首页 | 农村土地改革 | 法经济学茶座 | 法经济学论坛
     网站首页≯ 程序法≯
民事程序之一般理论

Abstract

The economic analysis of civil litigation has focused on the action of the

litigants and on the effects of substantive and procedural rules on their

behavior.This chapter focuses on the economic analysis of procedural rules and

how these rules alter the incentives of the litigants to file, settle and litigate

disputes. Such procedural rules affect the private costs and benefits of litigation

through altering the net expected value and loss faced by the plaintiff and

defendant. Procedural rules also affect the social costs and benefits of litigation

byaffecting the both the direct and error costs of litigation. The analysis in the

chapter is organized around the US Federal Rules of Civil Procedure, and in

reverse chronological order to reflect the economic analyses use of backwards

induction to examine civil litigation. Topics examined in depth include

sequencing rules, rules that affect the capitalization of litigation over parties

and claims, the rules of discovery, and juries.

JEL classification: K4, K40 K41

Keywords: Civil Procedure, Civil Litigation, Settlement, Trials

 

1. Introduction

 

The economic analysis of civil litigation, following upon the pioneering work

of Landes (1971), Gould (1973), and Posner (1973), has proven to be a fruitful

area. Economic analyses of litigation have focused both upon the actions of

parties in civil litigation and upon the effects of substantive and procedural

rules on the litigants’ behavior (see Cooter and Rubinfeld, 1989, for a previous

survey of the field). This chapter focuses on the economic analysis of

procedural rules. Papers in this field generally have examined the effects

procedural rules have on error costs and the direct costs of litigation. The

length of the reference list annexed to this chapter indicates the extent of

academic interest in studying the economic effects of procedural rules, and the

consequences of alterations to those rules.

 

2. Organizing Framework

 

The organizing framework for this chapter generally follows the US Federal

Rules of Civil Procedure (FRCP). The FRCP, as supplemented by statutory and

decisional standards concerning the jurisdiction of courts and the preclusive

effects of judgments, regulate most aspects of the civil litigation process, from

thegeneral statement of the purpose of the rules to the particularized standards

regarding process, pleadings, lawyer sanctions, discovery, the allocation of

legal costs, the relationship between multiple parties (including the regulation

and certification of class actions), motions, trials and judgments. Since their

promulgation in 1938, the FRCP have provided generally uniform rules for the

conduct of litigation in the US Federal courts, and similar rules have been

adopted in a majority of American states (for a comparison of federal and state

procedural rules, see Oakley and Coon, 1986). In addition, the Federal Rules

have been frequently revised, and the process of rulemaking itself has become

a topic for economic and legal analysis. However, we do not discuss the rules

and related topics in numerical order. Rather, because the economic analysis

of litigation must be forward looking, and thus proceed using backwards

induction, our chapter is organized in reverse chronological order.

 

3. Topics Covered

 

Topics reviewed in depth in this chapter include sequencing rules, rules that

affect the capitalization of litigation over parties and claims, the rules of

discovery, and juries. Related areas not addressed in detail in this chapter are

treated in other chapters of this volume and earlier volumes. These areas

include: other areas of civil procedure, including fee shifting (7300), the

litigation/settlementdecision (7400) and class actions (7600); the organization

of the courts, including jurisdictional issues (7100-7200); criminal procedure

(7700); the economics of crime and punishment (8000-8600); arbitration and

the private enforcement of law (7500); bankruptcy proceedings (7800); legal

error, rules versus standards, and accuracy in adjudication (Volume I, 790),

punitive damages (Volume II, 3700), the computing of damages, including the

allocationsof damages via contribution and indemnity rules (Volume II, 3500);

andthe production of legal rules and precedent (9000-9900). In addition, issues

relating to evidence and information in litigation, including the rules of

discovery, attorney client privilege, and advice for litigation, are examined in

a companion section on evidence (7900).

 

4. The Economic Analysis of Procedural Rules

 

Economic analyses of procedural rules generally have proceeded within

Posner’s (1973) framework, which conceives the purpose of such rules to be the

minimization of the sum of error costs and direct costs. See also Tullock (1975)

and Posner (1992) for a general analysis of procedure. Consistent with this

framework, Rule 1 of the FRCP directs that the rules be construed and

administered to secure the just, speedy, and inexpensive determination of every

action. Economic analyses of procedural rules focus on the effect such rules

have on the incentives of potential and actual litigants. Thus, the economic

analysis of procedural rules must begin with an economic model of litigation.

