‘Self-Enforcing’ Contracts Don’t Exist
Economists have claimed that contracts can be “self-enforcing” long before ‘smart contracts.’ But self-enforcement is an oxymoron, so why the circular logic?
By this point, it should be clear that “self-enforcing contract” is an oxymoron.
It doesn’t matter whether we’re talking about contracts in the so-called analog world, digital contracts in the digital world, or so-called ‘smart contracts’ (SCs) in the blockchain world.
Contractual self-enforcement is conceptually unstable because a contract, by definition, is a legally enforceable agreement.
As such, there is always the specter of third party enforcement bodies (e.g., C) that are extraneous to the contracting parties (e.g., A & B).
A ‘self-enforcing contract’ is like a ‘self-reading book.’ At first, it sounds good to automate a difficult underlying process, until one realizes that if the book is reading to itself, that defeats the purpose of the book — which is to be read by readers.
The same is true with contract. And the confusion around ‘smart contract’ self-enforcement only keeps growing.
There are a several reasons for this poor state of affairs.
- First, many conceptual building blocks for today’s ‘smart contracts’ turned out to be doctrinally wrong, expressly contradicting settled global contract law principles.
- Second, separately from legal doctrine, self-enforcing ‘smart contracts’ are conceptually unstable because they offer no theory for solving the main objective of contract (enforcement), other than fiat pronouncements that automating performance is de facto enforcement. Spoiler: it’s not.
- Third, supposedly self-enforcing ‘smart contracts’ radically restrict contract parties’ autonomy — in both legal & functional terms.
- Fourth, even as the myth of contractual ‘self-enforcement’ gets debunked and the real costs of conceptual incoherence continue to skyrocket, many people — including many lawyers — believe that its impossible or impracticable to reverse the confusion.
- Fifth, instead of ending the confusion by returning to a more simple, doctrinally-grounded, and functional understanding of … contract, many SC theorists propose even more complex and elaborate taxonomies — including categories like ‘smart legal contracts’ and variously-styled ‘legal wrappers’ for SCs. Because of conceptual fragmentation, confusion regarding ‘self-enforcing smart contracts’ is likely to get worse before it gets better.
For this installment of the ongoing critique of ‘self-enforcing SCs’ we’ll dive deeper into economic theory on so-called optimal contracts, incomplete contracts, and/or implicit contracts.
Are there ‘Optimal Contracts’ — ?
Economist Oliver Hart is one of the highest profile theorists to use the term “self-enforcing” contracts. So it makes sense to launch into the myth of contractual ‘self-enforcement’ through his writing.
Hart is an economist who — along with Bengt R. Holmström — received the Nobel Memorial Prize in Economic Sciences in 2016, for an attempt to formalize a comprehensive Contract Theory.
The basic problem that Hart (and Holmström) try to solve is how to “optimize” contracts — that is, how to make sure a contract between parties contains the best terms for incentivizing cooperative strategies that fulfill the parties’ respective interests/expectations.
From the start, it helps to break the optimal contract problem into two parts:
- There is an optimal substantive contract problem: what are good/“efficient” contract terms/arrangements/configurations in particular settings (e.g., executive compensation; loan agreements; etc.)?
- There is also a related, but separate optimal procedural contract problem: what are the ideal contract vehicles/forms for effectuating the “optimal [substantive] contract” — ?
From an economics perspective, the optimal [substantive] contract problem seems solveable if we can somehow gather enough information to understand what motivates parties in particular contracting settings.
Hart and Holmström’s optimal contract theory became so influential precisely because we could subject the substantive insights to rigorous empirical analysis across a broad range of settings, from corporate finance, to corporate governance, and employer-employee contracts.
Optimal Contracts With Full Information
In asymptotically-full information settings (e.g., executive compensation or large corporate mergers), it might be possible to conflate the substantive and procedural aspects of optimal contracts into a single problem of incentive alignment.
In his book, Firms, Contracts, and Financial Structure (Oxford University Press, 1995), Hart makes the following observation regarding the procedural implications of full substantive information flows.
Rhetorically, the parallels to ‘self-enforcing smart contracts’ are striking. We’ll come back to this shortly.
For now, please note how Hart and Holmström’s optimal contract theory breaks down in incomplete information settings and in multivariate contract settings, where parties may have different ‘irrational’ psychological and/or sociological reasons for contracting.
The best example of this is a basic residential rental contract.
- Rationally, it might make sense to move to another property that offers, say, 20% cheaper rent.
- If the cost of moving is less than the rent savings, it is economically rational for the tenant to move.
