Maastricht Chronicles

Sunday, February 22, 2004

I'll sell you a secret... 

I have this game I used to play with my little niece Clarisa when she was four or five years old. Playing with her, I would tell her, "Clari, Clari, do you want to know a secret?"

And, of course, being four or five, she'd get all excited and say "yes! tell me!"

Only the secret wasn't for free. "I'll sell it to you," I'd say. "100 bolivars."

At this point, Clari would get this kind of quizzical look on her face. Something was not quite right about that, and she could tell. Was 100 bolivars a high price or a low price for that secret?

Well, there was no way for her to tell. Maybe it was a really good secret, in which case those 100 Bs. (about 5 US cents) would be totally worth it. Just go, find Bs.100 and go for it. On the other hand, maybe the secret was something she really had no interest at all in. Maybe the secret was entirely worthless to her, and even Bs.100 was way too high a price.

The problem is, there is no way for her to tell what the worth of that secret was until she knew what the secret was. So she'd tell me things like "No, first you tell me, then I pay you," which might have made sense except then she would put me in a bad position: once she already *knew* the secret, what could possibly be the point of paying for it? I could not be sure I would get paid at all if I gave away my "goods" before having any guarantee of payment.

I don't think Clarisa really liked that little game very much - kind of an obnoxious game to play with your uncle, really - but imagine my surprise when I got all the way to fancy Ph.D. school and realized that the little silly game I used to play with my 4 year old niece actually perfectly encapsulates some of the key problems in the economics of innovation and technical change.

Clarisa's parents will be happy to realize that their 4 year old had already grasped the fundamentals of a paradox whose formal statement earned Ken Arrow a Nobel Prize in Economics a few years back. In a path-breaking theoretical paper written 42 years ago now, Arrow described formally some of the oddball features of knowledge that make it a very odd candidate for a good to be traded in the open market.

Knowledge, Arrow pointed out, is strange in that it has a mix of characteristics that place it somewhere in the gray area betwen private goods and public goods. Like a private good, knowledge can be "owned" through patents, trade secrets, trademarks and copyright protection. In some situations, knowledge can easily be appropriated and accumulated privately - I could have appropriated the "knowledge" contained in this essay simply by not posting it to the internet.

But as Napster made clear, knowledge is not like a traditional private good. In some ways, knowledge is non-rival. When I download a song from you over a p-2-p network, I am not depriving you of this song - my enjoyment of the song does not impair your enjoyment of it (unless you're a real snob!) Unlike a typical private good - like, say, a pizza - knowledge can be used many times without decreasing in value, and in some ways (such as when there are network effects at play) the more users who use it the more valuable it becomes.

More relevant to poor Clarisa, any imaginable exchange of knowledge as an economic good suffers from pervasive market failures. The possibility of trade in knowledge is, by definition, predicated on a knowledge assymetry. There has to be something I know and Clarisa doesn't know for the market to function at all - otherwise, I have nothing to sell her. But if she doesn't know what the secret is, she has no feasible way to gauge the economic value of the knowledge to her. Any guess to its value is just as good as any other guess. She's like a costumer at a record shop holding a shrink-wrapped new CD in her hands. She has no idea if she will like the music on it, but she knows that the only way to find out is by shelling out cash.

Obviously, this situation poses very serious theoretical problems to the very possibility of an efficient market in knowledge-goods. Remember that in standard economic theory, markets are said to be fully efficient (or Pareto optimal) only insofar as participants in market transactions have perfect information about the goods they are buying and selling. When it comes to buying and selling knowledge, the assumption of perfect information is not unrealistic, but in fact it's nonsensical. It undermines the whole point of the transaction: buying knowledge only makes sense if you don't know what it is you're buying!

Put differently, markets are fundamentally unable to measure the economic value of knowledge efficiently, because the very idea of knowledge as a scarce resource in need of efficient allocation is problematic.

