Until the IRS do us part: (Optimal) Taxation of households

In this paper we argue that a progressive tax system combined with individual
taxation of married couples can generate more revenue than the current household-based U.S. system, especially when the extra revenues do not induce negative labor supply effects through increased government transfers. A progressive system that taxes individuals rather than couples jointly leads to higher labor force participation and a higher level of human capital in the economy. Thus it creates more “fiscal space”, Laffer curves are higher and
welfare is potentially higher. In our model, which features one- and two-earner households, human capital and a decision about whether or not to participate in the labor market, we find that the peak of the Laffer curve is about 12 percentage points higher with an individual-based, progressive tax system than with the current U.S. tax system. The maximum revenue is attained with a system that is about 60% more progressive than the current system, and
at an average tax rate of 41%. Progressive taxation, when imposed on individuals rather than households, lowers the average tax rate for individuals with low income that are close to the participation margin. At the same time it creates a positive income effect on the labor supply of these individuals by reducing the net income of their higher earning spouses and limiting their net earnings potential in the case of a high temporary labor productivity.
Whereas steady state welfare under a joint tax system is maximized when the tax system features no progressivity at all, under individual-based taxation the optimal tax system features significant tax progressivity (albeit slightly less than the current U.S. status quo), and welfare gains relative to this status quo of 4.9% in consumption-equivalent variation.
Cohorts born during the transition realize significant welfare gains from this reform.