Concept
Judicial Sources of Policy
Illegality and public policy
Judicial Sources of Policy
When no statute speaks to a bargain, a court may still find a policy against enforcing it. Section 179(b) of the Restatement (Second) of Contracts authorizes the search: the court may rest on the need to protect some aspect of the public welfare, as reflected in judicial policies against restraint of trade, impairment of family relations, and interference with other protected interests.1 The previous section covered the first source of policy, legislation relevant to it under § 179(a), where the court reads a statute for its purpose and carries that purpose to the scale.2 Here the court has no text to read, so the question is how it can satisfy itself that the policy it invokes is real.
When no statute is quite on point, the court must be satisfied that a strong public policy can be articulated, and the guiding principle is protection of some aspect of the public welfare.3 Section 179 names two areas whose policies rest on precedents collected over centuries: restraint of trade and impairment of family relations.4 A policy applied in case after case has a documented source, much as a statute has, and that is the kind of grounding the strength-of-the-policy factor in § 178(3)(a) asks for. The third category is residual. It covers other protected interests and cross-refers to the penalty rule for liquidated damages in § 356, which this chapter leaves to the material on remedies.5 Some courts go a step further. In a case of first impression, with no statute, decision, or constitutional provision to guide them, they reason that a judicial determination can itself become an expression of public policy if it is so plainly right as to be supported by the general will.6
A policy rooted in precedent can still weaken. Until the 1970s, courts generally refused to enforce agreements between unmarried cohabitants because enforcement would weaken marriage as the foundation of a family-based society, and courts have since begun to uphold such agreements when the business part of the arrangement can be separated from the personal part.7 Doctrine that never changed would keep enforcing a social judgment the community had abandoned, so courts need room to adapt it. They hesitate to use that room without legislative guidance, however, and the disagreement over who should reshape the doctrine runs through every subsection below. The family-relations subsection tells the cohabitation story in full.
Four recurring clusters follow: restraints of trade, family relations, exculpatory clauses, and bribery and fiduciary breach. The first is the policy against agreements that unreasonably restrain trade, most often met in covenants not to compete.
Restraints of Trade
Section 179(b)(i) of the Restatement (Second) of Contracts lists restraint of trade, with a cross-reference to §§ 186-188, first among the judicial policies from which a court may derive a policy against enforcement.8 The policy is old. In John Dyer's Case, decided in 1414, a bond conditioned on a dyer's refraining from his trade in a certain town for half a year drew the judge's remark that the condition was against the common law and therefore void.9
A court takes the policy seriously today because a bare promise to stay out of a trade injures two interests: the free competitive market that serves society and the restrained person's livelihood.10 The person restrained can complain of the second harm. The public's loss of competition and of that person's services is a harm that neither party to the contract has a reason to raise, so the court guards it.
The doctrine stops short of banning every promise not to compete, because a buyer or an employer can have a legitimate interest in protection. Courts therefore separate a naked restraint, which stands alone and is illegal, from an ancillary restraint, a promise to stay out of competition that accompanies a legitimate transaction such as the sale of a business or an employment relationship.11 A naked restraint gives the court nothing to put on the enforcement side of the scale. An ancillary restraint carries the promisee's legitimate interest, and the Restatement limits how far that interest can go. The restraint is unreasonable if it is greater than necessary to protect the interest, or if the promisee's need is outweighed by the hardship to the promisor and the likely injury to the public.12 Courts often restate the inquiry as a three-part rule of reason: a restraint is reasonable only if it is no greater than the employer requires, imposes no undue hardship on the employee, and does not injure the public.13
Each part answers a different danger. The first keeps a covenant from becoming a shield against ordinary competition. An employer may seek protection from improper and unfair competition by a former employee but not from the ordinary kind, and the interest courts recognize is the employer's stake in information and relationships the employee acquired on the job.14 The second part protects the promisor, who faces the loss of a livelihood. The third protects the patrons and the market that lose the restrained person's services. Two principles pull against each other here, freedom to contract and freedom to work, and the court must find the balance between them.15 Because courts construe covenants against the party seeking enforcement, the employer carries the initial burden of proving the covenant reasonable and necessary to protect its business interests.16 The burden sits there because the employer knows what the restraint is meant to protect and can say so.
