What drives specialization — and why complaining about it doesn’t help
A short note circulated to Vassar, Shulman, and Owain Evans, sketching three drivers of academic and economic specialization, and a brief sociology of why intellectuals who complain about specialization haven’t actually managed to reverse it. The closing line refers to a search for an economics professor whose research overlaps closely enough to be worth cold-emailing.
Theories as to what drives specialization: Human population has been going up for quite a while now (since at least ~1300 AD). This will tend to increase specialization, because with more people there’s sufficient demand present for someone to make a living specializing in a wider range of things — though now that I think about it, this should be relatively minor compared to (2).
The world has grown vastly more interconnected since at least 1500 AD, and this tends to support specialization for the same reason as (1), but probably significantly more strongly because the effect is larger. Since 20K BC world population has gone from around 100K to 7B, but the size of the communities that can efficiently engage in the sort of comparative-advantage trades that support specialization has gone from around 100 to 7B. This is, of course, a gross approximation, because while trade within your inner monkey circle of 100 friends/co-workers/whatever seems to have remained on roughly the same order of difficulty for a fairly long time, trade between you and the rest of the world population has a ton of subcategories with different levels of difficulty that would make for some interesting analysis (“people you have a first-degree connection with,” “people you haven’t met but can get to in an hour,” “people you haven’t met but can get to in a week,” “people who don’t speak the same language,” “people who are actively hostile to you because you’re part of a group they don’t like,” etc.).
Within academia, which was largely what I was thinking about, we should expect to see the existing high degree of specialization because of an unusual constraint on demand: all research must be original, and so you can only “sell” it once. If there’s a sufficiently large market for cotton you can have ten million people who specialize in growing cotton, but you can’t have ten million people who all specialize in the same sorts of papers, because they’d all step on each other’s toes. There are also existing large bodies of knowledge which provide competition, because you can’t “sell” research that was already produced in 1950, while most other goods (with the possible exception of housing) have fairly standard use cycles.
My theory of why people complain about specialization but haven’t done anything about it: There are a sufficiently large number of problems resulting from specialization that the intelligentsia has basically all noticed them and reached agreement that this is an issue and that somebody should do something.
The obvious ways of doing something about it are (a) making a lot of noise so that you can try to form politically powerful organizations which will care about it, and (b) trying to encourage other people to de-specialize through such tools as the aforementioned noise, and providing scholarships, department support, etc. for people doing interdisciplinary stuff. This does not work very well, for basically the same reason that the War on Drugs doesn’t work very well: you can’t fight that powerful a force if you’re going to limit yourself to crude, fairly weak tools like propaganda and a very small chance of an arrest. And even if we took sufficiently drastic measures — shooting everyone who uses drugs, or firing the half of the tenured professors deemed most overspecialized — it would only help somewhat, because only some of the response would be in the form of doing what the government wants; the remainder would consist of things like coming up with more creative ways not to get caught.
I’ll scan the list of econ profs over the weekend, see which one has the most similar research, and pop them an email.