Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, ECONOMICS AND FINANCE ( (c) Oxford University Press USA, 2019. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 18 October 2019

The Positive Economic Theory of Tort Law

Summary and Keywords

Tort law is part of the common law that originated in England after the Norman Conquest and spread throughout the world, including to the United States. It is judge-made law that allows people who have been injured by others to sue those who harmed them and collect damages in proper cases. Since its early origins, tort law has evolved considerably and has become a full-fledged “grown order,” like the economy, and can best be understood by positive theory, also like the economy. Economic theories of tort have developed since the early 1970s, and they too have evolved over time. Their objective is to generate fresh insight about the purposes and the workings of the tort system.

The basic thesis of the economic theory is that tort law creates incentives for people to minimize social cost, which is comprised of the harm produced by torts and the cost of the precautions necessary to prevent torts. This thesis, intentionally simple, generates many fresh insights about the workings and effects of the tort system and even about the actual legal rules that judges have developed. In an evolved grown order, legal rules are far less concrete than most people would expect though often very clear in application. Beginning also in the 1970s, legal philosophers have objected to the economic theory of tort and have devised philosophical theories that compete. The competition, moreover, has been productive because it has spurred both sides to revise and improve their theories and to seek better to understand the law. Tort law is diverse, applicable to many different activities and situations, so developing a positive theory about it is both challenging and rewarding.

Keywords: liability, accidents, torts, damages, accident law, negligence law, intentional torts, strict liability, law and economics, nondurable precaution

History and Methods

“Tort” is the French word for “wrong.” It describes a body of common law that allows accident victims and those suffering from intentional wrongs to sue their injurers for damages. English judges began to develop tort law soon after the Norman Conquest when English judges still spoke French. During this early time, most torts were intentional because “accidents” were not a large problem when technology was simple. The most basic function of negligence law, the largest part of modern tort law, is to promote safety in the use of technology, and a negligent train wreck possessed no analog in the 14th century. In fact, negligence cases went from a tiny trickle in early times to a great flood in the 19th century as the Industrial Revolution improved technology and therefore increased the number of accidents and of negligence cases.1

Many believe that common law, including tort law, is somewhere written down, and that judges overstep when they decide cases without paying attention to written rules. Statutes are written down, and judges must follow them, but common law is entirely a creature of judicial decisions. A body of common law evolves over the centuries from the case decisions of generations of judges. Indeed, the common law was one of Friedrich Hayek’s prime examples of a “grown order” as distinguished from a “made order.” (Hayek, 1979; Leoni, 1991). A grown order, such as the economy, the common law, and the English language, possesses orderly features, but these have arisen from evolution as opposed to someone’s or some committee’s plan or design. A made order is the opposite. The orderly features of a building or a public park come from an architect’s plan. A good way to explain a made order is to find out what the architect or other designer intended to achieve, and why. This method of explanation is impossible for grown orders for the simple reason that no deliberate design exists, only an evolution. Thus, the purposes of the common law, including tort law, are hotly contested by scholars and hardly mentioned by judges (Grady, 1994b).

Given that tort law is a grown order, its rules and purposes can only be known through positive theory. On the level of rules, judges and lawyers examine the pattern of case results—liability or no liability—and seek to “gloss” or describe the legal rule that will help predict the results of future cases (Grady, 1994b, 1995). Thus, lawyers and judges will have “legal theories” of the rule for which a line of cases stands (Brewer, 1996; Gennaioli & Shleifer, 2007; Grady, 1994b, 1995; Llewellyn, 2008; Wambaugh, 1892).2 Oliver Wendell Holmes (1881), the first important American scholar of the tort system, celebrated that, with the common law, the case comes first and the gloss or rule only later.3

On a second level, the social purposes of tort law can also be examined only through the lens of “positive theory.” A positive theory gives a parsimonious account of what a grown order accomplishes and how it works. Especially when scholars first begin, different positive theories are possible and compete for acceptance. The competition between economic and philosophical theories of tort law has been especially productive of new and expanding knowledge. Karl Popper (2002a, 2002b) argued that a successful positive theory should have certain attributes. First, it should be “parsimonious,” explaining much with little. Second, it should be “falsifiable,” that is, a theory capable of being proved false. A successful positive theory will not be a mere tautology or “truism” that cannot be proved false. The second criterion, moreover, is related to the first. The simpler a theory, the more ways in which it can be proved false and paradoxically the greater the theory’s “truth value.”4 A faulty theorist complicates with so many assertions and qualifications that his or her theory becomes “immunized” against falsification, perhaps intentionally so. Of course, even a simple theory can be an unfalsifiable tautology, and this may be true of a leading philosophical theory of tort.5 Finally, a successful theory should provide fresh insight about its subject matter. Although this criterion is the most subjective, it is also the most important. Does the grown order in question possess surprising features that are better explained by one theory instead of another? Unless a theory generates fresh insights about its subject, what use is it?6

Modern torts scholars have now provided three positive theories: an economic theory and two philosophical theories. The economic theory basically says that tort law can be understood as a system of sanctions and penalties that minimizes the social cost of the behavior the law defines as torts (Landes & Posner, 1981, 1987; Posner, 1972). This social cost is comprised of two main components: the cost torts impose on victims and the cost incurred by injurers and victims to prevent torts and their harmful effects.

Probably the most important philosophical theory of tort is that of Ernest Weinrib (1983, 1987b, 1995). He claimed that intentional torts and negligence create a relationship of wrongdoing running from the injurer to the victim and which could be redressed and repaired by a court ordering the injurer to pay the victim damages. Tort is thus a system of “corrective justice” according to Aristotle’s original conception (Posner, 1981a). Weinrib excluded other explanations of tort, especially the economic theory, because these violated the philosopher Immanuel Kant’s ideas of morality and more specifically allowed injurers to treat victims not as “ends” in and of themselves, as Kant’s philosophy requires, but as mere “means” for accomplishing the injurers’ wrongful purposes (Byrd, 1989; Weinrib, 1983). Because of problems with this corrective justice theory, which will be briefly discussed, a second generation of legal philosophers has developed what they call the “civil recourse” theory, which more modestly claims that torts are “legal wrongs,” the result of “wrongdoing,” or, in some cases, simply legal constructs useful to induce correct behavior (Coleman, 2002; Goldberg & Zipursky, 2010, 2013; Hershovitz, 2010; Keating, 2012; Wells, 1990; Zipursky, 2003).

To summarize the basic points to be developed soon in more detail, Weinrib’s corrective justice theory is indeed a falsifiable theory of tort, but seems false instead of true. The newer philosophical “civil recourse theory” seems to be an unfalsifiable tautology and defective for that reason. Although Weinrib’s theory may be a false description, especially of negligence law, it is at the same time more impressive than the newer civil recourse theories because of its careful attention to Kantian ethics. In any event, the philosophical theories have been an important spur to knowledge because without them economists might have failed to notice and address important problems and issues of tort law. A consideration of tort law through the lens of these three positive theories, economic and philosophical, presents a useful framework for this summary because it recapitulates modern scholarly debates and shows the motivation of modern theoretical developments.

Tort law is comprised of three parts (Grey, 2001; Holmes, 1873). The first is intentional torts; the second, negligence; and the third part, strict or absolute liability.

Intentional Torts

The most basic intentional torts are truly ancient, coming from the common law that emerged immediately after the Norman Conquest in 1066.7 These are the modern torts that evolved from the writ of trespass vi et armis, literally trespass “with force and arms” (Prosser, 1971). There are six of these intentional torts: trespass to land; battery; assault; false imprisonment; trespass to chattels (damaging moveable property); and trespass de bonis asportatis (carrying away moveable property). At least one of these torts (trespass to land) is usually conceived to be a strict liability tort, and the others require varying degrees of fault (Dorfman & Assaf, 2015; Hylton, 2010a, 2016; Kelley, 2003). To each of these six torts a common set of defenses or “privileges” apply. These include self-defense; defense of property; private necessity; and public necessity. Thus, someone accused of battery, which is the tort of intentionally striking someone, might defend on the ground of self-defense, as could someone accused of trespass to land if he could show he was compelled to go onto the plaintiff’s land to hide from the violence of a third party who was seeking to maim him.

Economists have focused less on intentional torts, as compared with negligence and strict liability, but their positive theory of them has become clear enough. Intentional torts define behavior that is prima facie inefficient (Ellis, 1982; Fennell, 2006).8 By “inefficient” legal economists refer to Kaldor–Hicks inefficiency, which means that the cost borne by the victim from some conduct is greater than the benefit derived by the injurer.9 The reference point for these calculations is a system of property rights (Demsetz, 1967; Posner, 1980, 1981b).

Calabresi and Melamed (1972) and Richard Posner (1980, 1981b) developed the idea of “market bypass” to explain intentional torts and their privileges. Posner reasoned that if someone wishes to punch someone in the nose (as perhaps a prize fighter might wish to pummel his opponent), he must normally pursue this objective through the marketplace. It might seem tautological to claim that people cannot punch strangers in the nose without paying for their consent. Nevertheless, the theory becomes more interesting.

Posner based his reasoning on Ronald Coase’s (2013) idea that liability rules, even intentional tort rules, do not affect the allocation of rights when transaction costs (the costs of doing a deal) are low, because then people can allocate rights through contract, and the highest-valued possessor of a right will normally get it. Coase’s successors noted that when transaction costs are low, the law should make the original endowments very clear, so people need not do legal research or hire experts to tell them where the starting point for their contract is (Calabresi & Melamed, 1972; Ehrlich & Posner, 1974; Merrill, 1985). On the other hand, when transaction costs are high, the legal rule, whatever it is, will likely prevail and will not be superseded by contract, so this legal rule must usually be complicated to be efficient (Merrill, 1985). If people want a right different from what the law gives them as an endowment, normally, that is, when transaction costs are low, they must bargain with whomever has the right. Suppose, however, that transaction costs are high. In this situation, the person who has the right in the low-transaction cost case might have to put up with interference, but exactly when becomes a complicated rule defining privileges or defenses to intentional torts. This is a very simple theory that purports to offer fresh insight about intentional torts.

We can test the theory with legal cases (Hylton, 2012). In Ploof v. Putnam,10 the defendant owned an island in Lake Champlain and a dock attached to it. The plaintiff, his wife, and two minor children were sailing on the lake when a storm suddenly developed; he tied his sloop to the defendant’s dock to ride out the storm. This was ostensibly a trespass to land. For that reason, the defendant’s servant cast the boat off into the storm, damaging the plaintiff’s sloop and causing personal injuries to the plaintiff and his family. The court held the defendant was liable for his servant’s act because the plaintiff was privileged by private necessity to ride out the storm at the defendant’s dock. The economist’s theory of the case is that because the storm made transaction costs high between the plaintiff and the dock owner, the rule became complicated in a way that allowed the plaintiff to make use of the defendant’s property without the defendant’s contractual consent. The value of refuge to the plaintiff was much greater than what the defendant lost from being deprived of his privacy.

The converse of this situation arose in Vincent v. Lake Erie Transportation Co.11 The defendant possessed a contractual right to have its freight ship at the plaintiff’s dock until she was unloaded, but a storm developed on Lake Superior so that by the time the unloading was done the winds were blowing at 50 miles per hour. Although the defendant’s contractual right had become ambiguous when the unloading was over, the defendant ship owner nevertheless instructed its employees to keep the lines fast between the ship and the dock. This choice favored the ship owner’s interests because it saved the ship, but it badly damaged the dock because throughout the night the heavy ship kept bashing into it.

The court held that the ship had the right to stay at the dock, but was liable for whatever damage the ship caused to the dock. It was thus a “conditional privilege,” a privilege conditional on the ship owner’s paying the actual damages done to the dock. This rule, in combination with the rule in Ploof v. Putnam, encouraged the parties to work together to maximize social wealth, in this context, the sum of the values of ship and dock. If the dock was costly and the ship heavy but valueless, the parties would have the joint incentive to scuttle the ship away from the dock. On the other hand, if the ship was costly relative to the potential damage to the dock, the parties would acquire the joint incentive to save the ship at the expense of the dock. This was another situation in which the courts created a special rule because of high transaction costs. How do you bargain in a raging storm?

Suppose, however, no crisis of high transaction costs exists, but only a lack of wealth on the part of the invaders. In a third case, the defendants were London squatters who took over the plaintiff’s property and made a good showing that they needed the resource more than the plaintiff did.12 Here the court kept the rule simple by refusing to give the squatters a privilege and held them liable. The squatters’ problem did not arise from high transaction costs, as in the prior two cases, but from a mere inability to pay for the property they wished to inhabit. During the time of the squatters’ invasion the market for housing was operating efficiently.

Finally, in Greyvensteyn v. Hattingh,13 the plaintiff brought an action of trespass against the defendants for the damage done to his crops when the defendants drove locusts onto his land. On November 25, 1907, a swarm of locusts invaded the plaintiff’s farm in the Cape of Good Hope. The locusts were “voetgangers,” or young insects that had not yet acquired the use of their wings, and had trekked across the countryside on foot, eating the grass and crops on their way. The locusts entered the plaintiff’s farm from the north. The defendants, who were also farmers, had their lands to the south of the plaintiff’s land. The defendants stationed themselves at their land borders and shooed the locusts away from their farms, causing the locusts to eat more crops on the plaintiff’s land.

Here, although transaction costs were high, the rule did not need to become complicated because no surplus was available in this situation. There was the real possibility in Ploof that a valuable boat and the lives of its passengers would be sacrificed to preserve a mere right of privacy in the use of land; in Lake Erie Transportation a valuable dock might have been sacrificed to save a worthless ship. In Greyvensteyn, by contrast, the locusts would eat either the defendants’ crops or the plaintiff’s crops. From a social point of view, the loss would be the same either way. Although the case was superficially similar to Lake Erie Transportation in that the farmers used the plaintiff’s property to enhance their own interests, from an economic point of view no liability was required because, again, no surplus value was available from the interaction of the various farmers in defending their properties.14

A corrective justice theorist would be hard-pressed to explain all these cases using Kant’s philosophy. Indeed, the no-liability result of the Greyvensteyn case is inconsistent with any corrective justice theory based on Kant’s philosophy because the defendant farmers treated the plaintiff farmer as a means for protecting their own property, and not as an end, and yet were not required to redress the plaintiff’s loss. A civil recourse theorist would be at pains to identify which of these cases involved a “wrong” and which did not. By contrast, the economic explanation is parsimonious, falsifiable, and provides fresh insight. It asserts that when transaction costs are high, the law will disregard the normal property rights and institute other rules designed to incentivize social-wealth-maximizing behavior.

Other intentional-tort rules fail to support the philosophers’ theories. Children, people with mental illness, and others lacking “moral agency” are held liable for their intentional torts.15 Unlike with the criminal law, these torts are sometimes “intentional” only because the defendant knew that a prohibited “contact” (as to the plaintiff’s nose) would result from his act, often without knowing, or even having the ability to know, that the contact would be harmful or offensive.16 The courts say this liability gives the heirs of these defendants an incentive to control them.17 This type of reasoning is again prohibited by Immanuel Kant’s moral philosophy because it imposes liability on persons lacking moral agency only as a means to get their heirs to restrain them (Weinrib, 1985, 1987b, 2002). Thus, the heirs become a means to an end, and the defendants themselves are not treated as true Kantian ends. Kant would excuse these challenged defendants and hold them immune for their torts. Indeed, at least one moral philosopher, recognizing that the legal doctrine conflicts with moral philosophy, has proposed that the courts abolish their rule imposing liability on people lacking moral agency (Goudkamp, 2011; see also Goldberg & Zipursky, 2009).


The most important tort rule is negligence. It has five elements, all of which the plaintiff must prove to recover damages. The elements are duty; breach of duty; cause in fact; proximate cause; and actual damages. We start with breach of duty, because that is the most basic part of negligence law.

Breach of Duty

As already noted, the tort of negligence became prominent when harmful accidents became common. As the negligence explosion’s timing and constituent cases reveal, the prevalence of harmful accidents depends on the existence of advanced technology, which multiplies the potential harm from human error. Urban congestion also multiplies the possibilities for accidents.18

Calabresi (1970), Coase (2013), Pigou (1920), and Posner (1972) were pioneers in developing the idea that the social cost of accidents is comprised of the harm produced by them together with the cost of precaution needed to prevent them. A sensible objective is to minimize the social cost of accidents. When this happens, Kaldor–Hicks efficiency is also achieved.

Richard Posner (1972, 2014) developed the first positive economic theory of the negligence rule; indeed, his theory of negligence was also the first positive economic theory of any body of law. Posner argued that the center of the negligence rule was the Learned Hand formula, a judicial gloss of the breach-of-duty element of negligence developed by Judge Learned Hand in an admiralty case.19 In this case in which a barge broke loose in New York Harbor, crashed, and sank, the libellant (plaintiff in admiralty) alleged that the barge owner’s breach of duty was the unexcused absence of its barge attendant at the time of the accident. Learned Hand’s formula was that a court should find a breach of duty only when the burden or cost of an alleged untaken precaution was less than the reduction in risk it would produce, that is, when B < PL, where B is the cost of precaution and PL is the probability of an accident multiplied by the damages that would result should it occur (Grady, 1989). If courts behaved this way, Posner reasoned, they would induce Kaldor–Hicks efficient behavior because actors would acquire an incentive to use only those precautions that paid out their cost in reduced risk. Posner claimed that judges and even juries used the Learned Hand formula so that the negligence rule would indeed yield Kaldor–Hicks efficiency.

The next year, John Prather Brown (1973) published the first game-theoretic model of liability rules, soon followed by similar models (Diamond, 1974; Diamond & Mirrlees, 1975). Within ten years, Richard Posner and his collaborator, William Landes, had also adopted the game-theoretic approach (Landes & Posner, 1981, 1987).

The original economic theories of negligence law usually posited that negligence resulted from a conscious decision to use substandard precaution. The type of precaution omitted was almost always imagined to be “durable precaution,” that is, something like a fire escape on a hotel or a spark arrester on a steam locomotive. Once installed, durable precaution, as its name suggests, remains effective for a long time. This conception, however, abstracts away from the main arena of negligence law, which is the omission of “nondurable precaution.” Examples are failing to check one’s blind spot before changing lanes, failing to notice a pedestrian in a crosswalk, failing to notice a stop sign, and so forth. Such precautions are effective for a short period of time and must be used repetitively to yield safety. The early economic theories were game-theoretic proofs that a given liability rule, for instance, negligence with the defense of contributory negligence, would yield a stable social equilibrium of efficient and “reasonable” precaution by injurer and victim alike. The models generally demonstrated that the negligence rule would indeed induce zero or at least very little negligent behavior (Grady, 1990).20 The analysts could nevertheless see that judgments of negligence liability were common and often posited “legal error” or “actor error” as leading reasons (Landes & Posner, 1987; Shavell, 1987; cf. Dari-Mattiacci, 2005).

Why does so much negligence exist in the real world? The reason is nondurable precaution and the way courts treat it. In the United States some cases are tried to a judge alone, but most cases, if they go to trial, are decided by a judge sitting with a jury. With respect to the cases tried to judges alone, which in the United States are mainly Federal Tort Claims Act (FTCA) cases and admiralty cases (often involving ship collisions), the Learned Hand formula seems a good model for breach of duty determinations (Grady, 2019). In cases tried to a jury sitting with a judge, it has long been recognized that the jury has a large amount of discretion to find the defendant either guilty or not guilty of negligence (Abraham, 2001; Gilles, 1994; Grady 2019; Wright, 2003). Nevertheless, limits exist. In a jury-eligible case, a plaintiff will normally allege untaken precautions by the defendant, which would have saved the day. Before a case goes to a jury, the judge will often assess whether the alleged untaken precautions failed the Learned Hand formula by having costs greater than benefits. If so, the case will be dismissed before trial.21 In addition, after a case has been tried to a jury, and the jury has found for the plaintiff, the trial court or an appeals court can set the decision aside because the proven untaken precautions failed the Learned Hand test. Thus, in the United States the formula acts as an upper limit on liability in many jury-eligible cases.22 In other common law jurisdictions in which juries no longer decide negligence cases, the formula can be a decision rule for judges (Gilles, 2002).

A problem exists with respect to nondurable-precaution cases because the Learned Hand formula, when applied in the most straightforward manner, can be an unreliable measure of efficiency. It is impossible for people to use every cost-beneficial precaution with perfect consistency. Some potential legal errors requiring too much precaution are effectively limited by cost–benefit analysis because nondurable precautions can be imagined to be implemented as a part of a plan, and precaution plans are clearly analyzable under cost–benefit analysis (Grady, 2009). For instance, if a plaintiff argues that the defendant’s driver should have looked for potentially skittish horses before he dumped a load of trash into its truck, the court can find that a plan to use this precaution failed cost–benefit analysis and dismiss the case on that ground.23 The problematic jury decisions pertain to lapses in nondurable precautions that are part of plans that pass cost–benefit analysis and are even required by it. Legal decisions involving failures to use this type of nondurable precaution are almost certainly the largest part of the population of all negligence cases and claims. Examples are failures to check for opposing traffic when crossing an intersection; failures to retrieve all surgical sponges placed in a patient; and failures to dispense the correct drug or the correct dosage strength (Grady, 2019).

Although it is certainly important for drivers to check for opposing traffic, for surgeons to retrieve all sponges, and for pharmacists to dispense the right drugs, in the real world there are quite a few lapses in all three realms (Grady, 2019). Not all these errors are inefficient each time they are committed (Grady, 1988c, 1998). As the economist George Stigler said, “If you’ve never missed a plane, you’re spending too much time in airports.”24 Associating all human errors with inefficiency is a noneconomic way of thinking. Nevertheless, courts cannot easily measure whether the given error before them was efficient or inefficient. It often depends on how much the actor was investing by way of attention and focus on the task at hand. Moreover, how is a court to say which forgotten sponge was an inframarginal and inefficient error by a surgeon and which was an ultramarginal and efficient error? As a response to this problem, courts give juries broad power to find such a lapsing defendant either guilty or not guilty of negligence (Abraham, 2001; Grady, 2019). Juries do not always forgive, however. Thus, when a jury finds such a defendant liable, it is still possible that the error was efficient (Grady, 1998).

