 (noting that John Gotti, the head of the Gambino crime family, was acquitted all three times that Bruce Cutler defended him but convicted in a case in which the court barred Cutler from representation).

12. See generally DAVID B. BAUM ET AL., ADVANCED NEGLIGENCE TRIAL STRATEGY 83-98 (1979) (providing a sample opening statement presented in an actual products liability case); XAVIER FRASCOGNA, NEGOTIATION STRATEGY FOR LAWYERS (1984) (providing "a comprehensive treatment of the various patterns, principles and techniques that govern negotiation in the context of a law practice"); SIMON N. GAZAN, ENCYCLOPEDIA OF TRIAL STRATEGY AND TACTICS 75 (1962) (describing the importance of a trial lawyer's knowledge of psychology to defense jury selection strategy); CHARLES ROTHENBERG, MATRIMONIAL LITIGATION: STRATEGY AND TECHNIQUES (1972) (discussing strategies for selecting a jury that would likely be more favorable to a plaintiff seeking divorce). There are also entire periodicals devoted to legal strategy. See generally, e.g., BANKRUPTCY STRATEGIST, COMPUTER LAW STRATEGIST, CORPORATE TAX STRATEGY, THE JOURNAL OF STRATEGY IN INTERNATIONAL TAXATION, DEPOSITION STRATEGY: LAW AND FORMS, EMPLOYMENT LAW STRATEGIST, and INTELLECTUAL PROPERTY STRATEGIST.

13. See infra Part IV.

14. Another part of the explanation is the understandable reluctance of lawyers who engage in strategy to admit that they do so. The use of strategy is often condemned as unethical, see infra Part V, and the articulation of strategy tends to destroy its effect, see infra note 210 and accompanying text. Together, these two factors remove practicing lawyers-the persons most knowledgeable about strategy-from the discussion of its role in the legal system.

15. Sophisticated parties sometimes devise and manage the execution of legal strategies, but the large majority are devised and executed by lawyers. Our references solely to lawyers as the initiators throughout this Article should be understood to include parties to the extent of their participation.

16. But see 2 LASSWELL & MCDOUGAL, supra note 7, at 1067-73 (listing strategic decisions made by lawyers during litigation); Katz, supra note 7, passim (arguing that strategy operates in the realm of law in essentially the same manner that it operates in the realm of morality).

17. See, e.g., Professional Responsibility: Report of the Joint Conference, 44 A.B.A. J. 1159, 1161 (1958) (distinguishing vigorous advocacy from "muddy[ing] the headwaters of decision"). But see JEROME FRANK, COURTS ON TRIAL: MYTH AND REALITY IN AMERICAN JUSTICE 82-90 (1949) (describing adversarial presentation in terms of strategy and tactics).

18. See Donald Farole, Reexamining Litigant Success, 33 LAW & SOC'Y REV. (forthcoming Sept. 2000) (reviewing the literature). This literature often refers to "legal strategy" as the means by which skilled lawyers influence outcome, but it does not attempt to explain what the skilled lawyers do to influence outcome. See, e.g., Steven C. Tauber, The NAACP Legal Defense Fund and the U.S. Supreme Court's Racial Discrimination Decision Making, 80 SOC. SCI. Q. 325, 326 (1999) (referring to "skillful legal argument" and "compelling legal argument").

19. See supra note 2; infra Part III.B.1. Legal strategy is not, of course, the only thing that may be in the black box driving outcomes.

20. Professor Cynthia Williams has dubbed this view the "law as price" view of law and criticized it as an attitude toward compliance. Cynthia A. Williams, Corporate Compliance with the Law in the Era of Efficiency, 76 N.C. L. REV. 1265, 1267-68 (1998). The point we make here is unaffected by hers.

21. See Kaushik Basu, The Role of Norms and Law in Economics: An Essay of Political Economy 15-17 (Dec. 10, 1997) (unpublished manuscript, on file with the Duke Law Journal) (arguing that judges should be regarded as players in the economic game).

22. See, e.g., Meacham Corp. v. United States, 207 F.2d 535, 544 (4th Cir. 1953) ("[T]hey employed the New York lawyers to find a way through the stone wall of the statutes . . . .").

23. See infra Part II.

24. See infra Part III.

25. See infra Part IV.

26. Whether an issue constitutes the merits or is merely procedural or tangential is itself frequently the subject of legal strategy.

