Concept
Magnitude of the Risk
Risk-utility factors in judging unreasonable conduct
Magnitude of the Risk
Section 293 of the Restatement (Second) of Torts supplies the weight that goes on the risk side of the § 291 balance.1 Where § 292 measured the credit the conduct earns, § 293 measures what is set against that credit. It lists four factors: the social value the law attaches to the interests imperiled;2 the extent of the chance that the conduct will invade an interest of the plaintiff, or of a class to which the plaintiff belongs;3 the extent of the harm likely to be caused to the interests imperiled;4 and the number of persons whose interests are likely to be invaded if the risk takes effect in harm.5
The list exists because probability alone cannot say how serious a risk is. A one-in-a-thousand chance of scratching a fence and a one-in-a-thousand chance of killing a bystander carry the same probability and very different weights. Likelihood and gravity, in combination, fix the magnitude of the risk, sometimes called the “hazard” or “danger.”6 Each factor in § 293 answers a question the bare percentage leaves open: what is endangered, how likely the invasion is, how bad the harm would be, and how many people would suffer it.
The four factors act on one another. A modest chance of damage to one person's property may be a light risk, while the same chance of death to a crowd is a heavy one, and a large chance of a trivial inconvenience may weigh less than a remote chance of catastrophe. The court merges the factors into one judgment of how large the risk is, then sets that judgment against the utility credit built from § 292. Each factor is taught separately below, so its contribution is clear before the combined judgment is made, and a later section shows how the combination can be written as a formula.
Value of the Interests Imperiled
The first factor, § 293(a), looks at what the conduct puts at stake: the social value the law attaches to the interests imperiled.7 Section 292 asked what the conduct is for, and this factor asks what it endangers. A risk weighs only as much as the thing it threatens, so the court first names the endangered interest, whether a person's life and bodily security, a neighbor's fence, or something in between, and then asks how much the law cares to protect it. The same chance of harm weighs more when the interest is one the law prizes.
The law weights interests unequally because equal weights would let an actor trade one person's safety for another's convenience whenever the odds happened to match. If a one-in-twenty chance of killing a bystander counted the same as a one-in-twenty chance of denting a fence, the justification an actor needed would not depend on what was endangered, and a human life and a fence post would be interchangeable. Weighting by social value makes the required justification rise with the importance of the interest, so a threat to life needs a far stronger utility credit than a threat to a fence. The gravity of a potential harm thus has two parts, the extent of the damage and the relative societal value of the interest harmed, and one familiar statement of the second part is that people are more valuable than property.8
Section 292(a) used the same idea of social value on the other side of the scale. A weighty interest the conduct advances enlarges the utility credit, and a weighty interest the conduct threatens enlarges the risk. One case can place interests on both sides. In Eckert v. Long Island R.R., a rescuer ran onto a railroad track, risking his own life, a very weighty interest on the risk side; the value of the child's life he sought to save, and the necessity of the rescue, counterbalanced that danger.9 The child's life belongs in the utility credit and the rescuer's own life in the magnitude of the risk, and students often file an interest on the wrong side.
Only the imperiled interest moved the answer between Okafor's two sites, which is how to isolate this factor on an exam. The general account of unreasonable risk draws the same line: under some circumstances it may be reasonable to endanger property where it would be unreasonable, under the same circumstances, to endanger bodily security.10 The lighter risk at the second site is still a risk, and a shed worth a great deal, or work that earned almost no utility credit, could make even that drop unreasonable.
The ranking of interests is a weight in the balance and not a rule of decision. Life generally outranks property, and a risk to life taken to save property has often been held unreasonable, which has produced statements that such a risk is always unreasonable. No general rule forbids it, though, and courts have held it reasonable in the circumstances of other cases.11 The reason lies in where § 293(a) sits among the factors: the value of the endangered interest combines with the chance of invasion, the extent of the harm, and the number of persons exposed, and the sum is set against the utility credit. A remote chance of harm to a highly valued interest can still be reasonable when the conduct earns a large credit, and a near certainty of harm to a modest interest can still be unreasonable when the conduct earns none. The language of balancing interests names factors a court may consider and implies no mathematical correspondence between the two sides.12 With the weight of the endangered interest in hand, § 293(b) asks how likely it is that the conduct will invade that interest.