Hopper v. All Pet Animal Clinic shows the test cutting a covenant down. Dr. Glenna Hopper, a veterinarian at a clinic in Laramie, Wyoming, had agreed that after her employment ended she would not practice small-animal medicine within five miles of the city's corporate limits for three years.17 When the clinic's president heard a rumor that she was investigating a competing practice, the clinic discharged her, and she bought that practice and began operating it. Her old and new clienteles overlapped substantially, so the clinic had a real interest at stake.18 The Wyoming Supreme Court asked what the clinic actually needed to protect. Its pricing policies changed yearly, and practice-development information in a learned profession loses value quickly as technology and reference materials change, so a one-year limit secured both interests.19 The court held the one-year limit reasonable as a matter of law, held the three-year term a partially unreasonable restraint of trade, and remanded for an injunction running one year from termination.20 Years beyond the life of the protected information guard only against ordinary competition, which is why they fell. The case came to trial more than two years after the discharge, so a restraint measured from termination had already run its course, and the dissent objected that Dr. Hopper had beaten the system.21
A covenant that survives shows the same comparison coming out the other way. In O'Sullivan Films, Inc. v. Neaves, an employee who had risen from chemist to manager of research and development promised, for one year, not to sell or help others sell products and services to the manufacturer's customers in competition with it.22 The covenant had no geographic limit and so applied worldwide, yet the court upheld it. It found the time and activity limits narrowly drawn to keep the employee from exploiting his customer relationships in directly competing work, and it accepted the missing geographic limit because artificial leather film was a specialized niche market in which the company competed with other manufacturers worldwide.23 Dr. Hopper's covenant, with a modest five-mile radius, lost two of its three years, while the O'Sullivan covenant, with no territorial limit, survived intact. In each case the result follows from the fit between each limit and the interest behind it. Three years outlasted the value of what the clinic needed to protect, while a worldwide reach matched the reach of the manufacturer's competition.
How closely a court scrutinizes a restraint depends on the transaction it is attached to. Courts enforce a seller's promise not to compete with the business sold, for a reasonable period and area, most readily, because the buyer has paid for goodwill and the seller is well placed to poach customers, use trade secrets, and keep trading on the goodwill sold.24 A departing partner resembles a seller, so partnership covenants are usually easier to justify than employment covenants. Employee covenants meet the most resistance, because they threaten the employee's ability to earn a living and the employee seldom takes away anything more than personal skills.25 The label on the contract does not settle the category. The better question is whether the agreement transfers goodwill as a by-product of a sale or dissolution, and an agreement that does not merits greater scrutiny.26 In Valley Medical Specialists v. Farber, the Arizona Supreme Court treated a covenant among physician-shareholders as closer to an employment agreement than to a sale, because the concerns that justify a seller's covenant, chiefly the buyer's need to receive the goodwill it paid for, shrink when a physician leaves a medical group.27
The two promises have the same length and territory and come out differently because of what each one protects. A seller's covenant is valid only as far as the goodwill purchased, so a restraint that reaches territory where the seller had no goodwill, or lines of trade in which the seller was not engaged, is unreasonable even though it appears in a sale.28 An employee who learns no secrets and has no contact with customers presents the clearest case for refusal, because the covenant imposes undue hardship and deprives the public of the employee's skills.29 An exam answer that treats every employee covenant as void misses the rule of reason, and an answer that enforces whatever the parties signed misses the public-policy check.
The public prong can defeat a covenant even when the promisee has a real interest. In Valley Medical, the trial court found a three-year covenant covering about 235 square miles, and not limited to pulmonology, unreasonably broad and contrary to public policy.30 The Arizona Supreme Court agreed that public policy concerns outweighed the practice's protectable interests, citing the sensitive and personal nature of the doctor-patient relationship. Enforcement would have kept patients who wished to follow the departing physician from doing so. The court stopped short of declaring physician covenants void in every case, while insisting that they be strictly construed.31 Some legislatures have gone further and prohibited physician covenants by statute, and where a statute governs, the first route into § 178(1) applies and no balancing is needed.32 Without such a statute, an employment covenant can still be valid.33 The policy also reaches beyond jobs. In Davidson Brothers, Inc. v. D. Katz & Sons, Inc., the trial court found a forty-year covenant barring a supermarket on a downtown lot unreasonably long, an unreasonable restraint of trade, and contrary to the public interest because of a substantial public need for the store. The appellate court affirmed, holding the covenant so contrary to public policy that it should not be recognized as a valid, enforceable obligation.34
Three outcomes are therefore available. A court may enforce a covenant as written, enforce it with an adjustment, or invalidate it completely.35 If the harm to the public and the promisor outweighs the promisee's interest, the court may refuse to enforce the clause altogether. If public policy can be served by cutting the restraint down to what the interest requires, the court may shorten the time or territory or define the prohibited activity more narrowly.36 Hopper is the cut-down, O'Sullivan the enforcement as written, and Valley Medical and Davidson the refusals. Courts disagree about how far cutting down may go. The Arizona court tolerates striking grammatically severable unreasonable parts but disapproved rewriting a covenant into something the parties did not make. It observed that many overbroad covenants never reach a court and so frighten departing employees, and that employers who expect a court to trim their covenants to fit have little reason to draft carefully.37 A separate reading develops the detailed reasonableness analysis of ancillary restraints. This section concentrates on how the policy against restraint of trade supplies the policy side of the scale.