That juries can and do find people liable for efficient errors is the most important aspect of the negligence rule. The two consequential dangers are that people will inefficiently substitute durable for nondurable precaution and will inefficiently reduce levels of activities in which nondurable precaution is highly productive relative to potential durable precaution substitutes (Dari-Mattiacci & Franzoni, 2014; De Mot & Depoorter, 2011; Grady, 1995, 2009, 2013). The problem does not exist so much with respect to durable precaution. Think of installing a fire escape on a hotel or an apartment building. If this precaution is required by the Learned Hand formula, the fire escape will almost certainly be installed. As the economists’ game-theoretic models demonstrate, it is foolish to be negligent when you have the chance to reflect on your behavior and possess ample time to act on that reflection. That is why nondurable precaution cases are the biggest part of the population. With nondurable precautions, perfect consistency is uneconomic and usually impossible to achieve. Hence, negligence with respect to nondurable precaution is predictable and represents a huge flow of cases every year.

Negligent behavior will be common whenever and wherever the productivity of nondurable precaution is high and durable precaution is a poor substitute for it. In the present era of very advanced technology, we commonly see new devices (new durable precautions) that substitute for old nondurable precautions by reducing their productivity. Examples are warning lights on sideview mirrors that indicate that another vehicle is in the driver’s blind spot or a device that monitors objects ahead and automatically stops the vehicle when an object gets too close (Dari-Mattiacci & Franzoni, 2014; Grady, 2009). A more venerable example of a durable precaution substitute for nondurable precaution is automatic gates that lower to block a grade crossing when a train approaches, thus reducing the productivity of the engineer’s nondurable precaution of blowing the horn and looking ahead for obstructions. Nevertheless, basic technologies (steam boilers, trains, automobiles, airplanes, life-saving drugs) usually increase the productivity of nondurable precaution and thus also increase the amount of negligent behavior. Society is made better off by these inventions, and paradoxically so are its tort lawyers.

Even new safety technology usually increases the productivity of nondurable precaution. The invention of air brakes for trains, while reducing the number of accidents, also increased the amount of negligent behavior. When a train cannot stop quickly, it is not so important for the train crew to keep a constant lookout forward, because in many cases it will do no good. When trains possess effective brakes, the failure to keep a proper forward lookout will often be a negligent cause of disaster. Good brakes have nondurable precaution complements whose productivity they increase, and good brakes certainly do not substitute for nondurable precautions. Technology that reduces the productivity of nondurable precaution and thus substitutes for it, such as the automobile air bag or the blind spot monitor, usually comes much later than the original technology that created the need for it. We can thus distinguish between “compliance-using” technology, which increases the productivity of nondurable precaution, and “compliance-saving” technology, which reduces it (Dari-Mattiacci & Franzoni, 2014; Grady, 2009, 2013).25 While compliance-saving technology may someday sharply reduce the number of negligence cases, that day will not be tomorrow. The number of negligence cases at any time is the result of a race between unforgiving, basic technology and later compliance-saving technology.

Much of the complexity of the negligence rule is driven by the courts’ apparent wish to limit the bad consequences of liability for inadvertent lapses in nondurable precaution, which can easily represent efficient behavior (Grady, 1996, 1998). This legal complexity takes the form of causal and other limitations on negligence liability. The first of these is discussed next.

Cause in Fact

The cause in fact element of negligence requires that the same untaken precaution that constituted the breach of duty must have also prevented the accident, when we take an ex post view of the accident (Abraham, 2013; Grady, 1989, 2009). Because it focuses on expected harm, the Learned Hand formula inevitably takes an ex ante perspective on accidents. The question is whether the cost of some untaken precaution would have paid its way in reduced expected harm, or “risk” in the legal sense. Thus, the precaution plan of a driver to maintain a lookout for pedestrians is required by the Learned Hand formula because the cost of this plan is less than the risk reduced by it. Nevertheless, suppose the pedestrian stepped directly in front of the defendant’s moving car, effectively onto the front bumper, let’s say. Then, if the driver was inadvertently lapsing in his duty, the lapse will fail to yield negligence liability because of the cause-in-fact limitation. The accident would have happened whether the driver was lapsing or not.26

One legal philosopher has argued that the cause-in-fact doctrine is inconsistent with economic reasoning because economics deals with ex ante incentives, whereas cause in fact creates an ex post test (Wright, 1985). This is incorrect because although cause in fact does indeed use an ex post perspective on the accident, it ex ante limits liability in a predictable way. In Figure 1, the horizontal axis shows increasing precaution (P) levels as we move to the right. The R-curve is the marginal reduction in risk (the marginal PL from the Learned Hand formula), and it declines at increasing levels of precaution, thus reflecting diminishing marginal effectiveness. The C-curve is the marginal cost of precaution. Thus, social cost (total precaution cost combined with total risk) is minimized at P*.

The Positive Economic Theory of Tort LawClick to view larger

Figure 1. Negligent harm and unavoidable accident.

A breach of duty exists at any precaution level below P*, because at those precaution levels some untaken precaution will yield greater risk reduction (the additional area created under the R-curve) than cost (the additional area created under the C-curve). As just noted, the normal test of cause in fact is whether the untaken precaution would have prevented the accident. Thus, if the injurer had been using precaution level P1 and the victim alleged precaution level P* as the untaken precaution, the court (or jury)27 would assess whether the accident fell within the zone of negligent harm or the zone of “unavoidable accident.” The most common method is simply to ask whether precaution level P* would have prevented this accident, as when P* was the 25 miles per hour speed limit in a school zone, the defendant was driving at 35 miles per hour, and the child darted into the street ten feet in front of the moving car. An accident reconstruction expert should be able to testify whether the accident fell within the “negligent harm” zone or the “unavoidable accident” zone. (The expert would be instructed to say whether the driver would have been able to stop in time if the child darted out the same distance ahead of a car traveling at 25 miles per hour.) Sometimes, when causal evidence is less precise, courts will simply compare the sizes of the two zones to see which is greater (Grady, 2009). Large amounts of unavoidable accident are typically associated with modest levels of total risk, at zero precaution, and basic levels of precaution technology. Think instead of the possibility of a nuclear reactor meltdown. In this situation, total risk at zero precaution is huge, and the amount of unavoidable accident is small in comparison to the amount of negligent harm yielded by even a small lapse. Such a case can be easy from a cause-in-fact point of view (Abraham, 2013; Grady, 2009, 2014).

Suppose it is uncertain whether the untaken precaution would have prevented the accident or not. The typical legal test is whether it was more likely than not that the accident fell within “negligent harm” or within “unavoidable accident.” If the probability was only 30% that the accident fell within negligent harm and 70% that it fell within unavoidable accident, the plaintiff should be nonsuited. This may seem unfair or even inefficient, because the plaintiff who experienced a 30% chance of having been harmed by negligence is given a zero recovery (Stremitzer & Tabbach, 2014). Nevertheless, this rule can still create the proper overall incentives because a plaintiff is allowed a full recovery in the reverse situation when the probability was only 70% that the accident was caused by negligence. Thus, in the normal situation and over the long run the second type of case balances out the first type of case, which leaves the proper ex ante incentives (Grady, 2014; Hylton & Lin, 2013).

Imagine a precaution, however, that is very valuable, but always yields only a 30% chance of success. Such a precaution might be a chest x-ray for a patient suffering from a persistent cough that might reflect lung cancer, possibly treatable if discovered early enough. If such a patient was not given the x-ray, had cancer, and was blocked from legal recovery, doctors might lack the incentive to order the test, maybe in an HMO setting in which it could be privately beneficial for the HMO to let its lung cancer patients die as quickly as possible.28 In this type of “marginal-causation” case the courts often relax the strict “more probable than not” rule and allow the plaintiff to recover for the “lost chance” (Grady, 2014). Thus, the survivor would receive 30% of the damages (or perhaps even the full amount of damages as a deterrent for willful negligence). This type of proportional liability exists only in cases of marginal causation because the ordinary, “all-or-nothing” rule is less costly to administer across the usual cases and produces the proper ex ante incentives in them. This economic gloss of the legal rule for marginal-causation cases represents one of the first instances in which the positive economic theory of tort actually “predicts” court decisions that otherwise lack a good legal gloss (Grady, 2014; Malone, 1956).

From an economic point of view, the cause-in-fact rule is efficient because it reduces the liability from small amounts of under-precaution. Suppose an injurer was at precaution level infinitesimally to the left of P*. Without cause in fact, that injurer would face an expected liability in proportion to the total amount of unavoidable accident, which would then be within the liability zone. Such a rule would create an inefficiently great incentive never to be even the slightest degree negligent. Injurers would predictably respond by inefficiently substituting durable for nondurable precaution, by inefficiently reducing their activity levels, and by planning to use more than efficient precaution (Fennell, 2018; Grady, 1983, 1988b). Beginning with John Brown (1973), the early mathematical models of negligence left out the cause-in-fact limitation, which made it seem that the negligence rule would almost always induce due care, and never negligent behavior. In addition, these early models minimized the difficulties that can arise when multiple parties all possess good opportunities to prevent an accident, and liability must be divided among several concurrent injurers to induce efficient care by each (Grady, 1990; cf. Landes & Posner, 1980, 1987).

Consistent with the economic explanation of cause in fact, juries can be instructed to waive the requirement in situations of deliberate negligence, which is rarely efficient. Suppose, instead of 35 miles per hour, the driver was racing another driver at 90 miles per hour through a school zone where children were obviously present. Again, a child darts out ten feet in front of the car. Such a driver faces liability not only for negligent harm but also for unavoidable accident.29 This increased incentive to use care is appropriate because the driver has deliberately engaged in highly inefficient behavior and evidently needed a special incentive (Grady, 1988a). The driver would also be liable for punitive damages for the same reason (Ellis, 1982). The purpose of the cause-in-fact limitation is highly related to the problem of excessive liability for efficient, inadvertent negligence. Cause in fact is not needed when the actor deliberately engaged in inefficient conduct. Indeed, all limitations on negligence liability are strongly related to the possibility that the negligence in question was efficient (Grady, 1998). If courts could always be sure that the negligence before them was inefficient, it would not be so important to limit liability with causal and other restrictions.

The philosophers’ corrective justice and civil recourse theories say that the purpose of cause in fact is to establish a moral relationship of wrongdoing between the injurer and his victim (Goldberg & Zipursky, 2010; Weinrib, 1987a, 1995). Thus, a driver may be wrongfully speeding, but only when his speeding was the cause in fact of a victim’s injury is a relationship of wrongdoing established. It is this relationship, forged by cause in fact, that in turn creates the corrective justice duty of reparation. A major problem with this theory is that many breaches of duty are not “wrongs,” either in an economic or in any moral sense, because society would lose much more from the total avoidance of breaches of duty than from merely sanctioning them when they inevitably and productively occur (Grady, 1994a). Thus, under the philosophers’ theories, the surgeon who has excessively invested in sponge retrieval, but still lapsed, is supposed to need correction. Such a surgeon may indeed be liable, but the moral judgment is counterproductive (Holmes, 1997). If prospective surgeons credited this philosophy, they would never become surgeons, especially if they are moral individuals seeking to avoid wrongdoing. In addition, the philosophical theories misdescribe U.S. negligence law because U.S. courts give juries broad powers to forgive even instances of obvious negligence for which corrective justice,30 if it were operative, would require a correction (Grady, 2019). Instead, it seems more realistic to regard negligence law as creating a “stochastic tax” on inadvertent negligence, which is a policy highly consistent with the economic idea that much inadvertent negligence is efficient (Grady, 2019).