When it's a Bs.100 secret at stake, or a $14 CD, that may not be the end of the world. But when you're a business trying to decide whether to shell out thousands or millions of dollars to buy a production plant blueprint, or a software license, or a process algorithm, or an econometric analysis, or any other knowledge good, the stakes are somewhat higher, though the mechanisms underlying the market failure are exactly the same.

Now, this may seem like an incredibly obscure/esoteric detail of economic theory. And in standard economics programs you would hardly spend more than a few minutes even considering it. However, the program I'm enrolled in is gloriously, thankfully, stridently non-standard.

The schol of economic thought I'm being taught at Intech is variously known as Schumpeterian Economics, Innovation Economics, Structuralist Economics or Evolutionary Economics. Leave it to academics to think up four different names for (basically) the same idea. The school, pioneered by Joseph Schumpeter in the 1920s-1940s, is distinctive in that it puts innovation and innovative activity at the center of the economic system, analyzing the way innovation patterns determine the structure of the economy and studying the selection mechanisms that lead some firms to succeed and pushes others towards failure on the basis of their comparative technological level (that's the evolutionary part.)

One of the best teachers I've studied with here told us that the basic differences is that neoclassical economics has consisted, for 230 years, of the systematic ripping-off of Adam Smith's ideas. The "heretical transgression" of Schumpeterian economics is simply to start ripping off somebody else's ideas for once.

The differences between the basic outlooks of Schumpeter and Adam Smith are pretty fundamental. Smith was a theorist of how a free economy can function at all - how decentralized decisionmaking by millions of different economic actors all motivated by greed can lead to some sort of orderly and efficient allocation process rather than to sheer chaos. The vast majority of what neoclassical economists have been doing for 200-someodd years is simply refining and formalizing that insight about spontaneous coordination through the market, expressing it with increasing theoretical subtlety and mathematical sophistication.

By contrast, Schumpeter is concerned not so much with how the economy can function as he is with with how the economy can develop and grow. In innovation, Schumpeter thinks he's identified the ultimate driver of economic change. From that insight, he builds an economics centered on an analysis of innovative processes which he sees as the engine of economic dynamism. Intech, and some other institutes around these days, are all about expressing that insight with increasing theoretical subtlety and mathematical sophistication.

What's clear is that if innovation is a pervasive phenomenon and if most real economic advance can be traced back to innovative activity, then the market economy is kept alive not by its efficiency but by its inefficiency. That might seem like a somewhat odd statement, but Clarisa, on some level, understands it. If innovation drives the process of economic change but markets cannot, by their nature, accurately reflect the value of innovation, then the fundamental engine of economic growth is not the market's allocative efficiency but rather the generation of something that cannot be efficiently traded in an open market.

For this reason, Schumpeterian economic does away with the notion of market equilibria altogether. It's when markets fail to clear that economic growth can happen, not when markets clear. Economic growth springs from a succesion of knowledged disequilibria within an evolutionary selection environment, not from the allocative efficiency of the market. In fact, if you read traditional neoclassical economic theory, it's not entirely clear WHERE growth and dynamism springs from. If Adam Smith and Walras and Marshall and the rest of them were right, efficient markets should generate stable, pareto-optimal steady-state equilibria, not ongoing growth. The fact that the economy shows a marked tendency to continue to grow over time has taken a huge amount of work to explain in neoclassical theory, leading up to the complex formulations of Romer and Lucas and the New Growth Theory crowd. Certainly, it is odd to realize that a major theoretical reworking of the standard textbook models was necessary to make economic growth even intelligible within the neoclassical framework.

I don't want to dismiss neoclassical economics. God knows they've come up with mindbogglingly sophisticated, intricately detailed explanations for all kinds of economic phenomena and there's a real danger of throwing out the baby with the bathwater. In terms of statics analysis, neoclassicists have a huge amount of very interesting and important things to say. But I tend to agree with Schumpeter that "how can the economy grow and develop" is a more interesting question than "how can the economy function" - the dynamic question is much more interesting than the static one. So I'm thrilled to be at Intech studying this stuff.

This page is powered by Blogger. Isn't yours?

Weblog Commenting by HaloScan.com