Yet, we know from personal experience and large-scale sociological studies that people exhibit certain contract ‘stickiness’ and ‘homing’ instincts that contradict supposedly rational economic choices.
Optimal Contracts ≠ Enforceable Contracts
Like the idea of ‘unbreakable contracts,’ optimal contracts are hypothetical creatures that do not exist in the real world. In their substantive form, optimal contracts can be thought of as asymptotic economic ideals that may work under narrow sets of assumptions.
But as economists routinely note, we should be careful about generalizing.
Just because a contract is optimal for A & B in an economic sense, does not mean the contract is enforceable, nor does it mean that contract should be enforceable.
As between kids’ desire for pocket money, and employers’ desire for cheaper labor, an economically optimal contract may put those kids to work at ‘optimal’ wages for the employer. But we have child labor laws to block ‘optimal contracts’ like this, for strong public policy reasons.
It is extremely important to emphasize that Hart & Holmström theory of optimal contracts touches on, but is NOT focused on optimal contract enforcement modalities or processes.
The theoretical creature known as an optimal contract should NOT be confused with anything resembling an optimal [legal] contract form or template (eg, an ideal-type contract form that will supposedly accomplish the incentive-alignment objectives of an underlying deal).
This is because we don’t live in a world of full information and optimal arms-length contracting. And when legal disputes arise, even the most rational commercial parties often start to act in wildly irrational ways. Even their Nobel Prize citation makes this disclaimer:
[I]n this document we will mostly abstract from psychological and sociological aspects of contracting, and focus on the need to align conflicting interests among rational and selfish materialists. (emphasis added)
This is one reason why — despite our best efforts —we cannot predict legal outcomes with probabilistic certainty, not to mention absolute certainty, like this example from Hart:
The point here is that yes, economists can model how much it may cost for M2 to buy M1’s right to renegotiate an underlying M1 ↔️ M2 contract, but there is no way to preemptively legally block M1’s ex post decision to breach or ‘renegotiate’ the underlying contract.
Because we cannot know: (a) the parties’ future interests in relation to the contract, (b) the full spectrum of changed circumstances that both M1 & M2 (as well as 3rd parties) have experienced in relation to the contract; (c) what legal system(s) potential dispute(s) could percolate through, we cannot say that parties will not do X, or that courts will do Y.
Incomplete Contracts
Every economist knows a model is only as good as the assumptions and data that feed into them. The best economists decide to tackle the problem of incomplete information head on.
For Oliver Hart, this is the big idea of incomplete contracts. The basic observation is that many contracts (including extremely complex and sophisticated written contracts in corporate settings) are left intentionally ambiguous by the parties.
There are many reasons for that.
The parties may not deem those terms material, or the parties may want to afford themselves interpretative, legal, and/or negotiation leeway at a later point in time. By leaving certain terms ambiguous, the parties may also be sending signals to a third party who might be called to mediate a future dispute.
As we can see, an ‘incomplete contract’ is just an economist’s description of a contract with ambiguous or expressly open terms. Here’s an example of such an incomplete contract, from a standard BigTech website Terms of Use.
From a legal perspective, the key question of enforceability is largely unaffected by incomplete contracts. Incomplete contracts may or may not be enforceable; and the same is true for their constituent terms.
Because we don’t know what jurisdictions parties may choose to bring disputes in (& because of the doctrinal richness of contract law, globally), it is impossible to predict the likelihood of enforcement in any sort of rigorous way.
The relevance of Hart’s theory on incomplete contracts to ‘smart contracts’ should be obvious, but is worth accentuating:
- incomplete contracts may be optimal contracts (or if sub-optimal, still highly efficient) in part because they give parties more flexibility and control;
- incomplete contracts are the norm, rather than exception, and they are treated differently in different contexts;
- jurisdictional pluralism may be desirable (eg, Slawson (1971) “[A]t least some lawmaking is better accomplished in a decentralized manner. We therefore prefer that the economy be controlled privately to a large extent, and private control today means control largely by standard form.”);
- optimal contract theory & incomplete contract theory are not premised on any specific theory of enforcement, even though both sets of theories routinely invoke impressionistic accounts of, say, judicial enforcement (‘Court C will enforce Incomplete Contract Y; Court D will not enforce Optimal Contract Z’ — even though this is impossible to predict).
‘Self-Enforcing’ Implicit Contracts
The contradictory nature of ‘self-enforcing contracts’ is already clear from the context in which this term arose: optimal contract (theory) and incomplete contract (theory). But it becomes even more clear by reference to so-called implicit contracts.
