2 Torts

At Yale, I taught a small section (approximately eighteen students) three times; at UCLA, I taught a large section (approximately eighty students) every year, using the first eight editions of the Franklin & Rabin casebook (Reference Franklin1971 through Reference Franklin, Rabin and Green2006), which I chose for their material on tort law in action and California cases. I had had Guido Calabresi for half of my first-year torts course (intentional torts), which is part of the reason Yale hired me. But he had not yet published The Costs of Accidents (Reference Calabresi1970) and I was not influenced by it when I began teaching. It took me more than a decade to develop my own approach, which I articulated in a paper in The Politics of Law (Abel, Reference Abel and Kairys1982b) and later revisions (Abel, Reference Abel and Kairys1990b; Reference Abel and Kairys1998).

I believe my goal (like that of American law teachers’ alleged model, Socrates) was to unsettle students’ preconceptions. When a student complained that, however diligently he prepared for class, he never anticipated my questions, I took it as a compliment. My colleague Gary Schwartz (UCLA's resident authority on torts) once wrote that I could find nothing good to say about tort law. That's pretty accurate. Most tort doctrine is incoherent and most judicial opinions are poorly reasoned, conclusory, based on factual assertions for which judges adduce no evidence, probably because there is none.

My approach inevitably discomforted students, most of whom enter law school as naive pandectists, believing in a slot-machine theory of adjudication, a mechanical jurisprudence. Anxious to succeed in a novel environment, students grasp for a body of rules that, when applied to unambiguous facts, will dictate an outcome. They expect to be examined on those rules. But I wanted them to question the rules, examine the judges’ underlying justifications, refine their own moral intuitions and reach independent conclusions. Believing I should set an example by revealing my own political orientation, I began one semester by telling the class I was a democratic socialist – promptly losing their sympathy and my own credibility. To give material weight to my stress on meta-analysis, I divided the final exam into two equal parts: a factual hypothetical requiring a conventional legal analysis; and what students insisted on calling the ‘policy question’, which asked them to criticise the rules and choose among alternatives (fact situations included cougar attacks in Southern California mountains, ocycodone overdoses, 9/11, tobacco-related illnesses and gun-related injuries). I believe my approach was most successful at NYU, whose students were supremely confident, and least successful at CUNY, whose students were apprehensive about the bar exam (believing correctly that it dealt exclusively with black-letter law and knowing the school's pass rate was low). A significant proportion of my UCLA students routinely complained that I taught them no law (even though we read 800 pages of the casebook) because I questioned the reasoning of the cases we analysed. Some exhorted UCLA to ‘send the commie back to Russia’. (Because I was tenured, these hostile evaluations only delayed my merit-salary increases. But believing that students were likely to be far more biased against vulnerable untenured women and people of colour than they were against me – an old White guy whose politics they detested – I wrote a critique of our teaching evaluations (Abel, Reference Abel1990a) and agitated for changes, with no success.)

The casebook begins with a car driver who suffered an epileptic attack, injuring another, thereby raising questions about who should be responsible.Footnote ¹ It is an excellent vehicle for posing the alternatives of non-liability, fault-based liability and strict liability. I questioned the relationship between liability and moral culpability, arguing that the latter has been rendered largely obsolete by the rising cost of accidents, which forces plaintiffs to seek a deep pocket: faceless corporations, liability insurers, employers under respondeat superior and the wealthier of multiple defendants jointly and severally liable. I showed that the goal was not merely to transfer the accident cost from a plaintiff to a defendant, but rather to spread that cost. That conclusion led to comparing social insurance with private loss and liability insurance and discussing how the latter set premiums (often making crude distinctions based on age, gender and zip code). Private insurance is wastefully duplicative because potential victims will have to purchase loss insurance and potential defendants liability insurance. Adjusting their contributions (through subrogation and the collateral source rule) adds almost a dollar in transaction costs to deliver each dollar to the victim, whereas Social Security, by contrast, delivers that dollar for just eight cents. Finally, the case raised questions about the relative merits of criminal law, regulation and tort liability in promoting an efficient level of safety. Once students grasped these three goals (moral judgment, spreading and safety), they could begin to perceive the tensions between them.

I then turned to damages, which this casebook (like most) postpones to the end of the course, where it often receives cursory treatment because there are relatively few rules. I began with damages in the belief that students could not understand when tort law should offer a remedy unless they knew what that remedy was. Although the editors declared that the goal of tort law is to ‘return the plaintiff as closely as possible to his or her condition before the accident’, the case proves just the opposite.Footnote ² The plaintiff suffered a permanent serious injury to her leg while boarding a bus. The editors begin by explaining the single-judgment rule, offering several justifications but omitting the real reason: plaintiffs’ lawyers want to collect their contingent fees (25–50 per cent of damages). The rule creates an arbitrary distinction between plaintiffs who die moments before or after judgment and compels the jury to make impossible predictions about life expectancy, future earnings with and without the injury (increasingly speculative the younger the plaintiff) and future medical expenses (which tend to rise much faster than general inflation and are affected by scientific innovation). Even without these imponderables, salary replacement does not substitute for the experience of working; and reimbursing medical expenses does not exempt victims from medical procedures. Both predictions are further complicated by changing tax consequences, economic cycles and fluctuating rates of return.

Even if we could restore the plaintiff to the status quo ante, I questioned the desirability of doing so. After an accidental injury, should the state reproduce pre-existing inequalities of income, wealth, medical care and life expectancy associated with age, class, gender, sexual orientation, marital status, dependants, race and education? Should a wrongful-death case perpetuate the support dependants would have received from the decedent? Since most tort judgments are satisfied by liability insurance (often compulsory), the burdens of those unequal entitlements are shared by all insured, regardless of their income or wealth – a highly regressive form of taxation. It is morally repugnant to make the magnitude of the defendant's punishment for negligence (a capacious behavioural category) turn on the happenstance of the plaintiff's income and wealth. Even more troubling is the fact that entrepreneurs who make the decisions for accidents (in Calabresi's words) are compelled by the market to inflict greater risks on the poor than on the wealthy (the tort equivalent of the environmental justice critique of polluters).

