Smart Contracts Don’t Have to Be Dumb

Agreements enforced by computers freak people out. But how they work depends on how they're written.

Just ask Oliver.

Photographer: Scott Eisen/Getty Images

This year’s winners of the Nobel Prize in Economics, Oliver Hart and Bengt Holmström, have spent a lot of time figuring out how humans can make agreements that stand the test of time. Their work can also help us understand what many view as an inhuman and highly suspect innovation: the smart contract.

Hart and his collaborators pioneered the theory of incomplete contracts -- the idea that humans are simple mortals who lack the ability to negotiate a contract that covers every possible future state of the world 1 . This incompleteness suggests that smart contracts -- software that secures a relationship over a computer network 2  -- might be a dumb idea. What if I missed a payment on my auto loan because I came down with the plague? How would a computer interpret the unexpected situation with the warmth and compassion of a human?

Hart’s work suggests an answer: Smart contracts can deal with incompleteness the same way that real-world contracts do, by allocating decision rights ahead of time.

Suppose I want to pursue gainful income as an Uber driver, but can’t afford a car. One possible business partnership is that an investor rents me a Tesla Model S, with the agreement that I’ll use it to shuttle Uber passengers and we split the earnings.

The investor, being the asset owner, has more bargaining power in the event of a disagreement. If there’s any conflict, the investor could simply sell the car. Anticipating this possibility, I might preemptively divert the entire cash flow into my own pocket. Or I might try to lower the resale value of the car by driving into guardrails and stuff. This is a suboptimal financial contract.

In his book “Firms, Contracts, and Financial Structure,” Hart describes a better model for determining who has the right to decide what to do in the event of an unexpected disagreement:

The agent borrows money from the investor and promises to make certain repayments. If he makes them, he retains control of the asset. If he does not make them, control shifts to the investor. In other words, the theory of incomplete contracts and property rights -- when extended to the case of limited wealth -- can explain the use of debt financing.

Instead of renting a Tesla, I take out a loan to purchase one. As long as I make my payments on time, the car is mine. I can even upgrade the entertainment system to attract more passengers. But if I miss a payment, control goes to the creditor. The contract doesn’t have to specify what decisions are made; it only has to specify who gets to make them.

Real-world contracts, though, aren’t always great for guaranteeing transfer of control -- something that is easy if the asset is a corporation, but­ particularly tough, for example, if the asset is something mobile like a car. Before you can say “possession is nine-tenths of the law,” I’m already speeding my way to Mexico.

Here’s where smart contracts have advantages. In the first situation, software could automatically send my driving income to the investor 3 . In the second case, a smart contract could direct the Tesla to autopilot itself to the lender when I miss a payment.

True, automatic enforcement mechanisms aren’t always pretty. Consider a smart contract that turns off your car if you miss a payment, even if you need to take your child to the emergency room. But that’s not new! It’s exactly what a repo man is supposed to do. He just can’t act as quickly and efficiently as a computer.

We tend to think of smart contracts as deterministic automata, “existing simultaneously nowhere and everywhere and operating solely with the steadfast iron will of unstoppable code.” But they work that way only if that’s how we program them. It’s really easy to add a renegotiation function 4 . Maybe I miss a car payment, but I explain to the creditor that I fell ill, or my phone spontaneously combusted, or I’m in prison, or whatever. A sympathetic creditor might allow an extension, and instruct the car not to leave me just yet. The point is that the contract can selectively allocate control according to the state of the world.

A smart contract is a way of securing a relationship, just like a real-world contract. Such agreements help us cooperate when we might otherwise be opportunistic weasels. Their usefulness depends less on whether they’re written on paper or in code than on how intelligently they are designed.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

  1. Oliver Hart in “Firms, Contracts, and Financial Structure (1995): “First, in a complex and highly unpredictable world, it is hard for people to think very far ahead and to plan for all the various contingencies that may arise. Second, even if individual plans can be made, it is hard for the contracting parties to negotiate about these plans, not least because they have to find a common language to describe states of the world and actions with respect to which prior experience may not provide much of a guide. Third, even if the parties can plan and negotiate about the future, it may be very difficult for them to write their plans down in such a way that, in the event of a dispute, an outside authority can figure out what these plans mean and enforce them.”

  2. The term “smart contract” can be traced back to Nick Szabo in 1997: “The basic idea behind smart contracts is that many kinds of contractual clauses (such as collateral, bonding, delineation of property rights, etc.) can be embedded in the hardware and software we deal with, in such a way as to make breach of contract expensive (if desired, sometimes prohibitively so) for the breacher.” A smart contract doesn’t need to live on a blockchain, but putting it on a blockchain is a good way for contracting parties to be sure that no one has modified it after agreement.

  3. Uber already offers auto financing where payments are automatically deducted from the driver’s income.

  4. This past summer, some kids programmed a decentralized autonomous organization (“The DAO”) on the Ethereum blockchain and put $160 million into it (Disclosure: I put money into it, too). Then it got hacked and a bunch of money was stolen. The DAO’s creators did not have a provision for renegotiation, or even a cancellation function. Hopefully they won’t make this mistake again.

To contact the author of this story:
Elaine Ou at

To contact the editor responsible for this story:
Mark Whitehouse at

Before it's here, it's on the Bloomberg Terminal.