Transaction costs economics

Following Commons, Williamson (1975, 1985) makes transaction the unit of analysis, which he defines as a transfer of the right to use a good or service between technologi­cally separable activities.

Four assumptions define the “world of governance”: uncertainty, bounded rationality (borrowed from Herbert Simon), opportunism - “self-interest seeking with guile” (Williamson 1985: 47) - and asset specificity, which characterizes “durable investments that are undertaken in support of particular transactions, the opportunity cost of which investments is much lower in best alternative uses or by alternative users” (ibid.: 55). In this world, contracts are incomplete: they cannot be specified for every state of the world. And asset specificity implies bilateral dependence that generates a quasi-rent, part of which the partner may opportunistically appropriate. The problem is to design a contract that economizes on bounded rationality and safeguards against opportunistic behaviours. The institutionalist aspect of Williamson’s theory is his focus on the differ­ences of governance structures in their ability to adapt in an uncertain environment. The difference with Commons and Coase is that Williamson aims at “operationalizing” his theory, that is, providing refutable predictions. This comes to associate transactions dif­fering in their attributes (frequency, uncertainty and asset specificity) with governance structures differing in their costs and advantages in a way that minimizes transaction costs, thereby providing testable predictions as to the organizational form of specific transactions.

Williamson makes a distinction between three types of contract, the legal aspects of which come from a reinterpretation of Macneil (1974)'s typology of contracts. At one end of the spectrum is the market exchange with its “classical” contract: short term, com­plete, decentralized, anonymous, resolution of disputes before the courts, incentives and adaptation through prices. At the other extreme are the unified structure and its hierar­chical contract, with opposite characteristics: disputes resolved internally and incentives and adaptation through control. Between the two, the hybrid form corresponds to the “neoclassical” contract, and is defined by coordination between autonomous entities. As frequency, uncertainty and, above all, asset specificity increase, contracts are pre­dicted to go from classical to neoclassical to hierarchical. Transaction costs economics is considered by Williamson to be “an empirical success story”, early tests having been conducted since the beginning of the 1980s.

Williamson’s position differs from old institutionalism in that there are no histori­cal, social or political dimensions to institutions, which are viewed as contracts. The market itself is not viewed as an institution, as illustrated by the controversial sentence: “In the beginning there were markets” (Williamson 1975: 20), whereas the existence of markets presupposes the existence of law, rules, contracts, police, that is, of other insti­tutions. Williamson formalized Coase’s question of the emergence and choice of differ­ent governance structures, but consequently his theory remains an analysis of choice, with institutions viewed only as constraints. His four assumptions have no empirical support, but are formulated to design the problem of a choice of governance structure that comes down to a minimization of transaction costs. Commons’s (1934) transac­tion has lost its interaction between individual and collective levels; Coase’s opposi­tion between market and hierarchy is blurred; Simon’s (1955) bounded rationality is translated into a minimizing principle. Regarding law, Williamson bases his theory on a sociological analysis of law and focuses on private and incomplete ordering. He therefore neglects the real law of contracts and the role of the judge. Rather than ana­lysing the interaction between law and economics, he separates the two dimensions and focuses on the latter. Real law, as well as real markets and real firms (as institutions), is eliminated.