Chance of Invasion
The second factor, § 293(b), asks how likely it is that the stake will actually be lost: the extent of the chance that the conduct will invade an interest of the plaintiff, or of a class to which the plaintiff belongs.14 The section speaks of the extent of the chance, so probability is a matter of degree, weighed with the other factors. It sets no minimum below which a chance stops counting and no level at which the chance becomes decisive. The probability that counts is the one foreseeable when the actor chose to act, and the accident that later happened does not fix it.15
Probability enters as a weight and not as a threshold because a risk combines likelihood with consequence, and a cutoff on likelihood would erase the consequence. If a chance counted only once harm became more likely than not, an actor could run a 20 percent chance of killing a bystander on the ground that the odds favored the bystander, and repeat the gamble every day. That result would hollow out the balance in § 291, which sets risk against utility. Students most often go wrong by importing the preponderance standard, which tells a factfinder how convincing the evidence of past events must be. The likelihood of harm at the moment of acting is a separate question.
Tullgren v. Amoskeag Mfg. Co. states the point directly: “The test is not of balance of probabilities, but of the existence of some probability of sufficient moment to induce action to avoid it on the part of a reasonable mind.”16 The court gave everyday illustrations. One who crosses a railroad track may not reasonably anticipate that a train will be met, but because one may be coming, he must do what is reasonably required to find out. A driver rounding a sharp turn with an obstructed view will probably meet no opposing traffic, yet if the chance is great enough that ordinary people would drive differently, failing to do so is careless.17
A comparison of two sidewalks isolates this factor. Failing to repair a hole in a rarely used sidewalk might not be negligent, while the same failure on a heavily traveled sidewalk would be.18 The hole has the same depth and the same capacity to break an ankle in both places, so the gravity of the harm is constant and the only variable is how likely it is that someone will step into it. Where hundreds of people pass every hour, a fall is close to certain within a few weeks, and an owner who knows of the hole accepts that result. Where one person passes each day, a fall is a remote possibility. Heavy use also puts more people in the zone of danger, but the number exposed is its own factor under § 293(d), and the point here is only that more traffic raises the chance that the hole will injure someone.
A low chance can coexist with unreasonable conduct when the harm would be grave. A pedestrian who dashes across a little-traveled highway without looking is unlikely to meet a fast-moving car, yet the low probability does not rescue the conduct, because the potential harm is so great.19 Even recklessness, which requires the actor's consciousness of the risk, calls for something far less than certainty or even probability, and where the conduct has no social utility a chance below 1 percent may be enough.20 The rule also runs the other way. A risk is not necessarily unreasonable because harm is more likely than not to follow, nor reasonable because the chances are against it, so a very large risk may be reasonable in some circumstances and a small risk unreasonable in others.21 A surgeon who operates on a patient who will otherwise die runs a large chance of failure, and the value of the operation explains why that risk is reasonable.
A plaintiff can show the chance of harm in several ways. Evidence that similar injuries occurred earlier, and that the actor knew of them, can establish both the risk and its foreseeability. In one design case the plaintiff relied on eleven choking deaths of small children from small, brightly colored blocks in a single year. That history does not by itself show a high probability of harm, but it shows a real risk where slightly larger blocks could easily have avoided it.22
Section 293(b) also protects plaintiffs whom the actor could not have identified in advance. It speaks of an interest “of the other or of one of a class of which the other is a member.”23
The actor therefore needs no advance knowledge of the victim's identity, only a foreseeable chance of invading the interests of people in the plaintiff's position.
Extent of Likely Harm
The third factor, § 293(c), asks how bad the invasion would be if it happened: the extent of the harm likely to be caused to the interests imperiled.25 The chance and the harm vary independently, since the same one-in-a-hundred chance can attach to a bruise or to a death. Section 293(a) also looks at the interest, but it asks which interest is endangered, while § 293(c) asks how much of that interest the harm would destroy. Bodily security is a single interest, yet slight injury and great bodily injury or death sit at opposite ends of it. It may be reasonable to expose people to a danger of slight injury where it would be unreasonable to expose them to a danger of great bodily injury or death.26
Severity enters the balance because a risk is, in effect, a chance of a loss, and the size of the loss decides how much the chance is worth avoiding. Precautions cost something, so how much to spend depends on what they would prevent. A rule that looked only at the odds would treat a scratch and a death alike at the same probability. It would push actors to spend heavily against trivial harms, or let them take comfort in long odds when the worst case is catastrophic. Weighing the extent of harm lets the amount of care follow the stakes.
