Concept
Legislation as a Policy Source
Illegality and public policy
Legislation as a Policy Source
The preceding sections filled both sides of the scale that § 178 asks a court to use. This section asks where the court finds the public policy it puts on the policy side. Section 179 of the Restatement (Second) of Contracts answers with two sources: legislation relevant to the policy, and the need to protect some aspect of the public welfare.1 Public policy eludes precise definition and may mean different things in different contexts, but its sources include federal and state legislation and judicial decisions.2 This section takes up legislation. The next takes up the judicial sources, which are policies built up in precedent.
Legislation is the more reliable of the two sources. A statute is the form in which elected representatives state what the public interest requires, and Professor Friedman’s analysis found that public-policy defenses resting on specific statutes or regulations were twice as successful as those invoking a general notion of public policy.3 Even so, a statute does not hand the court a rule to apply mechanically. Professor Gellhorn drew the limit in 1936: judges are not bound to void every contract that falls within the general aura of the criminal law, but only those whose enforcement, after respectful study of the public policy involved, they are persuaded will disserve the general interest as the legislature has indicated it.4
A statute can do two different jobs in a public-policy dispute, and the doctrine is easier to follow when they are kept apart. In the first, the statute answers the enforceability question itself. It declares the term unenforceable, and under the first clause of § 178(1) the court applies it without weighing anything. In the second, under § 179(a), the statute never mentions the contract. It regulates a subject, and its purpose tells the court what the public interest on that subject is. The court then asks whether enforcing this term would work against that purpose and sets its answer against the interest in enforcement under the clearly outweighed standard. A silent statute is therefore an input to the weighing and does not end it. The strength-of-the-policy factor still governs, because what the statute adds depends on how clearly it speaks and how closely it bears on the contract.
The alternative would ask too much of legislatures. If a court could consult only statutes that expressly prohibited the contract before it, every legislature would have to foresee each private arrangement that might defeat a regulatory scheme and name it, and any arrangement the drafters missed would be enforceable however thoroughly it undercut the scheme’s aim. Reading a statute for its purpose closes that gap without licensing judges to invent policy, because the court looks to the statute’s text, the legislative scheme around it, and its history.5
Two settings show the method. Statutes that regulate conduct but say nothing about the contracts that violate them are the plain case. Licensing statutes, which rarely say whether an unlicensed person’s contracts are enforceable, are the common one.
Statutes Silent on Enforceability
Section 179(a) lets a court derive a policy against enforcing a promise or term from legislation relevant to that policy.6 The rule matters because legislatures seldom speak to enforceability. Statutes that proscribe conduct rarely say whether contracts involving the conduct may be enforced, and usury and gambling statutes, which characteristically declare violating contracts void, are the exceptions.7 A court will still generally refuse to enforce an agreement to do what the criminal law explicitly prohibits, because enacting a criminal code is itself a clear expression of policy.8 The harder case is the contract that violates no prohibition but frustrates a statute’s purpose. Shadis v. Beal, a Third Circuit decision, is the leading illustration, and its facts show how far the idea reaches.
