Procedure

Abstract

The purpose of this chapter is to provide a general and cross-national economic analysis of procedural law, both civil and criminal. The presentation here emphasizes two refining perspectives that have been neglected or de-emphasized in the prior literature: (1) the property of procedural law as a substitute for as well as a complement to substantive law; and (2) the tradeoffs involved in a second dimension of subtstitution, which is between private and public enforcement. Within this framework, economic analysis is compared and contrasted with normative criteria as supplied by conventional legal dogmatics.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

• (a) In its broadest sense, procedural law is the “law of law”. Procedural law is concerned with the making, application, and enforcement of substantive law. In this broad sense, procedural law would include constitutional rules of law production, such as constitutional arrangements among the branches of government or criteria for the enactment of legislation and administrative regulations. In a narrower sense, procedural law is concerned with specifying the conditions under which legal controversies are adjudicated by tribunals, including courts, administrative bodies, or other dispute-resolution entities such as arbitral panels. It is this narrower sense of procedural law to which this chapter is addressed primarily.

• (b) Conventionally, the law governing the adjudication of legal disputes includes the description of what disputes are ripe for determination by adjudication, what tribunals have the competency or jurisdiction to consider the dispute, how the applicable rules of substantive law are ascertained and applied, how the facts pertaining to the dispute are investigated and found, how the decision or judgment is rendered, what effect that judgment may have on subsequent disputes, and how that judgment may be reviewed, attacked, or reexamined. Procedural rules also govern the enforcement of judgments.

• (c) From the economic perspective, procedural law may be studied from either a macroscopic (aggregated) or a microscopic (individual cases) viewpoint. There is some literature studying procedure from the macroscopic viewpoint. A recent empirical case study on Austria has found a significant positive correlation between the increase in the real GNP/person and the volume of litigation.¹ Most of the existing law-and-economics literature takes a microscopic viewpoint, focusing on the combined effects of procedural and substantive rules on the incentives and behaviors of individual participants in both civil and criminal cases, including the parties to an adjudication and the individuals composing the tribunal.

• (d) From this viewpoint, the dominant economic model of legal procedure is what might be termed the “expected value” model of adjudication. This model explains litigation as the result of the predictions of both the plaintiff and defendant—actual or potential—as to the envisaged outcome in terms of the respective costs and benefits involved. Because in most such cases the existence of the legal entitlement is in dispute, the parties’ expectations are deflated by the probability of success in the claim, and hence the term “expected value” when referring to contested claims. The “expected value” model poses the problem as one of reconciling the differing views of the parties to a legal dispute, and of characterizing the interactive effects of the parties’ actions on each other. Therefore, a great deal of attention has been devoted to considering the conditions under which the parties’ expectations may converge, thereby allowing the dispute to be settled, that is, to be resolved by formal agreement before definitive adjudication. However, this approach obviously presupposes some type of default outcome in the absence of agreement. Furthermore the adjudicated outcome is not simply a matter of distribution between contesting parties, but can have external effects on social welfare. The conventional point of view thus poses the social problem of adjudication as minimizing the sum of two types of costs: (i) the “direct” costs of the adjudication itself; and (ii) the “error” costs associated with legally or factually erroneous judgments, which can extend beyond the immediate parties. Even at this primitive level, it is clear that there is a trade-off between direct costs and error costs. However, one cannot fully characterize the costs of error without considering the underlying substantive legal rules that are sought to be enforced.

  Thus, the economic analysis of procedural law does not draw such a sharp distinction between substance and procedure as do legal dogmatics. In the economic analysis of law, it is the combination of both substantive and procedural rules that determine the ultimate efficiency properties of the legal system. To illustrate, let us take the example from the law-and-economics literature of the distinction between a “property”-type rule and a “liability”-type rule. In essence, this distinction is one of remedy: a “property”-type rule is one enforced by specific order, while a “liability”-type rule is one enforced by money damages as ascertained by a tribunal. The relative efficiency properties of these two types of rules may depend upon procedural characteristics, such as whether economically compensable money damages may be determined by a tribunal within a tolerable range of direct and error costs. Note that such a decision may depend in part upon the properties of the procedural system as encouraging or discouraging strategic behaviors by the litigants, either in the course of the legal dispute or in the period before the legal dispute arose. If the procedural system is such as to discourage one or both parties from generating the necessary information ex ante, or revealing the specified datum ex post, then the characteristics of the procedural system may narrow the range of efficient substantive rules.

  For this reason, it is not necessarily true that a publicly provided procedural system can be directed toward minimizing the costs of legal disputes, either to the parties or to the society as a whole. Reducing the costs of legal disputes too low will encourage an oversupply of disputes, while raising the costs of legal disputes too high will discourage legal interactions. Social welfare can be reduced by making legal disputes either too easy or too difficult. These problems are magnified when considering forms of legal dispute-resolution that deviate from the inter pares model of private civil litigation, such as criminal prosecution or other forms of public law enforcement.

• (e) A more general consequence of this point is that procedural rules may be both complements of substantive law, as in conventional legal dogmatics, and substitutes for substantive rules. Thus, in a contractual situation, parties may be unable to agree on certain substantive terms (such as the price of an apartment to be purchased, or damages in the case of a contract breach), but may be able to agree in advance on a procedure that both would accept. This is the type of arrangement that is characteristic of private dispute-resolution arrangements, such as arbitration or, even more broadly, of negotiation-techniques. Whether parties will agree in advance to such arrangements depends in part on the properties of the procedural system that otherwise would apply.

• (f) The remainder of this paper presents, first a general analysis of the incentives for privately and publicly provided law enforcement (Part 2) and then an economic analysis of civil and criminal procedure along the commonly employed legal desiderata of “just, speedy, and inexpensive” adjudication (Part 3).

2 Law Enforcement and the “Expected Value” Model

In this section, we present a general economic view of law enforcement, beginning with case of private enforcement and passing to public enforcement, including criminal law enforcement.

