GRIPS 政策研究センター Policy Research Center


2015/4/1 ~ 2017/3/31

Income Inequality and Optimal Income Tax


The literature of optimal taxation has debated about the degree of progressivity that the income tax should have. Recently, because of the heavy government financing burden, major European economies (eg. France and Spain) have planned to increase the tax rates for high income earners. France even planned to set a 75% tax rate for top income earners. However, previous studies in the optimal income tax literature found a striking result that the marginal tax rate for the top income earners should be zero due to the economic efficiency.

The equity-efficiency trade-off is conventionally the center of the optimal income taxation literature, which is pioneered by Mirrlees (1971, Review of Economic Studies). Progressivity improves the equity by redistribution from the rich to the poor, but high marginal tax rates discourage the labor supply from productive workers. Saez (2001, Review of Economic Studies) developed a simple approach by using elasticities to derive optimal tax rates and showed that the optimal high income tax is positive and generally above 40%. Our project is to extend Saez’s approach and generalize it to a dynamic stochastic general equilibrium environment with heterogeneous agents, in which the income/wealth distributions match the data.

Furthermore, with taking into account a country’s household income dynamics and level of inequality, the method can be applied to different countries and we can provide a cross-country comparison of optimal tax system design.

Because Saez illustrates the optimal design of income tax in a static world, the behavior changes of individuals are not taken into account. With a dynamic model, we are able to consider the responses from individuals to a tax schedule over time that is important for constructing the optimal tax schedule. Some recent studies have tried to extent the discussion of optimal taxation to a dynamic environment. Golosov,Troshkin and Tsyvinski (2011, NBER working paper; R&R to AER) is one example. They suggest that the marginal tax rate for high income earners should be close to zero, which largely deviates from Saez’s finding. In this project, we plan to model both labor and saving decisions in a dynamic framework to investigate behavior changes and tax distortions for the design of tax system.

In addition, the equity-efficiency trade-off is always the focus of the optimal taxation literature, but the real income/wealth inequality is rarely discussed for the design of the optimal tax schedule. We plan to carefully take into account the inequality in the real world through a comprehensive data analysis.