Wednesday, November 16, 2011

Growth by saturation

When you compare rich and poor economies, you notice that many things differ between them One of them is that rich countries enjoy a much larger diversity of goods. It is not just that they carry more broads categories of goods, there are also more variations of a given good. For example, rich countries may have more car types and each model is sold with many more variations than in poor countries. Explaining this could be interesting.

This may be the motivation of Kozo Kunimune who builds a growth model with three essentially characteristics: there is constant growth in factor productivity, there is a ranking of goods with respect to the utility they provide, and consumers have a saturation point for each good. This implies that once they are saturated with a good, any "excess" production capacity can be dedicated to another good. As there is a strict ordering, goods are satisfied in succession and the number of goods defines the level of development. Also, the time needed to satisfy a new good decreases if the growth rate is constant, a feature that sounds empirically correct. But of course, there is no evidence whatsoever that we have such "lexicographic cum Leontieff" preferences. The model also violates elementary principles of Economics, such as local non-satiation. Not a particularly useful model.

3 comments:

T.V. said...

I haven't read the paper so my comment is purely based on your post -- but why should the mere fact that the paper adopts an assumption that conflicts with LNS imply that it is bad research?

It's true that LNS is at the core of economic theory, since it's the only condition needed (except of course for price taking, universal price quoting, divisibility of goods, etc) to obtain the first fundamental theorem of welfare economics. But exactly because it's so fundamental, and given your previous post on the value of non-conformism in academia, why wouldn't it be useful to see what happens exactly when dropping this LNS assumption ? Especially if the model's predictions would be consistent with reality, that would give some food for thought about the LNS assumption, no?

Also, I'm sympathetic towards this alternative assumption, which indeed conflicts LNS. I'm just considering the following example which has frequently crossed my mind when studying general equilibrium theory. Maybe I'm revealing my ignorance here, so correct me if I'm wrong, but LNS implies, roughly speaking, that people's desires are unlimited if only their budget would be large enough. But there's plenty of products of which I only need one -- I might buy the best cell phone in the world if I could afford it, I could even buy 3 of them, but even with no budget constraint at all would I buy 10. Of course, this is not a perfect example, because cell phones are non-divisible goods and the definition of LNS hinges on this divisibility. But since so many goods are non-divisible, that in itself might be a reason to try alternative assumptions to LNS. (I realize that my example also hinges on the fact that I have a time constraint, so that 'no budget constraint' is never possible.)

Of course, this example is purely personal and is not the hard evidence on this type of preferences that we should be looking for, but I'm trying to think of a good of which I would want to have an infinite amount if I could afford it and have time for it, and I can't think of any. Maybe it's just that I have atypical preferences, maybe I don't fully grasp the LNS assumption.

However, I still think that it's worth the effort to try modeling alternative assumptions.

Looking forward to your reaction.

Economic Logician said...

One problem is that this paper assumes that there is satiation for all goods. This is very wrong. While it will always be possible to find some good for which this holds true, you will never be able to claim that this is valid for a majority, let alone all.

Second, local non-satiation is necessary for a lot of equilibrium proofs. This [paper actually does not have any proof, so I would rely on any standard intuition for believing the results.

Economic Logician said...

T.V., I forget to answer your point about the benefit of non-conformism I discussed in an early post. Yes, but non-conformism should not be silly and senseless. That is he case here.