The third step does most of the work in Hopper and O'Sullivan, since each limit is measured against what the promisee needs to protect. The fourth step is where the public-policy angle appears most plainly, because it alone looks past the two parties to the patients, patrons, and market that neither of them represents.
Family Relations
Section 179(b)(ii) of the Restatement (Second) of Contracts names impairment of family relations, with a cross-reference to §§ 189-191, as the second judicial policy from which a court may derive a policy against enforcement.39 The cross-referenced sections cover promises that unreasonably restrain marriage, promises detrimental to marital relationships, and contracts affecting the custody of a child, which are unenforceable unless consistent with the child's best interest.40 The policy against restraint of trade protects a competitive market and the livelihood of the person restrained. The family-relations policy protects the stability of family ties and the welfare of children, and it enters the analysis the same way, as the public policy placed on the scale against the interest in enforcement under § 178.41
Family contracts raise a problem that ordinary commercial bargains seldom do. The person most affected by a term may not be a party at all. A child whose custody is arranged in advance has no seat at the bargaining table, and neither does a spouse who is the subject of a promise to divorce. The commitments are also ones a person makes before knowing what they will mean, such as agreeing to give up a child not yet born or to end a marriage at some future date. Enforcement would have a court compel the very choices about entering, leaving, or giving up a family relationship that the law otherwise lets people reconsider. The policy therefore protects two things at once: the interests of the children and spouses outside the bargain, and the freedom of the parties inside it to decide for themselves whether to be part of a family.
The leading illustration is re Baby M, where William Stern and Mary Beth Whitehead signed a written agreement under which, for $10,000, Whitehead would be artificially inseminated with Stern's sperm, bear the child, give it to Stern, and terminate her own parental rights.42 After Whitehead delivered a baby girl in March 1986, she refused to turn the child over permanently, and Stern sued for specific enforcement of the contract.43 The New Jersey Supreme Court held the contract invalid on two bases: direct conflict with existing statutes, and conflict with the public policies of the state as expressed in its statutory and decisional law.44 Those bases correspond to the two routes into § 178(1). The statutory conflict applies legislation without balancing, and the policy conflict weighs a policy that no single statute fully captures.45
The statutory conflict had three parts. New Jersey law prohibits paying or accepting money in connection with placing a child for adoption, requires proof of parental unfitness or abandonment before terminating parental rights or granting an adoption, and makes a mother's surrender of custody and consent to adoption revocable in private-placement adoptions.46 The contract paid money, committed the mother in advance to surrender the child, and called for termination of her parental rights by agreement, so it touched all three. The legislature's own reasoning shows why the court gave the statutes such force. Termination permanently separates parent from child, so the legislature severely limited the circumstances in which a consent could be irrevocable and required agency counseling first.47 The invalid terms were at the heart of the contract, so the entire contract was unenforceable.48
The policy analysis reaches the same result by a different path. New Jersey fixes custody by the child's best interests, and the contract's premise, that natural parents can decide before birth which of them will have custody, bears no relationship to that settled rule. Nothing in the contract called for any inquiry into the Sterns' fitness as parents or into the effect on the child of not living with her natural mother.49 The contract also offended the state's policy that the rights of natural parents are equal, since its whole purpose and effect was to give the father the exclusive right to the child by destroying the mother's.50 A court asked to enforce it would have settled custody on the strength of a private promise, and the child, who was not a party, would have had no protection.