When two actors take the same odds for the same small gain, the worst case separates them.
Probability and usefulness matched for the two actors, so only the worst case set them apart. A useful purpose does not cancel a severe risk either. A railroad employee who believes it is Sunday, when no trains run, and pulls up track on Saturday to repair it, with the result that the Saturday Flyer is wrecked, creates an unreasonable risk despite the social usefulness of repairing track. The risk was one of death and serious injury, not simply of slight injury or property damage.27 The repair earns a real credit under § 292, but a credit is a weight on one side of the scale and can be outweighed. The illustration also stresses that the danger fell on many persons, the separate factor in § 293(d) taken up next, and severity alone suffices to show why the repair credit could not carry the day.
Because the extent of likely harm sets how heavy a risk is, it also sets how much care a reasonable person must take, although the standard itself stays the same. The type and amount of care that reason requires are generally a function of the type, likelihood, and degree of harm, together viewed as the magnitude of the risk, that precautions of a particular cost may be expected to prevent.28 In the product setting, a great risk normally calls for great precautions, and a small risk ordinarily calls for only small ones.29 A defect that could cause steering or brake failure at highway speed demands the utmost care, while a flaw that could at worst cause a minor bruise warrants modest attention and expense. A later section on applying the balance develops this proportionality.
Two cautions keep the factor in proportion. Severity is weighed with the chance of invasion, the number of persons exposed, and the utility of the conduct, so a catastrophic worst case does not make conduct negligent automatically, and a remote chance of grave harm, taken for a weighty purpose with no safer route, can still come out reasonable. The reverse also holds: a fairly high chance of slight harm may be tolerable, as on a crowded sidewalk where people foreseeably jostle one another and now and then suffer a minor injury.30 The usual error is to reason that conduct is negligent because someone could be killed, when the question is how much the possibility of a death or a bruise adds to the weight of this particular risk.
Number of Persons Exposed
The fourth factor, § 293(d), counts the persons whose interests are likely to be invaded if the risk takes effect in harm.32 Section 293(b) asked how likely it is that the conduct will invade the interests of the plaintiff or of a class to which the plaintiff belongs, and this factor asks how many members of that class would suffer if the risk came to pass. A hazard that could strike one person weighs less, other things equal, than a hazard that could strike a crowd.
The reason is arithmetic. The harm a risk can be expected to cause is the sum of what it threatens to each person exposed, so the same chance and the same severity applied to a hundred people carry roughly a hundred times the weight they carry for one.
The classic illustration is again the railroad employee who, on a Saturday, wrongly believes it is Sunday, when no trains run, and pulls up the track to repair it just in time to wreck the Saturday Flyer. His conduct creates an unreasonable risk despite the social usefulness of repairing track, because the risk is one of death and serious injury and, more than that, a risk to many persons rather than to one or a few.33 The same source states the general proposition directly: it may not be unreasonable to endanger a single person though it would be unreasonable to endanger a great many.34 Two factors work together in the train wreck, since the severity of the harm comes from § 293(c) and the number of people exposed comes from § 293(d). The repair earns a real utility credit under § 292, but a credit is a weight on one side of the scale, and a risk of mass death outweighs it.
Heavy foot traffic in the concourse does two things at once. It raises the chance that someone will trip, which is the § 293(b) factor, and it raises the number of people who stand to be hurt, which is the § 293(d) factor. A careful analysis names each effect, because a risk to a crowd can be heavier even when the chance of harm to any one person is small.
Numbers multiply the other factors and do not replace them: a crowd exposed to a remote chance of trivial harm still bears a light risk. The value of the interests imperiled, the chance of invasion, the extent of the harm, and the number of persons exposed combine into a single judgment about how large the risk is, and that judgment is then set against the utility credit built from § 292. The next section shows how that comparison can be written as a compact formula.