Concept
What the Court Decides
Judge and jury roles in deciding fault
What the Court Decides
Section 328B of the Restatement (Second) of Torts gives the court four jobs in a negligence action. The court decides whether the evidence makes an issue on which a jury may reasonably find the facts to exist or not exist;1 whether those facts give rise to any legal duty on the defendant’s part;2 what standard of conduct that duty requires;3 and whether the defendant has conformed to that standard, in any case in which the jury may not reasonably come to a different conclusion.4
The four jobs sort into two kinds. Sufficiency of the evidence and conformity each run through a reasonable-minds gate: the court asks whether a jury could reasonably go either way on this record, and it takes the issue away only when the jury could not. Duty and the standard of conduct are legal calls. Each asks what the law requires of the defendant, and the court answers by applying legal rules, though it may leave the details of reasonable care for the jury to fill in. Sorting an issue into one kind or the other usually shows who decides it. If the answer turns on how reasonable people could view the record, the gate applies; if it turns on what the law demands, the court decides at the outset.
The subsections follow the order of § 328B.
Sufficiency of the Evidence
Section 328B(a) makes the court the first gatekeeper. Before the jury may find a disputed fact, the court decides whether the evidence makes an issue on which the jury may reasonably find that the fact exists or does not exist.5 The word that matters is reasonably. The judge asks what a rational jury could find on this record, whatever the judge personally believes happened. That test guards against two opposite failures. If every asserted fact reached the jury regardless of proof, jurors could find facts by speculation or sympathy. If the judge instead settled each contested fact by a personal reading of the evidence, the judge would take over the jury’s work. The reasonable-jury test removes facts that no rational jury could find and leaves everything else to the jurors.
In Anderson v. Liberty Lobby, Inc., the Supreme Court put the inquiry as whether reasonable jurors could find by a preponderance of the evidence that the plaintiff is entitled to a verdict. It borrowed an older formulation: whether there is evidence “upon which a jury can properly proceed to find a verdict for the party producing it, upon whom the onus of proof is imposed.”6 The burden of proof enters the test because a jury may find a fact only if the evidence could carry the standard that governs at trial, which in an ordinary civil case is a preponderance. The judge therefore measures the record against what a rational jury would need in order to find the fact. The Court added that this standard mirrors the standard for a directed verdict under Federal Rule of Civil Procedure 50(a): the judge must direct a verdict if, under the governing law, there can be but one reasonable conclusion as to the verdict.7
Because the question is what a jury could find, the judge reads the record from the standpoint of the party who wants the issue submitted. A jury inclined to find for that party could believe its witnesses and draw the inferences that favor it, so the judge tests the rationality of such a verdict by assuming the jury made those choices. A judge who credited the other side’s witnesses would already be deciding the facts. Boeing Co. v. Shipman states the corollary. A motion for judgment as a matter of law is not decided by which side has the better of the case, and it is not granted only when probative facts are completely absent. There must be a conflict in substantial evidence to create a jury question, and the jury, not the court, weighs conflicting evidence and inferences and determines the credibility of witnesses.8
The division of labor is easiest to remember as a pair of verbs. In Mahoney v. Beatman, a causation case, the court framed the judge’s task as a preliminary question: whether the evidence reasonably justifies submitting the issue to the jury. It then quoted a commentator’s statement of the split. The judge says whether the jury can reasonably find the fact, and, if the answer is yes, the jury says whether it does.9 The test cuts both ways. If reasonable people could disagree about the fact, the jury decides it for either side, and two juries on the same record might reach different results. If reasonable people could not disagree, the court decides as a matter of law, against the party whose evidence is too thin to make a jury question or for the party whose evidence admits of only one conclusion. A directed-verdict motion shows the symmetry: it assumes the opponent’s evidence is true and contends that the opponent still cannot recover, and either party may make it.10
Step three does the real work, and a hotel case shows evidence that fails it. A guest was cut by falling glass and later developed a skin cancer at the scar. One medical expert testified that trauma could cause such a cancer in about one case in a hundred; the other testified that trauma has no causal connection to cancer. The court held that the cancer should not have gone to the jury: “Possibilities will not sustain a verdict.”11 Even read in the guest’s favor, the testimony described a possibility, not a probability, and the cause of cancer was unknown even to the specialists, so jurors had no everyday experience from which to supply the missing link. A verdict could rest only on the inference that, because the cancer followed the cut, the cut caused it.
