Files in this item



application/pdf9952954.pdf (6MB)Restricted to U of Illinois
(no description provided)PDF


Title:Two Essays on Tax Evasion
Author(s):Arias, Omar S.
Doctoral Committee Chair(s):Gahvari, Firouz
Department / Program:Economics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Economics, General
Abstract:These two essays analyze theoretical and empirical issues in tax evasion. The first studies the optimal tax mix in a competitive economy of identical firms and heterogeneous taxpayers where linear commodity and income taxes can be evaded at a private cost and under the uncertainty of audit lotteries. We derive and provide an intuitive interpretation of a generalized: (i) 'Ramsey equation' for optimal commodity taxes, (ii) formula for the optimal income tax rate, (iii) the tradeoff between tax rates and audit probabilities, under multilateral tax evasion. We show that besides the standard redistributive role, commodity taxes provide insurance against the uncertainty of income tax audit lotteries whose distribution of benefits affect the optimal commodity tax structure. We find that optimality typically calls for a mix of taxes, at odds with the heavier reliance of developing countries on uniform commodity taxation on grounds of differences in tax compliance. Finally, the impact of tax evasion on income tax progressivity depends on its effect on the excess burden of taxation and how the proportion of income evaded varies across the population. The second paper re-examines the empirical determinants of income tax evasion using censored quantile regression methods on U.S. TCMP data with a specifically suited algorithm. Semiparametric tests reject key assumptions of maximum likelihood methods, pinpointing important biases in Tobit estimates. Quantile estimates are often more precise than MLEs in the samples of large evaders, apparently due to skewness and long tails in the evasion distribution. We invariably find a fairly elastic positive correlation of income underreporting with tax rates, and a positive association with income for the samples of large evaders. Heterogeneity tests indicate that taxpayers in the upper tail of the conditional evasion distribution are less responsive to changes in factors that affect the rewards and costs of noncompliance, including marginal tax rates. Further, the dispersion in evaded incomes is higher among filers facing lower taxes, filers with more visible business and farm income and/or less complex non-business returns, who therefore show more room for idiosyncrasies in evasion behavior. This may reflect differences in unobservable factors such as tastes for honesty, knowledge about the tax laws and concealment costs which are rendered invisible in conditional mean models despite their importance in explaining evasion behavior.
Issue Date:1999
Description:161 p.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1999.
Other Identifier(s):(MiAaPQ)AAI9952954
Date Available in IDEALS:2015-09-25
Date Deposited:1999

This item appears in the following Collection(s)

Item Statistics