The micro-economic approach is individualistic. All social interaction is factored down to the choices of individual actors. Then, the incentives operating on those individual actors are examined to predict their actions. In its most general form, this analysis focuses on the “opportunity costs” of the individual. This means that the individual’s incentives are compared with the next best use of the individual’s time and effort. To the extent that the consequences of an individual’s actions are felt by that individual, then the resulting costs are said to be “internalized” to that individual. To the extent that such consequences are not felt by that individual, then the resulting costs are said to be “externalized.” As the reader is aware from other chapters, much of the law-and-economics literature is concerned with the significance of this distinction between “internalized” and “externalized” effects. The Coase theorem challenged the significance of the distinction per se, instead arguing that the efficiency of legal entitlements was more profoundly affected by the incidence of transaction costs on individuals’ ability to reach socially optimal arrangements by private bargaining. Whereas the Coasian analysis suppressed the problem of costly enforcement of legal rights, this enforcement can be seen simply as another form of “transaction cost” standing in the way of optimal arrangements. Therefore, we can apply this same perspective to the analysis of law enforcement.

2.1 Private Enforcement

• (a) From the economic point of view as interpreted by Coase, substantive rules of law assign private rights and obligations in order to induce optimal arrangements of resource use. Thus, as applied to ordinary civil litigation between private parties, substantive law assigns the starting allocations of private property rights. However, given that transaction costs are strictly positive in any actual case, the initial assignment of legal rights may influence social welfare. Furthermore, both enforcement costs and the form of remedy may influence the efficiency properties of the substantive law. If enforcement is too costly, then the initial assignment of property rights will have no effect, and would be equivalent to the non-assignment of rights, which is likely to be inefficient.

• (b) Private enforcement of law seeks to align the enforcement incentives with the underlying substantive rule. Thus, the typical case is that the owner of the legal entitlement is given a right of action to enforce the entitlement against invasion by another (for example, the owner of a lot sues his neighbor to stop the strolling of his dog). To the extent that the owner-claimant “internalizes” the enforcement incentive by obtaining a private remedy, enforcement is obtained in proportion to its social value, as assessed by the claimant (in the case of property-type rules) or as assessed by the court (in the case of liability-type rules). In these cases, the individual will provide law enforcement, if this enforcement is cost justified on the basis of her own costs and benefits, which are equivalent to social costs and benefits (= full convergence of private and social incentives to bring suit).

• (c) However, there are several qualifications to this simplistic analysis, most tending toward the argument that private enforcement of law produces under-enforcement.

First, private and social incentives for enforcement may diverge. If individually provided law enforcement generates positive external effects to certain other individuals, to groups of individuals, or to the general public, these social benefits will not enter into the individual’s cost calculus. Such an “external” benefit to society may stem from enforcement that either clarifies a legal rule or entitlement or deters third parties from committing similar invasions. In these cases, the individual is likely to systematically disregard those social benefits (that do not occur to herself) and to underprovide law enforcement. This argument often is said to justify a degree of public subsidy to private enforcement by using general tax revenues to cover most of the costs of courts. Note, however that, whereas one can characterize analytically, with relative ease, litigation cases with “external” benefits to society, it is by far more difficult to identify concrete “real-life” applications. For example, “deterrence of third parties” through law enforcement (embodying a positive external effect to the public) can only be achieved if the potential wrongdoers exhibit a sufficient degree of responsiveness, and this is an empirical question.

There may also exists an inverse constellation, in which the individual does not experience the full costs of her litigation, such that social costs are larger than social benefits. These incentives will result in “excessive litigation.” Purely distributional conflicts, such as hereditary litigation, may serve as an example.

A special case of “external” benefits from individual litigation may arise if the defendant’s action has not only affected one single victim, but also various other individuals (as is the case with many environmental damages), so that each individual victim’s choice to enforce the law would generate benefits also for her peers. It is because of this “public good” aspect that the single individual may not be willing to provide law enforcements “for everybody” on her own costs.

Second, it is argued that, even with limited public subsidy, enforcement is costly, such that remedies measured only by the value of the property right are insufficient to induce adequate enforcement. For example, if one has a property right worth 10 €, but it costs 5 € to obtain legal vindication, there will be under-enforcement. In these cases, law enforcement will not be cost justified on the margin. Note that under this argument (meritorious) small claims are likely to remain unenforced irrespective of a possible divergence of private and social incentives for litigation. Even if the owner-claimant fully “internalizes” the enforcement incentive, he will rational refrain from law enforcement if this enforcement “eats up” the entire claim; this analysis is only exacerbated in case of external social benefits of litigation. This argument often is used to justifying various forms of fee-shifting rules (whereby the losing party has to indemnify the prevailing party) in the procedural system (examined in Sect. 3.1).

Third, shortfalls in private law enforcement of whatever source entail consequences on the behavior of potential trespassers to be deterred by litigation, as potential defendants would weigh the amount of compensation due with the probability that they are successfully sued. Therefore, it is argued that, even with limited public subsidy and fee-shifting, imperfect enforcement may justify damages that are above compensatory levels in order to produce optimal deterrence of wrongful acts. For example, if enforcement is shown to be taken in only 50 % of cases, this is said to justify awarding damages at double the compensatory levels, in order to present the potential wrongdoer with an “expected liability” equal to the extent of the damages. Of course, if such “punitive damages” are raised above expected harm, this feature produces too much enforcement, which is equally as undesirable as too little enforcement.

A combination of all three arguments is sometimes used to justify the “class action” device used in the United States to aggregate large numbers of similar claims into a single adjudication. In practice, the procedure has a somewhat coercive aspect, in that absent claimants generally are required to “opt-out” of the class, which is costly to them. This is argued to be justified as eliminating a “free-rider effect” that otherwise would produce under-enforcement as the individual claimants waited for each other to commence the case, and therefore reduce enforcement costs to the follow-on claimants. The class action devices also introduces agency costs by allowing the few named claimants’ lawyers to conduct the case on the part of the entire class, and to be paid first from the class recovery, often leaving little or no net recovery to the nonparty class members, and thus producing something akin to a “bounty” effect.