The court's concern about voluntariness shows why a family commitment made in advance is suspect. A decision made before the birth is uninformed, because the mother cannot yet know the strength of her bond with the child. A decision made after the birth is compelled by the earlier contract, the threat of a lawsuit, and the inducement of $10,000, so it is less than fully voluntary.51 Money sharpens the problem, because a financial incentive may, depending on the mother's circumstances, make her decision less voluntary still.52 On those premises the court described the arrangement as the sale of a child or, at the least, the sale of a mother's right to her child.53 The court then decided custody separately, on the child's best interests, awarding it to the father and restoring the mother's parental rights.54
The holding is limited to arrangements that pay the mother and bind her to surrender the child. The court found no offense where a woman voluntarily and without payment agrees to act as a surrogate, so long as she is not bound to surrender her child, and it left the legislature free to authorize surrogacy contracts.55 Legislatures and courts have since gone different ways. Michigan declares surrogate parentage contracts void as contrary to public policy and backs the declaration with criminal penalties, so the first route of § 178(1) is the whole analysis there.56 In Johnson v. Calvert, by contrast, the California court enforced a gestational surrogacy contract.57 The two decisions differ in the surrogate's relationship to the child and in what the money bought. In Baby M the surrogate was the genetic mother and the contract paid her to give up her own child. In Calvert the surrogate had no genetic connection to the child, and the payment was for gestational services and not for relinquishing parental rights.58 The trend has been to enforce gestational arrangements, and most states that permit surrogacy at all permit only the gestational kind.59
The family-relations policy does not always arrive with a statute that names the contract, so courts assemble it from statutes and decisions on related subjects. In A.Z. v. B.Z., the Massachusetts Supreme Judicial Court refused to enforce an agreement that would compel one donor of frozen preembryos to become a parent against his or her will.60 The court pointed to statutes that abolished the action for breach of a promise to marry and that bar a mother from agreeing to surrender her child before the fourth day after birth. It also pointed to decisions holding unenforceable a contract requiring a person to abandon a marriage and an agreement imposing a general restraint on marriage.61 From those sources it derived a policy that individuals shall not be compelled into intimate family relationships, because respect for liberty and privacy requires that the decision whether to enter one stay with the individual.62 The court treated that policy as outweighing the public interest in freedom of contract.
A promise that attacks an existing marriage meets the same policy. In Norton v. McOsker, a woman who had been in a 23-year adulterous relationship with a married man sued after he ended it, relying in part on his frequent promises to divorce his wife and marry her.63 The court upheld summary judgment against her and held that promise void as against public policy.64 The promise is the kind of agreement § 190 treats as detrimental to marital relationships. Enforcing it would use the court's authority to bring about the end of a marriage, and the spouse whose marriage was at stake never agreed to anything.65 The same opinion rejected his promise to support her for life on a different ground, because that undertaking was too general and nonspecific to be an enforceable promise.66 The first result rests on the policy protecting family relations and the second on a failure of definiteness, and an exam answer should keep them apart.
The cohabitation cases show this policy moving as the surrounding law moves. Until the 1970s, agreements between unmarried cohabitants were generally unenforceable on the ground that enforcement would weaken marriage as the foundation of a family-based society, and courts have since begun to uphold them when the business part of the arrangement can be separated from the personal part.67 Thomas v. LaRosa, from West Virginia, states that separability condition.68 The turning point was Marvin v. Marvin, in which the California court acknowledged that earlier decisions had invalidated contracts made on an immoral consideration but held that agreements between nonmarital partners to pool earnings and share property are enforceable.69 Reading its own precedents, the court found a narrower standard than their broad language suggested: a contract between nonmarital partners is unenforceable only to the extent that it explicitly rests on the illicit consideration of meretricious sexual services, meaning sex offered as the price of the bargain.70
Not every court moved at once, and the disagreement shows what drives the change. Hewitt v. Hewitt, decided in Illinois in 1979, declined to enforce a cohabitation agreement partly because it regarded revision of the law of marriage-like relationships as a task for the legislature, whose fact-finding facilities it thought superior to a court's. Marvin, by contrast, had treated the issue as one of contract law and found judicial policy-making appropriate.71 The Illinois court also reasoned that recognizing property rights between cohabitants would contravene the state's policy of strengthening marriage, and the Wisconsin court in Watts v. Watts called that reasoning an unsupportable inferential leap. It noted that Illinois still retained fault divorce and treated cohabitation as unlawful, while Wisconsin had abolished fault divorce in 1977 and criminal sanctions for cohabitation in 1983.72 The Wisconsin court found no sign that its legislature meant the Family Code to restrict a court's resolution of property or contract disputes between unmarried cohabitants.73 The two courts weighed the same interests and reached opposite results because the legislation behind the policy differed.
The line between the two agreements comes from decisions holding that a bargain between two people is not illegal merely because an illicit relationship exists between them, so long as the bargain is independent of that relationship and the relationship forms no part of the consideration.74 Premarital agreements followed a similar path. Under the traditional view they were void as contrary to public policy, because an enforceable contract fixing the financial consequences of divorce before the wedding was thought to dangle a lure before betrothed couples. Under the modern view, courts let the parties regulate the financial side of marriage if the agreement meets tests of procedural or substantive fairness, a doctrine this chapter does not develop further.75
The fourth step separates the two kinds of results collected here. A term that fixes a child's custody or parental rights in advance, or that commits a person to enter, remain in, or leave a family relationship, meets a policy the courts treat as strong. A separable money agreement between partners meets a policy that has weakened because the law around it changed. Where a statute speaks directly, as New Jersey's adoption statutes did in Baby M and Michigan's surrogacy statute does today, the court applies it and does not weigh.
