Teachers' Manual to Green, Nesson & Murray, Problems, Cases and Materials on Evidence, 3rd Edition.

Go to: Teacher's Manual Introduction Chapter 10 Contents Next Section

A. Allocation in Civil Cases (1021)

Problem - Civil Allocation in Idylia (1027)

The point:

There are three points:

(1) The concept of probability as a determining factor in the allocation of the elements of proof is primarily an expression of judicial efficiency. It is not efficient to expend judicial resources in the litigation of cases in which, even after litigation, no firm conclusion can be reached. It is generally sensible as an efficiency measure, therefore, to leave the loss in such cases on the party who suffered it. That party is the would-be plaintiff. Unlike the old system in Idylia, under the adversary system it is the plaintiffs decision whether or not to initiate a case. Allocation to the plaintiff of the burden of proving the most frequently determinative elements of a case inhibits the plaintiff from bringing his case to court unless he thinks that he will be able to convince the factfinder that he should win.

(2) For the inhibition to work, the burden imposed should be both a burden of production and a burden of persuasion. Although the two types of burden are logically distinguishable, they are functionally coupled.

(3) It makes sense to assign to the defendant elements which are seldom controverted.

Answer and Analysis:

There is ambiguity in the way this problem is stated that will have to be corrected in our next edition Damages were awarded in 1000 cases based on affirmative findings of liability and in an additional 300 cases based on the outcome of the coin flip.

This problem is designed to illustrate more clearly than the Cleary article does what the content of the concept of probability is and what its function is in the allocation of burden of proof.

In 600 cases the Tribunal winds up unresolved as to whether a given determinative element is present or not. Because the tribunal cannot come to a conclusion, one might think that the chances were 50-50 that the plaintiff should recover, and therefore the case might appropriately be resolved by flipping a coin or by awarding 50% of the damages sought. Such approaches do not generate verdicts which can be readily accepted as resolutions of what happened. They are, rather, expressions of the system's inability to come to a conclusion. That is why the hypothetical population of Idylia found the coin-flip resolutions unsatisfactory, and why coin-flip resolutions in our system would likewise by unsatisfactory.

In this connection, note what happened to a judge who tried the coinflip method:

The State Commission on Judicial Conduct yesterday announced that it had barred Alan I. Friess from ever again being a judge in New York. Mr. Friess was a Criminal Court judge for three and a half years before he resigned while misconduct charges against him were pending.

The charges stemmed from two incidents. In one, he asked courtroom spectators to vote on which of two litigants to believe in a harassment case. In the other, he decided the length of a jail term by the toss of a coin . . . .

The commission concluded that Mr. Friess "exhibited extraordinarily poor judgment, utter contempt for the process of law and the greatest misunderstanding of the role and responsibility of a judge in our legal system."

(From the New York Times of April 7, 1983)

As a general matter the state has no great interest in who bears the loss in cases in which the litigation process is unable to achieve an affirmative resolution, as long as the state.manages to avoid resolutions which undercut the credibility of its processes, like coin-flips. The most efficient result, from the state's point of view, would be to avoid the trouble of litigating in such cases at all.

This suggest using the initial allocation of assumptions about elements (i.e., do we start by assuming that there was an agreement or that there was not?) and the rule of decision about what happens if, after proof, the tribunal is still unresolved, as a means of inhibiting plaintiffs from initiating litigation in such cases.

The behavioral assumption is that plaintiffs will be deterred (to some degree) from initiating adversary process unless they think they can succeed in affinnatively proving each of the elements necessary to recovery. This logic would seem at first blush to lead to the conclusion that the burden of proving all elements necessary for recovery should be imposed on plaintiffs.

When, if at all, does it make sense to assign an element to the defendant? By assigning the burden of proof to the plaintiffs on every element we would assure that every case which is brought to trial will involve presentations of proof on every issue. In a contract action a plaintiff would have to address every element of claim or defense, and prove that the facts with respect to each issue were consistent with recovery. This suggests a modification of the allocation strategy. Some elements require significant time to prove, but are not often decisive. Assigning such elements to the defendant would cut down on the scope of trials without greatly diminishing the degree to which plaintiffs would be inhibited from bringing such cases to trial in which the system will be unable to achieve affirmative resolution.

In the Idylia problem, what would happen if the burden of proving all elements were assigned to the plaintiff? If plaintiffs could accurately assess the strength of their cases (which of course they cannot do with precision, but which they can do better than anyone else), then the consequence in Idylia would be that plaintiffs would not bother to initiate the cases in which they would eventually lose. Under this regime (operating ideally) the system would now litigate 1000 cases instead of 3000.

Each of those 1000 cases would, however, involve proof of all four of the elements of contract in Idylia (and many more if we take all of the elements of contract from Williston and Corbin). One of the elements, absence of mistake, is very seldom determinative (only 35 cases in 1000), and (assume) consumes considerable time and resource to prove. What would happen if the burden of proof with respect to it were assigned to the defendant? The defendant would now have to prove (to a degree sufficient to permit an affirmative resolution) that there had been a mistake. The consequence would be that plaintiffs would no longer be forcefully deterred from bringing cases in which they could not affirmatively prove absence of mistake. Plaintiffs might now be expected to initiate a total of 1015 cases, fifteen more than before. But this additional burden on the system would be offset by the fact that in less than fifteen of them (and in as few as five of them if the burden of proof on the defendant inhibits him from advancing the mistake issue because he will be unable to prove it affirmatively) would the issue of mistake have to be litigated at all.

The model is, obviously, highly idealized, particularly with respect to assumptions about the abilities of the parties to assess the strength of their proof. Nevertheless, the considerations of efficiency which are highlighted by the problem have been an important (though not well-understood) determinant in the allocation of elements between plaintiffs and defendants.

Go to: Teacher's Manual Introduction Chapter 10 Contents Next Section