Monday, December 18, 2006

Macro versus Micro Economics

I'm still getting used to Blogger Beta. Apparently there is no command to republish the index or republish the entire site. Instead, each item is stored separately in the database and gets republished when it is updated. Sounds good in concept and seems to work for posts. But I updated my profile and yet the update doesn't seem to appear. I hope that works in the near future. (I'm now writing later and it did work....interesting about the lag.)

In the meantime, here is a new feedburner feed for the site

* * * * *

Following David Brooks' suggestion in his Sidney awards, yesterday I read Alan Blinder's piece in Foreign Affairs: Offshoring - The New Industrial Revolution? Blinder is a good economist and this piece is fairly straightforward reading and makes one very important point - on the question of what services are "tradeable" and hence subject to off shoring --- that cuts across historic distinctions and it may very well be that getting a good education, the traditional remedy posed in light of the threat of competition from foreign workers who can produce goods more cheaply, may very well not insulate workers from offshoring for electronically delivered services this time around.

But I think of Blinder as a macro guy, one who is comfortable aggregating up ideas into a simple framework so that can be analyzed, not particularly disturbed by whether that type of aggregation distorts reality too much. I was trained as a micro guy, and below I'm going to comment on things that Blinder leaves out of his story that I think might matter a lot.

Let me focus on a few key issues. First, and here I'm biased by my old job and listening to the Network folks talk where they used the categories "voice" and "data" let's consider those as two different forms of electronic delivery of service. One can get services by interacting with a Web site such as,, or And then one can get service on the phone, talking with a customer service agent. They are both electronic services and there is some convergence between the two but I think for this purpose it is helpful to consider them separately.

The data type of service provision matches the Blinder description quite well. He talked about India being much more type of a threat in this domain than China, then I think it worth noting that written English is much more international than spoken English, which has regional elements, accents and patois. The off shoring of the Web work, in this view makes full sense, both the database content behind the scences and the content we all see, but the off shoring of the phone content is harder. is an example of the success and failure of both approaches - their Web site for building your own PC in a custom way is very well thought through and effective. Their "help desk" for solving technical problems, much more problematic.

If that distinction is sensible, there is an argument that cuts the other way than Blinder argues - voice services, which may not be quite as immune as face to face is from the Baumol's Cost Disease Problem (and if voice to data conversion techniques continue to improve there is the possibility of data mining from voice conversations bringing the above mentioned convergence that much closer) may require regional providers if not local providers to assure the customer feels comfortable in the conversation. Further, being an effective negotiator in voice conversation is potentially a skill that will be in high demand, especially as more of the population becomes comfortable with providing self-service in smart Web sites and hence the voice communication acts less as a substitute means for data communication and more as means for negotiating complex and idioysncratic situations for which the form driven data communication is ill suited.

Indeed, this will be possible if the data technology helps to address the routing problem of matching customers with their specific queries to people with the right expertise to address them. Solve the routing problem completely and I believe much of what Blinder argues about the off shoring of electronic services might prove opposite to how he argues.

Let me turn to a second issue - this is the question of ongoing (trust) relationships, what Arthur Okun coined as the invisible handshake, versus anonymous and one-time intereactions. Blinder's analysis does not differentiate between these two quite different type of transactions. The latter are likely to be commodity-like and the tendency would be for the supply of those type of activites to be off shored. I'm much less certain that true is for the former.

Here is a very special example of this and it is not so much about off shoring as it is about having insiders or outsiders provide a service. One may wonder why college textbook publishers haven't branched out into "online TA" services, since they presumably can leverage economies of scale better than we can at any campus and hence could have a network of such providers online at all hours of the day and night to helpf students. In turn, if they did this and were indeed more efficient at it, it would reduce our need for on campus TAs and hence take a bite out of the tuition bill. But as I've said, we haven't seen this market emerge and if I were betting I'd guess against it happening.

On the other hand, we might very well have online TAs from our own campus, students who took the course in the past from the same instructor who is offering it now and hence who can internalize the nuance and idiosyncracy of how the course is taught. So this is not fundamentally a strength or weakness of electronic delivery. Rather it is about how customized the content is and how to make such custom material well. Again, this seems to favor local knowledge and cuts againts off shoring. Admittedly, we are not offering both alternatives to the students and letting them choose. And the way instructors choose may depend on other factors than price. But let's also agree that students can now take online offering of these course and while some have done this, we are not witnessing wholesale switching to that mode.

Finally, and on this one Blinder does comment a bit, I want to consider the so called "blended" service provision issue - some of it is up close and high touch, some of it is electronic, and the pieces might fit together in as of yet hard to predict ways. Since we're talking about Blended Learning a lot now on campus, this particular issue has gotten a lot of my attention. And the question I'm asking myself is whether one can off shore the electronic piece if it must fit in with a customized and idiosyncratic face to face piece. Blinder talks about reading CT scans as such an off shored activity in Health Care, providing an example both involving some high level skill and one where the cost argument seems to swamp any other consideration of customization. Perhaps it is a portent of things to come. But I've got the feeling that in this case the service is treated in a commodity like way - indeed, that's the way doctors view all services provided by hospital labs. So, yes, that stuff can be off shored but customized stuff...

No comments: