Optimal Income Taxation with Adverse Selection in the Labor Market
Massachusetts Institute of Technology - Department of Economics MIT
April 17, 2012
MIT Department of Economics Graduate Student Research Paper No. 12-01
This paper studies optimal linear and nonlinear redistributive income taxation when there is adverse selection in the labor market. Unlike in standard taxation models, firms do not know workers’ abilities and competitively screen them through nonlinear compensation contracts. The equilibrium concept used is the Miyazaki-Wilson-Spence (MWS) one, adapted to a labor market with taxes. The government observes neither abilities nor the private market contracts and has to foresee the reaction of firms, in addition to workers. Adverse selection leads to different responses to taxes than in the standard Mirrlees (1971) model, because of the use of work hours as a screening tool by firms, which for higher talent workers results in a .rat race. Accordingly, the new optimal income tax formulas include corrective terms for the rat race and redistributive terms take into account the informational rents and cross-subsidies received by lower productivity workers. The most surprising result is that, if the government has sufficiently strong redistributive goals, welfare is higher when there is adverse selection than when there is not. This result is due to the rat race in which high productivity workers are caught, which limits their flexibility to react adversely to distortive taxation. I draw the link to policy praxis by discussing various policies that a government can use to endogenously affect adverse selection. The model also has practical implications for the interpretation, estimation, and use of taxable income elasticities, which are central to optimal tax design.
Number of Pages in PDF File: 57
Keywords: Adverse Selection, Labor Market, Optimal taxation, Rat Race, Redistribution, Screening contracts
JEL Classification: D82, H21, H23, H24working papers series
Date posted: April 26, 2012 ; Last revised: December 30, 2013
© 2014 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo6 in 0.312 seconds