2.2 Public Enforcement

• (a) Both the qualifications noted above and the counterarguments that they engender also surround the economic analysis of public enforcement, including enforcement by administrative action, by criminal prosecution, or by public civil action (whereby public agencies enforce the law “as a private party” under civil procedure which is more developed in Anglo-American than in continental European countries).

• (b) Like class actions, public enforcement is said to be justified by economies of scale in enforcement costs (i.e., an additional unit of law enforcement can be more cheaply provided by an increase in output of one agency than by a separate supplier) and by the external social benefit of enforcement. Two other arguments often encountered in support of public enforcement are (i) limitations on the efficacy of private civil remedies, due to insolvency of the defendant or the like, and (ii) the absence or ambiguity of assignment of private property rights to support private enforcement, as is in certain forms of environmental cases involving pollution to navigable waters or public lands.²

• (c) However, somewhat like class actions, public enforcement has been criticized as introducing new problems that could produce either too much or too little enforcement. In these instances, the use of public agents (usually state officials) for purposes of law enforcement in the interest of the public (embodying the principal) introduces problems of agency cost and public choice. “Agency cost” refers to the problem that public agents’ incentives may deviate from social welfare-maximizing incentives, and “public choice” influences can exacerbate such problems by selectively favoring certain types or times of law enforcement. Unlike private enforcement, public enforcement generally does not internalize either the benefit or the cost of enforcement to the enforcer. Public enforcement agents may be constrained by budgets to under-enforce, or by political or careerist pressures to over-enforce. These enforcement choices are, almost by necessity, surrounded with a certain side taste of arbitrariness (say, in case the police tickets only a few, but not all drivers for speeding or for parking violations). Furthermore, public enforcement priorities may reflect public-choice influences toward transferring the cost of enforcement from those obtaining concentrated private benefits of enforcement to the tax-paying population in general, and thus producing another form of wealth-transfer legislation (for example, lobbying by a certain city district for preferential police protection).

• (d) All of these features of public enforcement also enter the analysis of criminal law enforcement by imposing punishment. If the crime already has taken place, then it can be argued that the crime victim’s incentives may produce either under-enforcement or over-enforcement. Many people, at first blush, would expect over-enforcement of crimes, if punishment is left to the individual victim on the basis of an assumed vengefulness on her side. Some such motivation is likely to be present in the context of individually provided enforcement of the criminal law and may explain a certain amount of individual contributions to the enforcement of criminal law. However, this analysis disregard that punishment is costly (involving both various categories of “out of pocket costs” and intrinsic costs in terms of the “disutility” experienced when actually meting out punishment against the wrongdoer) and that the victim could save these costs if she leaves the wrongdoer unchallenged. In particular, if punishment does not produce a transfer for payment to the victim (perhaps because the criminal has no transferable assets or is subject to punishment by incarceration) it may be rational, at least, judged in a mere response setting, not to pursue the wrongdoer.³ Whereas the above-mentioned vengefulness may provide some incentives for law enforcement also in cases where the punishment, considered in itself, is not cost justified, the dominance of either incentive is an empirical question and not analytically predetermined ex ante. Moreover, even if the victim may explicitly wish the wrongdoer to be punished, she may prefer, due to the costs of punishment that someone else carry out the punishment so that she would not have to enforce the law herself. The relevance of the costs of punishment for actual choices can also explain why people on the street, when asked about their attitudes on punishing in a particular case, tend to be more vengeful (because of the lack of punishment costs) than as jurors on the bench (where they bear the intrinsic disutility of punishment).

For these several reasons, most procedural systems place criminal law enforcement in the hands of public agents. However, those agents still present the same problems of agency cost as in the more general case of public enforcement. Criminal law enforcement may be misused by the prosecuting agencies for political reasons against innocent individuals, as was the case frequently in earlier times, and still is, at some occasions, today. The current discussion in many continental countries is concerned with the inverse constellation, namely with possible shortfalls in the prosecution of criminal wrongs (which may result both from political bias to protect “friends,” or simply from shirking). Among those institutional instruments currently considered to cope with insufficient enforcement by the public agency is the initiation of criminal prosecutions by judicial order (“Rechtserzwingungsklage” = “Action to enforce criminal prosecution”). More generally, if the enforcement agency is politically accountable (say, if prosecuting agents are directly or indirectly elected, as is the case in some parts of the U.S.), the election pattern is likely to influence enforcement activities (namely with the goal of shifting more resources toward the period immediately preceding the elections) and, ultimately, crime rates—the result being a politically influenced cycling of the crime rate.

2.3 Settlement Under the “Expected Value” Model

The most highly developed feature of the economic analysis of procedure is the trial-versus-settlement decision. The economic models take into account that litigation is a strategically interactive process, whereby each side reacts to the moves of the other. The trial-versus-settlement model attempts to identify the conditions under which some cases settle before decision, whereas others go to final decision.

2.3.1 Settlement in Private Civil Litigation

• (a) The basic model of trial-versus-settlement postulates two opposing views of expected value (for the claimant) and expected loss (for the defendant). In each case, the “expected” result is a compound of each side’s assessment of the value of the claim times the probability of a plaintiff victory at trial, as adjusted by each side’s costs (which are analytically the costs of proceeding to the next stage of decision). In this basic model, the prospect of settlement is influenced by whether there is a “bargaining range” of overlap between the parties’ estimates. Specifically, if the defendant’s expected loss is greater than the plaintiff’s expected gain, then the case will settle between those points of the bargaining range, unless prevented from doing so by either strategic behaviors on the part of the parties or external influences. On the other hand, in the absence of a bargaining range, then the parties can not settle without taking additional moves to change each other’s estimates.⁴

• (b) Much of the literature in the economic analysis of procedure is concerned with characterizing the conditions that prevent settlement. Three (mutually complementary) basic models have been proposed: (i) “prediction failure,” (ii) “bargaining failure,” and (iii) the “external effects” model. Whereas under the first two models there exists a bargaining range, but the parties fail to settle because of mutually inconsistent, relatively optimistic estimates as to the outcome of the trial or because of mere distributional quarrels, the third model argues that it is the existence of asymmetric external effects, and thus, asymmetric stakes that prevents the existence of a bargaining range and forecloses settlement.

