Factors Associated With Noncompliance: Evidence from the Michigan Tax Amnesty Program.

Using data collected from individual tax amnesty participants in Michigan, I perform a series of regression analyses to identify factors associated with Michigan amnesty participation and compare these factors with findings from previous taxpayer compliance research. Analysis of amnesty participants provides the unique opportunity to examine previous nonfilers and underreporters (although amnesty participants are a subset of all noncompliant taxpayers). Separate regressions were run on the entire data set, five income stratifications of the entire data set (based on adjusted gross income), and a data base created to examine taxpayers who were likely unknown to the Michigan Treasury Department prior to amnesty. Subject to the confines of the data, this study suggests that noncompliance increases as income increases. Specifically, amnesty participants with twice as much adjusted gross income relative to other amnesty participants had about 51 percent more unreported income. In addition, taxpayers having some form of Treasury Department contact prior to amnesty (either directly or by information submitted to the Treasury Department) had 93 percent less unreported income than those amnesty participants who had no prior contact. Amnesty participants with the most opportunity to evade taxes (a composite of occupation, income level, and access to cash income sources) had 84 percent more unreported income than other taxpayers. Finally, increased noncompliance was associated with (1) sales or self-employed occupations, (2) returns prepared by practitioners, (3) geographic location of the taxpayers, (4) single males, and (5) family size.