27. See, e.g., Ann Davis, How a Lawyer Turned Tables in Tobacco Case, WALL ST. J., Oct. 4, 1999, at B1 (noting a speech by tobacco company lawyer William Hendricks III "about how prosecutors can be 'mortally wounded' when they decide to skirmish over legal and factual issues in court before indictments are handed up"); John Gibeaut, Another Broken Trust, A.B.A. J., Sept. 1999, at 40, 41 (describing the Indian trust case as "spinning out of control" as a result of discovery decisions and missed strategic opportunities). To the extent that cases are decided on the merits, they cannot "spin out of control" before their conclusion. Nor can an agreement to give discovery compromise them in any way.

28. See, e.g., Daniel A. Farber, Legal Formalism and the Red-Hot Knife, 66 U. CHI. L. REV. 597, 604 (1999) ("A formalistic system puts a great premium on people (particularly lawyers) who are able to invent clever ways to manipulate the rules to produce desired outcomes."); Katz, supra note 7, at 595 ("Lawyers routinely exploit law's formality.").

29. Texaco Inc. v. Pennzoil Co., 729 S.W.2d 768, 784-87 (Tex. App.), rev'd, 481 U.S. 1 (1987).

30. One of us has argued elsewhere that "[t]he outcome of a case can be predicted more accurately at the level of the whole case than at the level of the case-dispositive decisions within the case that supposedly produce the outcome." Lynn M. LoPucki, Legal Culture, Legal Strategy, and the Law in Lawyers' Heads, 90 NW. U. L. REV. 1498, 1527 (1996). This "whole-case realism" puts a damper on what can be accomplished through legal strategy. It is also worth noting that the predictions of whole-case realism referred to are not made from law, but from community expectations regarding outcomes.

31. See authorities cited supra notes 3-5.

32. See Texaco, 729 S.W.2d at 784-87 (describing the facts of the case).

33. The amount of the jury verdict was $10.53 billion. The amounts of the verdict, judgment, and bond are referred to below in amounts that range from $10 billion to $12 billion. The variation results from a reduction in amount by the Texas Court of Appeals, an increase in the amount through the accrual of interest, and rounding by various commentators.

34. THOMAS PETZINGER, JR., OIL & HONOR 391-92 (1987).

35. See Texaco, 729 S.W.2d at 784.

36. See id.

37. See id. at 866.

38. See Allanna Sullivan, Texaco's War with Pennzoil Ends Officially, WALL ST. J., Mar. 24, 1988, at 3.

39. We do not discuss the strategic manipulation of juries because the conventional view already acknowledges it. On jury manipulation in Texaco, see Walter O. Weyrauch, Unwritten Constitutions, Unwritten Law, 56 WASH. & LEE L. REV. 1211, 1232-37 (1999).

40. See PETZINGER, supra note 34, at 241-42.

41. See id. at 248.

42. See id. ("[Pennzoil's] lawyers figured that the loss of independence by one of Tulsa's biggest and most-loved companies might dispose the judiciary unfavorably toward Texaco's megadeal.").

43. See, e.g., Walt Harrington, Born to Run: On the Privilege of Being George Bush, WASH. POST MAGAZINE, Sept. 28, 1986, at W16 (noting that Liedtke's father was Gulf Oil's chief counsel in Tulsa and that Liedtke raised money in Tulsa for an investment with George Bush).

44. See STEVE COLL, THE TAKING OF GETTY OIL 383 (1987) (referring to the Oklahoma and Rhode Island suits as "sponsored private antitrust suits"); PETZINGER, supra note 34, at 249 (reporting that Pennzoil had given Fairlawn Oil Service $50,000 to file the Rhode Island case).

45. COLL, supra note 44, at 384.

46. See id. at 385-86.

47. See PETZINGER, supra note 34, at 261.

48. Leave probably would not have been granted for two reasons. First, Texaco would have filed an answer before the hearing, thereby joining the issue. Second, Pennzoil's dismissal was merely for the purpose of refiling in another court after losing the initial hearing in the Delaware court-an obvious instance of forum shopping.

49. See COLL, supra note 44, at 387 (discussing the failure to file the answer and noting that an experienced attorney could have dictated it in thirty minutes).

50. Id. at 386.

51. See PETZINGER, supra note 34, at 261 (referring to the rule permitting dismissal without notice as "seldom-used").

52. See id. at 384 (characterizing the decision to bring the Texas case as virtually an afterthought).

53. Pennzoil's case has generally been acknowledged to have been weak on the merits and to have succeeded only because it was tried to a populist Texas jury. See, e.g., Michael Ansaldi, Texaco, Pennzoil and the Revolt of the Masses: A Contracts Postmortem, 27 HOUS. L. REV. 733, 834-40 (1990).

