Tax enforcement is especially costly when market participants are difficult to observe. The benefits of enforcement depend crucially on pre-enforcement compliance. We de rive an upper bound on pre-enforcement compliance from the pass-through of newly enforced taxes. Using data on Airbnb listings and the platform's voluntary collection agreements, we find that taxes are paid on, at most, 24% of Airbnb transactions prior to enforcement. We also find that demand for Airbnb listings is inelastic, driving three key insights: the tax burden falls disproportionately on renters, the excess burden is small, and tax enforcement is relatively ineffective at reducing local Airbnb activity.
We thank Kate Cuff, Laura Grigolon, Enda Hargaden, Ross Milton, Rich Patterson, Justin Ross, Jay Shogren, Michael Smart, Sarah Smith, Fanglin Sun, Alisa Tazhitdinova, Quinton White, Michael Naretta, John D. Wilson, two anonymous referees, and seminar participants at the 2017 WEA Annual Meeting, 2017 NTA Annual Conference on Taxation, 2018 Public Choice Society Meetings, 2018 CPEG-PEUK Public Economics Workshop, 2018 International Industrial Organization Conference, 2018 Tennessee Empirical Applied Microeconomics Festival, 2018 Taxation and Regulation in the Digital Economy Workshop, 2018 Association of Private Enterprise Education Conference, 2018 Peer-to-Peer Markets Work- shop on Airbnb and the Accommodation Industry, 2019 CESifo Workshop on Taxation in the Digital Economy, Georgia State University, University of Louisville, and University of Alaska Anchorage for their helpful comments. We gratefully acknowledge support from the University of Alaska Anchorage, McMaster University, and the University of Louisville.
A supplemental appendix is available online at https://doi.org/10.1162/rest_a_00910.