Proximate Cause

Proximate cause is another major limitation on negligence liability (Grady, 2013; Moore, 2009; Shavell, 1980a, 1987). Two different proximate cause doctrines exist, and both potentially apply to all accidents. The first, the “reasonable foresight” doctrine, holds that no liability exists when the alleged untaken precaution had little systematic tendency to reduce the probability of the accident in question. In other words, if someone were to look at the accident ex post and were asked how best to prevent this type of accident in the future, the “last thing” this person would recommend would be to use the untaken precaution alleged as the defendant’s breach of duty. Thus, a merely coincidental relationship existed between the accident and the untaken precaution in question.

The second doctrine of proximate cause, which is called the “direct consequences” doctrine, cuts off liability when, between the time of the defendant’s inadvertent negligence and the plaintiff’s injury, someone else with a duty to the plaintiff either intentionally acted to produce the injury or intentionally omitted a highly effective precaution against it.

If we return to the school zone example, suppose that a driver is passing by a school at 35 miles per hour and inadvertently fails to see the children that would invoke the 25 mile per hour speed limit. Just at that point, a murderous adult pushes one of the school children into the road 30 feet ahead of where the speeding car was then located. Cause in fact might exist in that case, but both doctrines of proximate cause would eliminate liability.31

Res Ipsa Loquitur

The main method of showing negligence is for the plaintiff to prove an untaken precaution by the defendant, which was simultaneously a breach of duty, a cause in fact, and a proximate cause of the plaintiff’s harm. This method of proof is called “specific negligence.” What if the plaintiff does not know or cannot say exactly what the defendant’s untaken precaution was? In this situation, the courts allow the plaintiff to rely on the doctrine of res ipsa loquitur, which means in Latin “the thing speaks for itself.” The effect of this doctrine is to create a prima facie case for the plaintiff as well as a legal suggestion of the defendant’s negligence (Grady, 1994c, 2009).

The first such case was Byrne v. Boadle,32 the English precedent that gave the doctrine its name.33 The plaintiff was walking on a busy street past the defendant’s warehouse when a barrel of flour fell upon his head. He didn’t see what hit him, but witnesses later testified that the barrel fell from the defendant’s warehouse. The defendant argued that unless the plaintiff could prove a specific precaution untaken by him or his employees, the plaintiff could not recover. In rejecting this argument, Chief Baron Pollock said: “There are certain cases of which it may be said res ipsa loquitur [the thing speaks for itself], and this seems one of them.” As res ipsa has evolved, two key questions have become whether the instrumentality that caused the harm was within the exclusive control of the defendant and whether the accident was the type usually caused by negligence.34 These questions effectively ask whether the accident was probably within the defendant’s zone of negligent harm or was instead within the defendant’s zone of unavoidable accident, or even beyond the defendant’s R-curve.35

Byrne v. Boadle remains a prototype for res ipsa cases. As already noted, the most common type of negligence is a lapse in nondurable precaution. These lapses can easily be efficient in the sense that the cost is prohibitive to avoid them entirely. On the other hand, we can be more certain that fire escapes (a durable precaution) will be installed because it is rarely efficient to omit them when they are reasonable. The res ipsa cases embed a theory of when negligence by a type of actor is likely. These cases suggest that negligence by an actor is likely when the productivity of nondurable precaution is very high and when the possibilities for negligent harm are great relative to the risk of unavoidable accident. In both situations, and especially when the two conjoin, res ipsa cases are strong.

Byrne v. Boadle fits this theory to a tee. When you are lowering a heavy barrel over a busy sidewalk, using ropes and a jigger-hoist, the productivity of nondurable precaution is at a high level. Every check of the ropes, the fastening knots, the mechanism, pays out big in terms of reduced expected harm to the people below. Similarly, if there is a failure of precaution, the result can easily be death. Thus, the required level of nondurable precaution is high, leaving a modest amount of unavoidable accident. Moreover, because the rate of required nondurable precaution is so high, like a fine net, when a barrel slips through, the likelihood is that some omitted check or inspection was the cause.

Suppose instead it was one-pound bales of eiderdown that were being lowered over the sidewalk. This would be a bad res ipsa case because the feather bales are unlikely to do much harm, so the required rate of nondurable precaution is low and the amount of unavoidable accident correspondingly large relative to the possibilities for negligent harm.36 When a feather bale slips through, the inference is much stronger than with flour barrels that the bale fell between required precautions, not through a missed required precaution. In Figure 1, when you increase the risk (move the R-curve up), with the cost of precaution constant, you make a res ipsa case stronger because the risk of unavoidable accident remains the same, but the opportunity set for negligent harm increases. Thus, highly dangerous activities often create strong cases of res ipsa loquitur (Grady, 2009).

In addition to what has just been said, with res ipsa cases there is always an accident signature that may speak of the defendant’s negligence independently of the danger of the activity and the resulting required rate of nondurable precaution. In Anjou v. Boston Elevated Ry.,37 the plaintiff was a passenger on the defendant’s elevated railroad, and as he was walking through the station he slipped on a banana peel that was on the platform. The evidence was the banana peel was “black, flattened out and gritty.” The court held that this was enough evidence for a jury to find the defendant negligent because the condition of the banana peel demonstrated that the defendant had not inspected and cleaned its platform as often as due care required.38

The accident signature can combine with information about the required rate of nondurable precaution. Suppose an airplane crashes in perfect weather with the fuel tank empty, the fuel gauge showing empty, and several airports nearby. This is a very strong case of res ipsa loquitur because the accident signature shows the pilot failed to check the fuel gauge with sufficient frequency.39 In fact, it could equally be regarded as a strong case of specific negligence in which the defendant pilot’s omitted precaution was checking the fuel gauge. The clearest cases of res ipsa loquitur entail a highly dangerous activity combined with an accident signature that points to the defendant’s negligence as the probable cause.40 Conversely, res ipsa cases are weak in the opposite type of situation.41 This is another situation in which the positive economic theory of tort predicts case results more accurately and parsimoniously than even the ruling legal glosses.42


The most obvious purpose of the duty element is to preserve human liberty. The main distinction that runs through duty cases is the difference between “misfeasance” and “nonfeasance.” The basic rule is that negligence can exist only for an injurer’s risk-creating acts as opposed to “nonfeasances,” or failures to act. If failures to act fell under the Learned Hand formula, a doctor who merely observed a sick person might have to expend his energy in curing him up to the point at which further measures and precautions would finally become unproductive. No one with skill or other assets would be safe leaving home.

The preservation of human choice is the key to duty. Under negligence law, people can acquire legal duties to help others only through their voluntary conduct, as by acting in a way that creates risk to others, by voluntarily undertaking to protect others, or by establishing a special relationship with others. Without this element of free choice, it would be impossible for the duty doctrine, or for that matter the entire negligence system, to advance economic well-being. Definitionally, the very concept of economic welfare depends on people’s ability freely to choose their own life activities without punitive taxes on their socially beneficial decisions (Alchian & Allen, 1967; Epstein, 1979; Posner, 1980).43 It may be enough that tort law taxes physicians’ errors in commission, without also taxing their failures to act to help strangers.

In the absence of an affirmative act, the most common reason for a duty is a special relationship between the injurer and the victim, which gives the injurer a comparative advantage at precaution over some ranges. These special relationship duties tend to be most comprehensive when an economic conduit exists between the parties so that the potential victims can pay their injurers ex ante for the expected liability the relationship creates (Hylton, 2009; Levmore, 1986). Thus, a shop owner owes a strong duty to its customers to abate hazards on the premises. The defendant may inadvertently, and possibly efficiently, fail to notice a pool of water on the shop floor upon which a customer slips, but the shop owner need not inefficiently substitute durable precaution or reduce activity levels, because it can build a liability premium into the prices of shop merchandise. A physician owes a similar duty to his or her patient and can similarly compensate himself ex ante for the predictable flow of liability from efficient negligence. These legal duties become much weaker when no economic conduit exists and especially when the asserted untaken precaution may be a precaution the customers, that is, potential victims, cannot afford.44

Actual Damages

The existence of “actual damages” is an element of negligence. With intentional torts such as battery and assault, a victim may sue for insult or mental disturbance even if she was not physically hurt. The most basic “actual damages” are injuries to human life or limb and the destruction or injury of physical property. These damages clearly reduce social wealth. Problematical damages are economic loss and emotional injury.

Economic loss, such as lost profits or lost wages, is routinely recoverable if it accompanies and flows from a destruction or injury of physical property or from a physical injury to the plaintiff’s body. Thus, if through the negligence of a power company the electric supply to a metal producer is interrupted and melted alloy solidifies and is spoiled in the process, the recoverable damages include the spoiled product, as well as damage to the lining of the smelting furnace.45 Nevertheless, when the power continues to stay off, and the metal producer is unable to process further “melts,” the lost profits on those are unrecoverable; they are purely economic loss.46 Leaving the loss on the victim encourages it to mitigate, as by rearranging production schedules, so that the undone melts are fabricated at some other time. Similarly, if because of contractors’ negligence the side of a building under reconstruction falls off and harms property or people below, the contractors would be liable for the physical injuries done. Nevertheless, if merchants in the area lose trade because the police cordon off the area, the latter losses are not recoverable.47 One leading scholar has theorized that the reason for nonliability is that other merchants are likely to gain the sales that the cordoned-off merchants lose. It is a mere wealth transfer (Bishop, 1982). Since the cost of precaution is a real social cost, the courts behave as if it is a bad social bargain to incentivize a real cost—precaution cost—merely to prevent a wealth transfer. The more efficient solution is for the merchants to purchase business interruption insurance.

Damages for emotional distress are often awarded when they result from some physical injury also caused by the defendant’s negligence.48 The legal problem arises when the plaintiff’s emotional distress arises independently of any physical injury. One situation of recovery arises when a mother or other close relative observes the negligent killing of a child.49 This liability is appealing for two reasons. First, the emotional distress inflicted is a real social cost because it diminishes the happiness of the mother and may do more. Second, when a child dies, not many recoverable damages result, such as lost earnings or medical bills. The liability for emotional distress helps insure that people who may kill children possess an adequate incentive to use precaution. Another situation in which emotional distress damages are awarded with an associated physical injury is when the defendant’s negligence puts the plaintiff in a situation in which the plaintiff becomes the apparent cause of an injury or death to someone else, a situation that the courts regard as very distressing to the plaintiff.50 The courts have devised many restrictive rules to bar liability for claims that may be fictitious or exaggerated.51

The general rule of nonliability or highly restricted liability for purely economic losses and emotional distress again seems partly driven by the courts’ policy to limit liability generally for inadvertent and therefore possibly efficient negligence. These rules are relaxed and recoveries for economic loss and emotional distress become more likely when the defendant’s negligence was deliberate, or “willful and wanton,” and thus less likely to have been efficient.52

Defenses to Negligence

Assumption of the Risk

The original defense of assumption of risk applied to a broad range of activities in which a plaintiff saw a risk, appreciated it, and voluntarily encountered it. This defense was an absolute bar. When courts and legislatures transformed contributory negligence into comparative negligence, the traditional doctrine of assumption of the risk was in many jurisdictions limited to sports and recreational activities where a certain degree of heedlessness is a source of the thrill or benefit of the activity. The apparent objective is to prevent negligence liability from reducing the levels of these activities.53

Contributory Negligence

Under contributory negligence, the plaintiff’s claim is totally barred if the defendant can show that the plaintiff’s own breach of duty was also a cause in fact of his injury. The original game-theoretic models of the negligence rule found that contributory negligence was unnecessary to induce zero negligence (Brown, 1973; Dari-Mattiacci & Garoupa, 2009; Landes & Posner, 1987; Shavell, 1983, 1987). The modelers reasoned that a landowner adjacent to a steam railroad would assume that the railroad would install a spark arrester because the negligence rule gave the railroad an adequate incentive to do so. With that assumption, the adjacent landowner would expect to bear the cost of any fire that resulted. This expected liability in turn would induce the adjacent landowner to use proper precaution (plowing a safety zone next to the tracks), all without the need of contributory negligence. Thus, according to the early economic theories, the law apparently embodied a foolish conception, which was an odd claim for theories that purported to be positive.

