Discussion

Greg Kaplan opened the general discussion with a remark about asset return heterogeneity. In the authors’ model, a sufficient degree of dispersion in asset returns guarantees a well-defined equilibriumwealth distribution. Removing asset return heterogeneity would not imply a counterfactually low degree of wealth inequality but rather a degenerate wealth distribution with a mass point at infinity, he said. The authors clarified that their model is able to generate a Pareto tail in the wealth distribution by adding capital return heterogeneity to a standard S. R. Aiyagari (“Uninsured Idiosyncratic Risk and Aggregate Saving,” Quarterly Journal of Economics 109, no. 3 [1994]: 659–84) model with labor income inequality. In their calibration, the mean asset return is below the discount rate, but introducing sufficient dispersion in asset returns allows the model to generate the Pareto tail in the wealth distribution observed in the data, they argued. The rest of the discussion focused on a single topic: the importance of modeling the tax system and measuring the effective tax rate correctly. Frederic Mishkin noted that the effective tax rate can be influenced by tax avoidance. Actual tax collection from highmarginal-income tax brackets in the 1960s and 1970s, for example, was likely impaired due to widespread avoidance, he argued. The authors agreed that accounting for tax avoidance is crucial. They emphasized that their paper employs estimates of the effective tax rate from T. Piketty and E. Saez (“How Progressive Is the US Federal Tax System?AHistorical and International Perspective,” Journal of Economic Perspectives 21, no. 1 [2007]: 3–24), which takes care of this issue. James Poterba offered two comments related to the measurement of the effective tax rate on capital income. Such estimates are typically low, Poterba argued, because of two features of the US tax system: the deferral