A. Allocation in Civil Cases


12 Stan. L. Rev. 5-14 (1959)


Since all are agreed that procedure exists only for the purpose of putting the substantive law effectively to work, a preliminary look at the nature of substantive law, as viewed procedurally, is appropriate.

Every dog, said the common law, is entitled to one bite. This result was reached from reasoning that man's best friend was not in general dangerous, and hence the owner should not be liable when the dog departed from his normally peaceable pursuits and inflicted injury. Liability should follow only when the owner had reason to know of the dangerous proclivities of his dog, and the one bite afforded notice of those proclivities. So the formula for holding a dog owner liable at common law is: + ownership + notice of dangerous character + biting.

This rule of law becomes monotonous to postmen. Hence the postmen cause to be introduced in the legislature a bill making owners of dogs absolutely liable, i.e., eliminating notice from the formula for liability. At the hearing on the bill, however, the dog lovers appear and, while admitting the justness of the postmen's complaint, point out that a dog ought at least to be entitled to defend himself against human aggression. Then the home owners' lobby points out the usefulness of dogs in guarding premises against prowlers. Balancing these factors, there emerges a statute making dog owners liable for bites inflicted except upon persons tormenting the dog or unlawfully on the owner's premises. The formula for liability now becomes: + ownership + biting - being tormented - unlawful presence on the premises.

So in any given situation, the law recognizes certain elements as material to the case, and the presence or absence of each of them is properly to be considered in deciding the case. Or, to rephrase in somewhat more involved language, rules of substantive law are "statements of the specific factual conditions upon which specific legal consequences depend... Rules of substantive law are conditional imperatives, having the form: If such and such and so and so, etc. is the case, and unless such and such or unless so and so, etc. is the case, then the defendant is liable....'' Now obviously the weighing and balancing required to determine what elements ought to be considered material cannot be accomplished by any of the methodologies of procedure. The result is purely a matter of substantive law, to be decided according to those imponderables which travel under the name of jurisprudence.

This view of the substantive law may seem unduly Euclidean, yet some system of analysis and classification is necessary if the law is to possess a measure of continuity and to be accessible and usable.


Under our adversary method of litigation a trial is essentially not an inquest or investigation but rather a demonstration conducted by the parties.

Since plaintiff is the party seeking to disturb the existing situation by inducing the court to take some measure in his favor, it seems reasonable to require him to demonstrate his right to relief. How extensive must this demonstration be? Should it include every substantive element, which either by its existence or nonexistence may condition his right to relief? If the answer is "yes,'' then plaintiff under our dog statute would be required to demonstrate each of the elements in the formula: + ownership + biting - tormenting - illegal presence on the premises.

In the ordinary dog case this would not be unduly burdensome, but if the suit is on a contract and we require plaintiff to establish the existence or nonexistence, as may be appropriate, of every concept treated in Corbin and Williston, then the responsibility of plaintiff becomes burdensome indeed and the lawsuit itself may include a large amount of unnecessary territory. Actually, of course, the responsibility for dealing with every element is not placed on plaintiff. Instead we settle for a "prima facie case'' or "cause of action,'' consisting of certain selected elements which are regarded as sufficient to entitle plaintiff to recover, if he proves them and unless defendant in turn establishes other elements which would offset them. Thus in a simple contract case, by establishing + offer + acceptance + consideration + breach, plaintiff is entitled to recover, unless defendant establishes + accord and satisfaction or + failure of consideration or + illegality or - capacity to contract, and so on.

Observe that the plus and minus signs change, in accord with proper mathematical rules, when we shift elements to the defendant's side of the equation as "defenses.'' For example, if plaintiff were required to deal with capacity to contract, it would become + capacity to contract as a part of his case, rather than the - capacity to contract of defendant's case.

Defenses, too, may be prima facie only and subject to being offset by further matters produced by plaintiff, as in the case of the defense of release, offset by the further fact of fraud in the inducement for the release. The entire process is the familiar confessing and avoiding of the common law.


The next step to be taken is the determination whether a particular material element is a part of plaintiff's prima facie case or a defense. Or, referring back to the statement that rules of substantive law are "conditional imperatives, having the form: If such and such and so and so, etc., is the case ... then the defendant is liable,'' should the element in question be listed as an if or as an unless?

In some types of situations, the test has been purely mechanical, with the mechanics in turn likely to be accidental and casual. Thus, in causes of action based on statute, if an exception appears in the enacting clause, i.e., the clause creating the right of action, then the party relying on the statute must show that the case is not within the exception; otherwise the responsibility for bringing the case within an exception falls upon the opposite party. The principle is widely recognized, but the vagaries of statutory draftsmanship detract largely from its certainty of application. Returning to our dogs, two statutes will serve as illustrations.

If any dog shall do any damage to either the body or property of any person, the owner ... shall be liable for such damage, unless such damage shall have been occasioned to the body or property of a person who, at the time such damage was sustained, was committing a trespass or other tort, or was teasing, tormenting or abusing such dog. Mass. Ann. Laws ch. 140, 155 (1950).