All these objections to ‘special’ damages are compounded by ‘general’ damages for pain and suffering (Abel, Reference Abel2006). Here the pretence of restoring the status quo ante deteriorates into a parody of Bentham's hedonic calculus – the fiction that any pain can be cancelled by an equivalent pleasure. Jurors must simulate a market in sadomasochism, asking what they would require to be paid to experience the plaintiff's suffering. (The alternatives advanced by economists are just as unsatisfactory: willingness to pay (because the hypothetical is meaningless to someone who has not suffered the injury) and wage-risk premium (which assumes that the labour market is perfectly competitive and workers can choose not to work).) Using another case – a young man, about the same age and education level as my students, whose promising life was upended by horrible burnsFootnote ³ – I graphed what students said they would award for pain and suffering to show that it did not assume the normal bell-shaped distribution. Some voted for huge damages because nothing less could compensate for the catastrophic loss; others awarded little or nothing, for the same reason or because they felt the plaintiff's life had been changed rather than diminished. (These differences will lead lawyers on both sides to seek jurors whose demographics make them more or less empathetic towards the victim.) I then asked the outliers to deliberate as though they were jurors, exposing the difficulty they would have in reaching the consensus required for a verdict. Because there are no metrics (as this exercise showed), lawyers and judges use special damages to estimate and justify generals. But there is no reason to expect actual proportionality; and imposing one ensures that the inequalities inherent in specials (noted above) will infect generals and hence intensify unfairness in liability-insurance premiums and amplify unequal exposure to risk. Like the single-judgment rule, general damages are sometimes justified as defraying contingent fees – but this jettisons the fiction that generals are necessary to restore the plaintiff to the status quo ante.

This commodification – which Marx (Reference Marx1867/2017) saw as a defining feature of capitalism – transcends pain and suffering. In a secular perversion of the Faustian compact, tort law awards damages to those who recover from a coma or suffer diminished life expectancy, as though money can buy experience to make up for lost years. Paradoxically, there is no compensation for the greatest loss – death – and uncertainty about compensating those in a persistent vegetative state, because the dead and comatose cannot buy substitute pleasures. (Students are intrigued by the moral perversity that it is cheaper to kill than to maim – until I remind them that killing may be criminal.) Nor is there a principled answer to whether pain-and-suffering awards should survive death (benefiting dependants). And were we to compensate the dead for death, should we also compensate the living for being born (so-called wrongful-life cases)?

Even these intractable problems of incommensurability are dwarfed by the award of damages for harm to relationships through the injury or death of another. Which relationships should be protected: spouses, cohabitants, parents and children, siblings, other kin, friends, celebrities? If we consider the physical victim's characteristics in evaluating loss of enjoyment (for example, playing a musical instrument, engaging in athletic activities), should we also calibrate damages for impaired relationships in terms of the injured person's characteristics – physical attractiveness, sociability, cooking, cognitive skills, caregiving, sexual performance – emulating the ratings of online dating? Should we do this in so-called wrongful-birth cases for an unplanned pregnancy or a child born with a disability? What about a child who is simply below average (unlike those in Garrison Keillor's fictional Lake Wobegon, where all were above average)? Should there be an offset for benefit conferred? A duty to mitigate damages by aborting or surrendering the child for adoption?

This analysis of damages unsettles two of tort law's fundamental justifications: moral judgment (since there is no proportionality between the wrongfulness of the defendant's conduct and the quantum of damages) and compensation (since damages cannot and perhaps should not restore the plaintiff to the status quo ante). The discussion also constructs the essential foundation for addressing the third, arguably most cogent, justification: encouraging an efficient level of safety (what Calabresi calls primary accident cost reduction). The casebook introduces negligence with a Cardozo opinion denying recovery as a matter of law (despite a jury verdict for the plaintiff affirmed by the Appellate Division) to a boy badly burned when a wire he was dangling while walking across a pedestrian bridge touched a trolley company's high-voltage line strung underneath.Footnote ⁴ Cardozo wrote a seductive brief in the guise of rendering a judicial opinion, misrepresenting ambiguous facts in the light most favourable to the defendant:

I helped students demystify this reasoning (preparing them to read other misleading Cardozo opinions): the trolley wires were suspended at least eighteen feet above the ground (a fact he omits); the boy's wire was just eight feet long; gravity made this contact happen and would have rendered Cardozo's imaginary scenario virtually impossible (it also would have established contributory negligence); and the eighteen-inch parapet obscured the boy's view of the trolley wire. Although Cardozo maligns the plaintiff as mischievous or thoughtless (implying he had never done foolish things as a child), the defendant did not assert a defence of contributory negligence because the plaintiff's standard of care was that of a twelve-year-old boy. (Cardozo's Supreme Court colleague Holmes (Reference Holmes1881) famously declared that ‘the life of the law has not been logic; it has been experience’. He failed to add, however, that it was the parochial experience of appellate judges.)

The book then presented Learned Hand's influential formulation of negligence as B < PL (the cost of safety is less than the cost of the accident discounted by its probability).Footnote ⁵ The formula (conventional cost–benefit analysis) makes clear that the legal conceptualisation of damages determines what should be spent on safety. I also showed that the formula should be ΔB<ΔPΔL: actors must consider the entire range of safety precautions and the total population of possible accidents, each with its associated probability. (Indeed, it is meaningless to speak of the probability of a unique event, which is vanishingly close to zero. Tort law acknowledges this in the ‘eggshell plaintiff’ doctrine: defendants take their plaintiffs as they find them because the vulnerabilities of the particular plaintiff are just one instance of those in the population of potential plaintiffs.) Instead of the prospective statistical approach demanded by cost–benefit analysis, however, triers of fact (juries acting within limits set by judges) reflect retrospectively on what happened in the case before them. (This divergence between economic and legal reasoning lays the foundation for my later comparison of the relative merits of negligence and strict liability.)