 

5. The Economic Model of Litigation

 

The basic economic model of litigation has two parties, a plaintiff (p) that has

a potential claim against the defendant (d). The plaintiff’s estimate of the net

expected value (NEV)of the claim equals his estimate of the probability (P) that

he or she will prevail (Pp) multiplied by the expected award (Dp), net of the

marginal costs to the plaintiff of proceeding to the next stage of litigation (Cp).

The defendant’s objective function for estimating expected loss (EL) is

developed in parallel fashion, as equaling the defendant’s estimate of the

probabilitythe plaintiff will prevail (Pd) times the expected award (Dd) plus the

defendant’s marginal costs of proceeding to the next stage of litigation (Cd).

 

In the more developed models, several of the variables may be endogenously

determined. For example, both litigants’ selections of C may affect their

estimates of the outcome, and one litigant’s choice of C may affect the other

litigant’s estimate of its own cost of proceeding to the next stage. Most analyses

use sequencing models, for which the definition of marginal cost given in the

text is the most general case. In the most basic models, this cost is

conceptualized as the marginal cost of ‘trial’ over ‘settlement’, though

obviously it can be generalized to any of the multiple stages of litigation, and

can refer to cooperative, decisional, or unilateral withdrawal outcomes, as

developed by Cornell (1990). In processes without discrete formal stages, such

as the American discovery process and some Continental trial processes, the

most general case will be continuous updating of NEV and ELas each new item

of information (for example, each witness or investigation) becomes available.

 

In addition, the litigants can be influenced by effects external to the current

litigation - for example, by the precedential or preclusive effects of the

judgment in the current litigation on future litigation. In such cases, a current

plaintiff’s judgment will produce an external gain to the plaintiff (Gp) while the

defendant will suffer external loss (Ld). Likewise, a defendant’s judgment will

produce external gain (Gd) while the plaintiff will suffer an external loss (Lp).

 

The NEV and EL are the primary determinants of the incentives given to

litigants in the economic model, including the amount of effort expended

duringlitigation, the trial/settlement decision, the decision to file a suit, and the

decision to avoid behavior that would give rise to legal liability. Because

procedural rules alter the NEV and EL, these rules will have a central role in

determining litigation incentives and outcomes. Because decisions made at

earlier stages of litigation are dependent upon the parties’ expectations of the

outcomes (and their own and opposing parties’ decisions) during later stages

of the litigation, economic analysis begins from the last stage of trial and

judgment,and proceeds by backward induction. Thus, we list and discuss topics

in reverse chronological order.

 

6. Trial Courts

 

The last stage of a civil litigation in the court of first instance is the trial and

judgment (Cooter and Rubinfeld, 1990). This discussion suppresses the

possibility of appellate review, which in the US federal system is governed by

a separate set of Federal Rules of Appellate Procedure. Appellate review in the

United States is concerned primarily with the correction of legal error and

secondarily with factual error. For a discussion of the economic analysis of

those topics, see Volume I, Chapter 0790 in this volume.

 

7. The Choice Between Judge and Jury Trials

 

Trials and judgments are governed by Sections VI and VII of the FRCP (Rules

38-63). Subject to certain motions for the court to set aside the verdict (Rule 50)

or order a new trial (Rule 59), judgment is entered as given by the factual

findings and legal conclusions of a judge (see Rule 52) or the verdict of a jury

(Rule 58). The federal rules allow any party to demand a trial by jury of any

issue triable as a matter of federal constitutional right by a jury (Rule 38).

Although most American states also provide by state law or constitution for a

right to trial by jury in some or all cases, the state courts generally are not

bound by the federal constitutional right to jury trial, except in certain cases

where they are adjudicating federal claims for which federal law requires a

right to jury trial. Rule 39 allows the parties in these cases also to have a trial

by the court, by joint consent. Gay, et al. (1989) examine the choice between

judge and jury under the assumption that juries are ‘noisier’ adjudicators than

judges.Clermont and Eisenberg (1992) examine empirically whether the choice

of judge versus jury produces differing verdicts in product liability cases tried

under state law, finding that - contrary to the popular belief - judges, not juries,

are more generous to plaintiffs. Further, they do not find that these results are

explained by selection bias.