The reality is that much negligence exists in the world because much precaution is nondurable and actors cannot at reasonable cost reduce their levels of negligence to zero. Thus, when potential victims see that an actor has been negligent, or should expect that negligence is very possible, they possess duties under the doctrine of contributory negligence. If this doctrine did not exist, plaintiffs could willfully disregard their own safety and still collect damages (Grady, 1988a). For plaintiffs to use “corrective precaution” against defendants’ possibly efficient negligence can be highly desirable.

The traditional system of negligence and contributory negligence became complicated because it was (and in a few jurisdictions remains) an all-or-nothing system, at least on its face. It allocated total liability to one party or the other. The main problem was that it also led to strategic behavior by both victims and injurers (Grady, 1990). For instance, an aggressive cab driver might expect that pedestrians in crosswalks would inevitably see that he was muscling through them. If they continued into his path, they would be guilty of contributory negligence and collect nothing. Thus, the rule evolved that contributory negligence was no defense to willful or deliberate negligence such as the hypothetical cab driver’s conduct.54 A similar problem arises when a plaintiff has through inadvertent and possibly efficient negligence placed himself in a dangerous situation. If the defendant could then rely on the doctrine of contributory negligence barring the plaintiff’s recovery, he could strike the plaintiff with immunity. More dangerously, in many situations a defendant could assume that any pedestrian in the wrong place could never recover and could thereby relax his vigilance. To make the all-or-nothing rules work, the courts and juries had to assess the parties’ mental states and obscure issues such as which party possessed the “last clear chance” to avoid the accident.55

Comparative Negligence

Under comparative negligence, which courts and legislatures introduced as a reform, the plaintiff’s contributory negligence no longer bars the plaintiff’s recovery, but merely reduces it. Thus, a cab driver who muscles his way through a crosswalk expects to bear perhaps 50% of the liability for an injured pedestrian, and a pedestrian who fails to yield to such a driver expects also to bear a substantial proportion of his own losses. This rule largely eliminates strategic gambits (Grady, 1990).

Comparative negligence, however, comes at a great cost. The all-or-nothing system economized on administrative costs because many cases in which contributory negligence was obvious did not need to be tried but were simply dismissed before trial or were never brought by plaintiffs in the first place.56 Comparative negligence largely eliminates strategic gambits, but the system must process many more triable cases than under contributory negligence. Comparative negligence makes good economic sense for a very rich society committed to reducing accidents, many of which cannot be fully compensated.

Strict Liability

The third part of tort is strict liability, sometimes called absolute liability (Hylton, 2010b). The most famous case is Rylands v. Fletcher,57 which was decided by the English courts in the 1860s. The defendants had hired independent contractors to build a reservoir by which they wanted to create water power to run their nearby factory. The contractors noticed old mine shafts during the construction, but took no special precautions and failed to notify the defendants, who as just noted were the owners and took no part in the construction. When the reservoir was finished, the contractors brought water in wagons to fill it, and soon thereafter all the water escaped through the ancient mine shafts into the plaintiff’s mine. By the time of suit, the contractors were apparently out of business, so the only possible defendants were the reservoir owners, who had not been negligent because before the accident they were reasonably ignorant of the risk of building their reservoir close to mine shafts. The court of Exchequer Chamber held that the defendants were strictly liable, and the House of Lords subsequently affirmed that decision.

The Exchequer Chamber opinions, which seem the most important, analogized the case to the “alkali cases” in which a new industrial process regularly afflicted nearby Liverpool residents with noxious gases, and the courts held the factory owners liable (Dingle, 1982). The legal historian Brian Simpson convincingly argued that because of then-recent dam-bursting disasters, which caused significant loss of life, the Exchequer Chamber regarded Rylands v. Fletcher in the same light, even though in that case no lives were lost due to the good luck of miners being absent when the reservoir burst (Simpson, 1984).

Strict liability cases have since divided into two sets. The first set of cases involves accidental and dangerous, or if not dangerous at least massive,58 invasions of the plaintiff’s land. The second set of cases involves activities that do not necessarily invade the plaintiff’s property, but are “ultrahazardous” in the sense that the unavoidable accident is absolutely large and potentially catastrophic.59

Following earlier work by Steven Shavell (1980b), Landes and Posner (1981, 1987) theorized that the courts have applied strict liability to those accidents in which the most important objective is to influence the level of the defendant’s activity. The negligence rule typically leaves some risk on the backs of victims, and this externalization can yield an insufficient incentive for injurers to reduce the levels of their dangerous activities. The downside of strict liability is that it makes the defendant an insurer of the entire risk, which diminishes victims’ incentives to use their own precaution. Economists have sometimes complained that actual rule of strict liability fails to include a defense of contributory negligence and so might encourage moral hazard by victims, but this observation neglects two points. First, the courts have allocated to strict liability almost exclusively activities in which the actor possesses a comparative advantage at precaution in all ranges (Jones, 1992). It is cost-ineffective for neighbors of a nuclear reactor to use precaution against a meltdown or for a person walking along a country road to use precaution against debris thrown by a blasting operation.60 Second, courts have adopted the rule that participants in an activity cannot hold the actor strictly liable for their injuries.61 This bar for participants is more comprehensive than contributory negligence because it covers plaintiffs who have not been provably negligent, but who might have used some precaution that could be disincentivized by strict liability, for example, getting a good sleep on the night before setting dynamite sticks.62 In activities to which strict liability applies, participant-plaintiffs are more likely than any other type to possess good precaution opportunities. Hence, contrary to some economists, strict liability is indeed strict with both injurers and victims, which is appropriate for the “ultrahazardous” activities to which it applies.

Many economists have argued that strict liability induces the same amount of precaution as the negligence rule (Brown, 1973; Landes & Posner, 1987; Shavell, 1980b), but this also is incorrect. As noted above, juries can and do forgive negligence (Grady, 2019), but strict liability creates liability as a matter of law, decided by a judge without the possibility of jury forgiveness. Thus, the common law rule of strict liability typically creates more liability than negligence for injurers and victims. Strict liability is thus especially appropriate for the highly dangerous activities to which it applies.

The positive economic theory of tort law is relatively new, but has already added much to our knowledge of this body of law and promises to teach us more in the future. Moreover, the insights that have come from this theory will not be limited to its subject matter, but will certainly be extended to other areas of economic theory and policy.

Further Reading

Bishop, W. (1982). Economic loss in tort. Oxford Journal of Legal Studies, 2(1), 1–29.Find this resource:

Brown, J. P. (1973). Toward an economic theory of liability. Journal of Legal Studies, 2(2), 323–349.Find this resource:

Calabresi, G., & Melamed, A. D. (1972). Property rules, liability rules, and inalienability: One view of the cathedral. Harvard Law Review, 85(6), 1089–1128.Find this resource:

Coase, R. H. (2013). The problem of social cost. The Journal of Law & Economics, 56(4), 837–877.Find this resource:

Cooter, R., & Ulen, T. (2012). Law & economics. Boston: Pearson/Addison–Wesley.Find this resource:

Dari-Mattiacci, G., & Franzoni, L. A. (2014). Innovative negligence rules. American Law and Economics Review, 16(2), 333–365.Find this resource:

Fennell, L. A. (2018). Accidents and aggregates. William & Mary Law Review, 59(6), 2371–2446.Find this resource:

Gilles, S. G. (2002). The emergence of cost–benefit balancing in English negligence law. Chicago–Kent Law Review, 77(2), 489–586.Find this resource:

Grady, M. F. (1988). Discontinuities and information burdens: A review of The Economic Structure of Tort Law by William M. Landes and Richard A. Posner. George Washington Law Review, 56(3), 658–678.Find this resource:

Grady, M. F. (1989). Untaken precautions. Journal of Legal Studies, 18(1), 139–156.Find this resource:

Grady, M. F. (1994). Modern accident law does not fit corrective justice theory. Jahrbuch für Recht und Ethik/Annual Review of Law and Ethics, 2, 7–35.Find this resource:

Grady, M. F. (1994). Res ipsa loquitur and compliance error. University of Pennsylvania Law Review, 142(3), 887–947.Find this resource:

Grady, M. F. (1995). Legal evolution and precedent. Jahrbuch für Recht und Ethik/Annual Review of Law and Ethics, 3, 147–182.Find this resource:

Grady, M. F. (2009). Unavoidable accident. Review of Law & Economics, 5(1), 178–231.Find this resource:

Grady, M. F. (2014). Marginal causation and injurer shirking. Journal of Tort Law, 7(1–2), 1–33.Find this resource:

Grady, M. F. (2019). Justice luck in negligence law. Revus.Find this resource:

Grey, T. C. (1983). Langdell’s orthodoxy. University of Pittsburgh Law Review, 45(1), 1–53.Find this resource:

Grey, T. C. (2001). Accidental torts. Vanderbilt Law Review, 54(3), 1225–1284.Find this resource:

Holmes, O. W. (1870). Codes, and the arrangement of the law. American Law Review, 5(1), 1–13.Find this resource:

Holmes, O. W. (1997). The path of the law. Harvard Law Review, 110(5), 991–1009.Find this resource:

Hylton, K. N. (2016). Tort law: A modern perspective. Cambridge, UK: Cambridge University Press.Find this resource:

Landes, W. M., & Posner, R. A. (1987). The economic structure of tort law. Cambridge, MA: Harvard University Press.Find this resource:

Merrill, T. W. (1985). Trespass, nuisance, and the costs of determining property rights. Journal of Legal Studies, 14(1), 13–48.Find this resource:

Moore, M. S. (2009). Causation and responsibility: An essay in law, morals, and metaphysics. Oxford: Oxford University Press.Find this resource:

Posner, R. A. (1981). The economics of justice. Cambridge, MA: Harvard University Press.Find this resource:

Posner, R. A. (2014). Economic analysis of law (9th ed.). New York: Wolters Kluwer Law & Business.Find this resource:

Shavell, S. (1987). Economic analysis of accident law. Cambridge, MA: Harvard University Press.Find this resource:

Weinrib, E. J. (1995). The idea of private law. Cambridge, MA: Harvard University Press.Find this resource:


Abraham, K. S. (2001). The trouble with negligence. Vanderbilt Law Review, 54(3), 1187–1223.Find this resource:

Abraham, K. S. (2013). Self-proving causation. Virginia Law Review, 99(8), 1811–1854.Find this resource:

Alchian, A. A., & Allen, W. R. (1967). University economics. Belmont, CA: Wadsworth.Find this resource:

Bishop, W. (1982). Economic loss in tort. Oxford Journal of Legal Studies, 2(1), 1–29.Find this resource:

Brewer, S. (1996). Exemplary reasoning: Semantics, pragmatics, and the rational force of legal argument by analogy. Harvard Law Review, 109(5), 923–1028.Find this resource:

Brown, J. P. (1973). Toward an economic theory of liability. Journal of Legal Studies, 2(2), 323–349.Find this resource:

Byrd, S. (1989). Kant’s theory of punishment: Deterrence in its threat, retribution in its execution. Law and Philosophy, 8(2), 151–200.Find this resource:

