Working Paper: NBER ID: w26640
Authors: Youssef Benzarti; Jarkko Harju
Abstract: This paper uses quasi-experimental variation in payroll taxes to estimate their incidence and investigate how firms use their input factors. We find that higher payroll tax rates lead to large employment responses and have no effects on employee earnings. As payroll taxes increase, firms substitute away from low-skilled, routine and manual workers towards more productive workers and also reduce investments. Our results imply that, contrary to the canonical tax incidence model, firm-level production and input factor choices are affected by payroll taxes.
Keywords: Payroll taxes; Firm behavior; Labor economics; Tax incidence; Employment effects
JEL Codes: H20; H22; H23
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Reduction in less productive labor (J29) | Increase in productivity (O49) |
| Higher payroll tax rates (H29) | Reduction in employment (J63) |
| Higher payroll tax rates (H29) | Reduction in investments (G31) |
| Higher payroll tax rates (H29) | Decrease in fixed asset investments (G31) |
| Higher payroll tax rates (H29) | Increase in research and development investments (O39) |
| Higher payroll tax rates (H29) | No effect on employee earnings (J31) |