The Aggregate Consequences of Tax Evasion (with Alessandro Di Nola, Almuth Scholl and Anna-Mariia Tkhir)

There is a sizeable overall tax gap in the U.S., albeit tax noncompliance differs sharply across income types. While only small percentages of wages, salaries, taxable interest and dividends are underreported, the estimated misreporting rate of self-employment business income is substantial. This paper studies the aggregate consequences of tax evasion. To this end, we develop a dynamic general equilibrium model with incomplete markets and occupational choice and analyze how incomplete tax enforcement affects wealth inequality and aggregate productivity. In our model environment, heterogeneous agents face a progressive income tax and choose between being a worker and being self-employed. While workers cannot evade taxes, self-employed agents may hide a share of their business income but are confronted with the probability of being detected by the tax authority. Self-employed agents optimally determine the size of their firms by choosing capital and labor inputs, taking into account that detection becomes more likely as their firm grows. Our model replicates important quantitative features of U.S. data, in particular, the misreporting rate of self-employment business income, the fraction of self-employed agents, and the firm size distribution. Our quantitative results suggest that incomplete tax enforcement increases wealth inequality but has ambiguous effects on aggregate productivity. We quantify three channels how imperfect tax enforcement affects aggregate outcomes. First, tax evasion reduces the distortionary impact of taxation on aggregate savings and output. Second, the possibility to evade taxes induces low-productive agents to become self-employed. The distorted occupational choice generates a misallocation of resources to low-productive firms. Third, the probability of being detected by the tax authority induces firms to stay small with adverse effects on aggregate output.