Effects of Industrial Concentration on Earnings Outcomes, Combined Nontradable and Construction Sector
| Variables | HHI (1) | Mean Earnings (2) | 90/10 (3) | 90/50 (4) | 50/10 (5) | Gini (6) |
|---|---|---|---|---|---|---|
| log(HHI–m) | 0.344*** | |||||
| (0.0285) | ||||||
| log(HHI) | -0.184*** | 0.396*** | 0.0976*** | 0.298*** | 0.0148*** | |
| (0.0278) | (0.0691) | (0.0223) | (0.0538) | (0.00506) | ||
| Observations | 333,000 | 333,000 | 333,000 | 333,000 | 333,000 | 333,000 |
| R2 | 0.976 | 0.970 | 0.867 | 0.936 | 0.767 | 0.933 |
| Market FEs | Yes | Yes | Yes | Yes | Yes | Yes |
| CZ by year FEs | Yes | Yes | Yes | Yes | Yes | Yes |
| F-stat | 145.0 |
Source: Longitudinal Business Database and Form W-2 2005–2015
Notes: Table reports instrumental variables regression estimates of the effect of local industrial concentration, as measured by the HHI, on measures of earnings and inequality, constructed using earnings data from Form W-2, within the combined nontradable and construction sector, as defined by Mian and Sufi (2014). The first column reports the first-stage regression. In the subsequent columns, the dependent variables are the log of mean earnings (Column 2), the logs of the ratios of the 90th and 10th (Column 3), 50th and 10th (Column 4), or 90th and 50th (Column 5) percentiles of the earnings distribution, and the Gini coefficient (Column 6). Columns represent separate regressions, which include the indicated years of data and fixed effects (FE). Regressions are employment-weighted. Coefficients in Columns 2–5 represent elasticities, while the coefficient in Column 6 is a semi-elasticity. Sample sizes and statistic values have been rounded for disclosure avoidance.