The claim of Timothy Daley fell on the other side of the line. The reviewing court examined the evidence offered for him, called the question a close one, and concluded that, viewed favorably, it presented facts from which a jury could reasonably find or infer a causal relation between the defendant’s alleged negligence and his alleged injuries. It ordered a new trial so that he could try to prove his claim.12 The two cases differ in what the plaintiff’s own proof said about causation. In the hotel case it described a one-in-a-hundred possibility and met an expert’s denial of any connection, so no rational jury could find causation by a preponderance. In Daley the evidence supported a reasonable inference, and the court treated the closeness of the question as a reason to let a jury decide it.
When a trial court gets this ruling wrong, an appellate court can correct it. If the trial court wrongly set aside a jury verdict, the appellate court can reverse and enter judgment for the party in whose favor the jury ruled, with no need for a new trial.13 Later sections take up the motions through which these rulings are made.
This job concerns facts that are disputed or thinly proved. Section 328B(d), covered below, concerns a different situation: the facts are settled, and the only question is whether they permit more than one rational conclusion about the defendant’s care.
Existence of a Duty
Section 328B(b) gives the court the second job: deciding whether the facts give rise to any legal duty on the defendant’s part.15 A duty, as the negligence cause of action uses the word, is an obligation the law recognizes that requires the actor to conform to a certain standard of conduct for the protection of others against unreasonable risks of harm.16 The court answers whether the law imposes that obligation on a defendant in the plaintiff’s position. This is the first of the two legal calls, and it carries no reasonable-minds gate. The court asks what the law requires of this kind of defendant toward this kind of plaintiff, whatever a jury might make of the evidence.
Amaya v. Home Ice, Fuel & Supply Co. states the rule plainly: the existence of a duty owed by the defendant to the plaintiff is “in the first instance for the court, not for the jury.”17 The California Supreme Court relied on an earlier New Jersey decision for the reason. Before liability can attach, the court must determine the character of the duty the law imposes on the defendant under the facts, because it cannot be conceded that “the jury from their inner consciousness may evolve in every variety of tortfeasance a legal duty as the standard of liability.”18 The mechanism is easy to see. Duty answers a policy question: should the law impose any obligation of care on this kind of defendant toward this kind of plaintiff? Policy has to be uniform across cases, because it guides conduct before anyone is hurt, and it has to be reviewable on appeal. If each jury invented its own duty, two identical defendants in identical relationships could owe different obligations, and an appellate court would have no stated rule to examine, since a general verdict announces none. With duty settled, the jurors can be given what they need for meaningful assistance: a definition, so far as relevant, of the ultimate factual conclusions upon which, if found, the law gives the plaintiff a right to recover or the defendant a right of exoneration.19
Foreseeability is the formula most likely to tempt a student into sending duty to the jury. The argument runs that duty is often framed in terms of foreseeability, that reasonable minds could differ about whether the injury here was foreseeable, and that duty is therefore a fact question. Amaya answers that even where it would be difficult to say that reasonable minds could not differ about whether a duty should be imposed, the question is one of law for the court.20 So the gate that sends sufficiency and conformity issues to the jury whenever reasonable people could differ does not govern duty. Foreseeability measures how likely a harm is; duty asks whether the law should attach an obligation to that likelihood. A risk can be foreseeable and the law still impose no duty, and a fact pattern that mentions foreseeability does not turn duty into a jury question.