Exculpatory Clauses
Interference with other protected interests, the third category in § 179(b), shows up most often in the decided cases as the exculpatory clause.77 An exculpatory clause is a promise by one party not to hold the other liable for harm the other may cause in the future. It appears as a ski resort's release, a landlord's no-liability lease term, or a gym's waiver, each meant to shift the risk of personal injury or property damage away from the provider.78 An opinion in AT&T Mobility LLC v. Concepcion describes exculpatory contracts as a paradigmatic example of contracts courts refuse to enforce on public-policy grounds.79 That description should not be read as a ban. Courts enforce many of these clauses, and the work in this section is learning which ones fail and why.
The reason these clauses belong under public policy lies in what each side of the dispute stands for. Freedom of contract calls for enforcing a freely bargained agreement. Tort law holds a wrongdoer accountable for injury and deters careless conduct. A clause that absolves a party in advance sets one policy directly against the other.80 Tort liability works as a price on carelessness. A provider who must pay for the injuries it causes has a financial reason to spend on safety, and a release removes that reason by moving the cost of the provider's mistakes onto the customer, who may be left with no source of compensation.81 Courts have named two concerns behind the rejection of exculpatory agreements: that they may be produced by overweening bargaining power, and that they do not sufficiently discourage negligence.82 Because the customer signed and the provider wants the clause enforced, neither party will defend the deterrent function of tort law, so the court must weigh it.
The opposing considerations are real, which is why the cases divide. A provider that cannot cap its exposure by contract must raise prices, shrink the activity, or leave the market. Critics of the public-policy defense argue that striking such clauses raises costs and narrows choices for the very people it is meant to protect, as when landlords raise rents or manufacturers withdraw products that could generate large claims.83 The clash between the policy against insulating parties from their own negligence and the policy of freedom of contract remains unresolved, and courts enforce some clauses and strike others.84 The enforcement interest is the special public interest in keeping valuable activities available that § 178(2)(c) places on the scale, and the court asks whether it is clearly outweighed by the policy against enforcement.85
The order follows from where the contract-side policy is weakest. Freedom of contract is best served by enforcing a disclaimer that is voluntary, fairly bargained, and expressed clearly and conspicuously, so a release that fails those tests gives the court an easy reason to side with tort policy.86 Most releases are standard forms, so the drafter must have brought the clause to the other side's attention in explicit language, and unclear wording is read narrowly because the law does not favor provisions that relieve a party of responsibility for its own negligence.87 In Brooks v. Paul, a surgery patient's release appeared in smaller type beneath a notice about the doctor's decision not to carry malpractice insurance, named no particular type of claim, and ran three contradictory sentences, so the appellate court reversed summary judgment for the defendants.88 Coverage is the other half of the inquiry. In Galligan v. Arovitch, a lease excluded the owner's liability for the hallways and six other listed common areas, and the court reversed a judgment for the owner because a tenant's fall on the lawn was outside the listed locations.89 In Munn v. Hotchkiss School, the court held that a student's release of claims arising on a school trip to China was not broad enough to cover negligence at all.90
A release that clears the first two steps may still fall for improper bargaining. The overlap with unconscionability is close enough that some courts blend the two inquiries, and the full treatment of unconscionability belongs to another reading. In Bagley v. Mt. Bachelor, Inc., the Oregon court held a snowboarding season-pass release unenforceable by applying the factors for procedural and substantive unconscionability. It found adhesion in a consumer transaction enough for the first, and it found the release one-sided, harsh, and destructive of the operator's duty of reasonable care, which satisfied the second.91
When the release is clear, genuinely agreed to, and covers the injury, the competing policies meet in their starkest form and the court weighs all the circumstances.92 The first consideration is the scope of the clause. A court may accept a release of ordinary negligence and still refuse to enforce it to the extent it excuses gross negligence, recklessness, or intentional misconduct.93 The reason follows from what each policy protects. The worse the conduct, the stronger the interest in making the wrongdoer answer for it, and the weaker any claim that the injured person agreed to carry that risk by accepting a clause aimed at carelessness. Most courts likewise decline to enforce such clauses as to willful and wanton misconduct or intentional torts.94 A bank sued its security company after burglars broke in, alleging that the company knew its system did not work and did nothing, and the claim survived a motion to dismiss despite a contract clause excusing the company's own negligence.95
Where only ordinary negligence is covered, the court considers four further matters.96
- The importance of the service to the public, so that a release in a residential lease weighs far more heavily against enforcement than one in a skydiving contract.