 

In addition to explicit choices by the parties, several of the Rules allow the

judge to exercise certain powers even when the case is tried by jury. Rule 49

allows the court to require a jury to return a special verdict, which involves

explicit findings upon each of several elements of the claim at issue rather than

allowing the jury to return the more common general verdict (see Lombadero,

1996; James, Hazard and Leubsdorf, 1992, p. 377). Rule 50 provides for

judgment as a matter of law, which can be used by the judge to preempt or

overrule a jury upon the determination that, after being fully heard, there is no

legally sufficient basis for a reasonable jury to find for a party on an issue (see

McLauchlan, 1973). Rule 56 provides for an equivalent ruling in the pretrial

stage by summary judgment, on the basis of pretrial discovery material and

written submissions by the parties (see McLauchlan, 1977), and Rule 12

permits a judgment as a matter of law at the earlier pleading stage, in cases

where the plaintiff’s claim is legally deficient on its face.

 

8. Jury Structure and Decision Rule

 

The rules also regulate the size and decision rule used by civil juries, as well as

theprocess through which jurors are selected. Rule 48 specifies that, unless the

parties otherwise stipulate, a civil jury shall be not fewer than six nor more than

twelve, and that the verdict shall be unanimous. Economic analyses of jury

decision making include an examination of how juries process information

(Klevorick, Rothschild and Winship, 1984; Froeb and Kobayashi, 1996), and

the effect of jury size and alternative decision rules (Klevorick and Rothschild,

1979). Finally, Rule 47 (as supplemented by federal statutes) controls the

selection of jurors, including the procedures for examination of jurors during

the selection process, and the use of peremptory challenges (see, for example,

Schwartz and Schwartz, 1996).

 

9. Litigation Expenditures

 

A separate literature has bypassed explicit consideration of the jury versus

judge decision-making process by modeling the outcome of a trial as being

determined by the litigants’ expenditures on litigation. The most common

model of litigation expenditures is where the both litigants expend resources to

alter the probability of prevailing (see Posner, 1973; Goodman, 1978; Tullock,

1980; Wittman, 1988; Katz, 1988; Hay, 1995; Kobayashi and Lott, 1996). That

is, in these models, the marginal cost of future litigation stages Cpand Cd

(usually conceptualized as ‘trial versus settlement’) are endogenous choice

variables that determine the probability that the plaintiff will prevail P(Cp,Cd).

The equilibrium amounts of litigation expenditures depend upon the relative

stakes of the parties, and upon the relative merits of the case. The litigation

expenditures and probability determined by solving for the Nash equilibria are

used as the expected values for earlier stages of litigation. Other articles have

examined the relationship between the burden of evidence or decision standard

and litigation expenditures. In these models, litigation expenditures that allow

a litigant to meet the burden of proof or decision standard serve as a costly

signal of liability (see, for example, Rubinfeld and Sappington, 1987).

 

10. The Trial/Settlement Decision

 

Themajor decision preceding commencement of a trial is the decision whether

to settle the case or proceed to trial. The relative magnitudes of the plaintiff’s

NEV and the defendant’s EL are the primary determinants of whether a case

settles or whether trial occurs. As noted above, this modeling can be

generalized to any stage preceding a decision by the court, or a further

commitmentto incur litigation expense. Under risk neutrality, the NEVsets the

plaintiff’s minimum acceptable settlement offer and the EL the defendant’s

maximum settlement bid. In the absence of lawyer-client agency costs, a

sufficient (but not necessary) condition for litigation is the perceived absence

of a bargaining range, which occurs when the plaintiff’s minimum acceptable

offer is greater than the defendant’s maximum bid, that is, NEV > EL, or

equivalently:

 

Pp(Dp+ Gp+ Lp) ! Pd(Dd+ Ld+Gd) + Gd! Lp> Cp+ Cd

(1)

 

The above condition identifies two reasons why parties choose to forego

settlement and proceed to trial. The first is based on prediction failure. The

second is based on consideration of external effects, specifically precedential

or preclusive effects of the current litigation on future litigation.

 

                                            A. The Optimism Model

 

11. The Prediction Failure Model

 

Under the simplifying assumptions that the litigants have symmetric stakes (Dp

= Dd= D) and that there are no external effects (Ld= Lp= Gd= Gp= 0), the

condition for trial rather than settlement reduces to the following expression:

 

(Pp! Pd)D > Cp+ Cd                                                                         

                                                                                                         (2)

 

Thiscondition illustrates the optimism model of litigation. In such a model,

settlement fails to occur because of prediction failure - that is, settlement fails

because the parties have mutually inconsistent and relatively optimistic

estimates of the probability that the plaintiff will prevail (that is, Pp> Pd). The

litigants’ erroneous predictions lead them to behave as if there were no room

for a mutually beneficial settlement.