Medical malpractice illustrates additional problems generated by formulating negligence in economic terms. First, expert testimony is essential, raising questions about its availability (because of a conspiracy of silence) and reliability.Footnote ⁶ I asked students: if you were facing a serious health crisis, would you choose among possible diagnoses and treatments by consulting a random selection of ignorant laypeople? Second, informed consent poses a fundamental challenge: this doctrine seeks to protect the patient's autonomy, not to make the most efficient choice among alternative procedures, since the doctor is far more knowledgeable about medicine. Therefore, the justification must be deontological, not utilitarian. Yet tort law declines to embrace deontology unconditionally: although it entrusts the treatment decision to the patient, it insists on utilitarian criteria for causation (whether a reasonable patient with the missing information would have chosen a different procedure) and damages (the costs of following the unconsented procedure rather than the dignitary harm of being denied a choice, which cannot be quantified).

Students often simplify tort with the mantra: duty, breach (i.e. negligence), causation, damages. Yet courts rarely address the first, threshold issue. Particular duties of care historically were associated with unique dyads, but modern courts tend to assume a general duty of reasonable care. The situations in which courts still relax this duty are illuminating. Solicitude towards landowners may represent a feudal relic, but its preservation is a political choice: Norway (with just thirty-three people per square mile compared with 702 in the UK, where our legal doctrines originated) affirms the right not only to walk across all land, but even to camp on it for a night (if not within sight of a window). Courts defer to parents in raising children and prefer not to intrude into the ‘private’ sphere of families. Charitable immunity was a (misguided) subsidy (the poor should be grateful for shoddy care). Sovereign immunity originated in deference to the Crown.

Although there may be justifications for these lesser duties – families, charities and governments lack a profit motive; landowners profit only from business invitees – courts have questioned all these categories and contracted many of them. The ‘affirmative duty to act’, however, remains an intractable problem. Running on the Santa Monica beach, I generally have no duty to use reasonable care to rescue a drowning swimmer, but there are exceptions: commencing to act or a relationship with or reliance by the victim. Yet these concepts are hopelessly vague. Do I commence to act when I leave my house? Set foot on the beach? Stick a toe in the water? Do I have a relationship with a swimmer I see daily? Do I induce reliance by waving or exchanging a greeting? Is this misfeasance or nonfeasance? There are no answers because this is another situation in which incompatible moral discourses – utilitarianism and deontology – talk past one another. I may be able to save the swimmer at little or no cost (I usually do swim after my run). But the law refrains from imposing such an obligation out of respect for my autonomy. (Of course, autonomy is a political choice; a society could choose to favour community.) Given this unsatisfactory situation, I ask students to reflect on how social environments shape both the sense of obligation and behaviour: what distinguishes the neighbours in the Queens apartment house who ignored Kitty Genovese's cries for help from the altruism displayed following natural or man-made catastrophes (hurricanes, fires, earthquakes, terrorism) or in situations of common peril (backpacking in the Sierras).

The second issue in the student mantra – causation – is irrelevant to the two central goals of tort law: who caused the accident says nothing about the relative ability of plaintiffs and defendants to spread its costs; and we want to encourage an efficient level of safety whether or not risk has resulted in injury. It is the moral foundation of tort law that restrains it from imposing liability when the trier of fact lacks confidence that the defendant caused the plaintiff's injury. The book's example involves a city whose contaminated water supply was associated with an elevated level of typhoid cases, but it was impossible to determine whether any particular victim had been injured by the contamination or some other cause.Footnote ⁷ Allowing all typhoid sufferers to recover would unfairly punish the city for instances it did not cause; denying recovery to all would fail to motivate the city to be more careful; awarding all victims damages discounted by the probability of other causes would not spread their total accident costs.

Trapped by the incompatibility between science, which views causation probabilistically, and law, which must make dichotomous choices about liability, the New York court in 1919 adopted criteria that were either oxymoronic (reasonable certainty) or meaningless (reasonable possibilities). Although few cases explicitly address this problem, it is omnipresent: accident-avoidance costs (ΔB) can only be correlated probabilistically with accident reduction (ΔPΔL). Contemporary judges display no greater sophistication in dealing with probability, demanding a ‘reasonable degree of medical probability’,Footnote ⁸ rejecting ‘a payout scheme on the basis of a statistical chance’ because ‘to dispense with’ the requirement of ‘some degree of certainty’ is ‘to abandon the truth-seeking function of the law’,Footnote ⁹ complaining that ‘the use of statistics in trials is … unreliable, misleading, easily manipulated, and confusing to a jury’.Footnote ¹⁰ Just two years ago, in a landmark challenge to egregious gerrymandering, Chief Justice Roberts dismissed statistics as ‘sociological gobbledygook’.Footnote ¹¹ The judiciary is the last bastion of the shamelessly innumerate.

