Economists and lawyers know that almost every contract is incomplete. Contracts routinely omit or leave undefined important terms, and they rarely anticipate every circumstance that might frustrate performance. The reason is simple: It makes little sense to spend $1,000 drafting a provision that has an expected value of only $100. Parties therefore leave lower-value contingencies unaddressed.
Yet economists who analyze contracts have focused on a related but distinct issue: the many events for which one might want to account in a contract which cannot be verified by external authorities (e.g., courts). If performance terms cannot be verified, it would appear meaningless to include them in a contract.
Enter Robert E. Scott and George G. Triantis’ 2005 article “Incomplete Contracts and the Theory of Contract Design.” Real-world contracts, they observe, often address contingencies that courts cannot verify. This fact suggests that much of the economic literature on contracts is incomplete at best and wrong at worst. Here, I outline the paper’s core arguments: its account of the economic literature on contract design, its critique of that literature, and its alternative theory of incomplete contracting.
Why Contracts Are Incomplete
Many trades unfold over time. Purchases from Amazon, purchases of a home, and purchases of a business all involve a delay between promise and performance.
Trade across time creates a problem. In the intervening period, circumstances may change and one party may want to go back on his promise. Without assurances against reneging, no deal will be struck. For example, a seller will refrain from trade unless he is assured that the buyer will not back out at the last moment. Gains from trade then go uncaptured.
Contracts help solve this commitment problem. They define what each party has promised and make those promises legally binding. The prospect of legal sanctions makes parties less likely to renege. When buyers and sellers face less risk of being burned, they become more willing to trade. Ceteris paribus, enforceable promises promote exchange.
Parties may have various reasons to want enforceable promises. Economic theory has focused on one circumstance in particular: investments made before performance. As Scott and Triantis explain, “one or both of the parties may . . . make investments in anticipation of the exchange that will increase the exchange value by either (a) lowering the cost of performance or (b) raising the benefit from performance” (Scott & Triantis, p. 188).
Suppose, for example, that a cabinetmaker agrees to build shelving tailored to the unusual dimensions of a customer’s alcove. Once he has bought the lumber and made his first cuts, the half-finished piece is worth almost nothing to any other buyer. The delay between investment and performance gives the customer a chance to “hold up” the exchange and renegotiate more favorable terms. Legally binding promises deter hold-up by making the threat to walk away less credible.
In a world where investments precede performance and the future is uncertain, contract drafters worried about the possibility of reneging must balance ex ante efficiency and ex post efficiency. A contract that encourages the “optimal” amount of specific investments promotes ex ante efficiency. So-called “specific investments” or “reliance expenditures” are less likely to be made without contractual assurance. As a result, gains from trade may go uncaptured.
The problem is that such a contract may inhibit ex post efficiency. Once the future is known, the non-investor may find that performance no longer makes economic sense. Terms that promote ex ante efficiency may undermine ex post efficiency by “compelling exchange when there is no surplus to be gained” (189).
Parties to a trade could protect ex post efficiency through, for example, an agreement to renegotiate once the future is known. That flexibility can help avoid compelling “inefficient” trades.
But flexibility is no free lunch. The prospect of renegotiation “may expose the party that has made sunk cost investments to the risk of exploitation by the noninvesting party” (p. 189). The noninvesting party may threaten to withhold performance unless the investor accepts worse terms. A prospective homebuyer, for example, may have incentive to get the inspector to find “major defects” in a home. That threat—the threat of recall—would reduce the gains from trade, as the investor is less willing to make any upfront investment that enhances the value of the exchange. Contracts that promote ex post efficiency can therefore undermine ex ante efficiency.
Economic theory offers an idealized solution to this trade-off: the “complete, contingent contract” (p. 189). Such a contract specifies the efficient obligation in each possible state of the world. It elides the tradeoff by stipulating ex ante investments only when they create value and stipulating performance only when it creates value. Because the contract is already efficient in every state, neither party ever wants to renegotiate. In doing so a complete contract simultaneously protects incentives to invest beforehand and the flexibility to walk away from trades that turn out to be wasteful.
