Working Paper: CEPR ID: DP16797
Authors: Kevin Spiritus; Etienne Lehmann; Sander Renes; Floris Zoutman
Abstract: We analyze the optimal nonlinear income tax schedule when taxpayers earn multiple incomes and differ along many unobserved dimensions. We compare the tax perturbation and mechanism design approaches and find that both produce the same results if the number of incomes is equal to the number of unobserved characteristics. The mechanism design approach then requires slightly less stringent assumptions than the tax perturbation approach. We introduce a numerical method to determine the optimal tax schedule. Applied to the optimal taxation of couples, we find that the optimal isotax curves are close to linear and parallel. We make several additional contributions, including a test for Pareto efficiency and a condition on primitives that ensures the government’s necessary conditions are sufficient and the solution to the problem is unique.
Keywords: nonlinear; optimal taxation; multidimensional screening; household income taxation
JEL Codes: H21; H23; H24; D82
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Joint Income Tax Discounting Female Income (D15) | Social Welfare (I38) |
| Mechanism Design Approach (MD) (D47) | Optimal Tax Formulas (H21) |
| Tax Perturbation Approach (TP) (H29) | Optimal Tax Formulas (H21) |
| Mechanism Design Approach (MD) (D47) | Tax Perturbation Approach (TP) (H29) |
| Negative Welfare Weights (D69) | Self-Financed Pareto Improvement (D10) |
| Optimal Tax Schedule (H21) | Behavioral Responses of Taxpayers (H26) |