The plaintiffs in a civil rights suit essentially prevailed, and their counsel, Community Legal Services (CLS), moved for attorneys’ fees under the Civil Rights Attorney Fees Awards Act, 42 U.S.C. § 1988.9 The Commonwealth answered with the no-fees provisions in the contracts between CLS and PLSC and raised them as a defense to the award.10 The underlying claim was a federal civil rights claim and the contract appeared only as a defense to part of it, which is why the district court was correct to analyze the provisions under § 178.11
The Commonwealth defended the clauses by insisting on textual closeness. It argued that the Restatement’s formulation required the legislation either to expressly prohibit the contract or to require a performance or result that directly contradicts what the contract requires.12 The Third Circuit rejected the demand. Neither federal nor Pennsylvania law requires so direct and explicit a conflict, and such a rule would make a contract invalid only if it were illegal in a literal sense.13 The doctrine reaches further than that. Contracts of this kind are often called illegal, but the label is a misnomer because the contract is frequently not illegal in any strict sense.14 Section 179 does not require the legislation to prohibit or require an act inconsistent with the contract. It is enough that the legislature has made an adequate declaration of public policy inconsistent with the contract’s terms.15
The contrary rule would reward whoever can draft around a statute’s wording. In Shadis the party relying on the clause was the Commonwealth, the very entity the fee statute pressures to comply with the civil rights laws, and the court observed that the state had attempted to avoid the policy’s impact.16
Legislation works here as information about policy and does not decide the case by its own force. The Restatement commentary on which the Third Circuit relied treats the statute as significant for enlightening the court about a specific policy to which it is relevant, to be read in light of the whole legislative scheme, its purpose, and its history.17 In Shadis the inquiry turned on how the fee statute works. The Commonwealth argued that the minimum funding it gave CLS to qualify for federal money did more for the poor than fee awards could, and the court called that argument disingenuous. Advance funding gives a state no inducement to comply with the civil rights laws, while the prospect of paying fees does, and it may also give the state a reason to settle early.18 The fee award was the statute’s lever, and a clause that surrendered the fee removed it. The district court balanced the factors favoring enforcement against the policies weighing against it, as § 178 directs, and the Third Circuit held that its conclusion and its approach were correct.19 The no-fees provisions were void as contrary to public policy.20
An earlier decision that failed on the same test shows what the policy side of the scale needs. The Third Circuit distinguished one of its own prior holdings because that holding rested in part on the absence of any contrary policy in the relevant legislation and administrative regulations. With nothing on the policy side, the interest in enforcement prevailed. In Shadis there was such a contrary policy, and the state had tried to avoid its impact.21 The two decisions apply one test to different legislative inputs, and the question to ask is whether the surrounding legislation declares a policy that the term undermines.
Courts reach the same conclusion where a legislature protected a class of people or a regulatory scheme and said nothing about private agreements that would erode the protection. In Javins v. First National Realty Corp., the court reasoned that any private agreement to shift the duties the legislation imposed would be illegal and unenforceable, and it found the precedents on industrial safety statutes directly in point.22 Rent-control ordinances usually prohibit waiver in express terms. Courts have held purported waivers void even where an ordinance is silent, as an impermissible effort to supplant the public interest the ordinance protects, and one court put it this way: even an agreement that modifies the rent laws in a manner favorable to the tenant is of no effect.23
Legislatures also dictate terms. After New Hampshire required in 1957 that insurers include uninsured-motorist coverage in all automobile liability policies issued or delivered there, comparable requirements spread across the country, and the coverage itself had been developed in response to the severe problem of financially irresponsible motorists.24 A policy term that cut back what such a statute requires would be measured against that purpose.
Henningsen v. Bloomfield Motors, Inc. reads a statute’s permission from the other side. The Sales Act authorized agreements between buyer and seller qualifying the warranty obligations, but the court concluded that the legislature contemplated lawful stipulations reached freely by parties of relatively equal bargaining strength. The lawmakers, it said, did not authorize the automobile manufacturer to use its grossly disproportionate bargaining power to relieve itself from liability and to impose on the ordinary buyer, who has no real freedom of choice, the grave danger of injury that attends the sale of a defectively made automobile.25 A permission in a statute carries the purpose that justifies it, and it does not reach a term that defeats that purpose.