 

12. The Priest-Klein Selection Model

 

Economists also have modeled the process through which such mutually

inconsistent and optimistic predictions are generated. The Priest and Klein

(1984) model of case selection is based on the optimism model, and has two

major hypotheses. The first and more general hypothesis is that the cases

selected for trial are not representative of the population of disputes.

Specifically, under the assumptions of symmetric stakes and specific

distributional assumptions about the litigants’ estimates of P (specifically that

each litigant’s estimate of P is an independent draw from a unimodal

distribution centered around the true P), the basic Priest-Klein model predicts

that a disproportionate number of cases selected for trial will come from cases

that are close to the decision standard. The second hypothesis is a specific

prediction for observed case outcomes in the limiting case where the litigants

accuratelyestimate P. Priest and Klein show that under these circumstances the

distribution of filed cases around the decision standard becomes approximately

symmetric, and the generated plaintiff win rate approximately equals 50

percent.

 

A large number of articles have examined the limiting prediction of a 50

percent win rate, and have presented evidence where the win rate differs from

50 percent. See, for example, Priest (1985), Wittman (1985), Ramsayer and

Nakazato(1989), Eisenberg (1990), Hylton (1993), Thomas (1995), Waldfogel

(1995), Shavell (1996). See Kessler, Meites and Miller (1996) and Kobayashi

(1996)for recent surveys of the literature. While generally rejecting the specific

fifty percent hypothesis, this evidence may simply reflect the fact that the

assumptions underlying the limiting case do not hold. Further, empirical

studies examining the more general predictions of the Priest-Klein model have

found evidence consistent with model (see Kobayashi, 1996).

 

13. Settlement Failure and Discovery

 

Because the prediction failure model predicts settlement failure because

litigants lack information at the time of the final trial/settlement decision, it

suggests procedural reforms aimed at increasing pre-trial information. The

management of pre-trial information is addressed in Section V of the FRCP

(Rules 26-37). The primary rule is Rule 26, which contains the general

provisions governing civil discovery. Under Rule 26, litigants are required to

respondto requests for information by the adverse party, on the theory that such

requests and responses would increase pre-trial information and increase

settlement and accuracy in adjudication. However, economic models of

discoveryhave pointed out several problems with the rules. First, the rules shift

the costs of gathering information from the requesting to the responding party,

and thereby introduce both a potential moral-hazard problem and strategic

behavior into the demand for pretrial discovery. Because the costs of

responding to a discovery request are, in many cases, larger that the cost of

making a request, this externalization predicts that parties will request

information well past the point where the marginal value of information

outweighs the marginal costs of gathering the information. In addition, the

process of discovery can decrease, rather than increase, settlement by making

parties more optimistic (see Cooter and Rubinfeld, 1994). Finally, discovery

can potentially decrease the amount of information in litigation by providing

disincentives for the production and retention of information that might be

subject to extensive legal discovery (see Kobayashi, Parker and Ribstein, 1996;

see also Hay, 1994 and Sobel, 1989).

 

14. Amendments to the Discovery Rules

 

Theperceived problems of ‘overdiscovery’ predicted by the economic model of

litigation have been addressed in several amendments to the discovery rules. In

1983,Rule 26(b) was amended to embody an explicit cost-benefit test that could

be applied by the judge to limit discovery. At that same time FRCP 16 was

amended to expand the judge’s power to manage the pretrial process in general.

However,the economic model suggests that these managerial solutions may be

inferior to forcing each litigant to internalize the costs of its own information

requests, or at least to pay for the marginal cost of extensive discovery requests

(Cooter and Rubinfeld, 1994).

 

In 1993, Rule 26 was further amended, again in response to concerns

regarding litigants’ overinvestment in discovery activity. However, the major

change to the rules was to institute a regime of initial disclosures to be made

without any request for information from the adverse party (see FRCP Rule

26a;Brazil, 1978; Schwarzer, 1989; Cooter and Rubinfeld, 1995). Applying the

economicmodel, this change is likely to make matters worse. To the extent that

cost-shifting is the source of ‘discovery abuse’, these rules exacerbate the

problem by allowing the requesting party to externalize the costs of both

requestingand responding to requests for information. This effect is magnified

by abandonment of code pleading in favor of the broad notice pleading rules

contained in the FRCP (Section III, Rules 7-15; see Epstein, 1973a, 1973b,

1974, 1986; Posner, 1973; Katz, 1990; Bone, 1997). Further, to the extent that

this rule moves the costs of pretrial discovery into an earlier phase of the

litigation, this may have the effect of increasing total costs, and decreasing the

marginal costs of trial over settlement. The 1993 amendment was subject to a

local option in US federal districts, which has reduced some of the effects of the

rulechange. Empirical analysis has shown that the busiest federal districts have

opted out of the new rule (Kobayashi, Parker and Ribstein, 1996).