Calabresi, G. (1970). The costs of accidents: A legal and economic analysis. New Haven, CT: Yale University Press.Find this resource:

Calabresi, G., & Melamed, A. D. (1972). Property rules, liability rules, and inalienability: One view of the cathedral. Harvard Law Review, 85(6), 1089–1128.Find this resource:

Coase, R. H. (2013). The problem of social cost. The Journal of Law & Economics, 56(4), 837–877.Find this resource:

Coleman, J. L. (2001a). The conventionality thesis. Philosophical Issues, 11, 354–387.Find this resource:

Coleman, J. L. (2001b). Naturalized jurisprudence and naturalized epistemology. Philosophical Topics, 29(1/2), 113–126.Find this resource:

Coleman, J. L. (2002). Risks and wrongs. Oxford: Oxford University Press.Find this resource:

Coleman, J. L. (2007). Beyond the separability thesis: Moral semantics and the methodology of jurisprudence. Oxford Journal of Legal Studies, 27(4), 581–608.Find this resource:

Cooter, R. D. (1984). Prices and sanctions. Columbia Law Review, 84(6), 1523–1560.Find this resource:

Dari-Mattiacci, G. (2005). Errors and the functioning of tort liability. Supreme Court Economic Review, 13, 165–187.Find this resource:

Dari-Mattiacci, G., & Franzoni, L. A. (2014). Innovative negligence rules. American Law and Economics Review, 16(2), 333–365.Find this resource:

Dari-Mattiacci, G., & Garoupa, N. (2009). Least-cost avoidance: The tragedy of common safety. Journal of Law, Economics, & Organization, 25(1), 235–261.Find this resource:

De Mot, J., & Depoorter, B. (2011). Technology and torts: Memory costs, nondurable precautions and interference effects. International Review of Law and Economics, 31(4), 284–290.Find this resource:

Demsetz, H. (1967). Toward a theory of property rights. The American Economic Review, 57(2), 347–359.Find this resource:

Diamond, P. A. (1974). Single activity accidents. Journal of Legal Studies, 3(1), 107–164.Find this resource:

Diamond, P. A., & Mirrlees, J. A. (1975). On the assignment of liability: The uniform case. The Bell Journal of Economics, 6(2), 487–516.Find this resource:

Dingle, A. E. (1982). “The monster nuisance of all”: Landowners, alkali manufacturers, and air pollution, 1828–64. The Economic History Review, 35(4), 529–548.Find this resource:

Dorfman, A., & Assaf, J. (2015). The fault of trespass. University of Toronto Law Journal, 65(1), 48–98.Find this resource:

Ehrlich, I., & Posner, R. A. (1974). An economic analysis of legal rulemaking. Journal of Legal Studies, 3(1), 257–286.Find this resource:

Ellenberg, J. (2014). How not to be wrong: The power of mathematical thinking. New York: Penguin Press.Find this resource:

Ellis, D. D., Jr. (1982). Fairness and efficiency in the law of punitive damages. Southern California Law Review, 56(1), 1–78.Find this resource:

Epstein, R. A. (1979). Causation and corrective justice: A reply to two critics. Journal of Legal Studies, 8(3), 477–504.Find this resource:

Fennell, L. A. (2006). Efficient trespass: The case for bad faith adverse possession. Northwestern University Law Review, 100(3), 1037–1096.Find this resource:

Fennell, L. A. (2018). Accidents and aggregates. William & Mary Law Review, 59(6), 2371–2446.Find this resource:

Friedman, M. (1966). The methodology of positive economics. In M. Friedman (Ed.), Essays in Positive Economics (pp. 3–42). Chicago, IL: University of Chicago Press.Find this resource:

Gennaioli, N., & Shleifer, A. (2007). The evolution of common law. Journal of Political Economy, 115(1), 43–68.Find this resource:

Gilles, S. G. (1994). The invisible hand formula. Virginia Law Review, 80(5), 1015–1054.Find this resource:

Gilles, S. G. (2002). The emergence of cost–benefit balancing in English negligence law. Chicago–Kent Law Review, 77(2), 489–586.Find this resource:

Goldberg, J. C. P., & Zipursky, B. C. (2009). Intervening wrongdoing in tort: The Restatement (Third)’s unfortunate embrace of negligent enabling. Wake Forest Law Review, 44, 1211–1246.Find this resource:

Goldberg, J. C. P., & Zipursky, B. C. (2010). Torts as wrongs. Texas Law Review, 88(5), 917–986.Find this resource:

Goldberg, J. C. P., & Zipursky, B. C. (2013). Civil recourse defended: A reply to Posner, Calabresi, Rustad, Chamallas, and Robinette. Indiana Law Journal, 88(2), 569–610.Find this resource:

Goudkamp, J. (2011). Insanity as a tort defence. Oxford Journal of Legal Studies, 31(4), 727–754.Find this resource:

Grady, M. F. (1983). A new positive economic theory of negligence. Yale Law Journal, 92(5), 799–830.Find this resource:

Grady, M. F. (1988a). Common law control of strategic behavior: Railroad sparks and the farmer. Journal of Legal Studies, 17(1), 15–42.Find this resource:

Grady, M. F. (1988b). Discontinuities and information burdens: A review of The Economic Structure of Tort Law by William M. Landes and Richard A. Posner. George Washington Law Review, 56(3), 658–678.Find this resource:

Grady, M. F. (1988c). Why are people negligent? Technology, nondurable precautions, and the medical malpractice explosion. Northwestern University Law Review, 82(2), 293–334.Find this resource:

Grady, M. F. (1989). Untaken precautions. Journal of Legal Studies, 18(1), 139–156.Find this resource:

Grady, M. F. (1990). Multiple tortfeasors and the economy of prevention. Journal of Legal Studies, 19(2), 653–678.Find this resource:

Grady, M. F. (1994a). Modern accident law does not fit corrective justice theory. Jahrbuch für Recht und Ethik/Annual Review of Law and Ethics, 2, 7–35.Find this resource:

Grady, M. F. (1994b). Positive theories and grown order conceptions of the law. Southwestern University Law Review, 23(3), 461–468.Find this resource:

Grady, M. F. (1994c). Res ipsa loquitur and compliance error. University of Pennsylvania Law Review, 142(3), 887–947.Find this resource:

Grady, M. F. (1995). Legal evolution and precedent. Jahrbuch für Recht und Ethik/Annual Review of Law and Ethics, 3, 147–182.Find this resource:

Grady, M. F. (1996). Accident law seeks to minimize insurance effects. Michigan Law and Policy Review, 1, 11–42.Find this resource:

Grady, M. F. (1998). Efficient negligence. Georgetown Law Journal, 87(2), 397–420.Find this resource:

Grady, M. F. (2009). Unavoidable accident. Review of Law & Economics, 5(1), 178–231.Find this resource:

Grady, M. F. (2013). Causation and foreseeability. In J. Arlen (Ed.), Research handbook on the economics of torts (pp. 114–148). Northampton, MA: Edward Elgar.Find this resource:

Grady, M. F. (2014). Marginal causation and injurer shirking. Journal of Tort Law, 7(1–2), 1–33.Find this resource:

Grady, M. F. (2019). Justice luck in negligence law. Revus.Find this resource:

Grey, T. C. (1983). Langdell’s orthodoxy University of Pittsburgh Law Review, 45(1), 1–53.Find this resource:

Grey, T. C. (2001). Accidental torts. Vanderbilt Law Review, 54(3), 1225–1284.Find this resource:

Hayek, F. A. (1979). Law, legislation and liberty: A new statement of the liberal principles of justice and political economy. Chicago, IL: University of Chicago Press.Find this resource:

Hershovitz, S. (2010). Harry Potter and the trouble with tort theory. Stanford Law Review, 63(1), 67–114.Find this resource:

Hicks, J. R. (1966). The theory of wages. New York, NY: St. Martin’s Press.Find this resource:

Holmes, O. W. (1870). Codes, and the arrangement of the law. American Law Review, 5(1), 1–13.Find this resource:

Holmes, O. W. (1873). The theory of torts. American Law Review, 7(5), 652–663.Find this resource:

Holmes, O. W. (1881). The common law. Boston, MA: Little, Brown & Co.Find this resource:

Holmes, O. W. (1997). The path of the law. Harvard Law Review, 110(5), 991–1009.Find this resource:

Hume, D. (1921). An enquiry concerning human understanding. Chicago, IL: Open Court.Find this resource:

Hylton, K. N. (2009). Tort duties of landowners: A positive theory. Wake Forest Law Review, 44(4), 1049–1070.Find this resource:

Hylton, K. N. (2010a). 2009 Monsanto lecture: Intent in tort law. Valparaiso University Law Review, 44(4), 1217–1242.Find this resource:

Hylton, K. N. (2010b). The economics of public nuisance law and the new enforcement actions. Supreme Court Economic Review, 18(1), 43–76.Find this resource:

Hylton, K. N. (2012). The economics of necessity. Journal of Legal Studies, 41(2), 269–289.Find this resource:

Hylton, K. N. (2016). Tort law: A modern perspective. Cambridge, U.K.: Cambridge University Press.Find this resource:

Hylton, K. N., & Lin, H. (2013). Negligence, causation, and incentives for care. International Review of Law and Economics, 35, 80–89.Find this resource:

Jones, W. K. (1992). Strict liability for hazardous enterprise. Columbia Law Review, 92(7), 1705–1779.Find this resource:

Kant, I. (1996). The metaphysics of morals (M. J. Gregor, Trans.). Cambridge, U.K.: Cambridge University Press.Find this resource:

Keating, G. C. (2012). The priority of respect over repair. Legal Theory, 18(3), 293–337.Find this resource:

Kelley, P. J. (2003). Infancy, insanity, and infirmity in the law of torts. American Journal of Jurisprudence, 79, 179–252.Find this resource:

Kelley, P. J., & Wendt, L. A. (2002). What judges tell juries about negligence: A review of pattern jury instructions. Chicago–Kent Law Review, 77(2), 587–682.Find this resource:

Kuhn, T. S. (1983). Rationality and theory choice. The Journal of Philosophy, 80(10), 563–570.Find this resource:

Lakatos, I. (1970). History of science and its rational reconstructions. PSA: Proceedings of the Biennial Meeting of the Philosophy of Science Association, 1970, 91–136.Find this resource:

Landes, W. M., & Posner, R. A. (1980). Joint and multiple tortfeasors: An economic analysis. Journal of Legal Studies, 9(3), 517–555.Find this resource:

Landes, W. M., & Posner, R. A. (1981). The positive economic theory of tort law. University of Georgia Law Review, 15(1), 851–924.Find this resource:

Landes, W. M., & Posner, R. A. (1987). The economic structure of tort law. Cambridge, MA: Harvard University Press.Find this resource:

Leoni, B. (1991). Freedom and the law. Indianapolis, IN: Liberty Fund.Find this resource:

Levmore, S. (1986). Waiting for rescue: An essay on the evolution and incentive structure of the law of affirmative obligations. Virginia Law Review, 72(5), 879–941.Find this resource:

Llewellyn, K. N. (2008). The bramble bush: The classic lectures on the law and law school. New York, NY: Oxford University Press.Find this resource:

Malone, W. S. (1956). Ruminations on cause-in-fact. Stanford Law Review, 9(1), 60–99.Find this resource:

