Concept
When the State Contracts
Contracts Clause
When the State Contracts
A State that signs a contract keeps its power to legislate, and the Contracts Clause exists because those two roles can collide. The clause answers a history of legislative interference so numerous and extreme that the confidence essential to prosperous trade was undermined and credit itself was threatened; the restraining power of a central authority had to be interposed to secure the foundations even of “private faith.”1 A State that borrows money, sells land, or charters a company holds the lawmaker's pen and the promisor's obligation at once, which makes it the party best placed to repeat the abuse the clause was written to stop.
Three questions organize this section. Does the clause protect a State's own contracts at all? Which powers can a State bind by promise, and which can no promise reach? And when a State breaks a promise it had power to make, how closely does a court examine its reasons? The answers are yes; not every power; and less deference than a court gives a law that rearranges a private bargain.
Two pressures run through all three answers. Binding the State is what makes its promises worth accepting, because a promise that the promisor may cancel at will secures nothing. A State bound too tightly, however, could not govern. The first subsection settles coverage, the second marks the limit on what a State can promise, and the third supplies the standard for a broken financial promise.
The State's Own Contracts
The clause forbids any State to pass a “Law impairing the Obligation of Contracts,” and its words do not say whose contracts they mean.2 Fletcher v. Peck decided in 1810 that a State's own grant falls within the ban.3 The case grew out of the Yazoo land scandal. A later legislature rescinded a land grant that, it believed, had been obtained by corrupt practices, even though the land had passed to purchasers who knew nothing of the fraud.4
The Court held that a grant is a contract executed, and that its obligation continues after performance. The Constitution says “contract” without separating the executory from the executed, so it covers both. A law annulling conveyances between individuals, and declaring the grantors still seised of their former estates, would offend the Constitution as much as a law excusing sellers from conveying as they had promised.5 Georgia answered that its grantor was the State. The Court replied that the words of the clause are general and apply to contracts of every description, so any exemption for contracts made with the State “must arise from the character of the contracting party, not from the words which are employed.”6
The neighboring clauses supplied a second argument. Section 10 also forbids ex post facto laws and bills of attainder, and the rescinding act would have worked like one: it forfeited the estate of a purchaser for a wrong committed by those from whom he bought. The Court asked why a State forbidden to do that by ex post facto law or bill of attainder should be free to do it by a law annulling the original grant.7
The corruption did not save the statute. The legislature may have had ample proof of practices that would have justified abrogation as to those to whom the crime was imputable. But the grant conveyed a transferable fee simple, and purchasers of parts of it who knew nothing of the fraud stood where any purchaser of a legal estate stands.8 Fletcher thus settles a threshold point: a State's grant is a contract. It does not say how much justification could excuse impairing one, a question the third subsection takes up.
The principle did not stop at land. In Dartmouth College v. Woodward, New Hampshire tried to modify a charter that King George III had granted to the college, and the Court held that the clause forbade it.9 A charter is not a contract as the treatises on contract law use that word. The Court reached its result by asking what the clause was for and what the framers presumably intended.10 Since then the Court has used the clause to restrict a State's power to modify public charters and contracts as well as private ones.11 The same reasoning protected other grants a State had made, among them the tax exemption at issue in New Jersey v. Wilson.
Land grants and charters look like contracts; statutes look like law. A statute counts as a contract only when its language and circumstances show a legislative intent to create private rights of a contractual nature enforceable against the State.12 The Port Authority covenant in United States Trust Co. v. New Jersey is the modern example. The two States promised that the revenues and reserves securing their bonds would not be depleted by deficit-producing passenger railroads beyond a stated level of “permitted deficits.” The Court treated the covenant as the States' contract and held that its outright repeal impaired “the obligation of the States' contract.”13
Ordinary legislation is a different matter, because lawmakers must remain free to amend what earlier lawmakers enacted. Statutes of frauds and statutes of limitations are the old examples. They are valid unless they operate on contracts already made, and even then they are void only to that extent, because only then do they impair the obligation of those contracts.14 A general law consenting to suits against the State is ordinary legislation of this kind and not a contract; Beers v. Arkansas is the usual citation. The Port Authority opinion explains why a change in remedies hurts less than a change in terms. Parties may rely on adequate statutory remedies, but they are unlikely to expect state law to remain entirely static, so a reasonable modification of remedy statutes upsets expectations far less than a law adjusting an agreement's express terms. Repeal of the covenant, by contrast, was a serious disruption of the bondholders' expectations.15
Coverage by the clause is a different question from suability. In Hans v. Louisiana, a citizen of Louisiana sued his own State in federal court, alleging that a recent amendment to the State's constitution impaired the obligation of contracts.16 The Court answered that a State's obligations cannot be made the subject of judicial cognizance unless the State consents to be sued. It added a warning. Where property or rights are enjoyed under a State's grant or contract, they cannot wantonly be invaded; the State cannot be compelled by suit to perform, but an attempt to violate rights acquired under its contracts may be judicially resisted, and a law impairing the obligation of the contracts under which those rights are held is void and powerless to affect their enjoyment.17 A holder of a state grant may therefore be unable to sue the State for performance and still be able to defend what the grant already conveyed. Protection depends on whether the promise is a contract, and the right to sue to enforce it is a separate question.
