This dissertation comprises three essays that explore how human and social capital influence innovation and promote firm dynamism, particularly for small businesses. It studies how firms' uptake of projects is shaped by the local environments, such as superstar firms, talent clusters, and local social capital.

In the first essay, I study the labor channel underlying the agglomeration of innovation activity. It identifies the reallocation of human capital as a key channel of agglomeration spillovers for innovative firms. To measure agglomeration spillovers, I study how R&D labs in different local labor markets respond differently to scientific breakthroughs, which create large and unexpected shocks to innovation productivity in certain technology categories. I document four main findings. i), following scientific breakthroughs, affected labs in thicker local labor markets (i.e., commuting zones with more inventors innovating in a certain field) produce more patents and higher-quality patents, consistent with positive agglomeration spillovers. ii), the increase in patenting is mostly attributed to new hires rather than incumbent inventors. iii), the thick labor market effect is concentrated in states and industries where there is lower enforceability of non-compete agreements and labor is more mobile. iv), using textual analysis to identify lab-level exposure to scientific breakthroughs, I find that inventors are reallocated to labs that are more favorably affected by shocks, which helps labs in thicker labor markets to more easily bring in inventors working in the same niche fields and having a diverse knowledge base. Taken together, these results point to labor mobility as a key force in explaining why innovative firms cluster and suggest that the clustering of firms in thick labor markets can foster corporate innovation by facilitating the productivity-enhancing reallocation of human capital following scientific breakthroughs.

In the second essay, I identify the entry effects of top innovative firms on incumbent innovation. I exploit the inter-temporal variation in patenting activities of local inventors in chosen commuting zones that attracted the firm headquarters and in runner-up commuting zones that were finalists of location choice. Treated and control groups have similar trends prior to the entry, while the local inventors in the chosen zones apply 6.7% more patents, gain 16.8% more top patents, and receive 11.6% more citations. Entry effects are stronger among local inventors who are technologically or socially closer to the entering firm, after controlling for innovation incentives and labor mobility. Social closeness, isolated from technological proximity, consistently explains the innovation gains, which suggests knowledge diffusion is the important channel for local innovation productivity spillovers.

In the third essay, we investigate why small businesses exploit business opportunities better in some areas than others. In a sample of 1.2 million consumer-facing establishments, social capital predicts the uptake of risk-free loans controlling for close-by bank branches, income, and education. One standard deviation increase in the social capital metric accounts for 20 percent of the variation in uptake across zip codes, surpassing the impact of a bank branch within 1000 yards. Strong social capital benefits large, low-growth stores in less-dynamic areas, whereas bank branches benefit small, high-growth stores in more-dynamic areas. Virtual connections have the greatest effect on uptake in already advantaged locations.