Generally, I restrict this weblog solely to discussions of new economics research, with a bias toward theory; there is no shortage of good economics writing on the internet about policy, or the “economics of everyday life”, or economic statistics.
That said, I understand that theory is often esoteric. How, you may wonder, do some of these results apply to “real” economics, or to the “real” world? An editorial by a top-notch empirical economist, Ken Rogoff, is making the rounds on the internet today, and I think this is a great example to show the value of theory.
Essentially, Rogoff argues that growing inequality will in some sense self-correct. This is because “simply put, the greater the premium for highly skilled workers, the greater the incentive to find ways to economize on employing their talents.” Rogoff is considered quite liberal, and supports efforts to expand educational opportunity, but “it would be foolish, if not dangerous, to infer rising inequality in relative incomes in the coming decades by extrapolating from recent trends.”
Unfortunately, theory suggests this argument is not totally correct. As discussed on this blog earlier this year, Samuelson pointed out decades ago that there is, from the perspective of a firm in general equilibrium, no such thing as cheap or expensive factors of production. This is true even if the firm is a monopolist; all we require is free entry in factor markets. The argument is simple: every factor of production is paid their marginal products. A firm has no greater need to economize on high-skilled labor than they do on low-skilled labor or on capital: at the margin, lowering costs is lowering costs.
Constructively, what does this mean for inequality trends? It means that to decrease inequality, either marginal products of different workers need to equalize (perhaps implying more equality of educational opportunity), or factor markets without free entry need to loosen. The second argument strikes me as the important one. Intellectual property, occupation licensing requirements such as those in medicine, skilled immigration restrictions that are biased across sectors of the economy, and high-paying patronage jobs are a few among many distortions on the factor side. Theory tells us that the market will not self-correct these distortions, and therefore contra Rogoff, we should be worried about long-run growth in inequality.
A Fine Theorem 
Great point. “the greater the premium for highly skilled workers, the greater the incentive to find ways to economize on employing their talents” is basically saying that we are currently out of equilibrium. In equilibrium, if there were a better way to economize, wouldn’t it have happened already?
Is there a way to salvage Rogoff’s point? I think so. As time evolves and, say, structural parameters change in a way that would make some factor more productive, one thing that makes the factor price less likely to change drastically is that the technology will re-equilibrate to de-emphasize that factor. This isn’t decisive, but neither is it a point completely negated by theory. It’s a le Chatelier principle.
Agreed, but I don’t think Le Chatelier solves inequality: it just means that the rich class will get smaller, but no less rich as individuals.
I think what Rogoff is missing (and he’s in good company, since Hicks made the same mistake) with the reference to goods is exactly Samuelson’s point. When produce becomes expensive because of some weather event, this is analogous to some type of labor supply being artificially restricted. If a greenhouse can be built at a cost less than the rents being gained by the produce-owner, then there is a great incentive for an innovator to do so. But equilibrium labor inequality is not at all like that. Equilibrium labor inequality is not about skilled labor being nominally “expensive” and unskilled being nominally “cheap”, but rather about one being more productive than the other, even after optimally setting inputs into the production. It’s implausible that we’ve been out of equilibrium for the last 35 years as US income inequality has grown. So the story about how innovation affects inequality must be one about either differences in factor productivity or about frictions. Both of these lead me, at least, to the opposite of Rogoff’s assumption.