Working Paper: NBER ID: w31549
Authors: Henrik Kleven; Claus Thustrup Kreiner; Kristian Larsen; Jakob Egholt Sgaard
Abstract: A key contention in economics is the discrepancy between micro and macro elasticities of labor supply with respect to marginal tax rates. We revisit this question, focusing on the role of dynamic returns to effort among top earners. We develop a new model of earnings responses to taxes in the presence of dynamic returns. In this model, the returns to effort are delayed and mediated by job switches such as promotions within firms or movements between firms. Short-run micro elasticities are attenuated relative to the true long-run macro elasticity. We proceed by providing two main empirical analyses using rich administrative data from Denmark. The first part presents descriptive evidence on earnings and hours-worked patterns over the lifecycle that confirm the predictions of the theoretical model. The second part presents quasi-experimental evidence on earnings responses to taxes using discrete job switches. The empirical strategy is informed by the theoretical model, according to which job switches can be used to (partially) identify the macro elasticity of labor supply. The evidence shows that, at the top of the distribution, macro elasticities are much larger than micro elasticities due to dynamic compensation effects.
Keywords: Labor Supply Elasticities; Dynamic Returns to Effort; Tax Policy; Job Switches; Quasiexperimental Design
JEL Codes: C1; D6; E6; H2; H3; J2; J3
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Micro Elasticity (D11) | Earnings (J31) |
| Macro Elasticity (E19) | Earnings (J31) |
| Job Switches (J62) | Earnings (J31) |
| Dynamic Returns to Effort (I26) | Earnings (J31) |
| Job Switches (J62) | Dynamic Returns to Effort (I26) |
| Job Switchers (J62) | Long-run Macro Elasticity (E23) |