A New Approach for Estimating the Impact of Tax Policies by Race and Ethnicity

This working paper describes the method the Tax Policy Center (TPC) has used to enhance its tax microsimulation model to enable analysis of the distributional effects of tax policies by race and ethnicity. It also presents preliminary results using the enhanced model. Like microsimulation models used in the federal government and other research organizations, TPC’s model relies on a sample of individual tax returns as its primary data source. These data lack information on race and ethnicity because the IRS does not collect it from tax filers. TPC therefore developed a method to impute race and ethnicity in its tax data. The method replicates the tax units in the model database, with one copy for each race and ethnicity included in the analysis. Using an algorithm, those copies are then weighted to match aggregate statistics calculated from survey data for variables related to tax liability. TPC’s initial work matches statistics calculated from the Current Population Survey and the Survey of Consumer Finances. TPC’s method has the advantage that it can incorporate data from multiple surveys or other sources, and can be readily adapted to include updated or additional data. Preliminary results using the enhanced model indicate that, across all income categories, itemized deductions benefit tax units classified as White more than those classified as Black or Hispanic. Within the same income categories, tax units classified as Hispanic benefit less than those classified as Black or White.