- The kinds of people who use the service, meaning whether it is offered to the public at large or only to a specialized or qualified group.
- The provider's control over the injured person or the person's property.
- The effect of refusing enforcement on the public's ability to obtain the service at all or at a reasonable cost.
Each matter answers one of two questions: whether the signature reflects a real choice, and whether the provider is the party best placed to prevent the harm. Importance and users go to choice. A tenant who needs housing cannot decline a lease term as easily as a customer can decline a parachute jump, so the signature says less about acceptance of the risk, and a service offered to everyone presents the clause to people with no special reason to understand the hazard. Control goes to prevention. Once the customer's person or property is in the provider's hands, the customer cannot take precautions, and relieving the party with greater expertise and information from the burden of proper maintenance makes little sense. The last matter is the counterweight. If refusing enforcement would make the activity unavailable or unaffordable, the public loses the very thing the clause helped provide.97
The older and widely adopted formula from Tunkl v. Regents of the University of California covers the same ground in six parts. It looks to a business suitable for public regulation, a service of great public importance, an offer to any member of the public, a decisive bargaining advantage, a standardized adhesion contract with no option to pay for protection against negligence, and the placing of the customer's person or property under the provider's control.98 A release may offend the public interest even when some of those characteristics are missing.99 In Dalury v. S-K-I Ltd., a skier who collided with a metal pole in a lift-line maze had signed a season-pass release of all liability for negligence, and the Vermont court held the agreement void as contrary to public policy.100 The factors cut the other way in Jones v. Dressel, where the court found that a parachute jump did not involve a service affecting the public interest.101
Hanks v. Powder Ridge Restaurant Corp. shows the factors defeating a release. An adult injured while snowtubing had actually read a release that was conspicuous, unmistakable, and capitalized the word negligence several times. The court still refused to enforce it. The contract was adhesive, the run was open to all members of the public who reasonably expected safety, the operator controlled the run and rented the equipment, and letting the operator shift the risk of its own negligence to customers would not serve the public interest.102 Clear language therefore settles the first step, and it cannot carry a release through the second.
McCune v. Myrtle Beach Indoor Shooting Range, Inc. came out the other way on facts that look similar on the surface. A paintball player's eye was injured because her mask did not fit properly. The release she had signed, in capital letters, emphasized the risk of injury, acknowledged that she assumed it, and absolved the operator of liability whether or not negligence caused the injury, while expressly leaving gross negligence and wanton misconduct uncovered. The court affirmed summary judgment for the operator, finding the release explicit, limited to conduct no worse than negligence, and a voluntary assumption of risk. It added that it would not serve the public good if a business could not exclude liability by a reasonable and explicit disclaimer, because the risk of liability would preclude many activities.103 Both operators ran recreation and both used clear releases, so the difference lies in the other facts. The snowtube operator controlled the run and the equipment and offered an adhesive form to the whole public, while the paintball release carved out the worst conduct and the activity was voluntary and nonessential.
Courts do not agree on how much adhesion and public access count in recreation. Brigance v. Vail Summit Resorts, Inc. enforced a release against a skier whose foot was caught in a ski lift, because skiing is recreation rather than an essential service, the disclaimer was clear, and the skier signed voluntarily. The court gave no weight to the adhesive form or to the slopes being open to the public.104 Many courts decline to extend public-interest reasoning to nonessential recreation, and a few apply it, as Hanks and Dalury did.105 An exam answer should therefore walk through the factors on the stated facts and avoid deciding the case on the label recreation.