If ‘cause in fact’ confronts epistemological problems inherent in the incompatible discourses of science and law, proximate cause is law's self-inflicted wound (and has nothing to do with causation) (Abel, Reference Abel2002). The doctrine, which introduces a requirement of ‘reasonable foreseeability’, is redundant, since the Hand formula already considers the information cost of anticipating the accident as part of the cost of safety (ΔB). (At this point, I pretentiously wrote Occam's razor on the blackboard: entia non multiplicanda sunt praeter necessitatem.) Every year, I walked students painfully through the unholy trinity of cases, showing that a different narrative strategy could obviate the need for proximate cause: the accumulation of flammable benzene vapours in the hold of the Thrasyvoulos Footnote ¹²; the information cost of learning that bunkering oil spilled in the cold waters of Sydney harbour could catch fire and burn after impregnating cotton wasteFootnote ¹³; the information cost of inspecting the packages of every Long Island Railroad passenger (an inconvenience all too familiar to contemporary air travellers) or an alternative story about inadequately secured baggage scales.Footnote ¹⁴

As they did in causation, the criteria judges offered for proximate cause were meaningless: ‘direct’ cause (Polemis) ‘whatever that may mean’ (Wagon Mound); ‘remoteness in time and space’ (Andrews in Palsgraf) (rendered obsolete by Cheronobyl, which spread radiation from Italy to Norway and created waste that will remain dangerous for more than 10,000 years); ‘probable consequences’ (Wagon Mound) (all consequences are ipso facto probable). The judges’ justifications were equally incoherent: ‘general public sentiment’, ‘current ideas of justice or morality’ that one should not be liable for ‘all consequences’ of ‘an act of negligence, however slight or venial’ (Wagon Mound) (but there is no proportionality between negligence and tort damages); ‘all are agreed that some limitation there must be’ (Wagon Mound) (Why? I don't agree). I reserved my deepest scorn for Judge Andrews's rambling dissent in Palsgraf:

His solution was the ‘substantial-factor’ test:

After reciting this mishmash of platitudes in my most sarcastic tones, I threw up my hands in mock despair, declaring that, if arbitrariness and expedience were the best we could offer, we might as well go home. (This elicited cheers from the students.) Laws are the reasons the state gives for exercising power. Without persuasive reasons, law is mere ipse dixit, and judges and lawyers have no more legitimacy than laypeople. My students were bemused by this explosion. By this point in the semester, they had grown deeply cynical about judges’ efforts to articulate and justify rules (a cynicism I had encouraged). They preferred the Andrews opinion, which would have allowed the injured and innocent Ms Palsgraf to recover from the deep-pocket Long Island Railroad (which had been negligent in letting a passenger carrying fireworks jump aboard a moving train). And they sought comfort in jargon they could memorise to regurgitate on exams. Indeed, the academic support instructor at CUNY pleaded with me to teach the ‘substantial-factor’ test and, when I said I couldn't, undertook to do himself.

I tried to lighten my diatribe against the doctrine with a story. Clearing customs in New York on the way home from Europe, I was asked a routine question about what I did. I replied that I taught and, when pressed, that I taught at a university and finally that I taught at a law school. Apparently sensing some incongruity between my alleged profession and my appearance, the inspector wanted to know what I taught. When I mentioned torts, he followed up by asking whether I taught Palsgraf (proudly noting that his daughter was a law student). I repeated this incident to my torts class the next day, when I happened to be teaching Palsgraf. A student came up afterwards and told me that was no idle chit-chat. He had been a customs officer; this was standard protocol when you doubted the traveller's veracity. Every year thereafter, I told that story to students mystified by proximate cause, adding: remember Palsgraf if you have a stash in your backpack and don't want to be strip-searched.

The fact situations above showed how an act's effects could radiate widely through the physical world of space and time like ripples from a stone thrown into a pond (a metaphor many judges embraced). Causes ramifying through other media posed equally difficult questions. An event could transmit emotional consequences through the psyche: fear for self and others, trauma from witnessing another's injury, loss of consortium. Courts adopted bright-line rules, cabining recovery by reference to impact, zone of danger and the nature of the relationship rather than letting those variables influence the quantum of damages. Similarly, an event's economic consequences could ramify through the market. Here, too, courts promulgated bright-line rules, requiring that property be physically damaged, or compensating property owners for lost profits but not workers for lost earnings (since the former own the means of production – my one explicitly Marxist intervention). I argued that every invocation of proximate cause was simply a power grab by judges seeking to wrest decisional authority from juries as triers of fact in order to limit liability.

After covering the plaintiff's affirmative case, the book turned to defences. Should the law treat the actions of plaintiff and defendant symmetrically? This could not be justified by concern for safety: plaintiffs are motivated by self-preservation, defendants only by fear of liability. Nor could it be justified by spreading: defendants are more likely to spread the damages as a cost of doing business and to carry liability insurance than plaintiffs are to carry loss insurance (even though the transaction costs of claiming under the latter are lower). Therefore, the argument for contributory negligence must rest on a moral judgment that the parties are in pari delicto. But are those who injure themselves as culpable as those who injure others? If so, why do workers’ compensation and products liability disregard the plaintiff's fault? The diminished influence of moral judgment in tort law is evident in the shift from contributory negligence to comparative fault. Although the law lets people contract away their entitlement to sue before injury (just as they can settle claims after it), scepticism about freedom of contract sometimes persuades judges to reject these agreements, perhaps recognising that such agreements do not make plaintiffs safer or more likely to insure and may make defendants less safe and less likely to insure. (I used this discussion as an opportunity to raise ethical questions. Should a lawyer for a potential defendant draft an agreement not to sue, knowing it is unenforceable but may discourage meritorious claims? Alternatively, should a lawyer for the victim of a defective product advise the client to accept a settlement that is inflated (and will increase the lawyer's contingent fee) because it contains an agreement not to warn others endangered or injured by the product?)