 

                                            B. The External Effects Model

 

15. External Effects

 

Relative optimism is not the only way in which to generate the absence of a

perceived bargaining range. The existence of external effects can cause

litigation to occur even if the parties to the litigation agree on the likely

outcome of the case. Examples of external effects include the precedential and

preclusive effect of litigation, both of which serve to affect litigants’ prospects

in future cases (see Galanter, 1974; Rubin, 1977; Che and Yi, 1993 and

Kobayashi, 1996; see also Landes and Posner, 1979; Blume and Rubinfeld,

1982; Priest, 1987, 1980; Cooter, 1987; Galanter, 1987). For a more complete

discussion of precedent and legal change, see Chapters (9000-9900) of this

volume.

 

To illustrate this point, consider the case with symmetric stakes and where

the litigants agree on the probability the plaintiff will prevail, that is, Pp= Pd

= P. According to the simple optimism model, such a case will settle in order

to save the costs of a trial. However, even under these assumptions, litigation

maybe generated if the current case has implications for future behavior of the

litigants involved. The NEV of the lawsuit to the plaintiff equals P(D +Gp) !(1

! P)Lp! Cp, and the defendant’s EL = P(D+Ld) !(1 ! P)Gd+ Cd. The

condition for litigation NEV > EL becomes:

 

P(Gp! Ld) + (1 ! P)(Gd! Lp) > Cd+ Cp

(3)

 

Condition (3) shows that, even when the prospective outcome of a trial is

agreed upon by the parties, trials can be generated when the external gains to

the winning litigant outweigh the external losses to the losing litigant. In

contrast to the sources of trial in the prediction failure and settlement failure

models (where information costs prevent litigants from achieving a settlement

within an existing bargaining range), external effects cause trials by preventing

the existence of a bargaining range due to asymmetries in external effects.

Because no mutually beneficial bargain is foregone, these trials are not a

failure. In fact, the existence of external effects focuses on the positive and

valuable rulemaking function of trials.

 

16. External Effects and Procedural Rules

 

The model of trial based on external effects also suggests the relevance of a

different set of procedural rules - one that focuses on the effects of litigation on

future claims and third parties. The FRCP addresses the packaging of claims

and parties generally in Rules 13-14 and 17-25. These rules serve as inclusive

packaging devices, which operate by giving litigants incentives and

opportunities to join parties and claims in order to internalize external effects

and to achieve economies of scale in litigation. They include the permissive

joinder rules (Rule 18-20), impleader (Rule 14), interpleader (Rule 22),

intervention (Rule 24), the rule governing counterclaims (Rule 13), and the rule

governing class actions (Rule 23). For an economic analysis of counterclaims,

seeLandes (1994). For a discussion of various aspects of class actions, see Dam

(1975), Rosenfeld (1976), Bernstein (1978), Friedman (1996), and Chapter

7600 of this volume.

 

17. Common Law Packaging Rules

 

A second major category of packaging rules include the common law rules of

precedent, or stare decisis, and the rules of preclusion, or the law of res

judicata. The common law rules of res judicata include the doctrines of claim

preclusion (also known as merger and bar) and issue preclusion (also known

as collateral estoppel). Under the rule of claim preclusion, parties or their

privies (that is, predecessors or successors in interest) are precluded from

relitigating issues that were or could have been raised in the first action (see,

for example, Kobayashi, 1996; Landes and Posner, 1994). Claim preclusion

preventsthe splitting of claims based on the same out-of-court transaction, thus

functioning as mandatory joinder-of-claims rules by their prospective effect on

future litigation. In contrast, issue preclusion (collateral estoppel) is limited to

issues that were actually litigated, actually decided, and necessary to the

judgment in the first case. In addition, while application of claim preclusion is

generally limited symmetrically to persons who were either parties or

successors to parties on both sides of the initial litigation, collateral estoppel

can operate non-symmetrically, or, as it is termed in legal doctrine,

non-mutually, in the sense that collateral estoppel can operate in favor of

strangers to the first action, but can operate against only parties or their

successors(see Spurr, 1991; Hay, 1993). The existence of non-party preclusion

has led to the attempted use of settlement conditioned on vacatur (the

eradication of a prior public decision) as a means to eliminate the effect of

preclusion (see Fisch, 1991).

