This paper analyzes the relationship between marriage and taxes by examining federal income tax returns filed by cohabiting couples.  The federal income tax system is not neutral to marriage because the rate structure and the other features of the tax system are based on filing status and levels of income.  A couple can owe more or less income tax when they are married and file a joint return than if they are unmarried and file separate returns.  As a consequence, tax provisions aimed at providing marriage penalty relief will affect horizontal equity between cohabiting couples and otherwise equivalent married individuals because such policies change the relative tax liability faced by the two groups.  

In this study, we use IRS tax return data to identify couples who are cohabiting in the year prior to filing a joint return.  For these couples, we estimate the tax consequence of marriage by comparing the income tax from their separate returns to the tax they would owe if they were married and filed jointly.  This paper improves upon previous studies in several ways.  First, we use tax return data that not only provides all the variables needed to estimate tax liability, but also identifies future spouses.  This circumvents a common limitation of the existing marriage tax literature in which various sources of income and expenses as well as potential spouse’s earning capacity have to be imputed to estimate tax liability.  Because couples in our study already shared a residence, the observed incomes and expenses should more accurately depict the arrangements of married couples than the imputed data.  In addition, unlike studies that use survey data, we do not need to make assumptions about the filing behavior of unmarried couples since we observe each person’s filing status and we know who claims which dependents.

Preliminary results show that, under 2007 law, 48 percent of cohabiting couples in our sample would face a tax increase if they were married and filed jointly.  Furthermore, 37 percent of the couples would face a tax decrease and 14 percent would have no substantive change in tax liability.  We explain the results by analyzing the couples’ incomes, the presence of children, and the tax structure and designs of child-related tax credits.  Finally, we conduct policy simulation to provide insight on how marriage penalty relief policies and recent changes to child-related tax benefits affect marriage taxes for cohabiting couples.