I contend that the average wage within a given country is pinned down by the productivity of its best workers in a Kremer-style (1993) “O-ring” sector. In this sector, high-skilled workers perform tasks that depend on strategic complementarities in production.
Other less-skilled workers in that same country aren’t good enough to work in this “O-ring” sector–at least not without taking a massive wage cut. But in equilibrium, they can work in a conventional, diminishing-returns-to-labor “Foolproof” sector, earning only slightly less than the high-skilled workers in their own country. Importantly, high-skilled workers can move between the O-ring and Foolproof sectors. The key assumption of this model is that the wage of high-skilled workers must be equal across the two sectors, a simple invocation of the law of one price.
Under this model, within a given country, the less-skilled workers will earn only slightly less than the highly-skilled workers, since, after all, in the Foolproof sector high- and low-skilled workers are close substitutes. But across countries, a nation whose best workers are slightly lower in quality will be much less productive, since it means that workers in that nation’s “O-ring” or “weak link” sector will produce much less.
In his seminar presentation, Jones explained the implications of the model for low-IQ immigration. Long story short: Low-IQ immigrants do not reduce the productivity of high-IQ natives - any more than short immigrants reduce the height of tall natives. (See here for further discussion). IQ research has often been a rationalization for immigration restrictions, but that's largely because few psychometricians understand the principle of comparative advantage.