Hangjun He is a Graduate Research Fellow at the Penn Wharton Budget Model, where his work applies mathematical, computational, and machine learning methods to large-scale macroeconomic and financial models. He works on general equilibrium modeling of the U.S. economy for fiscal policy analysis, and he uses GPU-accelerated parallel computing to make these large models tractable. His dissertation made theoretical contributions to both public finance theory and computational methods. His research includes training neural networks with reinforcement learning to design efficient tax systems and examining how climate and entrepreneurial risks shape macroeconomic outcomes, including the equity premium puzzle. He received his Ph.D. and M.A. in Applied Mathematics and Computational Science from the University of Pennsylvania, where his dissertation was advised by Kent Smetters, and his B.S. in Mathematics and Applied Mathematics from Zhejiang University of Technology.
We estimate that the United States federal debt cannot rationally exceed roughly 210 percent of GDP as an outer limit. Under historical excess cost growth in healthcare, this outer limit is likely reached within 20 years; there is a 25% chance of reaching it in 14 years. Debt markets unravel earlier if beliefs about government repayment shift.
