So..I decided to start the course with the Coase theorem. Well, at first a small bit of game theory so (later) we could analyze in simple models problems of moral hazard , motivation (incentive contracts) and trust and prisoner’s dilemmas and coordination failures . But we kicked off the course listening to and doing a short keypoint exposition of Oliver Williamson’s Nobel Prize winning lecture on Transactions Costs economics.
In the discussion forums several students raised the question – what is transaction cost economics. Kreps’ and Baron’s Strategic Human Resource Management text has a brillliant short exposition in Appendix A a pdf of which is here
After talking about clean, clear, simple classical transactions they explain:
Textbook economics implicitly assumes that such transactions as these are simply more complex versions of classical transactions. But the complexities bear closer study. What complicates all these transactions (and others) is the confluence of
two distinct sets of factors.
One set of factors pertains to the parties involved. People are boundedly rational; they can’t anticipate everything that might ever happen, and they can’t perform complx tasks of optimization except at very high cost (if at all). People also have the potential for acting opportunistically; they may try to take advantage of others if the opportunity presents itself and economic or social forces don’t prevent this.
These human factors are relatively unimportant when the transaction is clean and straightforward. But they become quite important when a second set of factors come into play-namely, when the transaction is complicated because it involves time, uncertainty, and privately held information. When a transaction is complicated in this fashion, and when the parties to it are opportunistic, neither party can blithely trust the other’s goodwill in meeting contingencies as they arise.
To some extent the parties will try to anticipate what contingencies may arise later and to specify, at the outset, how those contingencies will be met. However, because the parties are boundedly rational, there are ex ante costs of negotiation. Moreover, simply making an agreement ex ante is not the same as fulfilling it ex post, and opportunism by either side may lead to breach of the agreement. Each party will incur ex post enforcement costs in trying to ensure that the other side meets its obligations under the agreement . And because of their bounded rationality, the parties will probably be unable or unwilling to anticipate up front all the contingencies that will arise later. As time passes and contingencies unfold, the parties to the agreement will renegotiate their arrangement to meet circumstances in other words, there will be ex post (re)negotiation costs.
Each party to a potential transaction anticipates (at least, in some rough sense) that the transaction will involve these different costs. When deciding whether to enter into the transaction, the parties consider the benefits of the transaction, net of the transaction costs. When the transaction costs are very high,the parties may forgo the transaction altogether. When the costs are low enough so that the transaction can be undertaken, the parties will look for ways to arrange the transaction so that the transaction costs are relatively low.
Transaction cost economics studies these transaction costs and, in particular, how they vary with the form of the transaction, by which we mean the formal and informal institutional arrangements for adapting, adjudicating, and enforcing the transaction as time passes. (We will flesh out this vague definition over the next few pages.)
The basic positive premise of transaction cost economics is that transactions will tend to take place in a form that minimizes the combined costs of the transaction. To understand all this, a concrete example may help.