“Prediction failure” refers to the case where the parties’ estimates of the likely outcome do not converge sufficiently to produce a bargaining range. An important variation of this type is the so-called “mutual optimism” model, where each side believes that it is more likely to prevail at trial. The possibility of such a failure is one justification for the American system of “pretrial discovery” of evidence from the opposing litigant or third parties, which may help to dissuade one or both sides from their mutual optimism, but of course such a procedure increases litigation costs and opens up the possibility for strategic behavior within the discovery process, either by imposing asymmetrical discovery costs on one or another party or through selective disclosure of information.

“Bargaining failure” refers to the case where the parties’ estimates converge or overlap but the parties nonetheless fail to settle because of strategic behaviors. To take the simplest case, suppose the parties actually agree with each other on the expected value and loss of the case. However, both parties still face incremental costs of proceeding to trial, and in this instance the true “bargain” is over how to allocate the mutual benefit of avoiding trial, which may or may not be symmetrical as between the parties. If one or another party bargains too hard over the division of that cost savings, or if one party is more or less averse to risk than the other, then there can be a bargaining failure. Or, the parties may behave strategically by failing to disclose their true estimate of the probable trial outcome, or their incremental costs of proceeding to trial. All of these can produce “bargaining failure.”

A third “external effects” model postulates that important classes of cases may have “external” effects (meaning here effects materializing outside of the concrete litigation on the immediate parties to the dispute) that produce asymmetrical stakes and thereby eliminate a bargaining range even when the parties’ estimates converge. One example where this can happen is the “repeat play” litigant opposing a “single play” litigant. For any of several reasons, such as the precedential or preclusive effect of an adverse or favorable judgment, or to establish a “tough” reputation to deter future litigation costs, the “repeat play” party may have more at stake than the “single play” litigant. Depending upon the relative magnitudes of the internal stakes versus external effects, such a situation can eliminate the bargaining range. One example might be where a bank litigates the validity of some provision in its standard loan agreement against one of its customers: the customer’s stake is limited to the case at hand, whereas the bank may have an entire line of business involving thousands of customers at stake.⁵

2.3.2 Settlement in Public and Criminal Litigation

• (a) Public litigation in general appears to be inherently more difficult to settle because most public litigation institutionalizes the asymmetry of stakes noted in the previous section through the intervention of the public enforcement agent. In the usual case of a public agency versus a private party, the public agent does not fully internalize either the benefit of a settlement or the cost of proceeding to trial, and therefore this type of litigation would appear to be influenced on the public side by an external effects model, with those external effects given by diffuse political or bureaucratic incentives.

• (b) There has been extensive study of the U.S. criminal “plea bargaining” process as a special case of settlement of public litigation, though much of this literature suppresses the agency cost problems associated with the settlement of public litigation. In the basic model, the agency cost problem is addressed in part by assuming that criminal prosecutions are limited by finite prosecutorial budgets, and so the problem is cast as rationing these limited resources to a highly elastic supply of criminal offenders. In most basic models, the prosecutor is assumed to be maximizing overall deterrent effect by offering plea bargains as a sorting device to distinguish innocent from guilty defendants (or more guilty versus less guilty defendants). The guilty (or more guilty) defendants accept the plea bargain, which provides a discount from the expected punishment at trial, but this result is rationalized as allowing a larger number of guilty defendants to be convicted at a given fixed cost. The difficulty with this analysis is that, in order to make the strategy credible, the prosecutor actually would have to go to trial against innocent (or less guilty) defendants, and, if a conviction were obtained, would have to seek a high penalty, for otherwise the risk-preferring guilty would masquerade as the innocent. Both of those implications would appear to violate prosecutorial ethics (even in the U.S.) and justice considerations. Ultimately, the argument for such an analysis would seem to rest importantly on the assumptions that the trial process rarely or never produces either false positives (erroneous convictions) and that false negatives (erroneous acquittals) are relatively uncommon. Either or both of these assumptions appear to be debatable empirically. Because of concerns about the asymmetric costs of error, criminal procedure systems tend to arranged so as to tolerate a relatively high level of false negatives in order to minimize false positives.

Another view of criminal plea bargaining may place it closer to the standard model of settlement in private civil litigation, under the assumption that prosecuting authorities internalize a cost (perhaps to professional reputation) of taking the innocent (or less guilty) to trial or failing to convict the guilty (or more guilty), and therefore choose to plea-bargain not for sorting or signaling purposes, but simply on the basis of predictions of trial outcome, as constrained by limited prosecutorial resources.

3 Economic Analysis of Civil and Criminal Procedural Rules

• (a) Virtually all civil and criminal procedural codes announce their objectives as securing the “just, speedy, and inexpensive” determination of legal disputes. These goals are interrelated. “Speedy adjudication” may also be inexpensive. And both speedy and inexpensive adjudication are regularly seen as aspects of “justice” in procedure, at least in the legal sense. However, an application of economic analysis demonstrates that these desiderata not only conflict with each other in some cases, but also are far more subtle than may appear at first blush. If “just” adjudication is seen in terms of reducing error costs, this reduction in error costs involves a tradeoff against reduction in direct costs of adjudication, which are more closely associated with the desiderata of “inexpensive” and “speedy” procedure.

