Does Schooling Cause Growth or the Other Way Around?

Barro (1991) and others find that growth and schooling are highly correlated across countries, with each additional year of 1960 enrollment associated with about .6% per year faster growth in per capita GDP from 1960 to 1990. In a model with finite-lived individuals who choose schooling, schooling can influence growth, but also faster technology-driven growth can induce more schooling by raising the effective rate of return on investment in schooling. We consider a variety of evidence to determine the strength of these channels, with two main findings. First, faster-growing countries have at most modestly flatter cross-sectional experience-earnings profiles, consistent with a minority role for the channel from schooling to growth. Second, we calibrate the model using evidence from the labor literature and employ UNESCO attainment data to construct schooling going back well before 1960. We find the channel from schooling to growth to be too weak to generate even half of Barro's coefficient under a range of plausible parameter values. The reverse channel from expected growth to schooling, in contrast, is capable of explaining the empirical relationship. We conclude that the evidence favors a dominant role for the reverse channel from growth to schooling.