When we reached assumption of risk, many students were exultant. They had started the course with a visceral mistrust of plaintiffs (like the one burned by McDonald's coffee),Footnote ¹⁵ whom they believed made frivolous claims (having been subjected to decades of propaganda by insurers, manufacturers, medical associations and local governments) (Haltom and McCann, Reference Haltom and McCann2004). A ‘vigorous young man’, who suffered a fractured knee cap (a serious injury in the 1920s) while riding ‘The Flopper’ to impress his future wife, sued the Coney Island amusement park.Footnote ¹⁶ Libertarians in the class cheered Cardozo for voicing discontents that had been intensifying all semester – and did so in Latin! ‘Volenti non fit injuria.’ (Students cherished such esoterica, sometimes sporting t-shirts emblazoned with ‘Res Ipsa Loquitur’.) Murphy's fall ‘was the very hazard that was invited and foreseen’. He ‘made his choice’ to join visitors ‘tumbling about the belt to the merriment of onlookers’. ‘The timorous may stay at home.’ Cardozo (proudly timorous) showed no sympathy for a young man who chose to join the ‘tumbling bodies and the screams and laughter’ for ‘merriment and fun’. To make his case, Cardozo told a characteristically partisan story. ‘Some quota of accidents was to be looked for in so great a mass’; true, but Murphy could not know how many. Indeed, the nurse at the emergency hospital maintained by Steeplechase (itself an admission of the riskiness of the amusement park) had cared for others injured on The Flopper; she testified that ‘none, however, had been badly injured or had suffered broken bones’. Cardozo's clincher was a profoundly misleading in terrorem argument: ‘One might as well say that a skating rink should be abandoned because skaters sometimes fall.’ But a decision for liability is not a prohibition; it only requires Steeplechase to pay the costs of its negligence. I asked the students: Who is better able to estimate and reduce the accident costs here, Murphy or Steeplechase? Who is more likely to insure or spread the costs? Is Murphy's ‘culpability’ in riding The Flopper equivalent to that of Steeplechase in designing and operating it? I anticipated the next set of issues by invoking libertarians’ own values, arguing that strict liability here would internalise the costs of such accidents among Flopper riders, which was both economically efficient and morally just. The best way to make Flopper riders ‘assume the risk’ of accidents is, paradoxically, to eliminate that defence.Footnote ¹⁷ (Those injured by the ride's negligence would recover, forcing Steeplechase to internalise its liability costs in the ticket price, which would be paid only by those who took the ride.)

One year, a libertarian student resisted my arguments by bravely maintaining that society should let Murphy's fractured knee cap go untreated if he could not afford medical care (there was no health insurance in the 1920s). I appreciated the student's principled stance (which made the Socratic dialogue more effective) and missed him in school the next year. He returned a year later and explained that he had gone hang-gliding the following summer, been caught by an unanticipated gust of wind when launching off a cliff and fallen hundreds of feet, breaking many bones. Lacking health insurance, he had relied on the government to pay his monumental medical costs. Shamefaced, he now agreed that those costs should have been internalised in the price of hang-gliding. I expressed my sympathy, regretting he had had to learn that lesson so painfully.

Many casebooks begin with intentional torts (although Franklin & Rabin ends with them, suggesting they are intended for a later elective). I frustrated students (who expected them on the bar exam) by choosing not to teach them at all because, although a staple of tort law before the Industrial Revolution, they now are rarely litigated. I assigned two papers: a psychological study of New Yorkers who swallowed personal affronts (Moriarty, Reference Moriarty1975) and an ethnographic account of why there were so few tort suits in a rural mid-western community (especially compared with contract claims) (Engel, Reference Engel1984). I asked students to interpret this behaviour. Those from rural backgrounds often confirmed that their neighbours felt claiming was inconsistent with an ethos of self-reliance. One of the most interesting responses came from an older African-American woman who served as a counsellor in an undergraduate women's dorm. A student had a problem: her roommate was beginning a relationship with a male student and asked whether she could have the double room for the night, suggesting that the other women sleep in the bathtub of the common lavatory. The night had been such a success that the woman had been asked to repeat this for a month. My student said (in effect): I don't know what's the matter with you White folk; no friend of mine would put up with that! This produced a lively discussion of how different people responded to harms under varying circumstances. I segued from the anecdotal to the statistical, assigning an overview of tort litigation (Saks, Reference Saks1992) and referring to work by Marc Galanter (Reference Galanter1986), the Oxford Socio-legal Centre (Harris et al., Reference Harris1984) and the RAND Corporation (Hensler et al., Reference Hensler1991) to argue (against many students’ strong preconceptions), as I had in a paper I facetiously referred to as ‘in praise of ambulance chasing’ (Abel, Reference Abel1987), that most tort victims do not claim, with the result that the legal system cannot achieve any of its goals: moral judgment, spreading or efficient levels of safety.

The casebook does not address strict liability until page 500. In 1868, English courts had to decide whether a landowner who excavated a millpond was liable to the adjacent mine owner when the water broke through abandoned shafts in the pond floor and flooded the mine.Footnote ¹⁸ The engineers who had noted the shafts arguably were negligent but lacked the resources to pay compensation; because they were independent contractors, the defendant was not vicariously liable. Teaching torts for the second time in 1969, I asked my Yale small section what they thought should happen. A student eagerly volunteered that strict liability was the obvious solution. A little surprised, I asked: ‘Why, Hillary?’ She said: ‘because otherwise the mineworkers would be jobless, and I'm always on the side of the workers.’ Yes, she was that Hillary. Every year I taught Rylands again, I asked students what they thought the rule should be before telling my story. But, after the 1960s ended a few years later, students became reluctant to express political views.

US courts followed Rylands, demonstrating that late-nineteenth-century judges had a choice between negligence and strict liability. Sometimes their language made it clear that they chose negligence in order to subsidise economic growth (as Horwitz (Reference Horwitz1977) argued). Refusing to apply strict liability to an exploding steam boiler in 1873, the New York Court of Appeals wrote:

(I pointed out that the neighbour did not share the ‘general good’ – an anticipation of trickle-down economics – and probably lacked the capital to put a steam boiler on his own land. The opinion perfectly illustrated Anatole France's (Reference France1894) famous quip two decades later: ‘The law, in its majestic equality, forbids rich and poor alike to sleep under bridges, to beg in the streets, and to steal their bread.’) Six decades after the New York case, the Texas Supreme Court also refused to follow Rylands, denying recovery to a landowner damaged by a neighbour's water because ‘without the storage of water … the great livestock industry of West Texas must perish’.Footnote ²⁰ (This repeats Cardozo's mystification: liability does not prevent the storage of water; it just requires storage to bear the cost of the accidents it causes; and, if the livestock industry is so great, it should be able to absorb those costs.) Just as judges could not explain why they sometimes preferred strict liability, so they could not offer a coherent rule for when they would do so. English courts offered a miscellany of historical examples (beasts, filth, stenches) and invoked the facts of Rylands – ‘bring[ing] on his land something which, though harmless whilst it remained there, will naturally do mischief if it escapes’, distinguishing between ‘natural’ and ‘non-natural’ uses. American courts did something similar, adopting strict liability in blasting cases and then generalising this in the Restatement's definition of ‘abnormally dangerous activities’: those with a ‘high degree of risk’ of ‘great’ harm, which cannot be eliminated by reasonable care, are ‘inappropriate’ to the place where they occur, not ‘common usage’, and of less ‘value to the community’ than the danger they create. I show students that these criteria are either unnecessary (if the dangers could be eliminated by reasonable care, the defendant would be negligent) or hopelessly vague (‘natural’, ‘high’, ‘great’, ‘common’, ‘inappropriate’, ‘value to the community’). Defendants held strictly liable could make those calculations better than courts trying to decide whether to impose strict liability.

Abnormally dangerous activities remain a minor historical anomaly. By the 1960s, however, American courts began to extend strict liability to victims of mass-produced products. This sparked a flurry of economic analyses of the relative merits of negligence and strict liability. Although most followed Posner (Reference Posner1973) in preferring negligence, I showed that economists’ primary criterion – efficiency – called for strict liability: the cost–benefit calculus mandated by the Hand formula can be performed more cheaply and accurately by strictly liable defendants than by triers of fact (juries controlled by judges); strict liability gives defendants a constant incentive to engage in research and development to discover new ways of reducing accident costs (something the trier of fact cannot do); strict liability eliminates false negatives (judgments incorrectly finding no negligence) and reduces the defendant's incentive to create them (by stonewalling, engaging in practices like defensive medicine or hiding information); plaintiffs who need not prove negligence are more likely to claim (further reducing false negatives); and internalising all accident costs in the price of the product encourages consumers to seek cheaper alternatives. Strict liability also reduces Calabresi's secondary accident costs – dislocations from bearing the costs of accidents – by placing those costs on the manufacturer or seller, who will spread them among all consumers. The net effect on tertiary (transaction) costs is indeterminate: more claims but each one cheaper to resolve because the most contentious issue – negligence – has been eliminated.

If the theoretical justification for strict liability is clear, however, doctrinal development again is incoherent, perhaps because judges still want to subsidise economic ‘growth’ (but never explain why the injured consumer should pay the subsidy) or remain uncomfortable with the moral implications of holding manufacturers and sellers strictly liable to consumers who may have been careless (though it is unclear why this is less acceptable than strict liability for workplace injuries). The Restatement explicitly preserves negligence, imposing liability on manufacturers for failing to choose a ‘reasonable alternative design’ or affix a warning, if either would have made the product ‘reasonably safe’. Courts have introduced doctrines of ‘ordinary consumer expectations’ (although manufacturers are better equipped to make design decisions), considerations of price (but laissez-faire economics believes consumer choice should be influenced by a price that reflects all costs, including accident costs), unavoidably unsafe products (but internalising accident costs will produce the efficient level of consumption) and unknowable defects (even though the purpose of strict liability is to create an incentive for manufacturers to learn about them).

I always ended the last class by enacting the goals of tort law: donning a rubber kitchen glove embossed with the Hand formula in large letters and reaching into a deep pocket (props provided by an earlier class). Although I did not assign my chapter in The Politics of Law, most students read it. There, I proposed three programmatic reforms, each grounded in a specific political economy. For conservatives wedded to the ‘free’ market and a minimalist night-watchman state, I recommended strict liability, an expansive definition of damages and incentives to encourage tort victims to claim in order to promote an efficient level of safety. For liberals, I argued that the US should embrace social democracy like all other advanced capitalist nations, spreading accident costs by guaranteeing universal health care and a minimum income (letting those with higher incomes or greater wealth protect them by buying loss insurance) while ending the commodification of experience (eliminating general damages). Because this response would do nothing for safety, it would have to be combined with strict liability or aggressive regulation (or both). Since the first two proposals perpetuate the unequal distribution of risk, I advocated a democratic socialist response, offering the example of worker-owned cooperatives, whose members share risk equally and collectively decide which risks to accept (Abel, Reference Abel1982a). This would extend the original principles of the French revolution – liberté, egalité, fraternité – from the public sphere to the private, one of the central contentions of critical legal studies.

What did I accomplish in forty years of teaching torts? Very few of my students practised tort law, and those who did would have been hard-pressed to translate anything they learned into action. Even fewer became judges or law teachers, for whom my critique of appellate opinions might have been relevant. The rest quickly forgot almost everything – as I did after my three years of law school. Perhaps they absorbed the value of understanding law in context and accepted that law is inextricably connected to politics. Some may retain an image of Learned Hand reaching into a deep pocket and a hazy notion that tort law has something to do with efficient levels of safety and spreading accident costs. I may even have convinced some free marketeers that efficiency requires the internalisation of all accident costs. But I doubt that anyone behaved differently as a result.

Did I help to sustain or inculcate a lifelong habit of sceptical inquiry? If so, is that desirable? Scepticism is a valuable antidote to cant – all the more essential in a Trump administration that conflates truth and falsity and idealises ignorance. But scepticism can easily descend into cynicism, nihilism and quietism. I never embraced subversion for its own sake. To inspire change, reformers must articulate attainable goals as well as limn more distant ideals. We are still far from universal health care and a guaranteed minimum income. Corporations continue to spend as little on safety as they can get away with: Boeing, e-cigarettes, pharmaceutical companies. As a crit, I tried to speak truth to power; is that enough?