 

18. Offensive Non-Mutual Collateral Estoppel

 

To see the effect of the application of non-mutual collateral estoppel, consider

the case where a single defendant faces two different plaintiffs sequentially.

Suppose for simplicity that both cases turn on a single issue that determines the

outcome. First consider offensive collateral estoppel. If the first plaintiff

prevails against the defendant, the defendant is precluded from relitigating the

decisive issue against the second plaintiff. Thus, a loss in the first case results

in the loss of both cases, and the external effect Ld= (1 ! P)D. In contrast, a

win by the defendant against the first plaintiff does not preclude the second

plaintiff from relitigating the issue, if the second plaintiff is neither a party nor

a privy to the first action. Thus, the only effect on the second litigation would

be the precedential gain Gd= (dP D), where dP equals the change in P in the

second litigation. If there are no novel issues of law, this gain is likely to be

close to zero. And under the assumption that the plaintiffs are not repeat

litigants, Lp= Gp= 0.

 

Substituting into the condition for litigation (equation (3)) yields:

 

(1 ! P)D(dP ! P) > Cd+ Cp

(4)

 

If the precedential change dP is close to zero, condition (4) is unlikely to be

satisfied. Thus, there will be little incentive to take such cases to trial. This

incentive for settlement has led some to object to offensive non-mutual

collateral estoppel on the grounds that such a rule will allow plaintiffs to

repeatedly extract large settlements from a common defendant. However, the

total effect on settlement amounts from such a rule is unclear (see Hay, 1993).

First, the effect of estoppel causes the repeat litigant to capitalize all future

litigation into the first litigation. This increase in the stakes causes the repeat

litigant to spend a larger amount of resources when he goes to trial. This has

the effect of lowering P, thus moving the bottom of the bargaining range

downward. Further, the repeat litigant is likely to take a harder bargaining

stance in the first litigation by anticipating its effects on future litigation.

 

19. Defensive Non-Mutual Collateral Estoppel

 

In contrast, consider the use of defensive non-mutual collateral estoppel with

multiple plaintiffs. Again, assume that Lp= Gp= 0. Now Ld= dP D and Gd=

PD. Substitution into the condition for litigation (equation (3)) yields:

 

PD ((1 ! P) ! dP) > Cd+ Cp

(5)

 

In contrast to the rule of offensive non-mutual collateral estoppel with

multipleplaintiffs, such a rule increases the likelihood of litigation. In addition,

use of such a rule would apply preclusion against a non-party that was

represented by a party in the first litigation with low relative stakes, which may

increase legal error and increase the amount of litigation (see Spurr, 1991).

Thus, the court’s rejection of the latter rule because of due process concerns is

consistent with the predictions of economic analysis.

 

                   C. The Bargaining Failure Model and Asymmetric Information

 

20. Bargaining Failure

 

Even within the basic optimism model of litigation, condition (2) is not a

necessary condition for litigation. Even if a positive bargaining range is

perceived to exist, the parties may fail to settle due to bargaining failure (see,

for example, Cooter, Marks and Mnookin, 1982; Gross and Syverud, 1991;

Shavell, 1993) or due to the existence of asymmetric information (see, for

example, P’ng, 1983, 1987; Bebchuk, 1984; Schweizer, 1989; Spier, 1992,

1994b; Daughety and Reinganum, 1993; Froeb, 1993; Wang, Kim and Yi,

1994). In the former case the litigants fail to agree on a specific bargain even

if both litigants agree that a positive bargaining range exists. In the latter case,

trials are a necessary cost of avoiding adverse selection problems. In both of

thesecases, the potential for costly trials represents a calculated cost of strategic

behavior aimed at obtaining a better expected settlement.

 

21. Bargaining Failure and Procedural Rules

 

The existence of asymmetric information in litigation again points to the

discovery rules (see the discussion in Part B, above). In addition, economists

have examined procedural rules that would increase the incentives for

settlement. Under FRCP Rule 68, the defendant can make an offer of judgment.

If this offer is not accepted by the plaintiff, and the subsequent judgment is less

favorable to the plaintiff, the plaintiff is responsible for the defendant’s costs

after making the offer (see Miller, 1986; Anderson, 1994; Anderson and Rowe,

1996; Chung, 1996). However, liberal and early use of Rule 68 is limited by the

requirementthat the defendant offer not merely settlement but judgment, which

can have a preclusive effect in subsequent litigation. Further, under the current

rules, it does not allow the plaintiff to recover marginal fees (although many

statutes provide for one way fee recovery by plaintiffs). In addition, economists

have analyzed contractual settlement devices such as settlement escrows (see,

for example, Gertner and Miller, 1995), and the use of mediation and

arbitration techniques (see Shavell, 1995).

