Research

Abstract: Does social security provide insurance for individuals facing uninsured idiosyncratic income risk? This paper argues that it does not in constrast with the exiting view. Using a stylized life cycle model with idiosyncratic income risk, I show that social security reduces post-retirement consumption dispersion, providing ex ante insurance. Yet, during working years, it can increase consumption dispersion, resulting in a negative insurance effect. This is attributed to the regressive nature of the payroll tax and  its distortion on the individual’s consumption and saving across income states, leading to an increased consumption dispersion. Quantifying the overall insurance effect in a quantitative overlapping generations model with idiosyncratic income risk and an incomplete market reveals that the negative insurance effect outweighs the positive one, indicating an overall negative insurance role of social security. Moreover, I find that the regressive payroll tax primarily drives the negative insurance effect. If the payroll tax were not regressive, it would provide significantly positive insurance, aligning with findings in the existing literature.

Abstract: This paper studies the implications of entrepreneurship and financing frictions on top marginal income tax rates. Raising the top marginal tax rate has a negative long-term effect on business income at the top through wealth distribution change. This paper argues that this negative long-term effect is potentially important in driving down the optimal top marginal tax rate. The baseline results suggest that the recent top marginal tax rates in the US are close to optimal for entrepreneurs.

Abstract: In this paper, we study the optimal Ramsey taxation of capital and labor income as the wealth share of entrepreneurs increases. We first document a rising trend of entrepreneurs’ wealth share in the United States using the Survey of Consumer Finance (SCF) data, then investigate how taxes for capital and labor should be adjusted. To achieve this, we build a tractable endogenous growth model that incorporates uninsurable idiosyncratic earning risks for both workers and entrepreneurs. The aggregate state of the model economy is well represented by the wealth share of entrepreneurs (A). Our result shows that (i) the optimal Ramsey tax for capital is positive and exhibits a U-shape over the range of A due to the interplay between idiosyncratic earning risks for workers and entrepreneurs, and (ii) roughly speaking, the sum of labor and capital taxes is determined by the government budget constraint, while the difference between capital and labor taxes is to yield sufficient human capital that achieves the “Ramsey allocation”. In the baseline calibration, the impact of an increase in A on the optimal capital and labor taxes appears to be marginal currently. However, if entrepreneurs’ wealth share continues to increase, policymakers become unable to ignore its effects.

Abstract: The unfunded social security has long been criticized for reducing capital stock and social welfare. In this paper, we study the aggregate and welfare effects of reforms that make social security tax (SST) age dependent. We characterize the optimal age-dependent SST schedule that maximizes social welfare and find that it increases capital and social welfare significantly. In addition, we propose a simple two-bracket SST schedule that captures the most of welfare gains from a fully age-dependent SST schedule. The welfare gains from the optimal schedule mainly comes from the increase in consumption level and better consumption smoothing over the life cycle, while the insurance effect of the age-dependent schedule is small. 

Abstract: Models with entrepreneurship can reproduce high wealth concentration at the top. The key assumption is the borrowing constraint, that is, households are unable to borrow enough assets to start a business or to invest at optimal level. More recent evidence, however, shows that borrowing constraints do not matter for the majority of households in the US. This paper seeks to generate high wealth concentration without assuming borrowing constraints. The baseline model that introduces heterogeneous abilities of entrepreneurs is able to match the wealth distribution while the model assuming same entrepreneurial ability fails. Besides wealth distribution, the baseline model generates other moments that are consistent with data.