London School of Economics EOPP: Economic Organisation and Public Policy Programme LSE
EOPP: Economic Organisation and Public Policy Programme

Current EOPP Research:
The Optimal Income Taxation of Couples

Henrik Kleven (with Claus Thustrup Kreiner, University of Copenhagen and Emmanuel Saez, UC Berkeley)

Abstract:

This paper explores the optimal income tax treatment of couples. Each couple is modelled as a unitary agent supplying labor along two dimensions: the labor supply of a primary earner and the labor supply of a secondary earner. The primary earner makes a continuous labor supply decision as in the Mirrlees (1971) optimal income tax model, whereas the secondary earner makes a binary labor supply decision (work or not work). We impose no a priori restrictions on the income tax system, allowing it to depend on the earnings of each spouse in any nonlinear fashion. This creates a multi-dimensional screening problem. Under our modelling assumptions, if, conditional on primary earnings, two earner couples are better off than one earner couples (for example, if second earners face different market opportunities), optimal tax schemes display positive tax rates on secondary earnings along with negative jointness whereby the tax rate on one person decreases with the earnings of the spouse. Conversely, if second-earner participation is a signal of the couple being worse off (for example, if second earners differ in their home production abilities), we obtain a negative tax rate on the secondary earner along with positive jointness: the second-earner subsidy is being phased out with primary earnings. The above results imply that, in either case, the distortion on the secondary earner is declining with primary earnings, which is therefore the general property of optimal tax schedules. We also prove that the second-earner distortion tends to zero asymptotically as the earnings of the primary earner becomes large. Although this result may seem reminiscent to the classic no-distortion-at-the-top result, our result rests on a completely different reasoning and proof.

Related Paper

The Optimal Income Taxation of Couples