Monday Nov 13, 2006

Transaction Costs and Search

Transaction Costs Economics (TCE) has gradually evolved its own specific tools of trade.

While works by Ronald Coase, Oliver Williamson and Douglass North have disclosed the basic tenets of TCE, a significant group of econometricians and neoclassical (or should I say analytic) economists have turned to the development of formal tools for analyzing transaction costs.

Earlier, econometricians focused on the structure of deals and whether core concepts of TCE such as specific investments, hybrid schemes, etc., had any conceptual validity. Neoclassical economists focused on applying the "costs of search" concept to the analysis of transaction costs.

Let me  motivate the neoclassical approach to formalizing transaction costs as search costs.

The cost of doing business can be imagined as the cost of searching for suppliers and customers, for example, the cost of finding good employees or good business partners.

As another example, consider real estate deals. A good deal of the cost of a real estate transaction is the search cost. In fact, the agents' fees themselves can be said to do with search costs.

Another case has to do with brands. Brands are themselves used to reduce transaction costs by reducing search efforts. A brand says what it is that you get and you know you'll get that. Unless you get what the brand promises, the brand value will be lost, and you'll start a search for another trustworthy brand. In a sense, the brand value is embedded in all the savings in search. If you have good brand value, people come to you instead of searching. You buy a certain brand of audio equipment, computers, etc., instead of searching for them every single time. Even in things as common as food, some retailers have exerted great effort to develop a brand. So, you go buying meat or oranges at Whole Foods or Trader Joe's, etc., because they just taste different and are more "natural" or "organic." You don't have to keep searching for the right orange or the right piece of beef. In the old times, the local butchers or grocery relied on their reputation to develop a kind of "brand." So, we went to Joe's butcher or Jack's grocery.   

So, does Internet search engines reduce such search-related transaction costs. I'm of the view that they bring mixed blessings.  

There are two problems here. First, there is a problem in imagining all transaction costs as search. Second, there is a problem in equating search in real transactions with search on the Web .... (To be continued) ...

...(I did want to continue from here but Richard Veryard anticipated some things I wanted to say.)

First, the concept of "search" does not capture all transaction costs. The cost of maintaining a certain level of expertise while can be seen as a way to avoid search costs is not itself a search cost although it may involve search costs such as search for methods that would lead to a successful maintenance. Furthermore, actual calculation of transaction costs in their true sense can be quite complex. So, modeling transaction costs as "search" costs is methodologically similar to the modeling of human beings as utility-maximizing machines. Both are gross abstractions.

Second, "search" on the Web hardly proves even a poor replacement for actual "search" in the real transactions. In real transactions, we rely on relationships. The fact that we rely much more on our real world relationships than on what we find on the web, proves my point.




« November 2006 »