Every person owning or harboring a dog shall be liable to the party injured for all damages done by such dog; but no recovery shall be had for personal injuries to any person when they [sic] are upon the premises of the owner of the dog after night, or upon the owner's premises engaged in some unlawful act in the day time. Ky. Laws 1906, ch. 10, at 25, Ky. Stats. 1936, 68a-5.

The Massachusetts statute was construed as imposing on a two and one quarter year old plaintiff the burden of establishing that he was not teasing, tormenting or abusing the dog, while under the Kentucky statute a plaintiff was held to have stated a prima facie case by alleging only that he was bitten by a dog owned by defendant, leaving questions of presence on the premises at night or unlawful activities in the day time to be brought in as defenses. The difference in result can scarcely be regarded as calculated but is typical. Unfortunately, the statute which states in so many words the procedural effects of its terms is a rarity.

Exceptions in contracts receive similar treatment. If the words of promising are broad, followed by exceptions, the general disposition is to place on defendant the responsibility of invoking the exception. Of course, many of the cases involve insurance policies, with all that implies. In Munro, Brice & Co. v. War Risks Assn., during World War I one underwriter insured plaintiffs' ship against loss due to hostilities and another underwriter insured it against perils of the sea except consequences of hostilities. The ship was lost, and plaintiffs sued on both policies. The King's Bench Division held that as regards the first policy plaintiffs must show the loss to have been due to hostilities, but that under the second policy merely establishing the loss was sufficient, leaving it to the underwriter to bring in loss by hostilities as a defense. Since evidence of the cause of loss was wholly lacking, the loss fell on the second underwriter.

Julius Stone commented as follows: "Every qualification of a class can equally be stated without any change of meaning as an exception to a class not so qualified. Thus the proposition `All animals have four legs except gorillas,' and the proposition `all animals which are not gorillas have four legs,' are, so far as their meanings are concerned, identical....

"If the distinction between an element of the rule and an exception to it does not represent any distinction in meaning, it may still remain a valid distinction for legal purposes. In that case, however, it must turn upon something other than the meaning of the propositions involved. It may turn, for instance, merely upon their relative form or order.''

So in a few kinds of cases the answer to the question of allocation is found in the structure of a statute or contract, perhaps with some tenuous reference to intent, either of the legislature or of the contracting parties. But what of the great bulk of the cases, involving neither exception in a statute nor limitation upon words of promising? What general considerations should govern the allocation of responsibility for the elements of the case between the parties?

Precedent may settle the manner in a particular jurisdiction, but precedent as such does nothing for the inquiring mind. Thayer was of the view that questions of allocation were to be referred to the principles of pleading, or perhaps to analysis of the substantive law, and "one has no right to look to the law of evidence for a solution of such questions as these....'' Books about pleading, however, have not been numerous in recent years, except for the local practice works; and aside from a brief but provocative treatment by Judge Clark they offer slight assistance. The substantive law texts, when they deal with the matter at all, tend to describe results rather than reasons.

Despite Thayer's strictures, his descendants in the field of writing about evidence, by assuming to deal with problems of burden of proof as an aspect of the law of evidence, have found themselves inevitably enmeshed in the problems of allocation and have contributed most of the literature on the subject, although in an introductory and incidental fashion.

Before trying to establish some bench marks for allocation, let us note, though only for the purpose of rejecting them, two which are sometimes suggested. (a) That the burden is on the party having the affirmative; or, conversely stated, that a party is not required to prove a negative. This is no more than a play on words, since practically any proposition may be stated in either affirmative or negative form. Thus a plaintiff's exercise of ordinary care equals absence of contributory negligence, in the minority jurisdictions which place this element in plaintiff's case. In any event, the proposition seems simply not to be so. (b) That the burden is on the party to whose case the element is essential. This does no more than restate the question.

Actually the reported decisions involving problems of allocation rarely contain any satisfying disclosure of the ratio decidendi. Implicit, however, seem to be considerations of policy, fairness and probability. None affords a complete working rule. Much overlap is apparent, as sound policy implies not too great a departure from fairness, and probability may constitute an aspect of both policy and fairness. But despite the vagueness of their generality, it is possible to pour enough content into these concepts to give them some real meaning.