 

                                    D. Lawyer-Client and Client-Client Agency Costs

 

22. Lawyer-Client Agency Costs

 

Finally, trials can be generated due to agency costs between lawyer and client

and by agency problems between co-interested clients. A large literature has

modeled the consequence of attorney compensation arrangements that give rise

to the agency problems in settlement (see, for example, Miller, 1987). Articles

examining the incentive effects of contingent fees include, Clermont and

Currivan, 1978; Danzon (1983); Lynk (1990, 1994); Miceli and Segerson

(1991); Thomason (1991); Dana and Spier (1993); Rubinfeld and Scotchmer

(1993); Miceli (1994); Watts (1994); Hay (1996, 1997). Potential solutions to

such problems include enforcement of ethical rules or the use of reputational

mechanisms (see Gilson and Mnookin, 1984; Smith and Cox, 1985; Yang,

1996).

 

23. Contribution, Indemnity and Agreements with Multiple Defendants

 

Similarly, litigation and settlement are affected by the rules which allocate

liabilityamong multiple defendants. These include the rules of contribution and

indemnity (see Easterbrook, Landes, and Posner, 1980; Landes and Posner,

1980;Hause, 1989; Kornhauser and Revesz, 1989, 1990, 1994; Klerman, 1996)

andagreements between plaintiffs and defendants (see Bernstein and Klerman,

 

           E. Effect of the Decision to File a Suit and the Decision to Avoid a Lawsuit

 

24. Incentives to Avoid Litigation

 

Finally, the NEV and the EL provide incentives during the earlier stages of

litigation. The EL, discounted by the probability that a suit will be filed and by

thelikelihood of settlement, provides deterrence of and incentives for avoidance

of behavior subject to legal liability (Polinsky and Rubinfeld, 1988a, 1988b;

Cooter and Rubinfeld, 1989; see also Ordover, 1978, 1981, and Hylton, 1990).

 

25. The Decision to File a Lawsuit

 

Similarly, the NEV of a lawsuit is a primary determinant of whether or not a

plaintiff chooses to file a suit. While many analyses of litigation limit their

examination of legal claims to those whose NEV is positive, this is not a

necessarycondition for a suit to be filed. Negative NEV suits may also be filed

in order to extract a settlement, and such strategies can be successful when the

threat to litigate is credible (see Rosenberg and Shavell, 1985; Nalebuff, 1987;

Bebchuk,1988, 1996; Katz, 1990; Klein, 1990; Miceli, 1993; and Bone, 1997).

For general analyses of the relationship between the cost of litigation and

private and social incentive to bring a suit, see Schwartz and Tullock (1975);

Priest (1982); Shavell (1982b); Menell (1983); Trubek, et al. (1983), Kaplow

(1986); Rose-Ackerman and Geistfeld (1987); Hazard (1989); Williams and

Williams (1994).

 

26. The Role of Sequencing of and Exit from Litigation

 

Perhaps one of the most salient features of the FRCP, and one that has been

onlyrecently addressed by the economics literature, is the effect of the rules on

the timing and sequencing of litigation. While many economic models treat

litigation as a timeless decision, the procedural rules lay out a sequence of

events that determine how litigation will proceed over time. The timing and

sequence of litigation has been shown to have important implications for the

NEV of a lawsuit and the behavior of litigants.

 

For example, the ability of litigants to sequentially litigate motions or to

separate liability and damage phases of trials have been shown to affect both

the cost of litigation and the value of the litigation if litigants are allowed to

exit. Sequencing with easy exit increases the ‘option’ value of litigation by

allowing plaintiffs to avoid conditional negative NEV outcomes by exiting the

litigation (see Cornell, 1990; Landes, 1993; Chen, Chien and Chu, 1997). In

addition, the ability to take motions sequentially can affect the credibility of

negative expected value suits (Bebchuk, 1996).

 

Under the FRCP, exit is controlled by Rule 41. Under Rule 41, the plaintiff

can unilaterally dismiss an action without order of the court at any time before

the adverse party answers. After this, dismissal requires mutual stipulation by

all parties, or an order of the court. The application of Rule 41 is a prime

example of the tension between the ex ante and ex post effects of procedural

rules. Ex post, it is unlikely that a defendant will prevent a plaintiff from

abandoning a claim. However, the willingness of the court or the adverse party

to allow the plaintiff to easily exit the litigation may be the primary reason the

suit was filed in the first place.

