The cognitive map of contract set out in Figure 1 -1 (Chapter 1) distinguishes between two branches of transaction cost economics: the governance branch and the measurement branch. The former is concerned mainly with organizing transactions in such a way as to facilitate efficient adaptations. The latter is concerned with the ways by which better to assure a closer correspondence between deeds and awards (or value and price). To be sure, these are not independent. The difference in emphasis is nevertheless real and needs to be highlighted, It is furthermore noteworthy that problems of governance and measurement both vanish if either bounds on rationality or opportunism are presumed to be absent.
Thus assume that parties to a trade do not experience bounded rationality. Assume, moreover, that this implies the absence of private information and that this competence extends to impartial arbiters. Governance problems then vanish, since comprehensive contracting is feasible. Opportunistic inclinations are simply of no account. Measurement problems likewise vanish, since a world of unbounded rationality is one in which measurement costs are zero. An opportunistic propensity to exploit private information is vitiated in these circumstances.
Assume instead that parties experience bounded rationality but are not opportunistic. Incomplete contracting does not then pose a governance issue, since the general clause device assures that appropriate adaptations will be implemented without resistance by either party to a bilateral trade. Similarly, costly measurement is not a problem if neither party to a trade attempts to exploit private information to the disadvantage of the other—which neither will do if opportunism is absent.
Repeated reference to bounded rationality and opportunism does not, however, without more, direct attention to the particular problems of economic organization that are most severe. Some transactions test bounded rationality limits more severely. Some pose greater hazards of opportunism. Which are they?
Just as the study of governance has benefited by efforts to identify the critical dimensions with respect to which transactions differ in governance respects, so likewise will the study of measurement benefit by efforts to develop the underlying microanalytics. Although the measurement branch of transaction cost economics has made considerable headway during the past decade (Barzel, 1982; North, 1982; Kenney and Klein, 1983), the relevant dimensions for ascertaining where the measurement difficulties reside remain somewhat obscure. Be that as it may, an effort to examine some of the un erying features will nevertheless be attempted.
1. Ex Ante Problems
The adverse selection problem referred to above is an illustration of an ex ante i ton w ere one party to the trade has private information that it can choose selectively to disclose, which asymmetry the other party cannot overcome except at great cost. The condition is a manifestation of a more general problem that is responsible for measurement difficulties, namely, idiosyncratic information. Many of the problems that George Akerlof (1970) treats in the’ context of “lemons” are precisely attributable to such an ex ante valuation condition. The seller of a used car can thus be presumed to have deeper knowledge than the buyer, which asymmetry introduces distortions into this market. And Groucho Marx’s refusal to join a club that would admit him reflects a condition of bilateral asymmetry: if they really knew what he was like, they wouldn’t admit him; and since they don’t know, they presumably have admitted many others of dubious reputation earlier.
The recent Kenney and Klein (1983) treatment of “oversearching” in the market for gem-quality uncut diamonds is another illustration of the phe- nomenon. Despite classification into more than two thousand categories, significant quality variation in the stones evidently remained. How can such a market be organized so that oversearching expenses are not incurred and each party to the trade has confidence in the other? The “solution” that Kenney and Klein describe entails more than just accumulating experience upon which to base “trust”. By assembling groups of diamonds—or “sights”— and subjecting the exchange to special trading rules, hazards of opportunism are more reliably attenuated.
2. Contract Execution
Information asymmetries of two kinds can be distinguished at the contract execution stage. The more familiar is where one party to the trade has more knowledge over the particulars than does the other. For example, a salesman’s success depends jointly on his sales efforts and stochastic state realizations. Although the salesman knows the former, he cannot be relied upon accurately to report them. Accordingly, if the producer can observe only output alone, then compensation is based entirely on sales. (That is the classic agency problem, where X = X(a,Q), where X denotes output, a is effort, and 0 is the state realization.) Complex incentive alignment problems are thereby posed (Holmstrom, 1979).
A second, less widely recognized type of asymmetry takes the form of King Solomon problems. Here each party to the transaction knows the full truth of what has occurred, but it is costly to disclose the facts to anyone other than an on- site observer. Those are the issues with which Alchian and Dem- setz (1972) were concerned in their discussion of team organization. If two or more workers must work coordinately and if their separate contributions cannot be ascertained by an ex post examination of the work product, then assignment of someone to oversee the work may be needed. Supervision purportedly arises in this way.
Unsurprisingly, many of the most interesting problems of economic organization involve both asset specificity and information asymmetry issues. Indeed, as Alchian has argued, the two are often inseparable (1984, p. 39).
Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.