Research

Solving and Estimating Heterogeneous Agent Model Using Reinforcement and Machine Learning (Job Market Paper)
[paper]
In this paper, I introduce an innovative approach leveraging Deep Reinforcement Learning techniques to solve and estimate heterogeneous models. Contrasted with conventional solution methods, the deep learning approach offers a global solution while retaining the entirety of nonlinearity. Moreover, it exhibits remarkable scalability, making it suitable for handling models with hundreds or even thousands of state variables. I also explore the integration of this novel solution method with amortized likelihood-free Bayesian inference, opening up new possibilities for advanced probabilistic modeling and estimation.

Working from Home, Household Production, and Durable Good Consumption
[paper]
I decompose the durable good consumption increase during 2020-2021 using a household production model with working-from-home. Durable good consumption is susceptible to business cycles; in past recessions, durable good consumption either decreased or slowed down. However, durable good consumption was very robust during and after the COVID-19 pandemic. I build a household production model with working from home and estimate the model using a Bayesian approach. Using Kalman smoother, we can then decompose the increase in durable good consumption into different channels. Working from home can account for up to one-third of the durable good consumption increase, and substitution between nondurable and durable can account for another one-third of the increase.

The effect of Trucking industry Deregulation on Truck Drivers’ Wage
[paper]
We assess the impact of intrastate trucking industry deregulation on the for-hire sector market. In the 1980s, several states removed regulation on intrastate truck transportation, which intensified competition, and had a tremendous impact on the market structure of the trucking industry and the labor market of truck drivers. We use the difference in difference specification to identify the effect of deregulation of intrastate trucking industry on the trucking market and for-hire sector drivers wage. We found that the deregulation reduced truck drivers wage by 7.64% and drivers whose wage is near the median of the wage distribution are affected the most severely; their wage was reduced by 12% - 15%. We also found that about 66.54% for-hire sector drivers either switched to other occupations or lost their job, and 69% of companies were forced out of business.