On The Margins

Wednesday May 18, 2005

[ Business ] Where Do Transactions Come From?

Singapore, Factory Floor

Carliss Y. Baldwin (Harvard Business School) and Kim B. Clark (Dean of Harvard Business School) have a paper, "Where Do Transactions Come From? A Perspective from Engineering Design," that lays out the basic concepts of transaction cost economics (TCE), heavily relying, in that respect, on the works of economist Oliver Williamson. While the paper does provide a review of TCE, their main focus is to describe what forces give rise to transaction boundries. So, it may actually be more useful for setting up the basic ideas of TCE from the perspective of the engineering and design communities, and in doing that, the paper has great value and is a must-read for every engineer, product manager or corporate strategist. Take the time to take a deep peek. You will not be disappointed.

Addendum (added May 19, 2005):

Reading more of the paper, I noticed that Baldwin and Clark, while borrowing heavily from Ronald Coase, Douglass North and Oliver Williamson, are imposing a narrow definition of what a transaction is (or ought to be thought of), leading to certain quite confusing conclusions such as "If transactions are more costly than plain transfers, then why have transactions at all?"

Besides the fact that it is not clear what they mean by "plain transfers" (probably they simply mean those transactions where no explicit compensation is involved, and yet an incentive may be), transaction cost economics (TCE) is about any costly transfer, not just about the "more costly" ones.

Indeed, "transaction" appears to be a much broader concept than "transfer." Transfer is simply a type of transaction where no compensation is involved (although some incentives may still drive it). For Clark and Baldwin, things are the other way around. They view "transfer" as being a broader concept, and "transaction" as only a transfer which involves "more" cost due to exchange of standardized goods that are countable for a consideration (or compensation).

Clark and Baldwin do make a valid observation that standardization is a pre-requisite of counting and compensation, but their narrow definition of transaction mistakes the fact that most "transfers" also involve a certain "standard" and that once a "standard" is involved, counting can take place, compensation can be implicit and the presence of incentives may drive the exchange (even if no compensation is paid for that particular exchange).

So, the concept of transaction is actually more broad than transfer (unless we understand transfer from a narrow engineering perspective), and transfer is simply a transaction for which an explicit compensation has not been paid (although, again, an incentive may have been driving it). I have now said this three times in three different ways. An example may be useful:

Take the family. Within the family there are many internal transfers of goods and services for which no compensation is paid but all those transfers of goods and services are transactions between members of the family and have associated transactional costs.

A transaction, I conjecture, is one of those things of which we say "you know one when you see one." Just like a chair. How do you define a chair unless you have a chance to see it and use it? Similarly, the concept of "transaction" as understood in transaction cost economics cannot be "defined" using some rule-based conditional model as Clark and Baldwin are having difficulty to do. (Conditional definitions are of this sort: "If X is Y,Z and W, then X is it.) To define transactions in narrow (rule-based) terms, as Clark and Baldwin have attempted, goes against the very grain of transaction cost economics and makes valuable (golden) conclusions like the one drawn by Ronald Henry Coase difficult to draw:

[A] firm will tend to expand until the cost of organizing an extra transaction within the firm become equal to the costs of carrying out the same transaction by means of an exchange on the open market or the costs of organizaing another firm— (emphasis from Clark and Baldwin)

Since within a firm, we are usually dealing with transfers, it is obvious from this quote from Coase that he was thinking of those transfers as transactions—from the perspective of transaction cost economics.

So, I prefer to return to Oliver Williamson's definition of transactions:

A transaction occurs when a good or service is transferred across a technologically separable interface. One stage of activity terminates and another begins.

Here, we simply need to know what we mean when we say "technologically separable" or even better, simply "separable," and I think that is where the crux of the matter lies and the place to which Williamson is pointing us: "One stage of activity terminates and another begins." He does not say anything about standardization, counting or compensation, ingredients Clark and Baldwin insist must be present to grant it the status of a transaction and to imbue it with transaction costs.

Transactions, Economics, Business, Engineering, Design, Transaction Cost Economics.

2005-05-18 14:35:58.0 -- Comments [0] ; Permalink ; Trackback.

[ Sun Microsystems Inc. ] The Santa Clara Campus' Gardens Come to Life

It is nice when something pleasant breaks us out of our habits of thought and behavior. Just a few minutes ago, locked into some thought while carrying my food from the cafeteria to my office, I noticed something rather special on my path. We have been on this campus for more than four years but all of a sudden something seems to have changed.

I don't know who takes the responsibility for the gardens in the Santa Clara campus of Sun Microsystems Inc. but one thing is for sure. In the last couple of weeks, the gardens have come to life and compare very well with those in the Menlo Park campus. If you're in the Santa Clara campus, make sure you walk around and take a good look at the care taken with the patches of green we have here.

Unfortunately, I don't have any picture of the gardens to share right now but if I have time, I'll post a few on my hiptop blog.

2005-05-18 13:27:59.0 -- Comments [0] ; Permalink ; Trackback.