Many theorists suggest that the concentration of human capital in a region helps improve productivity for firms and individuals by creating social learning chances for workers. Such human capital spillovers can generate innovations and new ideas which are the driving forces for sustainable long-term growth. My dissertation provides systematic empirical evidence for human capital spillovers using micro-level data from China.

In chapter 1, I provide motivation, a discussion of identification challenges, a literature review on human capital spillovers, a description of the data used in my dissertation, a description of the Chinese economy, and a preview of my identification strategies and findings.

In chapter 2, I investigate the effect of aggregate human capital on productivity in an indirect way. I compare the wages of otherwise similar individuals that live in cities with different level of college share. The resulting estimates indicate that workers working in cities with higher human capital do have higher wages than

otherwise similar workers in cities with lower human capital. Interestingly, I find that both skilled and unskilled workers benefit from the increase in human capital, but unskilled workers benefit more than skilled workers. This is may be due to imperfect substitution between skilled and unskilled workers.

In chapter 3, I take a constant-composition approach, which in theory sepa- rates imperfect substitution effect from spillovers by holding the skill composition in the workforce constant, to further investigate the existence and magnitude of human capital spillovers. The results show that the relationship between workers’ wages and city-level human capital remains positive and statistically significant. The es- timates from individual wage data indicate that a one percentage point increase in the share of college-educated workers in the population is associated with a 1.4 to 3.6 percent increase in wages.

In chapter 4, I take a direct approach to estimate the impact of aggregate human capital on productivity. Specifically, I apply a first differenced instrumental variable model to a balanced firm panel data to study the impact of an increase in the share of college-educated workers on firms’ total factor productivity (TFP). I find that one percentage point increase in college share in a city increases firms’ TFP by 0.8 to 2.1 percent. Private firms are more responsive to overall human capital than state firms, and the human capital spillovers are stronger in denser and larger cities. Chapter 5 concludes.