Lists: A Novel Experimental Method to Measure Tax Evasion

Abstract

This paper presents a novel experimental method to measure the propensity to evade taxes that overcomes sensitivity bias. Drawn from political science research, the method uses list experiments to estimate the prevalence of tax evasion in different population groups. Using this method, we found that 13 percent of the Canadian population admitted to evading income taxes and 29 percent admitted to evading consumption taxes. These estimates are comparable to estimates based on audit data. They provide a new source of reliable data about evasion, which can be combined with measures of psychological drivers of evasion at the individual level.