Powers Not Bargained Away
A State's own contracts are protected, but a State cannot promise away everything. A legislature that could bind its successors on health, safety, and the taking of property would let one session fix policy for all later ones. A dissent in United States Trust Co. v. New Jersey put the danger sharply: the Framers recognized that nothing would so jeopardize the legitimacy of a system that relies on the ebbs and flows of politics to “clean out the rascals” as the possibility that those same rascals might perpetuate their policies by locking them into binding contracts.19
The Court therefore asks a threshold question before it weighs any justification: did the State have power to create an irrevocable contract right of this kind in the first place? The inquiry looks to the State's power and not to the purpose or reasonableness of the later impairment. Its premise is that the clause does not require a State to adhere to a contract that surrenders an essential attribute of its sovereignty.20 Three situations show how the question comes out. Eminent domain and the police power cannot be bargained away; financial promises can be made; and a State that reserved the right to change a charter has created nothing to impair.
The eminent domain limit comes from West River Bridge Co. v. Dix. Vermont chartered the bridge company in 1795, and both sides agreed that the charter was a contract the State could not impair.21 Vermont then authorized its courts to lay out public highways over private property, including the property, easement, or franchise of any turnpike or other corporation, with compensation to the owners. On a petition by Joseph Dix and others, a county court extended a public road over the company's bridge, turned the bridge into a free public highway, and assessed compensation for the extinguished franchise.22 The Court held that the taking violated no contract. The parties' agreement that the charter was a contract, it said, determined nothing about the merits.23
The Court placed the taking inside the charter itself. Every contract carries conditions that come from a higher authority than its terms, the authority of the community to which the parties belong. Everyone is presumed to know them, they need never be written down, and every contract is made in subordination to them. The right of eminent domain is one.24 A taking under that right does not impair the contract. It recognizes the contract's obligation in full and claims only the fulfilment of a condition the contract always carried.25 A franchise to build a bridge and collect tolls stands toward the State's duty to promote the public good exactly as a citizen's land does under a patent.26 The practical stake was large. The country's policy on roads, mills, bridges, and canals rests on the power to condemn, and without it none of those improvements could be built.27 Justice McLean, concurring, reduced the mechanism to a line: the power acts upon the property and not on the contract, so a State cannot annul or modify a grant fairly made, but it may take the property for public use on paying compensation.28
A promise never to condemn changes nothing. A legislature may promise in a corporate charter not to exercise eminent domain, yet the promise will not stop the State from taking the corporation's property with just compensation, because the legislature never had power to convey that power away.29 The power has limits of its own. A State must exercise it in good faith, and the property itself, not merely its profits, must be applied to public use. A State could not take a charter and carry on the corporation's functions, and it could not take this bridge and run it as a toll bridge, because nothing would change but who collects the tolls.30 Vermont's road met the requirement, since the bridge became a free highway.
