“DODGING THE TAXMAN: Evidence on Firm Misreporting Behavior and Evasion Substitution”

Reducing tax evasion is a key priority for many governments, particularly in developing countries. A growing literature has argued that the use of third party information to verify taxpayer self-reports is critical for tax enforcement and the growth of state capacity. However, there may be limits to the effectiveness of third party information if taxpayers can substitute misreporting to less verifiable margins. We present a simple framework to demonstrate the conditions under which substitution will occur and provide strong empirical evidence for substitution behavior by exploiting a natural experiment in Ecuador. We find that when firms are notified by the tax authority about detected revenue discrepancies on previously filed corporate income tax returns, they increase reported revenues, matching the third party estimate when provided. Firms also increase reported costs by 96 cents for every dollar of revenue adjustment, resulting in minor increases in total tax collection. JEL codes: H25, H26, O23 We thank Lorenzo Casaburi, Raj Chetty, John Friedman, Edward Glaeser, James Hines, Lawrence Katz, Henrik Kleven, Wojciech Kopczuk, Michael Kremer, Asim Khwaja, Ben Olken, Andrei Shleifer, Joel Slemrod, and participants at the conferences of the American Economic Association, BREAD, Max Planck Institute, National Tax Association, NBER, NEUDC, PACDEV, Swiss Economists Abroad and in seminars at Brandeis, Columbia, Harvard, Lugano, Williams and the Centro de Estudios Fiscales for helpful comments and discussions. We thank Sajjad Goli, Andrea Lopez Luzuriaga, Marco Martinez del Angel, Parag Pathak, and Lucas Zavala for excellent research assistance. We are exceedingly grateful to the staff of the Centro de Estudios Fiscales and the Departamento de Control of the Ecuadorian Tax Authority for outstanding collaboration and to the Harvard Weatherhead Center, Harvard Taubman Center, and Harvard Business School for research support. George Washington University, [email protected]. Harvard University and NBER, [email protected]. Harvard University and NBER, monica [email protected].