I wish to summarize and discuss a 2000 legal article of Prof. Prof. Harold R. Medina of Columbia Law School, USA, entitled “THE CONTENT OF OUR CASEBOOKS: WHY DO CASES GET LITIGATED?”, for purposes of legal research of the visitors of this blog.
Using the law and economics model, Prof. Medina advances the view that the central paradox of litigation is the fact that taken together, the parties to a lawsuit are losers from the moment they enter the process of adjudication.
He urges lawyers and jurists to understand the behavioral dynamics that drive the real people we encounter in our profession. He theorizes that if our predictions make use of the narrow incentive structure admitted by law and economics, they will often fail (because) this impoverished model fails to capture a robust picture of human decision-making. He believes that the task of the Bench and the Bar is to adapt the system of dispute resolution to the world in which real humans exist, imperfections and all.
Prof. Medina writes that as disputants enter the litigation process, they are clear losers because “any division of the stake between them, whether it be one side taking all, or half-and-half, or anything in between, leaves the parties jointly in the same position as when they began their dispute: however they slice it, they will still have the entire pie to share. Once lawyers and courts and filing fees and witnesses and depositions and all the rest are brought into the picture, the pie starts getting smaller and smaller.
He adds that regardless of the contractual terms with their attorneys and even if represented on contingency, clients soon realize that they are signing away a significant amount of resources to their newly acquired legal representatives. Most parties quickly learn this lesson, he states, and a remarkably stable ninety-five percent of cases manage to get resolved well short of trial.
As applied to law, the critical economic insight regarding litigated cases comes from asking a simple question about the incentives that would lead people to actually seek a trial resolution in a case.
The key insight begins with a very simple model that assumes each party enters the
litigation process with an expected value attached to the claim of the plaintiff.
At its most simple, the model appears as follows: EVP= P x A minus CP.
In this simple model, EVP represents the expected value of the case to the plaintiff. As set forth in this account, the plaintiff’s expectations are a function of her probability of success (P), the likely award to be obtained (A), and the costs associated with prosecuting the claim (Cp).
This calculation can then also be expressed as the defendant’s expected loss from a plaintiff’s claim.
Here the scaled down version of the model appears as follows: EVD= P x A plus CD.
The defendant’s calculation is the mirror image of the plaintiff’s, with one critical difference: the costs are added to the defendant’s likely loss, whereas they are subtracted from the plaintiff’s likely recovery.
Thus, these streamlined equations reflect the fact that the costs associated with litigation are a joint loss to the parties and are subtracted from their joint welfare.
By combining these two equations, it is possible to isolate what is termed a “settlement zone” in which two parties with convergent expectations of the likely award and the probability of the plaintiff prevailing are able to negotiate a mutually advantageous end to the litigation. This may be represented as follows:
SETTLEMENT ZONE CREATED BY CP PLUS CD,
WHEN PARTIES AGREE ON VALUES OF P AND A
If parties can agree on the likelihood of P (the probability of the plaintiff winning) and A (the amount that will be awarded if she wins), what they are really arguing about is how to divide up the costs of litigation, CP and CD.
For this model to work, however, there must be a convergence of the estimated value of the case, which is a function of the likelihood that the plaintiff will prevail and the prospective damage award if she does indeed triumph.
So if there is agreement on both the probability of the plaintiff prevailing and the likely size of the ensuing award, cases should settle almost immediately, before much of the pie is eaten away by the transaction costs associated with litigation.
What then if the parties’ estimates of probable success or likely award do not converge? This turns out to be the arena for intervention of the American rules of procedure. The basic law and economics insight is to claim that the source of divergence between the parties must rest on incompatible assessments of either the facts or the law governing a particular case.
Since the parties (and society) are best served by promoting quick settlements that conserve the joint resources of the parties, the rules of procedure should attempt to intercede to remove the sources of division. The object of the rules of procedure is to foster the just, speedy, and inexpensive resolution of disputes.