Once the State takes a franchise and pays for it, the contract objection ends. The Constitution protects property through three provisions: the Contracts Clause, the Takings Clause, which forbids taking private property for public use without just compensation, and the due process clauses.31 Any remaining dispute over what the company is owed belongs under the Takings Clause. (Late in the nineteenth century the Court, following some state courts, also began reading the Due Process Clause to protect the liberty of private contracting and the sanctity of private property.32)
The police power obeys the same rule. In Stone v. Mississippi, a private corporation had paid the State a valuable consideration in money for a charter to conduct a lottery, and the State later repealed the charter. The Court said that “No legislature can bargain away the public health or the public morals.” Government is organized with a view to their preservation and cannot divest itself of the power to provide for them.33 A railroad that laid its tracks under a state charter took them subject to the State's power to provide for public safety, and the State owed no indemnity for the damage that regulation caused.34 Blaisdell stated the idea in general terms. The reservation of essential attributes of sovereign power is read into contracts as a postulate of the legal order, because the policy of protecting contracts presupposes a government with adequate authority to secure the peace and good order of society.35 The same opinion cautioned that emergency does not create power, though it may furnish the occasion for the exercise of power.36 The reserved authority is not conjured up by hard times. The State always held it and could not sell it.
Because the question is the State's power to make the promise, a promise that purports to forbid a later exercise of the police power never bound the State. The court does not go on to ask whether the later measure was reasonable and necessary; it stops at the promise.37 Running the reasonableness test on such a promise skips the threshold.
Financial promises stand on different ground. Any repayment obligation could be described as a surrender of the State's spending power, since money used to repay debts cannot be spent elsewhere, and the State may have to levy taxes to meet the debt. Yet the Court has regularly held States to their debt contracts.38 The reason is that a State needs credit. When a State borrows, it does not act as a sovereign that reserves a right to withhold payment. As an older opinion put it, “A promise to pay, with a reserved right to deny or change the effect of the promise, is an absurdity.”39 The Court has accordingly said that the power to enter into effective financial contracts cannot be questioned.40
The Port Authority covenant fell on the financial side of the line. The States had promised only that the revenues and reserves securing the bonds would not be depleted by deficit-producing passenger railroads beyond the level of “permitted deficits.” The Court called that promise purely financial, and so not necessarily a compromise of the State's reserved powers.41 The covenant did not stop the States from improving mass transit or promoting public transit; it barred them from using specified revenues to do so, and they remained free to use other money. The States had not contracted away their authority over transit.42 The label “financial” is therefore a test and not a password. The question is whether the promise restricts the spending of identified funds or restricts the exercise of a governing power. Passing the threshold makes the covenant a valid contract and nothing more, because a later impairment still faces the review described in the next subsection.
The third limit comes from the terms of the grant. If a State is to retain the ability to modify a charter or public contract, it must say so in the charter or in the enabling legislation.43 A State that reserved the right has written it into the bargain, so using it takes nothing the grantee was promised. That is why a reservation bears on the substantial-impairment step: a grantee that accepted the charter knowing the State could change it cannot claim the change upset its reasonable expectations. What a charter actually promises, as distinct from whether the State may change it, is a separate question; Charles River Bridge v. Warren Bridge is the classic case on reading a grant's silence narrowly.
Stricter Scrutiny
A State that breaks a financial promise it had power to make must justify the break. The Contract Clause is not an absolute bar to a State's later modification of its own financial obligations. As with laws that impair private contracts, an impairment may be constitutional if it is reasonable and necessary to serve an important public purpose.45 The Court will not, however, defer completely to the legislature's view of reasonableness and necessity, because the State's self-interest is at stake.46 United States Trust Co. v. New Jersey indicated that government interference with government contracts will be subjected to heightened scrutiny.47 The scrutiny is not necessarily fatal, but the State must prove that its action was both reasonable and necessary to effect an important public purpose.48
The case arose from a covenant that limited the Port Authority's deficits and so protected the general reserve fund securing its bonds from depletion.49 Pressure to spend Port Authority money on commuter rail grew. By 1973 the general reserve fund had reached $173 million, which raised the level of permitted deficits to about $17 million, but the PATH deficit stood at $24.9 million, and the Port Authority could not issue bonds for any new passenger railroad facility that was not self-supporting.50 The States repealed the covenant outright. The repeal eliminated an important security provision and impaired the obligation of the States' contract, and the States made no effort to compensate the bondholders.51 The Court saw in it a serious disruption of the bondholders' expectations.52
The stricter standard follows from who is deciding. When a legislature adjusts a bargain between private parties, it has no stake in who wins, and courts defer to its judgment of necessity and reasonableness, as they do in reviewing economic and social regulation generally.53 Keystone repeated the condition on that deference: courts properly defer unless the State is itself a contracting party.54 Even deference has limits. Keystone recalls that in Pennsylvania Coal the “particular facts” led the Court to hold that the Pennsylvania Legislature had exceeded its constitutional powers by prohibiting coal mining that would cause the subsidence of land on which certain structures stood.55 A State that has borrowed money stands on one side of the bargain and gains whenever its own obligation shrinks. A government can always find a use for extra money, especially when taxes need not be raised.56 If a State could reduce its financial obligations whenever it wanted to spend the money on what it regarded as an important public purpose, the Contract Clause would provide no protection at all.57 A worthy purpose is therefore necessary but never enough, and the court examines necessity and reasonableness for itself.