• (b) Closer observation shows that, as noted in the Introduction of this chapter, procedural systems and their constituent rules operate in conjunction with underlying substantive standards, and are characterized by interdependent strategic behavior by the parties and, to some extent, by the tribunals themselves. The basic goal of the procedural system is to make the substantive law efficiently enforceable, which does not imply the minimization of direct litigation costs only, nor even the sum of direct litigation costs plus “error” costs solely in the sense of erroneous factual or legal determinations in adjudicated cases. In addition, the procedural system “feeds back” on substantive law and influences both the supply of cases to the system and the supply of information available to resolve the cases that appear. Such a system would not be efficient, no matter how inexpensive and accurate it became, if, for example, it encouraged the bringing of cases that could have been more efficiently solved (or prevented entirely) by ex ante bargaining between the parties. Thus, a more inclusive statement of the optimization problem would be to minimize the total sum of transaction costs and other opportunity costs imposed by the legal system overall, both substantive and procedural taken together. At that level of generality, there is no completely satisfactory synthesis in the existing literature, because the problem is enormously complex. We can only look at pieces of the overall picture. In this section, we will examine some of the problems from the law-and-economics perspective, organized around the competing desiderata.

3.1 “Inexpensive” Adjudication

• (a) As noted above, the most “inexpensive” adjudication is the one that never arises in the first instance. However, this does not imply an extreme solution. It may be more expensive for contracting parties, for example, to anticipate each and every potential dispute that may arise, most of which will never arise. But the parties’ incentive to anticipate ex ante is given in part by the cost of ex post litigation. If ex post litigation were very cheap, then the parties would have little incentive to anticipate, and there would be a great of litigation, much of it concerned with trivial matters that are unworthy of anyone’s attention, including the immediate parties. If it were too expensive, then parties may not contract at all, and may not engage in any number of other social interactions. Obviously, the cost of litigation is not the only influence: clearer substantive law also tends to reduce both the supply of disputes and the cost of their resolution, perhaps more dramatically than any procedural rule.

• (b) To the extent that the costs and benefits of litigation were entirely internalized to the immediate parties, and distributed in such a way as to minimize strategic behaviors,⁶ then social policy might leave the entire problem to the parties’ private agreement and need not bother about the costliness of the procedure. This would be the purely “Coasian” solution. In the absence of some external effect on a third party or a defective bargaining process, there seems to be little basis for compelling private disputes to be litigated in public courts. This implies a policy that is open to contractual substitutes for litigation, such as arbitration, mediation, or the like.⁷ In fact, such a policy is followed by most procedural systems.⁸

  It would further appear that agreements between the parties, either before or after the litigation, generally should be respected by the public court system, no matter how “expensive” they appear, unless they also impose undue public expense. In other words, parties should be permitted to contract for their own procedure, within limits. This stands in contrast to the case of one contesting party’s unilateral request for expensive procedures, opposed by the other party. This instance may present a case of strategic or opportunistic behavior, but not necessary by the party who requests the “expensive” procedure. By analogy to the economic analysis of tort or contract law, the appropriate approach would seem to be a reconstruction of what the parties would have agreed to before the dispute arose, which may be somewhat easier to determine in a procedural than in a substantive context (see Sect. 3.3.2). Failing that approach, proportionality to the size of the stakes in the dispute is a rough guide to identifying excessive expenditures.

• (c) Setting aside the caveat noted above regarding “inexpensiveness” as a value in itself, one can discuss the legal desideratum of “inexpensive” litigation along different lines.

  In a first, quite immediate sense, this desideratum suggests the provision of a “lean” procedure, which does not only economize on those elements and costs “unnecessary” for a specific type of litigation, but also provides for the creation of different categories of courts (from “small claims” courts to “Supreme” or “Superior” courts), with procedures that vary in complexity and cost with the amount in controversy. Here again, some measure may be found in comparing litigation costs with the disputed stakes. For the economist such types of courts embody a distinct trade-off between direct and indirect costs of litigation. The above differentiation of various categories of courts is also of assistance in allowing (meritorious) low value claims to be litigated.

• (d) In the existing literature, the “inexpensiveness” of litigation is often interpreted as a question of equal access to the courts, and there is concern about raising the minimum cost too high, out of fear of pricing the non-affluent out of the courts. It is possible to develop an economic argument for some such institutional device in the light of the otherwise pending erosion of property rights. If only the affluent could litigate their claims, substantive property rights of the non-affluent would systematically be undermined. Still, it is a difficult institutional question how to ease this type of litigation. Different jurisdictions take different approaches to the problem. Under most European systems access to the legal system is granted by a complex system of either direct public provision of legal services to lower income individuals (at least in certain legally more demanding or high-valued cases) or by pro bono activities on the side of the legal profession. Regularly, these systems also rely on fee-shifting rules. In the United States, public legal aid is much more limited and legal pro bono activities largely are left to the selection of individual members of the legal profession. In the U.S., the contingent fee system⁹ (mostly encountered in torts cases) seems to function as a financing vehicle that sorts cases for threshold merit and spreads risk by inducing plaintiffs’ lawyers to build a diversified portfolio of cases. However, some jurisdictions are hostile to the contingent fee, including most of Europe. Most jurisdictions permit similar arrangements on the defense side by liability insurance, which is expressly designed to spread litigation risk.

• (e) In a further important sense the criterion of “inexpensive” litigation requires that the costs of the legal system incurred for the enforcement of a legal claim should not be prohibitive. One instrument to accomplish this goal is the implementation of a “fee-shifting-rule.” Under such a rule, the winning litigant is entitled to recover her attorney’s fees and out of pocket costs (including court fees). Under the “American rule,”¹⁰ in turn, there is no such indemnification so that each party bears her own costs.¹¹

The consequences of fee-shifting are to some extent ambiguous.

First, fee shifting cannot be seen independently of the substantive law. Its effect on case supply seems to be toward confirming the legal status quo, which itself is ambiguous. One could argue that fee-shifting has favorable efficiency properties for the enforcement of established law, to the extent that it compensates for the cost of enforcement that may otherwise reduce incentives for private enforcement below optimal levels. However, this assessment depends in part upon the correspondence between legal damages and economic damages: if the legal damages formula under-compensates, then fee-shifting may ameliorate but does not correct the problem; on the other hand, if the legal damage formula over-compensates, then fee-shifting makes the over-enforcement problem worse. It also depends upon how closely established substantive law corresponds with the economically efficient rule, as it tends to make established law more resistant to change.

