The Rise of Meritocracy and the Inheritance of Advantage

We present a model human capital accumulation with statistical discrimination on the background of agents. Firms do not observe the productivity of workers, they determine their wage after observing two signals. A direct signal on the productivity of the workers and a second signal on the income of the parents of the agent. The signal on background is useful (insofar the direct signal is not perfectly accurate) because parents invest in their children education, increasing their human capital. Knowing that richer parents invest more in education and that their children will be in average more productive, firms rationally discriminate favoring those with perceived good backgrounds. Thus, the income of an agent depends on both the objective information that society has on its human capital and the existing perceptions on her background. Our main result is that the accuracy of both signals have very similar effects in both inequality and intergenerational mobility. More accurate information on the background of individuals facilitates statistical discrimination increasing inequality and its persistence. More accurate information on the actual capabilities of workers leads to exactly the same result via an interesting feed back mechanism. It allows to discern good workers from the able ones, increasing their wage differential and inequality. But more inequality increases the uncertainty that firms have on workers characteristics, which itself increases the weight that firms give to signals, which further increases inequality, and so on. There is, though, a big difference in both types of signals in the manner that they affect investment. Accurate direct information on productivity is directly linked to the return to investment in human capital, as it is what connects human capital with income. The effects of information on parental background on investment are much more indirect, as it implies that your income affects your children independently from the investment you make in them. Using our model to interpret the data suggests that a country like the US (with relatively little intergenerational mobility, large degree of inequality and very large investment in education) might be characterized by having very precise direct signals on ability, resulting in high rewards to merit, and (conditioning on it) little rewards to have a good background. The US might be still a land of opportunities conditional in having the right ability. Of course, those abilities are to a large extent inherited, and that is the reason why the intergenerational persistence of income is so large. But notice that the reason why they are so inheritable might precisely be that merit is highly rewarded; that in a deep sense, the US may still be a very meritocratic society.