3 Legal profession

In June 1972, five men directed by the Committee for the Re-Election of the President broke into the DNC offices at the Watergate complex. The Nixon administration sought to cover up its responsibility. After this was exposed by the media, the Senate unanimously established a select committee, which held hearings in May–August 1973. Faced with overwhelming evidence, Nixon resigned in August 1974. Embarrassed by the fact that the perps from top to bottom were lawyers – Nixon, his personal attorney Herbert W. Kalmbach, Attorneys General John Mitchell and Richard Kleindienst, White House Counsel John Ehrlichman and John Dean, special White House counsel Charles Colson and Ehrlichman's aide Egil Krogh – the American Bar Association promulgated a rule in August 1974 directing all accredited law schools to require ‘instruction in the duties and responsibilities of the legal profession’, including ‘the history, goals, structure and responsibilities of the legal profession and its members’ and ‘the ABA Code of Professional Responsibility’.Footnote ²¹ The National Committee of Bar Examiners added an examination on legal ethics: the Multistate Professional Responsibility Examination (MPRE), consisting of sixty multiple-choice questions. (Students took a one-weekend cram course, which advised them to choose the second most ethical course of conduct if they did not know the pertinent rule. Everyone passed the MPRE, if not always the first time.)

Because I moved to UCLA the year it complied, I was asked to teach the course (colloquially known as Professional Responsibility or PR – feeding cynicism that it was really a public-relations exercise). I was eager to tackle a subject that seemed ideally suited for a sociological approach. In typical fashion, I produced an exhaustive bibliography (Abel, Reference Abel1980) and my own materials. I began the course by discussing where the ABA got the power to regulate law-school curricula, why it believed that mandating such a course would have prevented Watergate and what empirical research revealed about the influence of a PR course on lawyer behaviour (it didn't have any; Pipkin, Reference Pipkin1979; Zemans and Rosenblum, Reference Zemans and Rosenblum1981). My 1976 syllabus warned students choosing my section what to expect:

I elaborated these ideas throughout the syllabus. Criticism of the profession had a long and distinguished pedigree. Adolf Berle (Reference Berle1933) had written:

Citing empirical support for Berle's critique, I identified ‘the overwhelmingly unequal allocation of legal representation’ as ‘the central problem in a course on the legal profession’. If students ‘were beginning a study of the legal profession of another country’ (emphasis in original), they would want to know ‘the number of lawyers … the composition of the bar – by age, sex, race … the distribution of lawyers geographically, and by type of practice and specialization’. I directed their attention ‘to the pressures within the environment of practice, from clients, courts, and colleagues’ because ‘I strongly believe that “unethical conduct” is often no more than a rational response to the pushes and pulls of one's environment’. I began our examination of those environments with Carlin's classic ethnography of Chicago solo practitioners (Reference Carlin1962) because ‘until very recently, the sole practitioner represented the single most common form of practice’. I then turned to the other end of the status hierarchy, namely large firms, warning that

I reluctantly chose the latter (Smigel, Reference Smigel1969). Concerned that the subject of legal education ‘stimulates polemics from the left, responsible criticism from the establishment, and I suspect plain boredom from students’, I assigned only a three-page paper by Ralph Nader (Reference Nader1970). By contrast, I included several papers about ‘what determines [law graduates’] careers? Is it pure excellence … family background? Prior education? Connections?’ I also asked ‘whether the composition of the legal profession is changing with respect to race, sex, religion, etc.’ (It was just beginning to do so.) I observed that

I then turned to the question ‘of the way in which law and lawyers contribute to the structure of existing society, and more importantly, how they might alter it’. ‘[T]he existence and dimensions of unmet needs for legal services … has been largely accepted ever since Reginald Heber Smith [Reference Smith1919] published his revolutionary book, Justice and the Poor’, although, as recently as 1975, a member of the California State Bar Board of Governors had declared that the public had ample access.Footnote ²² Two years before Smith wrote ‘legal aid expenditures … totaled less than $200,000 throughout the country. There were 62 full time attorneys, who handled a little over 100,000 cases that year’. In 1976, by contrast, ‘the Legal Services Corporation's annual budget is almost $100,000,000, which supports nearly 3,000 full time lawyers who probably handle a million and a half cases a year’. Private efforts to reach the unserved faced their own obstacles. Karl Llewellyn (Reference Llewellyn1938) had criticised the restrictions on advertising four decades earlier:

I observed that ‘when [advertising] was debated at the annual meeting [of the California State Bar] last September, the delegates shouted down any proposals for change’. The Supreme Court extended First Amendment protection to lawyer advertising only in 1977.Footnote ²³ Before it did so, the California State Bar disciplined Jacoby & Meyers for calling themselves a legal clinic and Arthur Goldberg for practising as the Echo Park People's Law Center. Another barrier to access was lawyers’ fees. The American Trial Lawyers Association unsuccessfully challenged New Jersey's ceilingFootnote ²⁴ and the Supreme Court held that minimum fee schedules violated antitrust law.Footnote ²⁵

I also revised the topics and readings significantly. I began by questioning the ethical foundations of the ‘hired gun’ – the basic (but shaky) foundation of the adversary system. Virtually the only effort to justify it was by a member of the Harvard class of 1917 (Curtis, Reference Curtis1952). I substituted a cogent critique by a lawyer-philosopher colleague (Wasserstrom, Reference Wasserstrom1975) and a diatribe by a Black feminist entitled ‘The whorehouse theory of law’ (Kennedy, Reference Kennedy and Lefcourt1971). I then offered my synopsis of Larson's sociological theory of professions (Abel, Reference Abel1979b) and applied it to American legal education (Stevens, Reference Stevens, Fleming and Bailyn1971). I replaced Smigel's apology for Wall Street lawyers with a muckraking account (Hoffman, Reference Hoffman1973). I assigned Duncan Kennedy's (Reference Kennedy1970) notorious ‘polemic’ on legal education (bound to resemble Mao's Red Book), as well as a sociological study of its effect (Erlanger and Klegon, Reference Erlanger and Klegon1978). I offered my own highly sceptical view of legal ethics (Abel, Reference Abel1981). And I concluded with my speculation about what could be achieved by redistributing legal services (Abel, Reference Abel1979a), William Simon's (Reference Simon1978) questioning of the hired gun and Stephen Wexler's (Reference Wexler1970) account of lawyers organising for the poor. Each year, one of the most vigorous discussions was a response to a video of four lawyers talking about why they had been disciplined (Dubin, Reference Dubin1985). (This motivated me to write two books (Reference Abel2008; Reference Abel2011) using disciplinary cases to analyse lawyer misconduct in New York and California in the hope that this could help law graduates avoid similar pitfalls; but no one ever used the books to teach PR.)