54. See, e.g., Texaco Inc. v. Pennzoil Co., 626 F. Supp. 250, 255 (S.D.N.Y.), modified, 784 F.2d 1133 (2d Cir. 1986) ("This Court therefore believes that the compensatory damages which flow from this breach of contract, tortiously induced by Texaco, should in no event exceed $800 Million.").

55. Pennzoil's case against Texaco proceeded in Texas on the assumption that New York law applied. See Texaco Inc. v. Pennzoil Co., 729 S.W.2d 768, 858 (Tex. App.), rev'd, 481 U.S. 1 (1987) ("In the instant case, there is no dispute that the governing substantive law was New York law."). The strategic view recognizes that conflict of laws is only a marginally useful concept. Texas courts might state that they apply New York law, but they respond principally to the factors described in Part II.B, which are local. See Weyrauch, Unwritten Law, supra note 39, at 1233.

56. See PETZINGER, supra note 34, at 282.

57. See id. Jamail contributed another $10,000 to the campaign of the administrative judge with supervisory power over Farris. Jamail's other contributions-to seventeen other judges-totaled less than $13,000. See id. at 288.

58. See id. at 290-91.

59. See id. at 285-374; id. at 289 ("[E]very single pretrial ruling that Miller [Texaco's lead attorney] could think of had gone Pennzoil's way.").

60. But see generally Daniel H. Lowenstein, When is a Campaign Contribution a Bribe? (1996) (unpublished manuscript, on file with the Duke Law Journal) (taking the position that the kind of campaign contribution made by Jamail and accepted by Farris is arguably criminal).

61. See PETZINGER, supra note 34, at 288-91 (describing the court fight over the contribution).

62. See id. at 388-89.

63. See id. at 384-85, 389 (discussing Texaco's cautious decision).

64. See, e.g., Janet Elliott, Lasting Impact; Legal Anomaly; A Look Back at the Real Trial of the Century, TEX. LAW., Dec. 18, 1995, at 1 (describing and defending the decision).

65. See PETZINGER, supra note 34, at 403-04 (describing the jury deliberation on damages).

66. See id. at 321 (describing the illogic of Pennzoil's damage calculation and referring to it as "outrageous").

67. See, e.g., Joseph Sanders, From Science to Evidence: The Testimony of Causation in the Benedectin Cases, 46 STAN. L. REV. 1, 56 n.248 (1993) ("In hindsight, Texaco's failure to present damages evidence was unwise and could only be justified by a particularly uncomplimentary view of the jury's decisionmaking capacity."); How Texaco Lost Court Fight, N.Y. TIMES, Dec. 19, 1985, at D1 (stating that Texaco attorney Richard B. Miller acknowledges that his decision not to put on damage evidence "has since drawn a torrent of criticism"); Tamar Lewin, Pennzoil-Texaco Fight Raised Key Questions, N.Y. TIMES, Dec. 19, 1987, at A44 (quoting Columbia law professor Harvey Goldschmid as saying that "[o]ne area that was obviously a mistake was Texaco's failure to deal with evidence of damages at trial"). But see Texaco Inc. v. Pennzoil Co., 729 S.W.2d 768, 859-63 (Tex. App.), rev'd 481 U.S. 1 (1987) (concluding that the damage evidence presented by Pennzoil was sufficient to support the verdict).

68. PETZINGER, supra note 34, at 417.

69. The campaign appears to have made an impression on Chief District Judge Brieant. See Texaco Inc. v. Pennzoil Co., 626 F. Supp. 250, 252 (S.D.N.Y.), modified, 784 F.2d 1133 (2d Cir. 1986) (describing the court's receipt of numerous letters from Congressmen and Senators, along with amicus curiae briefs from several states).

70. See PETZINGER, supra note 34, at 430; see also id. at 434 ("A Texaco executive later swore in an affidavit that the refinery switch was a long-planned move intended as part of a plan to consolidate the assets of Getty Oil.").

71. See TEX. BUS. & COM. CODE ANN. ?24.005 (West 1999) (outlining, in the Uniform Fraudulent Transfer Act, situations in which debtors' transfers to creditors are fraudulent).

72. See Thomas Petzinger et al., High-Stakes Poker: Texaco and Pennzoil, with Truce Expiring, Again Talk of Settling, WALL ST. J., Mar. 14, 1986, at A1.

73. Bankruptcy would have been less effective if filed after imposition of the liens. For example, the liens would have constituted a default in some of Texaco's obligations, activating contractual default rates of interest.

74. See Texaco Inc., 626 F. Supp. at 251-52; PETZINGER, supra note 34, at 438.

75. See PETZINGER, supra note 34, at 440.

76. See Texaco Inc., 626 F. Supp. at 262.

77. See id. at 255 ("This Court therefore believes that the compensatory damages which flow from this breach of contract, tortiously induced by Texaco, should in no event exceed $800 Million . . . .").

