Identifying Treatment Effects on Productivity: Theory with An Application to Firms' AI Adoption
报告人： Moyu Liao（Nanjing University）
时间：2023-05-11 15:10-16:00
地点：Room 217, Guanghua Building 2
Abstract:
This paper provides a semi-structural econometric framework for analyzing the treatment effects on productivity. We embed the potential outcome framework into a dynamic firm model with firm-level productivity that can be affected by some discrete treatment. The treatment can either be externally assigned or chosen by the firm. We characterize conditions for non-parametrically identifying production functions. Then we provide conditions for identifying treatment effects objects for absorbing treatment and generalize them to non-absorbing treatment. Our study has general implications for empirical studies on evaluating treatment effects on productivity. We apply our method to study the effect of AI adoption on productivity growth. Our results robustly show that adoption of AI product has no significant effect on the firm-level productivity in China.
About the Speaker:
Moyu Liao is an assistant professor at Nanjing University. He received his Ph.D. from the Pennsylvania State University and a Bachelor of Economics from HKU. His primary research focuses on the intersection of econometrics and industrial organization. In particular, he is interested in identifying and estimating firm-level productivity, consumer demand, and treatment effect.
Your participation is warmly welcomed!
欢迎扫码关注北大统计科学中心公众号，了解更多讲座信息!