So, for instance, in a 2001 article titled Norms and the Theory of the Firm, Oliver Hart offers the following description of “self-enforcing contracts” —
In reality, of course, many economic transactions are sustained by self-enforcing (“implicit”) contracts, or norms of behavior, such as honesty or trust.
In this light, “self-enforcing contracts” (aka “implicit contracts”) can be thought of as manifestations of a broader web of social relations and/or expectations — like a part of the proverbial “social contract” that we’re all supposedly born into.
Please note that Hart’s usage of ‘contract’ & contract ‘enforcement’ terminology above has very little to do with contract law (the primary mechanism for enforcement of contractual obligations) or the core enforceability problem that contracts are designed to help solve.
Aside from Hart’s overbroad ‘social contract’ theories (see above), other economists have also described implicit contracts in ‘self-enforcement’ terms.
For example, see H. Lorne Carmichael, Self-Enforcing Contracts, Shirking, and Life Cycle Incentives (1989):
But Carmichael himself made clear that references to legal enforceability and enforcement were nothing more than heuristics and background assumptions.
In fact, so-called ‘self-enforcing’ contracts are a contradiction, by Carmichael’s own logic.
The contradiction lies in the fact that parties expect their contracts to be upheld (eg, performed); where one party refuses to perform, there is a need for an outside enforcement standard or process, because the parties’ internal self-enforcement mechanism has clearly failed.
Where the parties’ supposed ‘self-enforcement’ mechanism works as planned, it is not an enforcement mechanism, but a performance mechanism (it helps guide and nudge the parties towards fulfilling their performance obligations). In those cases, the parties do not have a breach dispute, and we may not even need to decide on whether there is a contract at all. There is no need to talk about enforcement or enforceability at all.
But when we talk about the need to ‘uphold’ contracts, we necessarily introduce the need for an ‘outside enforcement’ body (by Carmichael’s logic). At that point, the supposed contract is not self-enforcing (nor was it ‘self-enforcing’ at any point prior to the contract dispute).
So Many Contracts, So Little Enforcement
Let’s take stock of the many terms that are lumped into this myth of ‘self-enforcing’ contracts.
- optimal contract
- unbreakable contract
- incomplete contract
- implicit contract
- enforceable contract
- unenforceable contract
- smart contract
Now let’s add a few more terms from contract law for more context: (8) implied-in-fact contract; (9) implied-in-law contract; (10) open-term contract; (11) agreement-to-agree; (12) indefinite contract;¹ (13) …
Finally, let’s add a few more terms from blockchain discourse to round out this non-exhaustive list: (14) legal contract as opposed to ‘smart contract’; (15) ‘smart legal contract’ as opposed to ‘smart contract’; (16) legally-wrapped ‘smart contract;’ (17) ‘smart contract’ with a ‘legal backend;’ (18) …
The problem is clear by this point. Every attempt to qualify the term ‘contract’ actually leads to more confusion regarding the scope of … contract. It doesn’t matter if the adjectives are penned by Nobel laureates like Hart & Holmström or ‘smart contract’ advocates:
‘optimal contracts’ do not have optimal enforcement mechanisms; ‘smart contracts’ do not have smart enforcement mechanisms. Neither concept ‘solves’ the core contract enforceability problem; they restate and complicate the contract enforceability problem.
Why do we have this confusion to begin with? Who profits from ambiguity surrounding contract enforceability?
Answer: stronger contract parties and their lawyers.
K > ‘Optimal’ K > ‘Smart Contract’
So, is there such a thing as a ‘self-enforcing smart contract’ — ? The answer is … maybe.
Contractual self-en-force-ment is theoretically possible. But it requires autonomous law enforcement mechanisms (eg, RoboCop debt collectors & ex ante consent to the RoboCop’s use of force against debtors unable to pay their debts).
Who benefits from that future? Stronger contract parties.
But if you think those gross levels of ‘information asymmetry’ and unequal bargaining power are unfair — and in any event, shouldn’t be ‘unbreakable,’ autonomous, or ‘self-enforcing’ — you probably believe in the power of good ol’ … contract.
FWIW — both law & economics are on your side.
— — — — — — — — — — — — —
¹ Robert E. Scott, A Theory of Self-Enforcing Indefinite Agreements, 103 Colum. L. Rev. 1641, 1654–55 (2003) (reviewing cases and finding that courts enforced incomplete agreements where parties “were forced to cope with problems of hidden action and hidden information”).