78. See PETZINGER, supra note 34, at 448 ("If Texaco failed to reverse the actual judgment in the Texas appellate courts, the [New York federal judge] indicated he might do it himself, taking the entire case into his own hands."); see also Texaco Inc., 626 F. Supp. at 259 ("[E]ven if the Judgment ripened into a final judgment in this vast amount of money, and assuming all decisions down the road go against Texaco, Texaco would still have the federal claims that are pleaded in this action, which it could litigate on the merits."). In the New York action, Texaco pled five constitutional and statutory bases for the district court to nullify the Texas judgment. See id. at 251. To concern oneself with whether the claims were "right" or "in accord with the law" is to miss the point. A technically adequate basis exists for virtually any action a court may choose to take. The only effective limit is the cultural acceptability of the results.

79. See PETZINGER, supra note 34, at 449.

80. See Texaco Inc. v. Pennzoil Co., 784 F.2d 1133, 1157 (2d Cir. 1986).

81. See Pennzoil Co. v. Texaco Inc., 481 U.S. 1, 9 (1987) .

82. See KEVIN J. DELANEY, STRATEGIC BANKRUPTCY 137-41 (discussing Texaco's ability to post the bond). Although the amount of the bond exceeded world bonding capacity at that time, Texaco could have liquidated assets sufficient to post the bond in cash. And although those assets might not have fetched the best possible price under the circumstances, there is no suggestion that the proceeds of sale would have been insufficient to meet the full amount of the bond. See infra note 86. Texaco had 17 months from the verdict in November 1985 to the bankruptcy filing in April 1987 to initiate such sales.

83. See DELANEY, supra note 82, at 146 ("Texaco's favorable balance sheet left many performing 'linguistic contortions' trying to explain why the company was in bankruptcy.").

84. See id. at 144-46.

85. See, e.g., Lawrence Ponoroff & F. Stephen Knippenberg, The Implied Good Faith Filing Requirement: Sentinel of an Evolving Bankruptcy Policy, 85 NW. U. L. REV. 919, 984-89 (1991) (discussing the doctrine barring bad faith filings and whether Texaco filed in good faith).

86. See Texaco Inc. v. Pennzoil Co., 784 F.2d 1133, 1155 (2d Cir. 1986) ("In the present case there is no serious dispute that, should Texaco be required to liquidate its substantial assets, it would be able to pay Pennzoil's judgment in full."); DELANEY, supra note 82, at 146 (projecting Texaco's liquidation value somewhere between $22 billion and $26 billion); PETZINGER, supra note 34, at 441 (estimating a $26 billion net worth); Thomas C. Hayes, Pennzoil's Chances Unclear, N.Y. TIMES, Apr. 14, 1987, at D1 ("Pennzoil and Texaco agreed in court documents filed in Houston last week that Texaco's total assets, more than $34 billion, were more than adequate to pay $10 billion to Pennzoil if Texaco loses its appeals.").

87. DELANEY, supra note 82, at 167-68.

88. See id. at 167 ("Texaco continually argued that paying the damages was not financially impossible, but rather 'financially impracticable.'").

89. One possibility was fear that litigation over whether the filing was in good faith would itself delay the bankruptcy case. That appears to have occurred in the Manville bankruptcy. See In re Johns-Manville, 42 B.R. 654, 655 (Bankr. S.D.N.Y. 1984) (denying leave to take an interlocutory appeal regarding the issue of good faith filing because the appeal itself would further delay the case).

90. See 11 U.S.C. ?1129(b)(2)(B) (1994) (giving unsecured creditors absolute priority over shareholders as a condition for confirmation of a chapter 11 plan).

91. A second objective may have been to change fora, and with it, judges. But at this late stage of the litigation, there was probably little that a new judge could have done for Texaco except delay. Cf. Pennzoil Co. v. Texaco Inc., 481 U.S. 1, 24 (1986) (Marshall, J., concurring) ("[T]he odor of impermissible forum shopping . . . pervades this case.").

92. See DELANEY, supra note 82, at 146 ("Even Texaco admitted that the bankruptcy filing was mainly a strategic effort to delay paying the Pennzoil court award."); Robert H. Mnookin & Robert B. Wilson, Rational Bargaining and Market Efficiency: Understanding Pennzoil v. Texaco, 75 VA. L. REV. 295, 324-29 (1989) (suggesting that Texaco's managers sought delay in bankruptcy as a means of leveraging releases of their liability to Texaco's shareholders for the events leading to the judgment).