In 1983, I got faculty permission to teach the course in the first year – a fraught issue because my section then had a heavier course load. Even though classes met for only ten weeks and students completed the term paper two weeks before their other classes ended, the experiment was abandoned after two years (although UC Irvine successfully teaches a course to all first-year students). I published a revised version of my materials (Abel, Reference Abel1997), replacing many of the readings and adding coverage of lawyer–client interaction, women and minority lawyers, comparison with other countries, bar associations, cultural images of lawyers, and lawyers and social change.

After retiring in 2008, I taught twice on recall, offering a limited-enrolment seminar (to avoid the mandatory grading curve) focused on lawyers’ careers. I supplemented my book with newer studies of large firms, public-interest lawyering, gender, race and career choice (including the effect of educational debt). Each student interviewed four lawyers. We spent the first ten weeks discussing the readings to help students draft interview schedules and the last five responding to student presentations of their findings. The term paper used the readings to frame research questions and the interviews to answer them. This was by far my most successful experience in teaching PR. Students were highly motivated because the 2008 recession had greatly exacerbated career anxiety. (The seminar was overenrolled even though it did not offer PR credit the second time because I did not focus on the Model Rules.) Although students were apprehensive about interviewing, they and their respondents enjoyed the experience. The papers were excellent and the teaching evaluations enthusiastic. (In autumn 2019, the law school introduced practitioner talks about their careers; although the faculty member responsible solicited suggestions from colleagues, he seemed unaware that I had studied and taught the subject for forty-five years.)

I experimented with various modes of pedagogy and evaluation. In 1979, I stopped lecturing and devoted class meetings to discussion, handing out questions in advance and declaring: ‘If there is no discussion, there will be no class; one of my purposes is to destroy the myth that inevitably develops within each new cohort of law students that you have nothing to learn from your fellow students.’ By 1987, I required attendance, explaining that ‘I want you to clarify your own views through interchange with your classmates. If you do not attend, you cannot participate in this exchange, and your absence demoralises your classmates who are present’. I deducted points from the final grade for unexcused absences.

Students could write a research paper on any topic related to the legal profession or a take-home examination on ‘what is wrong with the legal profession and what, if anything, can be done to change it’ (but I dropped the latter after three years). We discussed the readings for the first ten weeks; students made oral reports on their term papers during the last five, eliciting feedback from others (and from me in writing). The term paper had to include ‘a thorough review of the existing secondary literature … and your own primary research using interviews … questionnaires, observation, archives, statistical data, etc.’ Students chose a wide variety of topics, including: Chinese Lawyers in Los Angeles; Stress at UCLA Law School; The Image of the Legal Profession in Fiction; A Study of the Distribution of Attorneys in Los Angeles; Law School: Mystification and Learned Helplessness; Everything You Wanted to Know about Legal Malpractice Insurance; A Survey of Los Angeles Attorneys in the Practice of Immigration Law; Unaccredited Law Schools; An Empirical Study of the International Legal Profession; An Analysis of the Arbitration and Grievance Procedures of Local Bar Associations; The American Indian Legal Profession; Legal Services in Prisons; A Little More Hemlock for the Socratic Teaching Technique; Cuban Popular Tribunals; Law School Preferential Admission Programs; Professional Interaction in Land Use and Development. (One daring student used graffiti in the women's bathroom stalls in response to the question: Why did you come to UCLAW? They included: ‘to be the hottest bitch in Hollywood’, ‘to avoid the real world for another three years’, ‘to ace out all those fuckin men who have serious insecurity problems’, ‘it sure beats being a secretary’, ‘to neglect my kids and alienate my friends’.)

Although I can immodestly claim to be a leading scholar on the legal profession, I was much less successful in teaching it than torts, where my knowledge was limited to the casebook. Students resented taking PR – the only course required in their last two years. If they had to take it, they wanted to be prepared for the MPRE. Accustomed to reading a few pages of cases intensively for each law-school class, they had forgotten reading entire books as undergraduates. Wedded to normative analysis, they had set aside childish things, like social science. Committed to becoming lawyers, they did not want to entertain doubts about the ethical foundations of their chosen profession. Attracted by the salary and status offered by large firms, they did not want to think about the hours they would have to work or the drudgery of document analysis. And maybe, paradoxically, I am better at the Socratic method of deconstructing appellate opinions than I am at teaching socio-legal analysis.

But I think the problem was deeper. I have always focused my energy on scholarship. I enjoy mastering new subjects, gathering data, engaging in analysis and writing. As a writer, I could choose a sociological approach, knowing there was an audience eager for such scholarship (if one that was smaller and less prestigious than that for doctrinal analysis). I prefer deliberation to spontaneity; I control what I write in ways I do not control the classroom. I tried to make teaching an extension of writing. That is most obvious in the legal-profession course, which both infused my scholarship and grew out of it. Yet, as the citations to my own work make clear, it was also true in torts. Students go to law school to learn a trade, acquire a credential and get a job. Faculty (who teach only five hours a week) are evaluated primarily on their scholarship. Many have commented on the divergence between these incentive systems. Mine drove me in a direction guaranteed to engender student discontent and hence my own.