Necessity works on two levels, and the Court found neither satisfied. First, total repeal was not essential, because a less drastic modification would have permitted the contemplated plan without removing the covenant's limits on using Port Authority revenues and reserves to subsidize commuter railroads. Second, the States could have pursued their twin goals of discouraging automobile use and improving mass transit without touching the covenant.58 A footnote showed how. Taxes on gasoline or parking could discourage driving, and their revenues could subsidize mass transit projects structured to be “self-supporting” within the covenant's meaning. Bridge and tunnel tolls could be raised for commuters and lowered at other times, leaving no excess revenue.59 The States replied that choosing among such alternatives belongs to the legislature. The Court answered that a State is not completely free to treat impairing its own contracts as one policy alternative among others, and that it is “not free to impose a drastic impairment when an evident and more moderate course would serve its purposes equally well.”60 The States also remain free to abrogate contractual rights through eminent domain on payment of just compensation, so a worthy project need not fail for want of a lawful path.61
Reasonableness asks whether the problem the State now invokes was a surprise when it made its promise. In the twelve years between adoption of the covenant and its repeal, public perception of the importance of mass transit undoubtedly grew with increased concern for environmental protection and energy conservation. But those concerns were not unknown in 1962.62 The logic is plain. Bondholders lent against the covenant, so a State that knew the pressure existed and promised anyway had already priced that pressure into the bargain, and repeal takes back something the lenders received in exchange for their money. Compare the kind of occasion that can justify interference. Blaisdell allowed that a temporary restraint on enforcement may fall within the State's reserved power when a great public calamity such as fire, flood, or earthquake makes it necessary.63 Even then, emergency does not create power or diminish the restrictions on power; it only furnishes the occasion for exercising power the State holds.64
El Paso v. Simmons shows that the review can be satisfied. Texas had sold land under contract for a small down payment, and a buyer who failed to pay interest could forfeit his right to the land but could reinstate his claim. The Court sustained Texas's later limit on that right, treating it as a reasonable response to effects the legislature had not foreseen.65 The outcome shows why the scrutiny is described as demanding without being fatal.
An alteration of a municipal bond contract passes only rarely. When United States Trust was decided, the only one the Court had sustained in the century was in Faitoute Iron & Steel Co. v. City of Asbury Park. New Jersey's Municipal Finance Act let a state agency place a bankrupt local government in receivership. A plan for composing creditors' claims needed approval from the agency, the municipality, and 85% in amount of the creditors, and it bound nonconsenting creditors only after a state court found that the municipality could not otherwise pay and that the plan was in the best interest of all creditors.66 Nothing like that supported the Port Authority repeal, which eliminated the covenant outright and offered the bondholders no compensation.67 The Court also rejected one argument that would have made the review toothless. The trial court had reasoned that any impairment less oppressive than the one in Kavanaugh was necessarily constitutional. The extent of impairment is relevant to reasonableness, the Court said, but it could not sustain the repeal simply because the bondholders' rights were not totally destroyed.68
The separate opinions sharpen what the majority did and did not decide. A concurring opinion stressed that the Court did not hold that any particular “less drastic modification” would pass constitutional muster, so the opinion identifies what failed without certifying a substitute.69 The dissent would have ended the inquiry once a State, acting in furtherance of its citizens' general welfare, made a policy choice that infringed contract rights but was not “plainly unreasonable and arbitrary.”70 It objected to the vocabulary of the majority's test. Reasonableness generally signals the most relaxed judicial review, while necessity has traditionally played a key role in the most penetrating, so joining them produced “a most unusual hybrid.”71 It also warned that if unforeseeability is the key to a “reasonable” decision, almost all prior cases must be repudiated.72 The majority's reasonable-and-necessary standard is the one to apply.
Steps four and five decide most cases, and they are where answers go wrong. An answer that praises the State's goal and stops skips the questions the Court actually asked.