Incomplete Contracts and Blockchain
The field of study in which I specialized during my PhD in economics is a small area called Contract Theory.
Contract Theory has generated interesting insights and implications for all kinds of transactional relationships, including P2P trade, long-term B2B relationships, and organizational governance.
The foundational principle of Contract Theory is an insight which can be leveraged by every blockchain startup: Contracts are inherently incomplete.
Why do contracts matter?
Early economic theory focused on the functioning of markets, in which a plethora of standardized goods are available and facilitate anonymous ‘spot contracting’, whereby a buyer hands a seller money at the same time as the seller turns over his product to the buyer.
In a spot contracting market, each party to the transaction can immediately observe and verify that the trade agreement is fulfilled.
In a more complicated transaction — for example, one for goods that are made-to-order or for an insurance policy — this instant verification feature does not exist. This is because, rather than being executed immediately, the transaction is executed over time.
The lack of instant verification introduces a set of complications that can be solved using contracts. Contracts are used to ensure that such a trade is executed as agreed upon by the transacting parties.
The ultimate aim of a contract is to turn a trade that requires actions to be taken in a sequential manner into a spot trade.
If contracts were complete, meaning that they could achieve a 100% guarantee that each party did exactly what she agreed to do under every possible contingency, then they would do exactly that: make it immediately observable and verifiable:
- To the buyer what she is receiving
- To the seller that she will be paid.
The problem: Contracts are incomplete.
There are a couple reasons why contracts are incomplete.
1. Contracts are only enforceable to the extent that events and actions can be verified by a third party.
There are many things we would want to include as terms to a contract, but cannot because it would be impossible for an arbiter to determine whether or not they happened.
For example, health insurers incur costs when their customers require more healthcare. It would benefit health insurers (decrease their costs) if they could stipulate in their contract that customers must follow a healthy diet. The trouble is that what a person eats on a day-to-day basis is not easy to observe or verify.
If the terms written cannot be third-party verified, then there can be no enforcement. Even if a health insurance company required that you follow a healthy diet, there is no way to prove that you did not. So, there cannot be any consequences for not following a healthy diet. No enforcement means there is no reason for the parties to abide by the contract if some individually better course of action, such as eating more bacon, exists. And there must be some individually better course of action, otherwise there would be no need for a contract.
2. Even if all the relevant information is verifiable, we would never write a contract that contains all possible contingencies.
People are boundedly rational. They are not able to conceive every individual predicament that could possibly take place between when a contract is written and when it must be executed.
Even if they could, it would be extremely costly to actually write down every single contingency. It would be a better use of time and money to leave such a contract incomplete and deal with the consequences in the relatively unlikely event that some rare contingency occurs and causes an unwanted consequence.
Health insurance, for example, usually covers some but not all procedures. When a new type of procedure is developed, it is not immediately accounted for in existing agreements between a health insurer and its customers. This often results in a conflict where a patient would like a procedure to be covered, but his insurer would like for it to not be covered. These disputes usually go through an arbitration process before it eventually becomes established whether that procedure will be covered by the health insurer in general or not.
The contract or agreement between the health insurer and customer is incomplete. This is because there are many contingencies that cannot be specified in advance. Knowing that this is the case, insurance companies develop and implement established arbitration procedures; having these procedures in place helps reduce uncertainty for potential customers, thereby facilitating trade.
What does this mean for blockchain?
The incomplete contracting perspective leads directly to one major conclusion: The problem with contracts is not what is in them, it is what is not in them.
This means that when it comes to trade facilitated by contracts:
The greatest gains in efficiency will not in fact come from the ability to automatically execute previously verifiable terms.
This is because those are cases where existing institutions already work relatively well. And in order to create real efficiency gains:
Blockchain startups should focus on improving the quality of information so that more terms can be used in contracts.
For example, while it was historically very difficult to account for driver behavior in car insurance contracts, technology has expanded the information that can be extracted, recorded, and made verifiable. As testament to this, insurance companies, like Progressive, are now offering discounts to individuals who agree to use devices (Progressive’s is called Snapshot) to monitor their driving.
Blockchain startups that innovate along these lines, will make progress on contracting relationships that suffer from lack of verifiable information and in turn create real efficiency gains.
Startups may also be able to leverage technology for contingency planning. This would allow for contracts to cover more of the possibilities where relationships can go wrong, making those contracts more complete. Perhaps big data and AI can help with this.
For those startups aiming to simply utilize smart contracts rather than to disrupt existing institutions, keep in mind the limitations of contracts.
If designed correctly, your system can overcome these challenges by improving the quality of information, maximizing enforceability, and putting measures in place to cope with incomplete contracts.