Human capital and knowledge spillovers in the United States and Brazil

This research examines whether knowledge spillover exists in the United States and Brazil, as proposed by Lucas, Jr. (1988), Romer (1990) and Acemoglu (1996). Specifically, this paper uses a quasi-Mincerian approach to verify wether the concentration of  college-educated individuals employed in the business service sector contributes to increased productivity in other sectors of the economy. We estimate the return to education in major market sectors using data from the 2008 U.S. Current Population Survey (March supplement) and from the 2008 Brazilian household survey (PNAD).  The rationale for considering Brazil and the U.S. is that Brazil and the United States are positioned differently in the development process, so by comparing the results for these two economies one can learn about the importance of knowledge spillover in economies that are either far or close to the technological frontier.

Our results provide evidence that knowledge spillover has an important impact on wages in Brazil and in the United States. Using quase-Mincerian wage equations, we show that the individual wage is influenced positively and significantly by sectoral knowledge concentration. There is also evidence to support that the knowledge spread intensity exercises a positive and significant effect on  wages. We take these results as evidence to support the theoretical model of Romer [1990], as well as the existence of human capital externalities in the Brazilian and in the U.S. economies. The results also provide evidence that the Brazilian economy experiences increasing returns to scale to schooling. However, the United States experiences diminishing returns to schooling.

Whereas in Brazil the return to schooling increases from 3.1 percent to four years of schooling to 32.7% with 17 years of schooling, in the United States the return to schooling change from 2.0 percent with 5.5 years of schooling to 12.2 percent with 20 years of schooling. For the Brazilian, data we observe a turning point at 5.2 years of schooling, where the initially decreasing returns become increasing. A explanation for the difference between countries regarding returns to scale in accumulation of human capital may be a result of their distinct development stages. On the one hand, it is notable that Brazil has enormous gains to be made from increasing average schooling in general, and higher education in particular.

Most important, however, is that despite this distinct pattern on the returns to schooling, both countries show similar and strong knowledge externalities. This finding is even more important in the United States, where individuals seem to benefit more from knowledge accumulation in the sector in which they are employed than of their own education. However, this finding does not imply that individuals do not benefit from their own accumulation of human capital, because the average educational attainment is very high across sectors in the U.S. and, therefore, a person can only benefit from knowledge externalities if he/she manages to acquire the skills (education) necessary to find a job.