The role of reporting institutions and image motivation in tax evasion and incidence

Abstract

We investigate effects of tax reporting mechanisms on evasion and incidence in experimental double auction markets where counterfactual reporting and market outcomes can be studied after convergence. There are two control conditions: (i) markets without taxes and (ii) markets where taxes are automatically levied. These are compared to (iii) markets with seller-reporting only and fines paid if low-probability audit discovers evasion, to (iv) markets with both seller-and buyer-reporting and a higher audit probability due to any gap in the numbers reported by the seller and her customers, and to (v) markets where, in addition, buyer-reporting is costly. The latter two mimic varying reporting incentives in the so called third-party reporting in tax enforcement. We find that 20% of the sellers are truthful when only sellers report, but that 80% and 66% of them are truthful under costless and costly third-party reporting, respectively. Pricing, incidence, and reporting patterns in all treatments can be explained by a model of lying costs with image concerns based on Gneezy et al. (2018). JEL Codes: H21, H22, H26, D40, D44, D91