Palsgraf v. Long Island Railroad Co. shows the court at work on a duty question. A package that looked like any other held fireworks, and nothing in its appearance gave notice of its contents. It fell and exploded, and the shock threw down scales at the other end of the platform many feet away, where they struck the plaintiff.21 Cardozo held for the railroad, reasoning that the guard’s conduct, whatever it was toward the package’s holder, was no wrong toward a plaintiff standing so far away. He added that the range of reasonable apprehension is at times a question for the court and, if varying inferences are possible, at times a question for the jury.22 That remark marks the point where the duty inquiry shades into the reasonable-minds gate. Where the record leaves no room for varying inferences about whom the conduct endangered, the court settles the relation between risk and plaintiff itself.
Duty must also stay distinct from breach, because the jury’s later role depends on the separation. A separate opinion in Castaneda v. Olsher draws the line this way. When the court decides duty, it does not ask whether this plaintiff’s injury was reasonably foreseeable in light of this defendant’s conduct. It asks, more generally, whether the category of negligent conduct is sufficiently likely to result in the kind of harm experienced that liability may appropriately be imposed. The jury, by contrast, considers foreseeability in two fact-specific settings: in deciding whether the particular defendant’s conduct was negligent in the first place, and in deciding whether the negligence was a proximate cause of the injury. Whether the defendant’s specific acts satisfied the standard, which is to say whether there was a breach of the duty owed, is a question of fact for the jury.23 Jurisdictions differ on whether foreseeability belongs in the duty analysis at all; the Restatement (Third) rejects any role for it there.24 On either approach, the court decides whether a duty exists.
Setting the Standard of Conduct
Section 328B(c) gives the court the third job: determining the standard of conduct that the defendant’s legal duty requires.26 Duty establishes that the defendant owed the plaintiff an obligation of care, and breach asks whether the defendant met it. The standard of conduct is the yardstick between the two, the measure of how careful the defendant had to be. Like duty, it is a legal call, and in most cases the court’s answer is brief.
Courts usually impose a duty without prescribing a rigid, specific, inflexible course of conduct. In most cases they fix no standard of care more precise than that of a reasonably prudent person under like circumstances.27 The jury then receives a short general definition of negligence and of the reasonable-person test, because the attributes of the reasonable person, and the factors that make a risk unreasonable, are matters of common sense and community values that jurors bring to the evidence.28 What a careful owner does about ice on a walk, or a careful driver at an intersection, is worked out case by case. A concurring opinion in Toschi v. Christian described the resulting division in two steps. If the court has formulated a standard of reasonable conduct that applies to the case, the jury’s sole concern is whether the defendant’s conduct met it, a question of fact. If the court has not, the jury must first formulate the standard and then decide whether the conduct met it.29 The court’s choice of standard and the jury’s application of it are therefore separate decisions, even though a general instruction lets the jury make both. The later subsection on judging breach takes up the application step.
A court sometimes replaces the general formula with a more specific measure, and a statute is the familiar source. In Clinkscales v. Carver, a boulevard-stop case, Justice Traynor explained that a statute providing only a criminal proceeding does not itself create civil liability. The civil suit rests on the common-law tort of failing to use a reasonable person’s care at the stop, and the statute matters because the court adopts its formulation of conduct as the standard for civil liability. The adoption remains a judicial act. When a legislative body has generalized a standard from the community’s experience and prohibits conduct likely to cause harm, the court accepts the formulated standard and applies it.30 Where no such standard exists, the case goes to a jury that must decide not only what the facts are but also what the unformulated standard of reasonable conduct is.31 The rules for when violating a statute establishes negligence belong to the separate treatment of negligence per se; here the statute matters only as one source of a court-adopted standard.
Courts also modify the reasonable-person standard for particular kinds of defendants. Most jurisdictions measure a child by what it is reasonable to expect of children of like age, intelligence, and experience.32 The court decides which standard governs, and the jury applies it to the facts. A blind pedestrian shows the same adjustment running in the other direction: one court said that a blind person may not rely wholly on the other senses to warn of danger but must use the devices usually employed to compensate for blindness.33 Beyond these adjustments, courts will sometimes declare that certain conduct in certain situations is negligence as a matter of law, is presumptively negligent, or is evidence of negligence, and the jury is instructed accordingly.34 Each device narrows what the jury has left to decide, and a declaration of that kind overlaps with the conformity ruling taken up next.