As a society, we invest heavily in the creation of the public good known as decisional law. We build courthouses, staff them with respected leaders of the communities called judges, stock them with bright clerks, and demand that their experiential wisdom be reduced to written form. The resulting case law forms the heart of the common law enterprise and is publicly available to counsel to inform their assessments of the strength of the claims put forward on behalf of their clients.
In addition, we allow for a relatively quick reality check of the legal basis for a plaintiff’s claim through the rules on motion to dismiss. In some circumstances we may even allow for interlocutory appeals, mandamus, or certification of a case to a state appellate court, all for the purpose of providing an early look at the governing legal principles.
Factual disagreements are more difficult. Here the true rendition of the factual strength of a party’s claim lies not in the public domain but almost certainly in the private knowledge of the litigants themselves. As long as the parties have private information about their side of the case, the prospect of settlement may be significantly compromised.
Here too the rules of procedure seek to intercede. Rather than draw on a body of knowledge that is maintained in the public domain, as with published decisional law, the combined effects of notice pleading and liberal discovery serve to create a limited public domain of shared information between the parties.
The scope of discovery is the single most distinctive feature of American procedure and its scope and cost not only draw attention but also typically shock foreign litigants who find themselves in American courts. But under the economic model of litigation, the costs of discovery serve two important functions.
First, the fact that parties face significant costs in the litigation process expands the potential settlement zone and creates a greater possibility of mutually advantageous settlement, even if the parties do not have perfectly matched assessments of the likely prospects were the case to go to trial.
More significantly, the costs of discovery are justified to the extent that they bring the parties’ assessments of the case into line at some point prior to trial. Under this approach, discovery not only allows for a trial to be “on the merits” if the parties are unable to settle, but “the investment in mutually shared information makes settlement much more likely”.
Once the parties have discovered all the information relevant to the claims and defenses in the case, and once they have read from the same decisional law and tested the application of the law through motions to dismiss and motions for summary judgment, there is no reason to believe that the parties should not settle.
We must understand how people actually behave under conditions of stress and uncertainty. Why then do cases go to trial once the lawyers have tested the law and discovered the facts? First, there is the possibility of parties just getting it wrong. Second, there are always new areas of law, new claims, new conceptions of rights and duties. Third, it may be the case that parties diverge in their estimations of likely trial outcomes because the law is unsettled in the particular domain in which their dispute arose. Thus, there are two potential explanations for cases going to trial. The first is mistake and the second is uncertainty in the state of the law.
To conclude that parties are being helped to settle in an efficient manner, we need to assume that they will integrate the “shared knowledge of the facts and the law” in such a way as to further their achievement of “shared assessments of the case”. In other words, we need to have a behavioral theory of “how parties make decisions in conditions of uncertainty as they go about the process of acquiring the costly information about the relevant law and facts through the litigation system”.
The concern is not over the benefits of liberal pleading and court-supervised discovery compared to some more formalized common law pleading regimes from days gone by. The concern is over the assumption of how parties will respond to the new regime. The challenge is to understand how people integrate information. People individually and even aggregated through market transactions simply do not see the world through the lenses offered up by the expected-value economic calculus.
For example, people value losses more than gains and that they will invest more heavily in seeking to avoid a loss than realize a gain, even of equal value. Perhaps as a consequence, people value what they have over what they may aspire to have. This is known as the endowment effect and is a robust effect, even if the goods are of equal value.
There is also the problem of “framing” — the effect that presenting the same information as a matter of gains or losses has on the valuation. In one study, the first group was asked how much should be awarded to make whole the victim of the accident. The second group was asked how much they would have to be paid to accept the harm suffered by the victim. The only difference in asking the question in one or another way is to pose the inquiry as backward-looking (ex post relief) or as forward-looking (ex ante valuation of the harm). In either case, the hypothesis is that the value should be the same. But the responses to the questions did not prove such hypothesis.