 

27. Lawyer Sanctions

 

The FRCP address this inability to commit through Rule 11, which allows the

court to impose sanctions on parties and their attorneys for filing frivolous

suits. This rule addresses the commitment problem by allowing the motion for

sanctions to survive the dismissal or mooting or the original claim. Thus, Rule

11 allows the motion for sanction to survive as separate action with the

sanctions as the incentive to litigate. Economic analyses of sanctions under

Rule 11 include Kobayashi and Parker (1993); Polinsky and Rubinfeld (1993);

Bebchuk and Chang (1996); and Bone (1997).

 

However, recent amendments to Rule 11 may have reduced the value of

Rule 11 as a commitment device. In the 1993 Amendments to the FRCP, the

rules were amended to allow the a litigant twenty-one days to withdraw his

pleading upon being served with a Rule 11 motion by the adverse litigant. In

essence, the new rules allows for free exit, and thus weaken the commitment

and deterrent effect of the rules. While the amendments to Rule 11 were

designed in large part to decrease the volume of ‘satellite’ litigation over

alleged Rule 11 violations, the effect of the rule change may be to increase the

number of filed cases, the rate of Rule 11 challenges, and total litigation costs

(Kobayashi and Parker, 1993).

 

28. Fee Shifting

 

Economists have also examined the use of two-way fee shifting (often referred

to in America as the ‘English’ rule, in contrast with the general ‘American’

rule that each party bears its own fees, regardless of outcome) to deter low or

negative value lawsuits. Because two-way fee shifting imposes extra costs on

the losing party, it disproportionately affects low probability lawsuits (see

Shavell, 1982a). However, the FRCP has not incorporated two-way shifting of

fees and costs, and state experiments have not proven successful or long lived

(see Snyder and Hughes, 1990; Hughes and Snyder, 1995). In addition, almost

all fee-shifting statutes require the imposition of a judgment. Thus, the use of

fee shifting would not necessarily reduce the option value of litigation. That is,

if exit is routinely allowed prior to judgment, fee shifting will not serve as a

deterrent. Indeed, to the extent that two-way fee shifting would increase the

variance in final outcomes, use of such a system would increase the option

value of litigation, and thereby increase the supply of filed cases (Cornell,

1990). For a more complete discussion of fee shifting, see Chapter 7300 of this

volume. See also, Dewees, Prichard and Trebilcock (1981); Braeutigam, Owen

and Panzar (1984); Reinganum and Wilde (1986); Bowles (1987); Katz

(1987);Plott (1987); Donohue (1991); Graville (1993); Spier (1994a), and Katz

and Beckner (1995).

 

                                           F. Procedural Rulemaking

 

29. The Process of Procedural Rulemaking

 

Finally, both the process of procedural rulemaking and the adversarial system

itself have been the subject of economic analyses. Indeed, the controversy over

the relative merits of party-controlled adversarial presentation and a more

‘managerial’ system of litigation was at the heart of the controversial

amendments to the discovery rules, and has been the subject of debate in the

academic literature (see McChesney, 1979; Tullock, 1980, 1988; Langbein,

1985). For a recent report of experimental results, see Block and Parker (1996).

For a public choice treatment of the common law of procedural rules, see

Macey (1994).

 

The controversy over the recent amendments to Rule 11 and 26a have

drawn attention to the process through which the rules are revised (see

Kobayashi and Parker, 1993; Kobayashi, 1996), and have highlighted the lack

of empirical evidence on the performance of procedural rules and systems (see,

for example, Galanter, 1983; Walker, 1994). Indeed, the amendment to Rule

26a, which allows individual district courts to opt-out of the discovery rules, the

Civil Justice Reform Act of 1990, which encouraged local variations in each

district, and variation in local Federal Court rules and state court rules all

provide substantial challenges to the traditional centralized system of rule

making that existed in the federal system prior to 1990 (see Kobayashi, Parker

and Ribstein, 1996; Solimine and Pacheco, 1997; and Subrin, 1989).



中国法律经济学网登载此文出于学术研究之目的,绝不意味着中国法律经济学网赞同其观点或证实其描述。以上内容仅供研究者学习与交流,无意侵犯版权。如有侵犯您的利益,请告知。我们将尽快删除。

加入日期:2007/1/11 15:02:31浏览次数:2727
发表评论
名号:
内容:
验证: 6427
法律经济学网
联系站长: 柯华庆 lawgame@263.net 京ICP备09028584号
北京市昌平区中国政法大学法学院(102249)
本网站由卡卡鱼网提供技术支持 网站总访问量:1996356