The Code contains numerous special rules applicable to the income taxation of persons related by marriage, birth, or adoption. The collective importance of these related-party provisions to the income tax system is hard to overstate. This paper suggests a new approach to their analysis. Its thesis is simple: the rules of the Code depend, in significant part, on the assumption that taxpayers are self-interested. In general, we are unwilling to transfer property or rights to income to others simply to avoid tax. Where this assumption proves or is likely to prove incorrect, the Code makes adjustments to its otherwise applicable rules. Sometimes such adjustments facilitate transfers between interpersonally committed individuals. More commonly, they shut down avoidance techniques. Some of these latter adjustments apply whenever the assumption of selfishness fails, regardless of the formal relationship between the parties. Most, however, apply only in the context of a specified formal relationship - marriage, parent/child, or owner/business.
The paper tests the utility of this thesis by comparing the income tax treatment of heterosexual married couples with that of gay couples in committed long-term relationships - which I refer to as married and spouses solely for ease of reference. It is clear that gay couples are not married for tax purposes, nor are they spouses within the meaning of the Code. Gay marriage by itself never invokes any related-party rules - taxpayer-favorable or anti-abusive. Part I explores problems created by the fact that spouses, whether heterosexual or gay, commonly share income, expenses, assets, and liabilities. Part II explores problems created by less complete failures of the assumption of selfishness in the broader intra-family context - parents, children, siblings, and more remote relations. None of the anti-abuse rules enacted to deal with either set of problems applies to gay spouses. As a result, gay couples should be able to arrange their affairs so as to pay federal income tax at significantly lower effective rates, on average, than identically situated heterosexual married couples.
Part III addresses the question of solutions. Ultimately, the paper concludes that the only way to ensure that gay couples will be taxed no more favorably than heterosexual married couples is to list gay marriage as one of the proxy relationships that automatically invokes pertinent anti-abuse rules - in other words, to treat gay marriage as marriage for federal income tax purposes. Even this by itself, however, will not be enough. A formal spousal relationship must be made available to gays, and it must be attractive enough to induce gay couples to undertake it voluntarily - just as heterosexual couples marry despite the tax costs. In the absence of an attractive formal status that then invokes related-party anti-abuse rules, well-advised gay couples are, and will continue to be, permitted to pay systematically lower federal income taxes than heterosexual married couples - a result unlikely to be acceptable to a majority of Americans in the long run.