Such ambiguities also exist with respect to the procedural consequences of fee shifting in a narrower sense. The “loser pays” fee-shifting system may tend to discourage settlement by raising stakes, and tends to increases the returns to more certain claims relative to other claims, especially for more risk-averse (usually meaning lower income) claimants, who otherwise are selectively more discouraged by fee-shifting. In addition, fee-shifting system have some tendency to encourage what economists call the “moral hazard”(also characteristic of some insurance systems), which means that individuals are likely to decide to spend more if they believe that someone else ultimately will have to pay. To the extent that one side’s litigation expenditures also induce the opposing party to raise its expenditures, this effect could be magnified by use of litigation expenditures as a strategic weapon. For these reasons, most fee-shifting systems embody a legal limitation to “reasonable” expenditures (regularly based on lawyers’ tariffs). Although the limitations provided in most European systems work smoothly, one must admit that the criterion of “reasonable” expenditures has sparked many legal disputes on cost issues, raising the direct costs of administering the system.

Still, there are more straightforward aspects of fee shifting. Due to the obligation to indemnify the prevailing party, fee-shifting discourages potential litigants from filing non-meritorious claims (meaning claims with low probability of winning). In turn, fee-shifting encourages meritorious claims, because it grants the plaintiff the prospect of being fully compensated for his costs (to the extent that the officially recognized tariffs represent the true market value of the services supplied) and to receive the reward. Therefore, economic analysis indicates that the European alternatives to the American contingent fee plus the “American rule” (under which parties bear their own costs) may selectively favor more certain claims over more speculative claims. This distinction may be justified by the differing approaches of the two systems toward case decisions as primary sources of law: in the U.S., most tort law is “common law”, i.e., based entirely on case decisions rather than statutes; in such a system, it may be necessary to encourage a wider diversity of cases in the first instance, in order to provide more raw material for the formulation of legal doctrines. One drawback to the European systems could be the higher direct costs of administration: public assistance cases involve public bureaucratic and political costs of budgeting, screening, and organization; fee-shifting systems involved more direct cost of judicial administration. The American contingent fee is a private market system that involves little or no direct public administration or supervision.

3.2 “Speedy” Adjudication

• (a) The desideratum that procedure should be “expedient” is close to universal recognition in legal doctrine both in the realm of civil and criminal law. Most legal systems even have a saying that “justice delayed is justice denied.” Still, on a more abstract level, the criterion of “speedy adjudication” is more ambiguous. Litigation can be “inexpedient” because one side deliberately delays the procedure or because both sides prefer a slower pace of their litigation. At least under European continental procedural systems, the inexpedience of a trial can also be the result of mismanagement of the case on the side of the judge, be it that he fails to implement an effective schedule of the proceedings (or is himself an obstacle because of bad preparation or illness), or fails to enforce the procedural instruments available to cope with unilateral attempts from one side for undue delay.

• (b) Much of what has been said about “expense” also could be applied to the supposed objective of expeditious litigation. Here again, in private civil litigation, the main problem for procedural rules would be to distinguish between purely strategic delaying tactics and legitimate (or mutually agreed) development of a case. There is an obvious conflict between a “speedy” disposition and a “just” one. But to some extent, there also is a direct conflict between a “speedy” adjudication and an “inexpensive” one, if the desire for “speed” in itself forces the litigants to incur unnecessary litigation expense that could be avoided (for example, in cases where both sides may prefer a slower development, because other events may avoid the need for a definitive adjudication). Unless such mutually agreed delays affect third parties, as by depriving other litigants of access to public judicial resources, there does not seem to be any public interest in pushing private litigation along.¹² Moreover, while it seems perfectly acceptable that procedural systems should be arranged so as to minimize the use of dilatory tactics as a form of predation by one party against another, it is equal possible that expeditious tactics also can be used as a weapon of predation. Therefore, institutional rules also must cope with this (inverse) constellation.

• (c) Criminal cases may appear to differ in this respect, given the emphasis on “speedy trial” guarantees in criminal procedure. However, in most such instances, this is a right in the accused against dilatory tactics by the prosecuting government. One might still argue, although somewhat attenuated from the economic perspective that there is some independent external interest of the general public in expeditious disposition of criminal charges, even in the absence of the defendant’s objection or perhaps because of concerns that the defendant’s consent may be extorted by the government in ways that cannot be directly observed.

• (d) Currently, we observe increasing calls for more expeditious dispositions, both in the European system and in the United States. When assessing the quality of this reform, one has to be aware of the heterogeneous nature of “inexpedience” in proceedings outlined above. Insofar as this reform is directed against judicial ineffectiveness, the preservation of judicial independence limits the deployment of direct performance-dependent (“carrot and stick”) incentives with respect to judicial behavior.¹³ One possible explanation for the recent reform may lie in the incentives of the judges and judicial bureaucracies themselves, as case dispositions and intervals to disposition are one of the few quantifiable “outputs” of the judicial system.¹⁴

3.3 “Just” Adjudication

• (a) It is commonly recited that procedural codes should ensure a “just” outcome. While the general concept of what constitutes justice can be controversial in substantive terms, it is the case that aspects of justice in procedure command more common agreement. From the economic perspective, a “just” outcome is most closely connected with the idea of reducing error costs.

• (b) Procedural justice generally has three aspects: (i) correct application of substantive law; (ii) correct determination of the objective facts (the “material truth”) of the case; and (iii) procedural fairness in the formulation and application of the procedural rules themselves. Our discussion below focuses on the second and third aspects of procedural justice.

