Because it is represented as a self-executing and self-enforcing agreement for transferring physical and immaterial asset, a smart contract seems to mitigate transaction costs (esp. 5 enforcement costs) which affect “ordinary” (incomplete) contracts. However, though smart contracts reduce several types of transaction costs, others arise; for instance,

• Ex-ante transaction costs: Computer code are an opaque language to most human beings. It may imply costs for contractual parties to understand the “language” and then to realize smart contractual clauses.

• Transaction costs for running of a smart contract: Smart contracts, because they perform complex operations, require “gas” (computational power). Consequently, smart contracts’ computer code cannot be too long, explicit and comprehensive in order to save computational power.

• Ex-post transaction costs: Smart contracts and the technology behind them is not 100% infallible (for instance, see DAO incident). In other words, the prose if-this-then-that in smart contracts can fail.

• Transaction costs related to inflexibility: because their underlying code structure is not interpreted, there is no room for legal concepts like force majeure, material breach, trade customs, and good faith that give some degree of enforcement discretion in the case of a semantic contract—the flexibility of contracts can be desirable, especially when there is a range of unforeseen circumstances, but it cannot be formalized in a blockchain.

• Transaction costs of consensus. With consensus mechanisms (e.g. proof of work), users of blockchain and, by-product, smart contractual parties can alter the rules. In particular, unilaterally the majoritarian participants on the blockchain can decide to fork (e.g., DAO incident). Since they are aligned to preferences of majority, which may change over the time, the logic of groups’ collective behaviour in blockchain’s democratic processes may create inefficient, unreliable and unpredictable rules, e.g. these rules may serve the interests of some or all components of the “dominant elite”, while they damage other minority groups.

This incomplete list of transaction costs shows that, thought smart contracts reduce several transaction costs, smart contracting could impose transaction costs that are more severe and intractable than those it seeks to solve. Hence, the final impact of smart contracts on transaction costs needs to be analysed to fruitfully re-shape economic rules and institutions. This paper makes an effort in this direction.