The doctrine has limits, and some are set by how explicitly a court demands that the legislation speak. In Terrien v. Zwit, the Michigan Supreme Court held that even if state policy favored home day care, no statute explicitly precluded enforcement of a residential-use covenant, so nothing justified calling the covenant contrary to public policy.26 That is the explicit-bar standard the Third Circuit rejected in Shadis. Jurisdictions also differ on how freely a court may look for policy at all. Some follow Mamlin v. Genoe, under which a court may constitute itself the voice of the community only when a policy is so obviously for or against the public health, safety, morals, or welfare that there is virtual unanimity of opinion about it. Others allow public policy to be drawn from legislation, administrative rules, regulations, judicial decisions, and in certain circumstances even professional codes of ethics.27
The statute itself sets two further limits. A statute that merely favors an activity the term burdens is weaker evidence than one that declares a policy the term defeats. The kind of statute also matters, because a regulatory statute designed to protect the public supplies policy, while violating a revenue-raising measure generally does not prevent a contract from being enforced.28 Licensing statutes, taken up next, are the standard place where that line gets drawn.
A policy drawn from legislation also reaches only the term it contradicts. Section 179 itself speaks of the enforcement of promises or other terms,29 and a single agreement can combine quite different terms. The Locke/Warner agreement, for example, paired a non-exclusive first-look deal paying $250,000 per year for three years with a separate $750,000 pay-or-play directing deal.30 A policy against enforcing one term says nothing yet about the other.
Reading the decisions that do this work requires separating holding from discussion. The holding is the narrow rule needed to resolve the case. In one exculpatory-clause decision it is that a parent is not empowered to bind a minor child under eighteen to a pre-injury exculpatory agreement, and that rule is considerably narrower than the court’s discourse on the law.31 Statutes yield policy only as broad as they state, and decisions bind only as far as their holdings reach.
Shadis also shows that policy can come from a government other than the one whose law governs the contract. The policy there came from a federal statute, and the contract term surfaced only as a defense inside a federal civil rights claim, a setting the court said made its use of § 178 correct.32
Hubner v. Spring Valley Equestrian Center shows what happens when the legislature has spoken specifically about the very risk in dispute. The New Jersey Supreme Court measured an exculpatory agreement in the riding industry against the state’s Equine Activities Act. It held that using such an agreement to escape liability the Act authorizes was so likely to upset the Legislature’s balance of risks and costs that, absent further direction from the Legislature, enforcement as a defense in those circumstances had to be denied as contrary to the policy the Legislature expressed.33 A specific legislative balance settled what a general impression about whether releases are favored would not have. The reason is institutional: a court that applied a general policy over the statute would be redoing the balance the legislature struck. The legislature’s later effort to clarify its intention about the wide scope of exculpatory clauses in the riding industry shows who holds the last word.34
The third step does the decisive work in the decisions above. The Commonwealth’s clause defeated the fee incentive, and the day-care covenant, on the Michigan court’s reading, defeated nothing the legislature had declared. The fifth step keeps the statute in its proper role, because the policy it supplies must still clearly outweigh the interest in enforcing the term.
Licensing Statutes
Licensing statutes are the standard setting for § 179(a), which lets a court derive a policy against enforcing a promise or term from legislation relevant to that policy.36 A legislature that requires a license has regulated a subject, yet most licensing statutes say nothing about whether an unlicensed person’s contract can be enforced.37 When a statute does speak, the court obeys it, as in K & K Services, Inc. v. City of Irwindale, where the statute barred a contractor from bringing or maintaining any action on a contract unless it had been duly licensed throughout performance. Silence leaves the court without that direct answer, so it must work out from the statute’s purpose whether refusing enforcement would serve the policy the license embodies. The result then goes on the scale that § 178 requires.
The purpose of the license decides the question, and the two groups into which courts sort licenses are shorthand for the two answers. A license that exists mainly to raise revenue gives no ground for denying enforcement. A medical doctor who fails to obtain a city business license can still enforce a contract for professional services.38 A license designed mainly to protect the public by regulating access to a trade or profession is treated differently, because an unlicensed person generally cannot enforce a contract made in violation of it. Someone not admitted to practice law in the jurisdiction, for example, cannot enforce a contract for legal services.39
The two results rest on opposite reasons. A protective license is a screen. The legislature has decided that customers cannot safely judge a practitioner’s competence for themselves, so it demands proof of minimum qualifications before the work begins. The screen works only if skipping it carries a cost, and losing access to the courts is a heavy one, because a practitioner who could collect despite skipping the screen would give up nothing by evading it. A revenue license contains no screen to protect. It certifies nothing about skill, so refusing enforcement would deter no unqualified practice and would hand the customer free services as a windfall for a lapse unconnected to any harm.
