The many faces of fault in contract law: or how to do economics right, without really trying.

• Author: Epstein, Richard A.

Modern law often rests on the assumption that a uniform cost-benefit formula is the proper way to determine fault in ordinary contract disputes. This Article disputes that vision by defending the view that different standards of fault are appropriate in different contexts. The central distinction is one that holds parties in gratuitous transactions only to the standard of care that they bring to their own affairs, while insisting on the higher objective standard of ordinary care in commercial transactions. That bifurcation leads to efficient searches. Persons who hold themselves out in particular lines of business in effect warrant their ability to achieve uniform standards, while individuals who seek favors from their friends are incentivized to choose them carefully given the subjective standard of care. These results, moreover, derive from the Roman conceptions of care brought into the Anglo-American law through the 1703 decision in Coggs v. Bernard, and are shown to have surprising durability in dealing with agency, medical malpractice, occupier liability, guest statute, and frustration cases. Often the efficient standard of fault is given only to those who do economics without really trying.

INTRODUCTION: FROM FAULT TO NEGLIGENCE--AND BACK

Any symposium devoted to the role of "fault" in the law of contract is likely to span not only the law of contract but also the adjacent and overlapping field of tort. The uncertain boundaries of this Symposium stem from the persistent ambiguity in the definition of fault itself. In many modern iterations, "fault" is the equivalent of the term "negligence." But that is definitely not the case if the definition of fault in tort law tracks the Hand formula, which compares the burden of precaution (B) with the expected losses, as defined by the probability of loss (P) multiplied by the expected severity of the loss (L), (1) without any reference to the role of custom that Hand discussed in his earlier opinion The T.J. Hooper. (2) Nonetheless, since the publication of Richard Posner's influential early article, the rigid Hand formula has been taken to be the dominant, if not the sole, test of tort liability, to which all other liability rules are subordinate. (3)

In this Article I reject on both normative and positive grounds any purported equivalence between the Hand formula and the idea of fault in contract law. The term "fault" in contract law offers a broad signal that one of a range of standards of blameworthiness applies, depending on context. In its broader sense, the term "fault" is paired with blameworthiness as well as negligence. (4) But in a narrower sense--found in the thesaurus (5)--"fault" and "negligence" actually fall into separate domains without direct overlap. The list of synonyms for fault includes "error," "weakness," "responsibility," "liability," and "burden." Add "blameworthy" into the mix, as well as the terms "guilty," "culpable," and "at fault." "Negligence" for its part does connote "fault," but with a more focused set of meanings that implies some want of care: "carelessness," "inattention," "laxity," "slackness," and "disregard." To add to the confusion, some sources conflate the notion of negligence and culpability. (6)

In some sense, the inexactness of these common definitions should come as no surprise. Ordinary people do fairly well in their lives without ever making the linguistic differentiations that are mother's milk to lawyers. But for our purposes, the notion of fault must be refined to reflect the differences between intentional harm, willful indifference, recklessness, gross negligence, failure to respond to known dangers, the failure to investigate to identify hidden hazards, and the failure to guard against great perils. It is therefore foreordained that some element of fault is in the law of contract. The task of the law of contract is to identify which standard applies in what context and why.

The problems here are twofold. First, contract law covers all agreements in which people agree to either perform or abstain from certain actions, with few, if any, subject matter restrictions. This huge class of enforceable agreements is highly heterogeneous. The standard of care for the nondelivery of goods need not be the same as the standard of care for complex partnership transactions. Second, the level of variation within the law of contract also depends on how broadly we define the sphere of contract law relative to its near neighbor, tort. That overlap is most acute in cases that involve the destruction or loss of property arising out of a consensual transaction. Modern law tends to treat these cases under tort law. I defend the earlier view that uses contract law to govern these cases, reserving tort law for harms that occur between strangers. The greater the fraction of the legal terrain governed by contract law, the greater the heterogeneity in fault standards.

In order to work out the arrangements between these various crosscurrents, I proceed as follows. I argue that as a first principle, this tort-contract line should be placed between (1) physical injuries that arise between neighbors and strangers and (2) physical injuries that arise between parties who are bound together by a prior consensual arrangement that could, in principle, allocate the risk of loss between the parties. Part I sketches out the reasons, based on the comparison of sporting events with sharp boundary lines, why negligence should tend to be the odd man out, with strict liability and intentional harms doing the bulk of the work in both stranger and consensual arrangement cases. Part II argues that physical harms that arise in the course of consensual arrangements should be treated under a contract law framework. It also explains why we rightly expect a greater variation in the use of fault in the consensual cases than we do in the neighbor and stranger cases. Part III then turns to the Achilles' heel of the common law: the proper treatment of gratuitous transactions, first for bailments and agency relationships, and then, briefly, in medical malpractice, occupier liability, and guest-statute contexts. Throughout this Part, I focus a great deal on Roman law conceptions of fault, identifying how modern courts have made use of them, and suggesting that an even greater use would have been beneficial, resulting in a lot less confusion over what should be the proper standard of fault. Finally, Part IV examines the influence of the seminal case of Coggs v. Bernard (7) in traditional frustration cases and compares the approach that allows for multiple standards of fault with the modern tendency to collapse all questions on the standard of care into a cost-benefit formula, concluding that the earlier approach is superior to the modern one--even on economic grounds.

1. TORT LAW

   Historically, tort law dealt primarily with a well-defined class of cases in which one person sought to hold another person liable for the physical invasion of his person or property. Defined in this fashion, the law of tort deals chiefly with trespass and nuisance, fire, animals (including cattle trespass), (8) and liability under the rule of Rylands v. Fletcher. (9) In these cases, I have long defended the view that negligence principles need not be used at all. (10) After all, the defendant seeks to gain privately by his actions, and is willing to throw any collateral losses onto the plaintiff. One way to deal with this is through a tort system that asks whether the defendant has taken the right standard of care. But it is both simpler and more expedient to fasten the liability on the party whose invasive conduct initiated the interaction, unless the plaintiff took that risk--a contract defense to a tort action--or otherwise misbehaved, at which point it becomes important to consider possible schemes for the division of loss to reflect the inputs on both sides.

   One piece of evidence that suggests the efficiency of these standards is their use in consensual arrangements, such as games that involve boundary lines--fair and foul in baseball, in or out of bounds in basketball, and so on. These output rules are adopted to maximize the gains from participation in the game. In each of these cases, the sharp line determines the consequences of the play, so that matters of luck and skill are irrelevant to the result. The system works because good outcomes correlate strongly with higher skill levels over the long run. Indeed, when the output-based rule does not apply, the shift is typically not to negligence, but to some form of intentional harm--the beanball in baseball, the flagrant foul in basketball--that elicits some criminal-like responses. These examples suggest that similar output rules should be used to determine the rules of the road or to resolve boundary disputes. As for the former, we need only assume that a single common owner will set the rules of the road to maximize his own revenues, which will happen only if it maximizes the welfare of the persons who sign up for the system. (11) Hence the use of output-based rules, not care levels, for determining liability in traffic accidents, at least in routine transactions. (12)

   This bifurcated system should likewise have much resonance in dealing with boundary disputes between strangers, where once again negligence should be the odd man out. But historically, the law took a different path. Vaughan v. Menlove (13) is generally credited with introducing the objective standard of care in negligence cases. Yet most suggestively, the defendant in Vaughan insisted that the appropriate rules for liability should be drawn from the law of bailments (14)--cases where one party delivers a chattel with a promise for its return at some future date. The unavoidable element of divided control in bailment cases makes the simple boundary-crossing rules used in boundary disputes and highway accidents a poor guide for the ultimate decision.

   In...