Working Paper: NBER ID: w28847
Authors: Elizabeth Ananat; Shihe Fu; Stephen Ross
Abstract: Racial social isolation within and across workplaces may reduce firm productivity. We provide descriptive evidence that African-Americans feel socially isolated from Whites. To test whether isolation affects productivity, we estimate models of Total Factor Productivity for manufacturing firms allowing returns to local area concentrations of economic activity and human capital spillovers to vary with the racial and ethnic composition of both the establishment and the local area employment. Higher own-race exposure for establishment workers to workers at surrounding establishments strengthens the relationship between productivity and both employment density and concentrations of college educated workers. Effects for human capital spillovers are largest for firms with more patents and more research and development spending. Looming demographic changes suggest that this drag on productivity may increase over time.
Keywords: agglomeration economies; race-specific spillovers; firm productivity; social isolation; human capital
JEL Codes: J15; J24; L11; R12; R23; R32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| racial social isolation (J15) | firm productivity (D22) |
| higher exposure to own-race workers (J79) | relationship between productivity and employment density (O49) |
| higher share of same-race workers (J79) | stronger relationship between productivity and local area employment density (O49) |
| no exposure to same-race workers (J79) | impacts of employment density and share of college graduates on productivity diminish to near zero (D29) |
| same-race exposure (J15) | productivity in industries with higher levels of patent activity and R&D spending (O39) |