The ordinance did the standard-setting work in advance, since it records from experience how much protection a stairway needs. The lighting dispute has no such source, so the details of care are left to jurors, whose varied backgrounds are the best available proxy for what careful owners in the community actually do. The first tenant’s case reaches the jury with a fixed yardstick, and the second reaches it with a general one.
A court that leaves the standard general thereby delegates to the jury the job of formulating it. The concurrence in Toschi gives the choice between fixing a standard and leaving it to the jury to the appellate courts and ties it to one question: whether the jury would be at an advantage in arriving at a standard by virtue of being a cross-section of the community, and therefore representative of community views and attitudes.35 The test has a mechanism. Where the issue is how much care ordinary life calls for, community experience supplies the input, and a cross-section of the community supplies that experience. Where a legislature or a line of decisions has already distilled the experience into a rule, the jury offers no advantage over the rule.
Two arguments compete over how that question should come out. The case for fixed rules is consistency in recurring situations, a position associated with Holmes: when courts see the same facts again and again, they should settle the question once as a matter of law instead of leaving each jury to decide it. One source observes that this position is most satisfactory in cases that constantly recur on precisely the same facts, and that modern mass-tort and class-action cases, in which one product or behavior is alleged to have injured many people, raise recurring questions of just that kind.36 The limit appears in Andrews v. United Airlines, Inc., where the court ruled that whether the airline breached its duty to maintain safe overhead bins was a question for the jury, even though the bins are identical on every aircraft of that model and the question is likely to recur.37 Recurrence is a reason courts weigh, and it does not by itself produce a fixed rule.
The case for leaving the standard general rests on the variety of human circumstances. A dissent in Kentucky Fried Chicken of California, Inc. v. Superior Court argued that judging the reasonableness of conduct case by case gives a more precise measure of liability than a categorical rule. It quoted the observation that “[s]tandards of care are typically relative; rules of law are basically absolute,” so that forcing conduct into absolute legal classifications is impracticable whenever the circumstances admit of materially conflicting inferences.38 A fixed rule is written before the facts arrive, and the next case may differ in just the way the rule did not anticipate.
The two positions answer one trade-off in opposite directions. A fixed rule buys predictability and uniform results, at the price of fit, because no rule written in advance can anticipate every variation. A general standard buys fit, at the price of consistency, because different juries may land in different places on the same conduct. A court fixes a rule when legislative or judicial experience has already generalized the community’s judgment, as with a boulevard-stop statute, and leaves the general standard to the jury when no such experience exists. Either way, the lasting division of labor is the one a separate opinion in Castaneda v. Olsher states: the court addresses the category of conduct, and whether the specific defendant’s acts satisfied the standard is a question of fact for the jury.39
Conformity as a Matter of Law
Section 328B(d) gives the court the last job: deciding whether the defendant has conformed to the standard of conduct, in any case in which the jury may not reasonably come to a different conclusion.41 To conform is to be as careful as the standard requires, so this job reaches the breach question that the jury ordinarily decides. The closing clause is the condition that moves the question from one decision-maker to the other. Where reasonable jurors could answer either way, the jury decides whether the defendant was careful enough; where they could not, the court decides. This is the second reasonable-minds gate, and it sits at the end of the chain, after the facts have been settled and the standard set.
The relationship to § 328B(a) needs to be exact, because the two jobs apply the same gate to different questions. Under (a), the dispute concerns what happened, and the court asks whether the evidence lets a rational jury find that a contested fact exists. Under (d), the facts are established or the inferences from them are clear, and the court asks something else: taking those facts as they stand, is there only one rational conclusion about whether the defendant used reasonable care? One case can call for both rulings, in that order, because the court cannot decide what the facts add up to until it knows which facts a jury could find.