The same question sorts two requirements that look alike on the surface. A federal district court in Ohio held void a contract made without the license that state law required of property managers, treating the requirement as a police or regulatory measure. In Benjamin v. Koeppel, a New York court considered a statute requiring attorneys already admitted to practice to register with a court office and pay a fee. It found the statute primarily a revenue measure and held that failure to pay the fee did not make the attorney’s contracts unenforceable. What each missing step certifies explains the difference. The property manager without a license has skipped the only check on qualifications, while the admitted lawyer has already passed the check that protects clients and owes only a payment on top of it. A lawyer who was never admitted therefore stands with the property manager, and a lawyer who was admitted but fell behind on the fee stands with the doctor.
Trade-Winds Environmental Restoration, Inc. v. Stewart shows the purpose question answered from the statute’s own words. After Hurricane Katrina, an unlicensed contractor performed mold remediation. The owner paid part of the invoice, refused the rest, and won summary judgment on the ground that the contract was null and void for want of a license. The mold-remediation licensing laws state that their purpose is to set qualifying criteria in a field where unqualified individuals may injure or mislead the public, and the court treated that stated purpose as a policy choice the legislature had already made. The contractor answered that voiding the contract would chill disaster cleanup, which puts a public interest on the enforcement side. The court weighed against it the interest in protecting the public from unlawful acts of contractors, and it noted that the executive branch, which can temper licensing laws in an emergency, had issued no such order. The court voided the contract on the missing license alone, because the license screens qualifications before the work starts and enforcement would have rewarded the evasion of that screen.
The line between revenue licenses and protective licenses is one indication of legislative purpose, and courts no longer treat it as an absolute test.40 Modern courts also worry that denying all recovery can give the defendant too large a windfall and impose too high a penalty. Some therefore allow recovery where the missing license caused the defendant no harm and posed no grave threat to the public. Some allow it where the party pleading illegality was not a member of the public but worked in the same line of business as the plaintiff, and substantial compliance with the licensing law has also supported recovery.41 The reasoning follows from the license’s aim. A protective license shields the public, so a competitor who invokes it gains a windfall without advancing that aim. Town Planning & Engineering Associates, Inc. v. Amesbury Specialty Co. shows a court running this weighing against a plaintiff who lacked certification as an engineer. The court considered limiting recovery to quantum meruit, a reasonable-value measure below the contract price, and concluded that a “vector of considerations” pointed toward the contract price. One consideration was whether the characteristics that gave the plaintiff’s work its value to the defendant were the same as those that made it a violation of law. In Trade-Winds the license was aimed at the very service the contractor sold, so that question pointed toward denial. A missing license opens the inquiry and does not by itself close it.
Courts shape the result to the policy in other settings too. In a restrictive-covenant case, the court held that enforcing a one-year durational limit was reasonable and did not violate public policy. Because the covenant’s three-year term imposed a partially unreasonable restraint of trade, it remanded for modification of the judgment to enjoin the defendant for one year.42
Even a protective licensing statute does not bar everything. A person may recover for work for which no license is needed, and a construction contractor unlicensed in a particular field can promise the services of licensed plumbers, electricians, and engineers without violating public policy.43
One consequence of refusing enforcement concerns money already paid. Although the unlicensed professional may be barred from recovering, if the client has paid, the unlicensed party can generally defend a restitution action by the client, though courts are not unanimous.44 The section on restitution after refusal develops that point.
The second step does most of the work. A revenue purpose ends the matter in favor of enforcement, and a protective purpose shifts the burden to the party who wants to collect despite the missing license.