Do Input Quality and Structural Productivity Estimates Drive Measured Differences in Firm Productivity?

Firms in the same industry can differ in measured total factor productivity (TFP) by multiples of 3. Griliches <p> (1957) suggests one explanation: the quality of inputs differs across firms. Labor inputs are traditionally <p> measured only as the number of workers. We investigate whether adjusting for the quality of labor inputs <p> substantially decreases measured TFP dispersion. We add labor market history variables such as experience <p> and firm and industry tenure, as well as general human capital measures such as schooling and sex. We also <p> investigate whether an innovative structural estimator for productivity due to Olley and Pakes (1996) substantially <p> decreases measured residual TFP. Combining labor quality and structural estimates of productivity, the <p> one standard deviation difference in residual TFPs in manufacturing drops from 0.70 to 0.67 multiples. Neither <p> the structural productivity measure nor detailed input quality measures explain the very large measured <p> residual TFP dispersion, despite statistically precise coefficient estimates