Merrill, T. W. (1985). Trespass, nuisance, and the costs of determining property rights. Journal of Legal Studies, 14(1), 13–48.Find this resource:

Moore, M. S. (2009). Causation and responsibility: An essay in law, morals, and metaphysics. Oxford: Oxford University Press.Find this resource:

Pigou, A. C. (1920). The economics of welfare. London: Macmillan & Co.Find this resource:

Popper, K. (2002a). Conjectures and refutations: The growth of scientific knowledge. London: Routledge.Find this resource:

Popper, K. (2002b). Popper: The logic of scientific discovery. London: Routledge Classics.Find this resource:

Posner, R. A. (1972). A theory of negligence. Journal of Legal Studies, 1(1), 29–96.Find this resource:

Posner, R. A. (1980). The ethical and political basis of the efficiency norm in common law adjudication. Hofstra Law Review, 8(3), 487–508.Find this resource:

Posner, R. A. (1981a). The concept of corrective justice in recent theories of tort law. Journal of Legal Studies, 10(1), 187–206.Find this resource:

Posner, R. A. (1981b). The economics of justice. Cambridge, MA: Harvard University Press.Find this resource:

Posner, R. A. (2013). Instrumental and noninstrumental theories of tort law. Indiana Law Journal, 88(2), 469–526.Find this resource:

Posner, R. A. (2014). Economic analysis of law (9th ed.). New York: Wolters Kluwer Law & Business.Find this resource:

Powell, L. F. (1957). Contributory negligence: A necessary check on the American jury. American Bar Association Journal, 43(11), 1005–1062.Find this resource:

Prosser, W. L. (1971). Handbook of the law of torts. St. Paul, MN: West.Find this resource:

Pylman, D. J. (2010). Res ipsa loquitur in the Restatement (Third) of Torts: Liability based upon naked statistics rather than real evidence. Chicago–Kent Law Review, 84(3), 907–942.Find this resource:

Rizza, D. (2014). Arrow’s theorem and theory choice. Synthese, 191(8), 1847–1856.Find this resource:

Shavell, S. (1980a). An analysis of causation and the scope of liability in the law of torts. Journal of Legal Studies, 9(3), 463–516.Find this resource:

Shavell, S. (1980b). Strict liability versus negligence. Journal of Legal Studies, 9(1), 1–25.Find this resource:

Shavell, S. (1983). Torts in which victim and injurer act sequentially. Journal of Law and Economics, 26(3), 589–612.Find this resource:

Shavell, S. (1987). Economic analysis of accident law. Cambridge, MA: Harvard University Press.Find this resource:

Simpson, A. W. B. (1984). Legal liability for bursting reservoirs: The historical context of “Rylands v. Fletcher.” Journal of Legal Studies, 13(2), 209–264.Find this resource:

Stremitzer, A., & Tabbach, A. D. (2014). The robustness case for proportional liability. B. E. Journal of Theoretical Economics, 14(1), 371–395.Find this resource:

Wambaugh, E. (1892). The study of cases: A course of instruction in reading and stating reported cases, composing head–notes and briefs, criticising and comparing authorities, and compiling digests. Boston, MA: Little, Brown.Find this resource:

Weinrib, E. J. (1980). The case for a duty to rescue. The Yale Law Journal, 90(2), 247–293.Find this resource:

Weinrib, E. J. (1983). Toward a moral theory of negligence law. Law and Philosophy, 2(1), 37–62.Find this resource:

Weinrib, E. J. (1985). The insurance justification and private law. Journal of Legal Studies, 14(3), 681–687.Find this resource:

Weinrib, E. J. (1987a). Causation and wrongdoing. Chicago–Kent Law Review, 63(3), 407–450.Find this resource:

Weinrib, E. J. (1987b). Law as a Kantian idea of reason. Columbia Law Review, 87(3), 472–508.Find this resource:

Weinrib, E. J. (1995). The idea of private law. Cambridge, MA: Harvard University Press.Find this resource:

Weinrib, E. J. (2002). Corrective justice in a nutshell. University of Toronto Law Journal, 52(4), 349–356.Find this resource:

Wells, C. P. (1990). Tort law as corrective justice: A pragmatic justification for jury adjudication. Michigan Law Review, 88(8), 2348–2413.Find this resource:

Wright, R. W. (1985). Actual causation vs. probabilistic linkage: The bane of economic analysis. Journal of Legal Studies, 14(2), 435–456.Find this resource:

Wright, R. W. (2003). Hand, Posner, and the myth of the Hand formula. Theoretical Inquiries in Law, 4(4), 145–274.Find this resource:

Zipursky, B. C. (2003). Civil recourse, not corrective justice. Georgetown Law Journal, 91(3), 695–756.Find this resource:


(1.) Any technology that increases the productivity of “nondurable precaution” also increases the amount of negligent behavior because lapses in nondurable precaution often create negligence liability. Thus, the invention and deployment of the passenger train increased the occasion for frequently checking whether the brakes are still working. The association between the increasing volume of negligence cases and the unfolding of the Industrial Revolution is described in Posner (1972) and Grady (1988c, 1994c).

(2.) A 19th-century legal scholar and the first dean of the Harvard Law School, C. C. Langdell, argued that theories about the legal rule established by a line of cases were the same as positive scientific theories. Nevertheless, once Langdell possessed a reasonable legal theory about the legal rule, he sometimes used the theory normatively to criticize nonconforming cases, an enterprise impossible in physics or chemistry. For an insightful description of Langdell’s ideas, as well as an outstanding introduction to 20th-century legal scholarship, see Grey (1983).

(3.) Holmes (1870) wrote, “It is the merit of the common law that it decides the case first and determines the principle afterwards. . . . It is only after a series of determinations on the same subject-matter, that it becomes necessary to “reconcile the cases,” as it is called, that is, by a true induction to state the principle that has until then been obscurely felt. And this statement is often modified more than once by new decisions before the abstracted general rule takes its final shape.” Although Holmes, unlike David Hume (1921), believed in the possibility of induction, another view is that reconciling cases of different results yields a “conjecture” of a legal principle, which may or may not withstand the test of time and further cases. See also Grady (1995).

(4.) Some have criticized Karl Popper because he laid so much stress on falsifiability, but since Popper’s theory is itself a positive theory of good positive theories, he may have wanted to keep his own theory simple so that it could be replaced by a better theory, which was his notion of how knowledge grows. Later theories of when a positive theory is good are more complicated and for that reason may be immunized against falsification. More complicated theories of theory choice are discussed in Kuhn (1983), Lakatos (1970), and Rizza (2014). See also Friedman (1966), which is a relatively parsimonious theory and otherwise consistent with Popper’s ideas.

(5.) As will be discussed below, the philosophers’ civil recourse theory is an apparent tautology. See Posner (2013). The civil recourse theory has simple (Goldberg & Zipursky, 2010) and complicated (Coleman, 2002) versions, but both seem unfalsifiable through the examination of conflicting cases, or otherwise.

(6.) Some philosophical theorists write as if the purpose of theory is to “justify” tort law rather than to gain fresh insight about its workings and social effects. See Coleman (2001a, 2001b, 2007). The aspect of tort that most needs justification from a philosophical point of view is why negligence law often and predictably imposes liability on people for acts that were socially beneficial, for instance, the efficient errors that outstanding medical professionals make in large numbers over long careers. See Grady (1994b).

(7.) More modern intentional torts also exist, such as “intentional infliction of emotional distress,” intentional interference with contractual relations, and the like, but these are beyond the scope of this article.

(8.) In legal circles, “prima facie” means that something is true unless something further is shown. Hence, battery is inefficient on its face, but if done in self-defense, it could be efficient.

(9.) A good definition of Kaldor–Hicks efficiency was given by an English judge in a tort case 40 years before the birth of either economist Kaldor or economist Hicks. In 1862, Baron Bramwell of the English Court of Exchequer Chamber said:

It is for the public benefit that trains should run, but not unless they pay their expenses. If one of these expenses is the burning down of a wood of such value that the railway owners would not run the train and burn down the wood if it were their own, neither is it for the public benefit that they should if the wood is not their own. If, though the wood were their own, they still would find it compensated them to run trains at the cost of burning the wood, then they obviously ought to compensate the owner of such wood, not being themselves, if they burn it down in making their gains. So in like way in this case a money value indeed cannot easily be put on the plaintiff’s loss, but it is equal to some number of pounds or pence, £10, £50 or what not: unless the defendant’s profits are enough to compensate this, I deny that it is for the public benefit he should do what he has done; if they are, he ought to compensate.

Bramwell, B., in Bamford v. Turnley, 122 Eng. Rep. 25, 33 (Exch. Ch. 1862), rev’g, 122 Eng. Rep. 25 (Q.B. 1860). Bramwell argued that strict liability should apply to railroad fires, but his argument neglected consideration of the moral hazard this rule would create for adjacent landowners, a problem that the accepted rule of negligence does not create.

(10.) 71 A. 188 (Vt. 1908).

(11.) 124 N.W. 221 (Minn. 1910).

(12.) See London Borough of Southwark v. Williams, [1971] Ch. 734 (C.A. 1970).

(13.) [1911] A.C. 355 (Cape of Good Hope P.C.).

(14.) The rule of no liability in this case is similar to, but more dramatic than, the rule prevailing for purely economic losses. For instance, when a building falls because of the contractors’ negligence and closes down a commercial section of a big city, the New York courts (and most courts) do not allow recovery for the lost business if that was the only loss suffered. Economists have explained this rule by observing that the business lost by closed-down merchants will likely be gained by other merchants. Instead of real social cost that results in most negligence cases in which “actual harm” exists, as when a defendant damages property or injures a person or kills a person, in a case of “pure economic loss” no increased scarcity results from the negligence. It would not make good sense from an economic point of view to induce persons to invest in socially costly precaution merely to prevent wealth transfers. That is in any event the courts’ policy. See Bishop (1982).

(15.) See Ellis v. D’Angelo, 253 P.2d 675 (Cal. App. 1953) (four-year-old defendant who rudely pushed his babysitter was held liable for substantial damages resulting from her fall); McGuire v. Almy, 8 N.E.2d 760 (Mass. 1937) (woman with mental illness liable for striking her attendant with lowboy leg); Williams v. Kearbey, 775 P.2d 670 (Kan. App. 1989) (14-year-old with mental illness civilly liable for killing junior high principal and wounding three others).

(16.) See, e.g., Garratt v. Dailey, 279 P.2d 1091 (Wash. 1955).

(17.) See, e.g., McGuire v. Almy, 8 N.E.2d 760 (Mass. 1937).

(18.) The first cascade of negligence cases came from London carriage accidents in the 1790s when street traffic first became congested.

(19.) United States v. Carroll Towing Co., 159 F.2d 169 (2d Cir. 1947), rev’g, Conners Marine Co. v. Pennsylvania R. Co., 66 F. Supp. 396 (E.D.N.Y. 1946).

(20.) An additional aspect of these early models was the omission of the cause-in-fact limitation on negligence liability. See, e.g., R. D. Cooter (1984). An infinitesimally negligent actor would thus expect to face liability for risk that reasonable care would not have prevented. This aspect of these models caused their makers to predict zero negligence in the real world.

(21.) See, e.g., Parsons v. Crown Disposal Co., 936 P.2d 70 (Cal. 1997) (plaintiff could not get to jury after trash truck frightened his horse and injured him because driver’s failure to look for horses before dumping bin was not condemned by cost–benefit analysis).