As Judge Clark remarks, "One who must bear the risk of getting the matter properly set before the court, if it is to be considered at all, has to that extent the dice loaded against him.'' While policy more obviously predominates at the stage of determining what elements are material, its influence may nevertheless extend into the stage of allocating those elements by way of favoring one or the other party to a particular kind of litigation. Thus a court which is willing to permit a recovery for negligence may still choose to exercise restraints by imposing on plaintiff the burden of freedom from contributory negligence, as a theoretical, though perhaps not a practical, handicap. Or the bringing of actions for defamation may in some measure be discouraged by allocating untruth to plaintiff as an element of his prima facie case, rather than by treating truth as an affirmative defense. And it must be apparent that a complete lack of proof as to a particular element moves allocation out of the class of a mere handicap and makes it decisive as to the element, and perhaps as to the case itself. In Summers v. Tice plaintiff was hunting with two defendants and was shot in the eye when both fired simultaneously at the same bird. The court placed on each defendant the burden of proving that his shot did not cause the injury. To discharge this burden was impossible, since each gun was loaded with identical shot. In Munro, Brice & Co. v. War Risks Assn. the absence of proof of the cause of the ship's loss meant that the party on whom that burden was cast lost the case. In these cases the admonition of Julius Stone is particularly apt: "the Courts should not essay the impossible task of making the bricks of judge-made law without handling the straw of policy.''


The nature of a particular element may indicate that evidence relating to it lies more within the control of one party, which suggests the fairness of allocating that element to him. Examples are payment, discharge in bankruptcy, and license, all of which are commonly treated as affirmative defenses. However, caution in making any extensive generalization is indicated by the classification of contributory negligence, illegality, and failure of consideration also as affirmative defenses, despite the fact that knowledge more probably lies with plaintiff. Certainly in the usual tort cases, knowledge of his own wrongdoing rests more intimately in defendant, though the accepted general pattern imposes this burden on plaintiff.


A further factor which seems to enter into many decisions as to allocation is a judicial, i.e., wholly nonstatistical, estimate of the probabilities of the situation, with the burden being put on the party who will be benefited by a departure from the supposed norm.

The probabilities may relate to the type of situation out of which the litigation arises or they may relate to the type of litigation itself. The standards are quite different and may produce differences in result. To illustrate: If it be assumed that most people pay their bills, the probabilities are that any bill selected at random has been paid; therefore, a plaintiff suing to collect a bill would be responsible for nonpayment as an element of his prima facie case. If, however, attention is limited to bills upon which suit is brought, a contrary conclusion is reached. Plaintiffs are not prone to sue for paid bills, and the probabilities are that the bill is unpaid. Hence payment would be an affirmative defense. Or again, "guest'' statutes prohibit nonpaying passengers from recovering for the negligence of the driver. If most passengers are nonpaying, then the element of compensation for the ride would belong in the prima facie case of the passenger-plaintiff. If, however, most passengers in the litigated cases ride for compensation, then absence of compensation would be an affirmative defense. In the payment-of-a-bill situation the probabilities are estimated with regard to the litigated situation, payment being regarded generally as an affirmative defense, while in the guest situation they are estimated with regard to such situations generally and not limited to those which are litigated, status as a nonguest being a part of plaintiff's prima facie case. No reason for the shift is apparent, and it may be unconscious. The litigated cases would seem to furnish the more appropriate basis for estimating probabilities.

Matters occurring after the accrual of the plaintiff's right are almost always placed in the category of affirmative defenses. Examples are payment, release, accord and satisfaction, discharge in bankruptcy, and limitations. A plausible explanation is that a condition once established is likely to continue; hence the burden ought to fall on the party benefited by a change.

In the cases of complete absence of proof, a proper application of the probability factor is calculated to produce a minimum of unjust results, and the same is true, though less impressively, even if proof is available.

Problem X-1

Civil Allocation in Idylia

In the past Idylia operated without an adversary system of justice. All civil disputes were brought to the Tribunal of Inquiry, which thoroughly investigated and reached factual conclusions about each and every element that the Idylia legislature specified as essential to recovery. In some cases the Tribunal of Inquiry would remain uncertain even after exhaustive investigation. These cases would be decided by coin flip. The system was enormously expensive and the results unsatisfying, particularly in the coin-flip cases.

You are commissioned to allocate the proof of elements between plaintiff and defendant in contract suits and to provide a coherent theoretical justification for your allocation.

The elements for breach of contract in Idylia are as follows: agreement, consideration, breach, and absence of mistake. Records of the Tribunal of Inquiry show that in the last year (consider it representative) there were 3,000 disputed contracts. Estimates are that this is approximately 5 percent of all contracts. Damages were awarded in 1,000 cases.

Number Number

of cases Number Number of cases

in which of cases of cases in which

the element in which in which Tribunal

was the Tribunal decision wound up

major decided was for uncertain

disputed for no and flipped

issue recovery recovery a coin

---------------- ------------ ------------- ----------------

Agreement 850 325 275 250

Consideration 400 200 150 50

Breach 1,700 440 970 290

Absense of mistake 50 35 5 10

---------------- ------------ ------------ ----------------

3,000 1,000 1,400 600(1)*

What do you recommend, and why?

1. 300 of these came out for the plaintiff.

div1.gif (1531 bytes)
Home | Contents | Topical Index | Syllabi | Search | Contact Us | Professors' Pages
Cases | Problems | Rules | Statutes | Articles | Commentary