The court holds this power because a jury’s discretion has limits set by reason. A jury’s job is to choose among conclusions the evidence supports. When the evidence supports only one, a verdict for the other side cannot be a judgment about the evidence; it can only reflect sympathy or guesswork. A judge may rely on experience and legal knowledge to conclude that conduct so clearly was or was not negligent that no reasonable person could reach another conclusion, and much of the detailed law on the reasonable person comes from appellate decisions reviewing rulings of that kind.42 In the law’s vocabulary, a question that has only one right answer is a question of law for the judge.43
A supermarket spill illustrates the ruling. A customer reaching for an item in a busy store knocks over a bottle of cooking oil, which breaks and spills, and the next shopper slips on the oil within a minute and a half. A sympathetic jury might call the store unreasonable for not cleaning up and find it negligent. But if the timing is not in dispute, no fact remains for the jury to find, and a judge may grant the store’s renewed motion for judgment as a matter of law on the ground that it would be unreasonable to require anyone to clean up a spill of liquid in a supermarket aisle within ninety seconds.44 The ruling leaves the store’s general obligation to keep its aisles reasonably safe untouched. It concerns only whether this much elapsed time could amount to a failure of care.
Cases involving children show that a lowered standard does not remove the court’s power. Once the court has settled that a child is measured against children of like age, intelligence, and experience, the jury ordinarily decides whether the child met that standard. If the only conclusion the evidence permits is that the child behaved unreasonably in view of the child’s estimated capacity, however, the judge may direct a verdict against the child, as for an adult. Norfolk & Portsmouth Belt Line R.R. Co. v. Barker, in which a ten-year-old was injured while trying to jump onto a moving train, illustrates that rule.45 The jobs under (c) and (d) therefore work in sequence. The first lowers the yardstick for a child, and the second asks whether, measured against it, a jury could reasonably find the child careful.
A ruling for either side is an exception to a strong default. A dissent in Kentucky Fried Chicken, quoting earlier California decisions, described negligence as a question of fact for the jury, an inference drawn from the circumstances of each case, which the court may withdraw only where the deduction to be drawn is inevitably that of negligence.46 The mechanism is that every change in the surrounding circumstances shifts the point at which care becomes reasonable, so a conclusion that fits one record seldom fits the next, and a court that ruled often would put its own judgment in place of the community’s. The dissent’s word inevitably and the Restatement’s phrase “may not reasonably come to a different conclusion” describe one narrow test, and it applies to a finding of conformity and to a finding of nonconformity alike.
A dispute over a drug label shows the power acknowledged and then declined. In Macdonald v. Ortho Pharmaceutical Corp., the label for an oral contraceptive warned of abnormal blood clotting that can be fatal and of death from clotting in vital organs such as the brain, but it never used the word stroke. The plaintiff suffered permanent disablement, the jury found the warning inadequate, and the trial judge entered judgment notwithstanding the verdict for the manufacturer. The reviewing court acknowledged that a court may decide conformity as a matter of law, but it observed that judicial intrusion into jury decision-making in negligence cases is exceedingly rare, and that few questions are more appropriately left to lay common sense than whether a written warning gets its message across to an average person. Reading the evidence most favorably to the plaintiffs, the court could not say that the jury’s decision was so unreasonable as to require the opposite conclusion as a matter of law. Jurors could reasonably have found that omitting any reference to stroke minimized the warning. The court reversed the judgment and remanded for entry of judgment for the plaintiffs.47
What separates Macdonald from the supermarket spill is the shape of the record. The spill facts left one rational answer about the interval. The label’s undisputed text still left room for two rational views about what an average consumer would take from it, because adequacy turns on comprehension, and on comprehension jurors’ own experience counts. Undisputed facts do not by themselves justify a ruling as a matter of law; the real question is whether they permit only one rational conclusion about care. A judge who would have decided differently from the jury has not thereby shown that the jury could not reasonably have decided as it did.