3.3.1 Justice by “Material Truth”

• (a) The concept of “material truth” is that there is an objective truth of the factual grounds for decision in a given case. It should be noted that this concept does not play an equally central role in all systems of procedure. The Anglo-American procedure is more concerned with determining the facts as they are submitted to the tribunal by one or both parties, especially in private civil cases. Within that system, it is permissible for the parties to stipulate to a set of facts, which are submitted for judicial decision. Nevertheless, the facts submitted to the tribunal generally are claimed to correspond with actual events, and the explicit decision of purely hypothetical cases is prohibited.¹⁵

• (b) If we focus only on the fact-finding aspects of adjudication, there is some debate within the law-and-economics literature over the extent of the external social interest in minimizing factual error costs, particularly in private civil disputes. Most of the cost of such error falls on the immediate parties. Provided that the level of inaccuracy in adjudicative fact-finding is not so extreme as to virtually force parties to shift away from adjudication to substitute forms of dispute-resolution (such as vendettas or feuds, organized crime, and so on) with negative social consequences, there may be a broad range of tolerable accuracy levels. Presumably, to the extent that parties are permitted to contract away from public courts to peaceful alternatives such as arbitration, the accuracy levels of public adjudication are disciplined by competitive forces.¹⁶ However, in areas of public monopoly in procedure, such as criminal cases, it may be more important for the procedural system to embody internal controls on factual accuracy, and this is consistent with some of the features that distinguish criminal from civil procedure, as in the more extensive procedural protections for the accused designed to minimize false positive errors (erroneous convictions).

• (c) Moreover, in all systems, the “search for truth” is limited by competing criteria of cost and efficiency. Most systems focus factual inquiries, whether by judges or parties, on those facts identified as pertinent or relevant by the applicable substantive law. Thus, parties may wish to show, outside of the narrow “pertinent facts of the case” that they are particularly worthy or virtuous individuals, but generally speaking, such considerations are excluded by substantive law and, therefore, not the object of proof in procedure. In “common law” systems where case decisions are conceived as one source of law production, the criterion of relevancy may be more loosely applied, so as to permit the evolution of legal rules through case law. This feature may be one of those contributing to a higher direct cost of adjudication in such systems, but it is not clear whether these costs compare favorably or unfavorably to the costs of fact-gathering for the purpose of generating legal rules by legislation.

• (d) Further, in all systems, the “search for truth” is not self-executing: that objective must be implemented through the design of mechanisms within the procedural rules themselves for generating factual information and resolving disputes. In particular, the procedural code influences the incentives for all persons involved (including the tribunal and witnesses, as well as the parties) to contribute toward that end. Obviously, the parties are self-interested and opposed, and can be expected to engage in strategic behaviors and selective disclosure of evidence. However, the same may be true of nonparty witnesses, who may have interests pertaining to the case or to their testimony. Nor are individuals associated with the tribunal itself immune to incentives: both professional judges and lay judges (jurors) may have incentives to deviate from “truth,” depending upon the institutional structure surrounding their work. For example, if judges were evaluated and promoted solely on the basis of their “throughput” of cases, then judges’ incentives would be more strongly aligned toward solving the case quickly rather than accurately. On the other hand, it has been argued that judges with little or no prospect of promotion, secure tenure, and minimal case-processing requirements may have an incentive to work as little as possible, and when they work, to spend more time on “interesting” cases, which also does not seem favorable to achieving high levels of accuracy.

• (e) Procedural systems differ markedly in their assignment of roles and incentives to the various actors in the fact-finding process. In particular, there is a contrast between more “adversarial” procedures versus more “inquisitorial” procedures. While most actual procedural systems embody some mix of these two types, traditionally the Anglo-American system emphasized “adversarial” elements while the European continental system emphasized “inquisitorial” elements. The adversarial system relies primarily upon the competitive interests of the opposed parties to bring out factual information, while the inquisitorial system places heavier reliance on the neutrality, professional training, and questioning rights of the adjudicating judge to ensure a full revelation of “material truth.”

• (f) The choice between these two types of procedural approaches appears to involve a complex set of tradeoffs between direct and error costs of adjudication, and there is very little empirical knowledge on the subject. Theoretical literature suggests that inquisitorial procedures may reduce direct costs while maintaining a tolerable but not maximal degree of accuracy. On the other hand, adversarial systems, though probably more expensive in terms of direct costs, may achieve higher factual accuracy under certain conditions, though it is not clear that such a tradeoff would be economically efficient. One difference to be observed between the systems is the difference in attention given to fact-finding in the original instance versus appellate review: inquisitorial systems involve more extensive appellate review of fact-finding, whereas adversarial systems tend to focus appellate review on questions of law only.

• (g) The available empirical evidence is very limited, but is consistent with the view that, leaving aside the cost aspects of litigation, the relative efficiency of adversarial versus inquisitorial procedures varies with surrounding conditions. This evidence consists of experimental research in which “material truth” could be supplied artificially.¹⁷ The method of investigation involved simple case scenarios with hidden facts known only to one of two opposing parties, with the roles of both parties and “referee” played by experimental subjects (mostly University students) according to stylized rules of purely “adversarial” (completely passive referee; questioning by parties only) versus “inquisitorial” (active questioning referee, but no questioning rights in parties) procedure. Referees were then required to award all or part of a contested stake to one or both parties, in accordance with a simple given rule of law. Under the full revelation of “material truth,” the accurate decision was to award the entire stake to one party, and nothing to the other, under the given rule of law. While the experimental subjects concentrated on their private incentives (parties to “win” the monetary stake; referees to render the “accurate” decision, which maximized their payoff), the experimenters also observed the incidence in which revelation of the important hidden fact occurred under the respective questioning systems. The results were that relative revelation rates depended upon the degree of information asymmetry with which the parties began: where the hidden fact was exclusively on one side, without any hint to the other, then inquisitorial procedure achieved a higher rate of revelation; however, when the less-informed party started with a slight “clue” to pursue, then adversarial procedure achieved a higher rate of revelation, both relatively and absolutely. One institutional interpretation of these results is that adversarial procedures perform much better when preceded by the opportunity for the type of pretrial “discovery” that is characteristic of American civil procedure. However, this aspect of American procedure is widely believed to raise the direct costs of adjudication, and may also tend to suppress the ex ante production of information prior to the dispute. Therefore, once again the efficiency properties are ambiguous.¹⁸

