Tax Rates and Tax Evasion: Evidence from “Missing Imports” in China

Raymond Fisman and Shang-Jin Wei


Tax evasion, by its very nature, is difficult to observe. We quantify the effects of tax rates on tax evasion, by examining the relationship in China between the tariff schedule and the ‘evasion gap’ which we define as the difference between Hong Kong’s reported exports to China at the product level and China’s reported imports from Hong Kong. Our results imply that a one percentage point increase in the tax rate is associated with a 3 percent increase in evasion. Furthermore, the evasion gap is negatively correlated with tax rates on closely related products, suggesting that evasion takes place partly through mis-classification of imports from higher-taxed categories to lower-taxed ones, in addition to under-reporting the value of imports.


This paper studies the responsiveness of tax evasion to tax rates. Much of the work in the theory and empirics of taxation has taken tax collection as given and often costlessly executed. This simplification is unlikely to be realistic: even within the United States, where tax collection is considered to be relatively efficient, about 17 percent of income taxes are estimated as unpaid (Slemrod and Yitzhaki, 2000). One particularly important issue is understanding the relationship between tax rates and tax evasion. A number of models have evolved to incorporate tax evasion, but these models fail to provide any prediction regarding the uniform impact of tax rates on evasion. In the pioneering work of Allingham and Sandmo (1972), the relationship between tax rates and evasion is positive, but this depends on particular assumptions of risk aversion and the punishment for evasion. A broader review of the literature reports that, more generally, theoretical predictions of the effect of tax rates on evasion are highly sensitive to modeling assumptions (Slemrod and Yitzhaki, 2000).1 Furthermore, even if the effect of tax rates on evasion may be signed, there is still a need to assess the magnitude of this effect. Hence, empirically examining the effect of tax rates on evasion would be very useful from the perspectives of both theory and policy. This has proven to be a challenging task due to the difficulties in measuring evasion, which by definition is not directly observed.