(22.) In the United Kingdom, where juries no longer decide negligence cases, the formula can be a decision rule for judges. See Gilles (2002).

(23.) See Parsons v. Crown Disposal Co., 936 P.2d 70 (Cal. 1997).

(24.) Stigler’s idea about the efficiency of sometimes missing your flight is discussed in Ellenberg (2014).

(25.) The distinction between compliance-using and compliance-saving technology closely parallels John Hicks’s (1966) distinction between “labor-using” and “labor-saving” technology. Nondurable precaution is labor, and durable precaution is capital.

(26.) That the driver’s speed brought him to the place of the accident at exactly the wrong time is not a reason for liability, though this frequently rejected argument is usually seen to fall under the proximate cause limitation, discussed below. See, e.g., Cunillera v. Randall, 608 N.Y.S.2d 441 (App. Div. 1994).

(27.) Juries are not instructed on the Learned Hand formula (Kelley & Wendt, 2002), and are not bound by it, but courts exclude from juries those cases in which the plaintiff’s theory of the untaken precaution exceeds the demands of the Learned Hand formula or similar cost–benefit analysis (Grady, 2019). Juries are often asked whether the untaken precaution was a “substantial factor” in producing the accident, but even if they answer “yes,” their decision can be overturned by a court that later finds the untaken precaution was not a “but–for” cause of the accident, that is, would not have probably prevented it.

(28.) See Herskovits v. Group Health Cooperative, 664 P.2d 474 (Wash. 1983), which seems to be an example of both this HMO’s practice policy and the law’s response to it.

(29.) See, for example, Gray v. Esslinger, 130 P.2d 24 (N.M. 1942) (to catch rare and elusive bus, plaintiff’s decedent tried to run across street in front of car, but car was speeding so fast it struck him; jury should have been instructed that decedent’s contributory negligence was no defense to willful and wanton negligence by defendant).

(30.) See, e.g., Minnegren v. Nozar, 208 Cal. Rptr. 3d 655 (Ct. App. 2016) (jury allowed to forgive motorist who negligently failed to yield right of way to plaintiff and crashed into her).

(31.) See Wiener v. Southcoast Childcare Centers, Inc., 88 P.3d 517 (Cal. 2004) (no liability for childcare center when its fence surrounding playground was weak, but murderer intentionally crashed car through it, killing plaintiffs’ children).

(32.) 159 Eng. Rep. 299 (Exch. 1863).

(33.) The English courts had previously applied a similar doctrine to cases in which a passenger was injured aboard a defendant’s public conveyance, such as a stagecoach or a train. Byrne extended these common carrier cases to a broader range of similar accident situations. See, for example, Christie v. Griggs, 170 Eng. Rep. 1088 (K.B. 1809) (allowed plaintiff’s claims to reach the jury on mere proof that stagecoach wrecked); Skinner v. London, Brighton & South Coast Ry., 155 Eng. Rep. 345 (Exch. 1850) (allowed plaintiff’s claims to reach the jury on proof that he was a passenger on defendant’s train when it wrecked).

(34.) Another element of the doctrine is that the plaintiff should not have contributed to his own injury. For glosses of the doctrine, see, for example, Carpue v. London & Brighton Ry., 114 Eng. Rep. 1431 (K.B. 1844); Scott v. London & St. Katherine Docks Co., 159 Eng. Rep. 665 (Ex. 1865); Larson v. St. Francis Hotel, 188 P.2d 513 (Cal. Ct. App. 1948). In each of these cases the plaintiff didn’t know and couldn’t find out what the respective defendants’ untaken precautions were.

(35.) This third possibility is that the accident wasn’t under the defendant’s R-curve at all, as when the defendant had no opportunity to prevent the accident, but someone else may have possessed such an opportunity.

(36.) Cf. Higgs v. Maynard, Har. & Ruth 581 (C.P. 1866), where the court held that a plaintiff walking through an alley and struck by a shard of glass falling from the defendant’s warehouse could not get to a jury on a res ipsa theory. The cause of the falling glass was a worker’s inadvertently allowing a ladder within the warehouse to fall against a glass window. The required rate of precaution with ladders within warehouses is much less than that for heavy barrels being lowered from warehouses over sidewalks.

(37.) 94 N.E. 386 (Mass. 1911).

(38.) Compare Goddard v. Boston & Maine R.R., 60 N.E. 486 (Mass. 1901), where the court nonsuited the plaintiff because there was no evidence whether the banana peel on which the plaintiff slipped was gritty or fresh. See also Gordon v. American Museum of Natural History, 492 N.E.2d 774 (N.Y. 1986) (no negligence could be inferred because wax paper on which plaintiff slipped was fresh).

(39.) Compare Newing v. Cheatham, 540 P.2d 33 (Cal. 1975).

(40.) See, for example, D’Anna v. United States, 181 F.2d 335 (4th Cir. 1950) (defendant’s stunt plane, performing maneuvers over populous city, dropped gas tank on plaintiff’s fruit stand; accident signature entailed evidence that defendant had not inspected small pin that held gas tank in place); Sharp v. LaBrec, Inc., 642 N.E.2d 990 (Ind. App. 1994) (defendant’s crane bumped 11-ton boiler into plaintiff’s decedent; accident signature included evidence that defendant’s employees had placed insufficient “cribbing” underneath crane outriggers, which were resting on muddy soil, and also that this cribbing had failed during fatal lift).

(41.) See, for example, Velez v. Sebco Laundry Systems, Inc., 178 F. Supp. 2d 336 (S.D.N.Y. 2001) (defendant entitled to summary judgment against plaintiff’s res ipsa claim that clothes dryer window fell on her foot).

(42.) The current legal glosses of res ipsa are notoriously imprecise. See Pylman (2010) and sources cited therein.

(43.) The philosopher Immanuel Kant sensibly opposed legal duties of beneficence (see Kant, 1996), but many of his legal philosopher followers support these legal duties, so long as they are modest. See, for example, Weinrib (1980). Immanuel Kant correctly said that beneficence is a moral duty.

(44.) See, for example, Saelzler v. Advanced Group 400, 23 P.3d 1143 (Cal. 2001) (tenants could not afford second shift of security guards that might have prevented criminal assault).

(45.) Spartan Steel & Alloys Ltd. v. Martin & Co. (Contractors) Ltd., [1973] Q.B. 27 (C.A. 1972).

(46.) Spartan Steel & Alloys Ltd. v. Martin & Co. (Contractors) Ltd.

(47.) See 532 Madison Avenue Gourmet Foods, Inc. v. Finlandia Center, Inc., 750 N.E.2d 1097 (N.Y. 2001). This no-liability scenario is often repeated. See, for example, Rickards v. Sun Oil Co., 41 A.2d 267 (N.J. Super. 1945) (no liability to island merchants when defendant negligently crashed into drawbridge and isolated them from their customers).

(48.) See Bartolone v. Jeckovich, 481 N.Y.S.2d 545 (App. Div. 1984) (plaintiff who suffered minor physical injuries from defendant’s negligence able to recover substantial damages for emotional distress when physical injuries precipitated mental illness).

(49.) See Dillon v. Legg, 441 P.2d 912 (Cal. 1968).

(50.) See Clomon v. Monroe City School Board (La. App. 1986) (defendant’s school bus driver persistently and negligently let disabled schoolchild out in dangerous place, and plaintiff ran him down); Kately v. Wilkinson, 195 Cal. Rptr. 902 (Ct. App. 1983) (defective steering wheel on defendants’ ski boat caused plaintiff to repeatedly crash boat into a child and to kill her).

(51.) Williamson v. Bennett, 112 S.E.2d 48 (N.C. 1960) (plaintiff unable to recover for emotional distress when defendant negligently crashed into her and she falsely and unreasonably assumed she had struck a child and allegedly suffered emotional distress because of that assumption).

(52.) See, for example, J’Aire Corp. v. Gregory, 598 P.2d 60 (Cal. 1979) (plaintiff able to recover for lost profits when defendant, through deliberate negligence, delayed installation of air conditioning in remodel of airport, thus causing plaintiff’s airport restaurant substantial economic losses); Ultramares Corp. v. Touche, 174 N.E. 441 (N.Y. 1931) (defendant accountants would be liable for purely economic loss to third-party lender who relied on negligent audit if negligence was deliberate or willful and wanton); Ochoa v. Superior Court, 703 P.2d 1 (Cal. 1985) (defendant liable for plaintiff mother’s emotional distress when defendant’s juvenile hall employees deliberately withheld proper medical treatment to plaintiff’s dying son who was obviously in need of it).

(53.) See Knight v. Jewett, 834 P.2d 696 (Cal. 1992) (plaintiff unable to recover for negligently inflicted injury in touch football game).

(54.) See, for example, Gray v. Esslinger, 130 P.2d 24 (N.M. 1942) (to try to catch a rare and elusive bus, plaintiff’s decedent ran across street in front of car, but car was speeding so fast it struck him; jury should have been instructed that contributory negligence was no defense to willful and wanton negligence by defendant).

(55.) See, for example, Williams v. Harrison, 497 S.E.2d 467 (Va. 1998).

(56.) See, for example, Markwell v. Swift & Co., 272 P.2d 47 (Cal. App. 1954) (car hop who forgot about hazard created by defendant at drive-in restaurant barred from recovery as a matter of law). This social saving could be overstated, however, because especially toward the end of the contributory negligence era, courts allowed juries to forgive contributory negligence and to return “compromise verdicts,” thus anticipating the comparative negligence rule. See Powell (1957).

(57.) Rylands v. Fletcher, 159 Eng. Rep. 737 (Exch. 1865), rev’d, L.R. 1 Ex. 265 (Exch. Ch. 1866), aff’d, L.R. 3 H.L. 330 (1868).

(58.) See Attorney General ex rel. Rhondda Urban District Council v. Cory Brothers & Co., [1921] 1 A.C. 521 (H.L.) (defendants strictly liable when their large coal slag heap tumbled across property line onto plaintiffs’ premises).

(59.) As in the discussion of negligence, “unavoidable accident” is the risk remaining when reasonable or efficient precaution has been used by the actor. Because with many strictly liable activities the zone of what would be “negligent harm” is often large relative to the zone of unavoidable accident, many strict liability cases could also be successfully prosecuted as res ipsa loquitur cases. Plaintiffs prefer the strict liability theory because res ipsa loquitur negligence can be forgiven by a jury, whereas a strict liability case is not subject to jury forgiveness. As noted below in the text, the exemption from jury forgiveness is one of the principal advantages of strict liability because it creates more incentive to use care in especially dangerous situations.

(60.) See Sullivan v. Dunham, 55 N.E. 923 (N.Y. 1900) (defendant blasters strictly liable for blasted tree that struck and killed plaintiff’s decedent who was walking down a country road some distance from the blast).

(61.) See Central Trust & Savings Bank v. Toppert, 554 N.E.2d 820 (Ill. App. 1990) (no strict liability when plaintiff’s decedent participated in setting the fatal dynamite charge); Kent v. Gulf States Utilities, 418 So. 2d 493 (La. 1982) (no strict liability when plaintiff’s decedent brought long aluminum handle of rake he was using into contact with defendant’s high-tension wires); Doherty v. Ohio State University, No. 89AP–746, 1990 WL 86772 (Ohio App.) (no strict liability when plaintiff graduate student was operating apparently inadequate chemistry apparatus when it exploded).

(62.) Participant-plaintiffs would do better to sue on a negligence theory, which is usually also available, but subject to comparative negligence.