Association for Public Policy Analysis & Management

The 2013 Supreme Court ruling in U.S. v. Windsor forced the federal government to recognize same-sex marriages that some states had granted since 2004. Same-sex couples who had already married by 2013 had previously filed federal tax returns as two single individuals, but beginning in 2014 they were required to file as married regardless of their state of residence. I exploit this novel source of variation in federal tax rates and household income induced by U.S. v. Windsor among same-sex married couples to, first, ask whether joint taxation affects hours worked and labor force participation, and, second, to separate the income and substitution effects of taxation within a collective model of labor supply.

I use the 2012-2015 waves of the American Community Survey, which is the most recently available data source with sufficient observations of same-sex couples, and the first of the Census Bureau surveys to explicitly identify same-sex married couples. In addition, I extend prior research by predicting which couple member is most likely to be the higher earner, rather than relying on the assumption that the husband is the higher earner in opposite-sex married couples.

I find that predicted lower (secondary) earners generally respond to taxation along the extensive margin, while predicted higher (primary) earners generally respond along the intensive margin. I estimate a Hicksian elasticity of 0.382 among predicted higher earners, which is largely offset by an income elasticity of -0.391, resulting in a small Marshallian hours elasticity among predicted higher earners. These results are comparable to other estimates in the literature and are robust to a number of alternative specifications and sample selection criteria. (JEL: H24, J22, D10)