With the respect to evidence law, it is characteristic for many European countries that the respective legal rules are not uniform, but enshrined in the separate procedural acts (say, in an act on civil procedure, on criminal procedure, and on administrative procedure). In contrast, the Anglo-American system includes a highly developed body of evidence law doctrine that regulates the forms of evidence that may be presented by the adversarial parties. This body of law applies both to civil and criminal procedure, though it contains some special rules applicable to each type. Aside from the standard of relevancy, this body of law focuses extensively on the acceptable forms of evidence, preferring extemporaneous live testimony based upon witnesses’ first-hand sense impressions to any other form, such as written testimony or deposition, documentary evidence, or opinion testimony by experts. To some extent, the development of this body of procedural law may have been influenced by the extensive use of lay juries as the fact-finding institution in Anglo-American procedure, which necessitates a compact “trial” stage in order to minimize the burden of jury service on citizens. However, it also reflects the parties’ primary responsibility for evidence production and presentation in the adversarial system, through the incentives provided to the parties. Thus, the famous “hearsay” rule in Anglo-American law operates not so much as a rule of exclusion of “hearsay” (out-of-court) statements, but rather as an incentive for the party who seeks to benefit from the statement to bear the initiative and costs of actually producing the witness in court, which also permits the opposing party to subject the evidence to adversarial testing by cross-examination.

3.3.2 Justice as Procedural Fairness

• (a) Notwithstanding the differences across procedural systems in the mechanisms used to achieve fact-finding, there is a remarkable degree of commonality across systems in the content of “procedural fairness,” such that most systems embody the idea of symmetrical access to procedural rights between opposing civil litigants, limitations on favoritism to certain classes of litigants, and recognition of the need to provide expanded rights to the criminal accused. This feature is in some contrast to divergent notions of substantive “fairness” that can be observed in many areas of substantive law. Economic analysis can help to explain why procedural fairness has a wider degree of universality.

• (b) The fairness or equality properties of legal rules are likely to suffer when the distributional consequences of such rules are transparent upon enactment, as in the case, for example, of tax laws. Both interest groups and decision-makers can easily predict the unidirectional wealth-transfer effects of such rules, and this accounts for much modern legislation of this type. However, the more general and multidirectional is the prospective applicability of the rule in question, the more likely it is to be fair. Thus, for example, general rules of contract law are more likely to be fair and neutral than rules of tax law or regulatory law, because virtually anyone could be on either side of a given issue of general contract law.

  There is an analogy here to the idea of “justice as fairness” put forth in the political philosophy of John Rawls, who introduces the concept of the “veil of ignorance.” According to this argument, principles of justice in the arrangement of a society will be promoted by deciding behind a “veil of ignorance” that prevents each individual from knowing their own personal characteristics (e.g., social standing, age, sex, income, health, intelligence, and so on). In that instance, even self-interested choices of social arrangements will be unbiased to personal attributes and therefore “fair.”

  Constitutional economics has extended this concept to characterize a setting of rule-choice that mimics the properties of the “veil of ignorance,” by postulating two conditions: (i) an extended time dimension of the rule (its longevity); and (ii) the generality of the rule’s application. In combination, these conditions impede the decision-maker’s ability to forecast the rule’s distributional consequences for given individuals.

• (c) Now let us apply these ideas to procedural rules. If such rules are long-lived and general in application, as they tend to be, then their formulation is more likely to reflect an idealized concept of justice as fairness. In the case of civil procedural rules, a given individual (including the rule-maker) would be unable to predict whether he or she ultimately would be a plaintiff or defendant, and therefore would be less like to support or promulgate rules systematically favoring one type of party over another. In this context, arguably it is in everyone’s self-interest to agree on procedural rules that are fair in general, though it does not exclude the possibility of special-interest rules in certain predictable classes of special-interest cases.

• (d) We can now compare this principle against observed instances of general procedural rules. One example is provided by Article 6 of the European Convention on Human Rights, which embodies, inter alia, the principles that no one should be the judge of her or his own cause, and that both parties to a dispute should be heard before a final judgment (audiatur et altera pars). We observe similar basic norms in most procedural systems. These are completely self-interested provisions, from the standpoint of someone who does not know the role they may play in relation to the procedural system.

  More peripheral procedural provisions (for example, the right of appeal, or the exact timing of hearings) may be more debatable, and may require the distinction of criminal from civil procedure.

  Many provisions in criminal procedure focus on the goal of minimizing false positive errors (i.e., convicting the non-guilty), which reflects some asymmetry of procedural rights between the prosecution and the defendant. Under an extended time frame and generality, plus the severe and asymmetrical costs of error, these provisions may coincide with self-interest. Many individuals may believe that they are relatively unlikely to be criminal defendants. However, under a highly extended time frame, the cumulative probability rises. Furthermore, given the focus of such provisions on protecting the innocent, a long-term view may evaluate the benefits of such protections as outweighing the costs in general. In this respect, one might contrast such rules as appellate review and a heightened standard of proof with a rule excluding improperly seized evidence, which does not differentially benefit the innocent or contribute transparently to improved accuracy, and may well have the opposite properties. Thus, whether or not such a rule is justified on balance as a matter of social policy (say, as influencing the incentives of police and prosecutors in more positive directions), we should not be surprised to see some diversity of approach across different procedural systems.

  Nor should we be surprised to find some diversity in more detailed rules across different systems of civil procedure. For example, a particular system may devote more resources to fact-finding in the first instance, and thus limit or dispense with appellate review on facts. These types of rules are more likely to be dependent upon local conditions and traditions, and in fact this is what we observe.¹⁹

• (e) The distinction between generally shared notions of substantive versus procedural justice redirects our attention to the substitutability between substantive and procedural law, and helps to explain why many constitutional provisions are procedural in nature. Substantive constitutional provisions may be less common simply because it is easier to forecast their distributional consequences and therefore more difficult to agree on their content. Hence, in constitutions we observe a higher proportion of rules governing the production of substantive rules (for example, representational standards, required pedigrees for legislation, and so on), rather than substantive rules themselves. In this sense also, procedure substitutes for substance.