The idealized solution is just that: idealized. Real-world transaction costs keep parties from writing the complete contract. Scott and Triantis identify two important types of transaction costs.
First, parties incur costs at the “front-end stage.” They must devote resources to “anticipating future contingencies and writing a contract that specifies an outcome for each one” (p. 190). As Scott and Triantis explain, “On the front end, the parties might not foresee all possible contingencies or they would have to incur prohibitively high negotiation and drafting costs to partition all contingencies sufficiently to provide for efficient obligations in each case” (p. 191).
Second, parties incur costs at the “back-end stage.” These include “the costs of observing and proving the existence (or nonexistence) of any relevant fact after uncertainty has been resolved” (p. 190). Scott and Triantis write:
On the back end, contracts that provide optimal obligations for all contingencies may be too costly to enforce because they require the court to distinguish among too many possible states of the world, some of which may be known only to one party or known to the parties but not the court. (p. 191)
Both front-end and back-end costs can explain why contracts contain so many gaps. A contract may contain gaps because it was too costly to anticipate and write down all of the relevant contingencies. It also may have gaps because it was too costly for the court to verify what happened and enforce the written promises. Contracting parties who expect costly terms to be unenforceable would not include them.
The Unverifiability Assumption
Scott and Triantis observe that economic theorists have taken contractual incompleteness as a given for decades. Theorists often assume back-end costs are prohibitive. In particular, they assume that some states of the world are not verifiable by a court. For example, the level of demand for a good at the time of delivery may be plain to both buyer and seller but impossible for an outside court to confirm. This means that even if a contract anticipates such an outcome—say, a price set low if demand turns out low and high if demand turns out high—the court could not enforce the contract’s obligations. Because the court has no way to determine which state actually occurred, it therefore has no way to determine what promises to enforce.
By Scott and Triantis’ logic, economic theorists have assumed incompleteness rather than explained it. This is likely due to the theorists’ focus. Instead of asking why contracts are incomplete, theorists have analyzed a related but different question: how can parties design contracts given that some terms are by definition unenforceable (read: unverifiable) by courts?
In such models courts are neither gap-fillers nor enforcers. They have no role to play. Renegotiation therefore plays a key role. When some terms are unenforceable, renegotiation is the primary way to promote ex post efficiency. Renegotiation lets parties bargain after uncertainty resolves.
Recall, however, that when terms are unenforceable, renegotiation can also inhibit ex ante efficient investments. Reopening the bargain ex post means reopening the division of the surplus. As the non-investor can capture part of those returns without having borne any of the cost, the investor anticipates being “held up” and invests too little. In such a world, the problem of contract design boils down to allocating ex post bargaining power in a way that protects both ex ante investment incentives and ex post efficiency.
Theorists’ use of unverifiability has allowed them to examine renegotiation’s implications for contract design. But Scott and Triantis argue that the assumption of unverifiability is not innocuous. It yields predictions that are at odds with observed contracting practice.
First, standard “contract theory simply posits that some factors—occurrence of contingencies and performance of obligations—are ‘not verifiable’ by a court” (p. 195). Unverifiability means that it is impossible for a court to tell whether some actions, investments, or states of the world actually materialized. For example, the level of effort that a franchisee put into keeping her restaurant clean is assumed to be unverifiable by a court.
The theory predicts that parties to a contract will never include terms that are conditioned on unverifiable states of the world. Such terms would be superfluous. They define obligations in a world that the court cannot enforce. Including them yields no benefit. And since including them is costly, no rational party ever would.
Scott and Triantis point out that the evidence refutes this prediction. Ostensibly “unverifiable” contingencies are actually commonplace. For example, terms such as “best efforts,” “reasonable care,” and “good faith” appear often in commercial contracts. Their prevalence suggests that unverifiability is not a realistic assumption.
Second, Scott and Triantis argue that the verifiability assumption also misunderstands what courts actually do when they enforce contracts. The concept of unverifiability, they note, implies that a court’s job is to find an objective truth. But that’s not how civil litigation actually works.
Consider criminal law. It uses an absolute measure like “beyond a reasonable doubt.” Thus the prosecution must establish guilt against a benchmark of near certainty.
