Q: Would you agree with the definition of an institution as a set of rules designed to reduce transaction costs?
A:I believe that some institutions have arisen specifically to reduce transaction costs, usually to solve asymmetric information problems though this is not necessary.
North (1991) defines institutions as those man-made rules that determine transaction and productions costs together with the standard constraints of economics. He also argues that institutions determine the cost of transaction – by raising the benefits of solutions or by increasing the cost of cheating – thus making economic activity possible.
Akerlof’s (1976) paper on the function of the lemons market points to the importance of having an institution exist to solve the lemons problem. Grief (1993) in his study of the Maghribi trading coalition explains the functioning of an institution that arose to solve an information problem. The solving of this information problem in turn guaranteed the existence of economic activity (trading) by reducing the transaction cost that would have otherwise been too high to support the economic activity.
Institutions can, however, also be seen as the end result of an evolutionary process which does not necessarily benefit everyone, or increase overall welfare (Sugden 1989). In such an argument, the reason for the existence of certain rules is not they solve information problems or reduce transaction costs necessarily but simply that enough people follow them. Every individual who obeys these rules stands to gain given the expectation that every other individual will also obey these rules.
Sugden (1989) proposes the idea of institutions as spontaneously arising rules that may or may not be efficient solutions. He argues that certain rules arise in society simply because other people follow it. Also, anyone following this rule expects to do as well, if not better, than anyone else following other competing rules. Viewed this way, institutions as a collection of rules don’t necessarily have to reduce transaction costs. In fact, Sugden argues that in many instances, institutions arise that are not pareto –efficient. They “need not be well adapted to the problems of coordination they resolve”.
So, while the definition of institutions as a means of reducing transaction costs is reasonable, it is not complete. Some institutions certainly do not solve transaction costs (the layout of the keyboard I am typing on for example) – for these rules, one must extend the definition to include the concept laid out in Sugden’s 1989 paper.
I amuse myself here; Your amusement is intended but boundedly so
3 comments:
hmm well written. what if i also look at transaction cost (delay, cost of time/argument) of arriving at a decision (a decision of allocation, design, anything). so, for eg. the keyboard design becoming a standard, reduces all future transaction costs in deciding the design of keyboards. in that sense, all norms no matter how innocent and unplanned, become transaction (delay) cost minimizers
Transaction cost is the cost associated with any transaction - usually between a buyer and a seller. So the keyboard example I gave is not really appropriate to this.
The origin of any institution may not have resulted from a cost-minimizing perspective; so we may be left with a high cost institution (if it is the only one that survives). So in the sense of "does this rule minimize cost compared to any other rule" you cannot then say that institutions are transaction cost minimizers.
When one rule becomes a standard, to say that now that standard has solved the problem of deciding which rule to follow thus minimizing costs in the future is alright, but that rule did not arise to do this.
Also, you're expanding the definition of transaction cost which is making things more complicated. But then you have to judge the evolution of that rule too by the same definition. So during the evolution of any rule there is confusion which possibly adds to transaction cost. By this definition, okay, I guess your point is valid.
But you still need to add the two dimensions of transaction cost.
Post a Comment