Smart Contracts and the Cost of Inflexibility

The blockchain revolution is upon us. In a seemingly endless wave of coverage focusing on everything from financial services to healthcare to supply chain management, commenters are predicting fundamental changes to technology and business models.

Smart contracts – electronic contracts executed on a blockchain – are an important part of this story. Though electronic contracts have been common for some time, smart contracts are radically different.

A smart contract contains the terms and the enforcement mechanism for a transaction. Parties to a smart contract translate the terms of their agreement into a series of if-then rules built into computer code, rather than committing the terms to paper as in a traditional contract. This code connects directly to the parties’ digital infrastructure, monitoring for the satisfaction of conditions and then sending commands to trigger payment or other responses. Smart contracts thus enable parties to create a regime of self-enforcing rules to govern their relationships and property.

This process stands in stark contrast to traditional legal contracting. Parties to a traditional contract must wrangle with fuzzy human language instead of precise computer code. They must rely on courts to enforce breached agreements and engage in costly litigation. It’s not surprising that the rise of smart contracts has heralded speculation (and excitement) about the imminent death of contract law itself.

But legal contracts should be more than a series of if-then rules. They serve as risk-management tools that deploy flexibility in critical ways. Flexibility is not incidental to agreement formation but reflects the reality that it’s expensive, and usually impossible, to fully define an agreement at its outset.

In a comment recently published in the University of Pennsylvania Law Review, I argue that smart contracts preclude flexibility, making agreement-formation more expensive in all but the simplest transactions. I outline my argument below, highlighting specific aspects of smart contracts that exacerbate these costs.

First, “smart contract” is a misleading phrase. It’s typically used to refer to any blockchain program, whether or not the program creates an agreement. Since blockchains enable users to execute all kinds of computer code inside a network of linked computers, even purely operational functions – such as assigning a value to a variable – have been viewed as “contracts” coordinating the behavior of multiple parties.

To steer clear of this semantic debate, I define “smart contracts” as blockchain programs that facilitate the movement of money, goods, or services. This narrows our analysis to blockchain transactions, as opposed to more operational functions, and creates a sharper contrast with traditional legal agreements.

How traditional contracts are efficient

Applying Ronald Gilson, Charles Sabel, and Robert Scott’s work on braiding, I argue that there are two sources of flexibility in traditional contracts, both of which create important efficiencies in the contracting process: semantic flexibility and enforcement flexibility.

Semantic flexibility comes from the inherent ambiguity of human language. Performance standards, such as “good faith” or “commercially reasonable,” are critical tools for parties contracting in uncertain or volatile environments. For example, it’s impossible to exhaustively and precisely define what good faith performance entails, but the term is imposed on every contract as a way to constrain opportunistic behavior.[1] Likewise, a party to a service contract will prefer to generally define its desired performance, stating that an agent should use commercially reasonable efforts to accomplish a goal, rather than attempt to exhaustively specify every action the agent should take.

Ambiguous drafting postpones interpretation of a term until there’s more information about the parties’ performance, lowering the costs of negotiation and drafting. It also shifts those costs to a court, which applies interpretation rules to give substance to the term. Over time, courts provide stable yet context-dependent meanings for ambiguous terms, creating “public goods” that accrue to all contractors.

A second source of flexibility comes from enforcement options. As parties become more secure in their business relationship, they have less need to strictly enforce every term of their contract. They may seek to informally modify the written terms of their agreement, waiving certain obligations or proactively taking on new ones. Parties are willing to make small concessions to preserve a valuable long-term business relationship rather than immediately resort to litigation that destroys their relationship.

How smart contracts are inefficient

Smart contracts lower certain transaction costs but significantly raise others. For example, smart contracts can be drafted from open-source template libraries, harnessing network effects that have fueled other open-source software movements. Smart contracts also minimize the costs of monitoring and enforcement, so they can be drafted to govern transactions that are too small to justify a full-fledged legal document (so-called micro-transactions).

However, smart contracts necessarily restrict the types of flexibility discussed above. This stems from three features: precision, decentralization, and anonymity.

First, smart contracts are constructed entirely from computer code, which can only tolerate precise if-then instructions. A smart contract cannot contain a term that has one meaning at the time of execution and then takes on another meaning later. Modifications are possible, but only if those specific modifications were contemplated at the outset of the agreement.  Unplanned modifications, responding to events that could not be predicted during the drafting stage, are impossible. Likewise, there is no room to vary enforcement of a smart contract term. The contract code, once loaded onto a blockchain, simply executes as written. Flexibility in responding to breach is not an option.

Second, smart contracts are decentralized. They reside on a blockchain, which spreads monitoring and enforcement among many participants, or “nodes,” in the system. Decentralization is a needed check against abuse or manipulation – smart contracts forge a deep link between parties’ business information systems, creating attractive targets for opportunistic nodes. But decentralization also makes ambiguous drafting extremely costly. To enable resolution of ambiguous terms, parties would need to transmit volumes of information about their relationship, transaction, industry, and so on to any number of nodes in the system. Since parties typically must pay transaction fees to interact with a blockchain network – such as “gas”  paid to the Ethereum blockchain– transmitting this information would be impractically expensive.

Finally, smart contracts are geared to anonymity (though so-called “permissioned” blockchains have explored identity-based arrangements). However, parties’ history of transactions is an important asset. It lowers the cost of contracting, enabling parties to shift from formal to informal governance mechanisms over time, leveraging their relationship as a source of trust. It also serves as a form of competitive advantage, creating unique inter-firm knowledge that isn’t easily copied by others in the market.

As a result, it will be unfeasible – and unwise – to deploy smart contracts in all but the simplest transactions, such as when performance can be fully defined, or where modifications will never be necessary.

Smart contracts won’t replace lawyers

Despite excitement about smart contracts in a variety of industries, my comment expresses skepticism that the technology will seriously threaten the legal profession. The types of transactions typically handled in legal practice aren’t easily translated into the ex ante precision required by smart contracts.

I suspect that smart contracts will function similarly to past generations of electronic contracting. Commercial parties will be unwilling to submit their entire legal relationships to smart contracts. Instead, they’ll draft the most fundamental or complex terms in a master agreement, a traditional legal contract that manifests their intent to be bound by specific, well-defined future transactions. However, because smart contracts are more sophisticated than previous technologies, I expect master agreements to contain fewer and fewer terms over time. Practitioners should thus understand the advantages of smart contracting – and its limitations – so they’ll be prepared to handle these kinds of dual contract structures as they emerge.


[1] See, e.g., Restatement (Second) of Contracts § 205 cmt. d (Am. Law Inst. 1979) (“A complete catalogue of types of bad faith is impossible.”)

This post comes to us from Jeremy M. Sklaroff at the University of Pennsylvania Law School and The Wharton School. It is based on his recent article, “Smart Contracts and the Cost of Inflexibility,” 166 U. Pa. L. Rev. 263 (2017), available here.  

1 Comment

  1. Bob Seeman

    Well explained. Over time, artificial intelligence smart contract systems like, an English-based system needing no computer programming, may allow smart contacts to handle more meaning and flexibility. In the future, lawyers may write more precise plain-language English contracts to create more powerful smart contracts.

Comments are closed.