By contrast, courts in civil trials do not have an absolute standard to which the facts can be compared. Instead, civil courts evaluate evidence in relative terms. For example they consider the “preponderance of evidence” or “the balance of probabilities.” This means that civil courts weigh the persuasiveness of the plaintiff relative to that of the defendant, rather than to an external standard like “beyond a reasonable doubt.” A fact is therefore effectively “verified” when one party can out-evidence the other. Verifiability does not require that the underlying state of affairs be demonstrated with objective certainty.
Third, Scott and Triantis question the common assumption that enforcement costs are exogenous. Contract theory’s unverifiability assumption treats the costs of enforcement as either zero or prohibitive. But as the authors point out, the cost of enforcing contract terms via litigation depends on each party’s litigation strategy. How much each wants to spend on enforcement via litigation depends on how much each expects their counterparty to spend.
This matters for two reasons. If the expected enforcement costs are prohibitive, the parties may choose to omit the term altogether. Alternatively, parties may further tailor the contract in anticipation of future litigation.
The Logic of Rational Incompleteness
In light of these three problems, Scott and Triantis offer a more general explanation for incomplete contracts that does not depend on unverifiability. In their framework, courts matter.
Writing more complete contracts offers several benefits. As in prior economic theory, more complete contracts protect the incentives to make investments beforehand. Unlike that earlier work, Scott and Triantis argue that more complete contracts also reduce back-end costs. Recall that back-end costs involve the costs of litigating:
- Any states that are omitted from the contract; and
- Any obligations that turn out to be inefficient after the state of the world is realized.
When contracts are more complete, litigation costs are lower because courts have fewer gaps to fill and need to exert less interpretive effort divining what the parties would have agreed to if they knew the future. But writing more complete contracts is costly. Parties must spend resources:
- Identifying possible future states of the world; and then
- Writing efficient obligations for each known state of the world.
Parties to a contract must therefore balance the marginal costs of more complete contracts against the marginal benefits. The equilibrium contract sets the marginal cost of additional specification equal to the marginal incentive gain, plus the marginal reduction in expected ex post enforcement costs.
This alternative framework is important for three reasons. First, it treats unverifiability as a special case in which back-end costs are prohibitive. Second, it makes incompleteness endogenous, rather than taken as given. Third, unlike the standard economic theory of incomplete contracts, it can account for the abundance of vague terms in commercial contracts.
Vague terms help save on ex ante drafting costs: “By using broad standards such as ‘best efforts,’ the parties defer this task to the litigation stage” (p. 197).
Incompleteness by Design
Scott and Triantis’ piece clearly belongs in the law & economics canon. It is among the clearest summaries of incomplete-contract theory to date. Moreover, unlike earlier work in economic theory, Scott and Triantis offer a simple framework for explaining why some contracts are more complete than others.
But most importantly, they rightly call for models of incomplete contracts in which courts can matter. Doing so promises to yield more accurate predictions about when and where contracts will be more or less complete. In their account, contractual gaps are purposeful rather than folly.
Further Reading
- Robert E. Scott & George G. Triantis, “Incomplete Contracts and the Theory of Contract Design,” Case Western Reserve Law Review, Vol. 56, No. 1 (2005).
- Robert E. Scott & George G. Triantis, “Anticipating Litigation in Contract Design,” Yale Law Journal, Vol. 115, No. 4 (2006).
- Oliver Hart & John Moore, “Property Rights and the Nature of the Firm,” Journal of Political Economy, Vol. 98, No. 6 (1990).
- Alan Schwartz & Robert E. Scott, “Contract Theory and the Limits of Contract Law,” Yale Law Journal, Vol. 113, No. 3 (2003).
- Sanford J. Grossman & Oliver D. Hart, “The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration,” Journal of Political Economy, Vol. 94, No. 4 (1986).
- Jean Tirole, “Incomplete Contracts: Where Do We Stand?,” Econometrica, Vol. 67, No. 4 (1999).
- Henry A. Thompson, “AI and the Law,” Kyklos, Vol. 79, No. 1 (2026).