The factors point more reliably toward invalidity for services people cannot easily do without. A charitable research hospital that required a release of liability for its employees' negligence as a condition of admission lost in Tunkl, and exculpatory clauses covering medical malpractice are almost uniformly unenforceable.106 Courts also refuse to enforce releases that shield product sellers, child care centers and nursing homes, public schools, and employers, and carriers and public utilities have traditionally been unable to avoid liability for their own negligence by contract.107 In Henningsen v. Bloomfield Motors, Inc., the New Jersey court held an automobile manufacturer's disclaimer of the implied warranty of merchantability so inimical to the public good as to compel invalidity. The lawmakers had not authorized a manufacturer to use grossly disproportionate bargaining power to escape liability for a dangerous product.108 Justice Stevens, dissenting in Carnival Cruise Lines, Inc. v. Shute, traced the same logic through the federal admiralty cases, which long held passenger-ticket exemptions unenforceable because they usually reflect disparate bargaining power and undermine the public interest in deterring negligence.109
The same weighing explains a split over a parent's advance release of a child's claim. The great majority of states that have ruled hold that a parent has no authority to release a minor's potential claims in advance in connection with a commercial activity, and decisions about community-run and school-sponsored activities divide.110 Galloway v. State invalidated releases for a program in which young children were told to cross a highway without supervision. In Zivich v. Mentor Soccer Club, Inc., the Ohio court enforced a parent's release for soccer play, because enforcement encourages volunteer programs for children and respects a parent's liberty to make life choices for them.111
The residential-lease dispute in O'Callaghan shows the question of who should draw these lines. A divided Illinois court enforced a clause in a form lease that exculpated the landlord from the consequences of its own negligence. The majority reasoned that freedom of contract is basic, that landlord and tenant had been treated as a matter of private concern, and that nothing showed the tenant had tried to negotiate the clause or look for an apartment elsewhere.112 It also found that thousands of competing landlords offered a variety of housing, that a housing shortage is a sporadic and transitory condition, that judicial policy should rest on a durable basis, and that other states had handled the problem by legislation, so the subject suited legislative rather than judicial action.113 The dissent applied the factors directly. The clause touched tenants' physical well-being, they lacked equal bargaining power, the testimony was that an apartment would not be rented to anyone who quibbled over the form, and the same clause appeared in practically all urban form leases, so the choice was illusory.114
The legislatures have since largely answered the question. Most states now provide by statute that clauses in residential leases exculpating the landlord from liability for negligence are ineffective, and a few use variant rules, such as Texas, which requires a waiver by the tenant to be made knowingly, voluntarily, and for consideration.115 Once a statute speaks, the first route into § 178(1) governs and the court applies the legislation without weighing. A clause purporting to release liability for violating statutory safety standards is likewise unenforceable.116 Where no statute exists, the court is making policy, and the majority's caution about doing so in O'Callaghan is the same institutional concern that runs through this chapter. The next subsection turns to a different protected interest, agreements that induce a breach of a duty owed to the public or to a principal.
Bribery and Fiduciary Breach
The residual category of § 179(b)(iii) cross-refers to §§ 192-196, and one protected interest there is the loyalty an agent owes a principal.118 Section 193 treats a promise that induces an agent to violate that duty as unenforceable.119 A fiduciary owes undivided loyalty. A person who acts for someone else, such as a purchasing agent buying for a store or a public official acting for a city, must decide in the interest of the party served and not in the interest of whoever pays for the decision. The previous subsection dealt with a clause that excuses a party's own future carelessness. This one concerns agreements that corrupt someone else's judgment, in two settings: commercial bribery, where the loyalty runs to a private employer, and bribery or self-dealing by public officials, where it runs to the public.
The rule has two reasons, and they protect different people. The first protects the principal. An agent who has been paid by a seller stops asking which goods best serve the employer and starts asking which seller paid, so the employer receives whatever the bribe bought at the price the bribe allowed. The second protects the market. A sale won by a payoff takes business from competitors who offered better goods or lower prices, and they can match the payoff only by adopting the practice themselves. Both harms fall on people absent from the bargain, because the seller and the corrupted agent each gain from secrecy, and the principal learns of the bribe, if at all, after delivery. That leaves the court as the only participant able to withhold its help. Withholding does more than decline to reward one wrongdoer: a seller who cannot collect on a bribed contract gains nothing by bribing in the first place.
Sirkin v. Fourteenth Street Store supplies the commercial pattern. A seller bribed a store's purchasing agent to obtain an order for hosiery, delivered the goods, and sued for the price after the store refused to pay.120 The New York court denied recovery, taking guidance from a state statute that made it a misdemeanor for a seller to offer a commission, discount, or bonus to a purchasing agent and for the agent to receive it, with a penalty of up to $500 and up to a year in prison.121 That statute said nothing about whether a bribed contract could be enforced, so the court inferred the policy from the legislature's evident hostility to the practice, the method the earlier discussion of silent statutes described.122 A party that has performed may recover when the offense is merely malum prohibitum and denial of relief would be wholly out of proportion to the requirements of public policy. Sirkin was not that case, because the court made quite clear its own and the legislature's hostility to commercial bribery.123 The court reasoned that “nothing will be more effective in stopping the growth and spread of this corrupting and now criminal custom.”124
The consequence is harsh, and the court accepted it. The store kept the goods without paying for them, and the dissent accused the majority of judicial legislation and of unjustly enriching the store.125 The agent who took the money is accountable to the employer for it, so the employer can likely recover the bribe in addition to keeping the goods.126 Denial lands on the seller who stood to profit from the bribe, which is the party whose incentive the rule exists to remove.
McConnell v. Commonwealth Pictures Corp. tests how far the rule reaches when the contract sued upon looks clean. A movie distributor promised McConnell $10,000 plus a percentage of gross receipts if he negotiated a contract between the distributor and a producer, and McConnell did the negotiating by paying the producer's agent a $10,000 bribe.127 The distributor knew nothing of the bribe, and McConnell argued that, unlike the seller in Sirkin, he was suing on a contract that had not been obtained by bribery.128 The New York court refused enforcement anyway, holding it beside the point that he might make out a prima facie case, meaning a claim that looks complete on its face, without the bribery ever being exposed.129 The clean-looking contract failed because of what the distributor's promise bought. It promised to pay for the deal with the producer, and the bribe was the means by which McConnell produced that deal, so the obligation he sued on was the price of the wrongdoing's result. The corrupted agent also worked for the producer, a stranger to the lawsuit, which shows that the defendant need not be the party whose agent was bought. The court fenced its holding as well. A minor wrong during the performance of an otherwise lawful contract does not strip the performing party of its rights, and the claim fails only where the obligation sued upon has a direct connection to the illegal transaction.130 McConnell meets that limit because the percentage paid for the very deal the bribe delivered, while an unrelated violation somewhere in the story would not.
A student who sees an enormous profit, or the federal government at the end of a chain of sales, may assume that those facts alone trigger the rule. Black Industries v. Bush shows that they do not. Black Industries was a middleman who was to keep the difference between the manufacturer Bush's price and the price paid by buyers who put the parts into products for the government during the Korean War, a margin that reached 84.09 percent on anvils.131 Bush stopped short of completing the order and defended on the ground that the profits were passed to the government and the public in higher prices.132 The court refused to void the contract because it fit none of the recognized categories: it was not a contract to pay someone for inducing a public official to act, not a contract to do an illegal act, and not a contract contemplating collusive bidding on a public contract.133 The first and third categories involve agreements that directly affect government activities, and here neither party had dealt with the United States, so the profit was not earned by inducing government action or by interfering with competitive bidding.134 In Sirkin and McConnell someone's loyalty was bought and the contract sued on was the product or the price of the purchase. In Black Industries nobody's judgment was bought, and the government was only the final customer. A large profit describes the amount of a bargain, and a government customer describes who ultimately buys. Neither shows that anyone's duty of loyalty was breached, which is the interest this subsection protects.
The same logic applies with added force when the person bribed is a public official, because the loyalty owed then runs to the whole public. Bribery of a public official can be raised as a public-policy defense in a contract action, and the concern now has a global dimension.135 In World Duty Free Co. v. Republic of Kenya, the investor, Mr. Ali, acting for World Duty Free, entered a 1989 contract with the Kenyan government to build and operate duty-free shops at the Nairobi and Mombasa airports. He later brought an arbitration claim after Kenya expropriated his interests, and during the proceeding he testified that he had paid a $2 million “personal donation” in cash to President Moi to obtain the contract.136 The tribunal accepted Kenya's public-policy defense, reasoning that bribery is contrary to the international public policy of most, if not all, States, so that claims based on contracts of corruption or contracts obtained by corruption cannot be upheld, and it awarded no damages for the alleged breach.137 This is called a zero-tolerance approach, because a claim resting on a contract obtained by corruption fails entirely.138 The reasons track the commercial case with the public in the position of the principal. Corruption weakens public accountability, diverts public resources from priorities such as health, education, and infrastructure, and disadvantages law-abiding firms when contracts go to the highest briber.139 Legislation reflects the same concern. The Foreign Corrupt Practices Act, first enacted in 1977, prohibits payment of bribes to foreign officials by individuals and corporations with United States ties, and legislation of that kind supplies policy for the weighing under § 179(a) even though it does not say what becomes of the contract.140
A public official can also breach the duty owed to the public without taking a bribe, by holding a financial interest in a contract with the official's own government. Such contracts are against public policy at common law and are prohibited by statute or charter in many jurisdictions, and the ban is categorical: it disregards whether the contract was fair to the government, how deeply the official was involved, and whether improper influence was actually exercised.141 The categorical form follows from what is being protected. Influence over a government decision leaves no trace once the decision is made, so a rule that waited for proof of actual influence would protect the public only in the cases that happened to be caught, while a ban on the interest itself removes the temptation along with the problem of proof. The contract is treated as void, or as voidable at the city's option, and because it is specifically prohibited or contrary to public policy, ratification, waiver, and estoppel will not save it.142 Quasi-contractual relief, meaning recovery of the value of what was delivered, may be denied as well, although some courts allow it, particularly where the contract is merely voidable, and a later section takes up that question.143
The second and third steps do most of the work. They explain why McConnell loses even though the distributor never bribed anyone, and why Black Industries wins even though the profit was large.