During my full-time research assistantship at the Department of Economics, National Taiwan University (July 2023 – February 2025), I had the opportunity to work extensively with large-scale administrative datasets from Taiwan’s government agencies. This experience broadened my research interests into applied economics, where rigorous quantitative methods meet real-world policy questions.
Administrative data and empirical methods
A distinctive feature of this line of research is the use of comprehensive administrative records—including income tax returns from the Ministry of Finance and pet registration data from the Ministry of Agriculture—that cover the entire population of Taiwan. These data allow us to study economic behavior at a scale and level of detail that survey data alone cannot provide, while also demanding careful data linking, cleaning, and econometric analysis.
My research projects
Labor market sorting and wage inequality
With Prof. Kuan-Ming Chen and Lin-Tung Tsai, I analyze wage inequality in Taiwan using the Abowd-Kramarz-Margolis (AKM) model. This framework decomposes log wages into worker fixed effects (capturing individual productivity) and firm fixed effects (capturing employer-specific wage premiums). By examining how high-productivity workers sort into high-paying firms, we quantify the contribution of assortative matching to overall wage dispersion. This paper has been accepted for publication in the Taiwan Economic Review.
Education economics: college and major preferences
With Prof. Kuan-Ming Chen, Yu-Chang Chen, and Chi-Chao Hung, I study how high school students in Taiwan make decisions about which college and which major to attend. Using panel data from the college entrance examination system (240,000 students), we estimate a heterogeneous random utility model with more than 2,300 parameters via maximum likelihood. I developed an analytical-form Fisher scoring algorithm that reduced computation time from two months to two minutes. This paper is under revision at Economic Inquiry.
Intra-household cooperation in tax filing
Using income tax return data, we investigate whether household members strategically coordinate their tax filing decisions to minimize collective tax liability (Pareto-efficient behavior) or whether they default to individual optimization, potentially leaving money on the table. This project sheds light on the extent of intra-household communication and cooperation in financial decisions.
Pet ownership and fertility
Taiwan has experienced a sustained decline in fertility rates alongside a rapid increase in pet ownership. Are pets substitutes for children, or complements? By linking pet registration records from the Ministry of Agriculture with tax data from the Ministry of Finance, we analyze the nationwide dynamics of fertility and pet ownership over the past decade to disentangle these competing hypotheses, with implications for population policy.
Methodological connections
These applied projects are deeply connected to my core expertise in quantitative methods. The college preference project involves structural estimation of discrete choice models rooted in random utility theory—the same theoretical framework that underlies Item Response Theory in psychometrics. The labor market project requires handling high-dimensional fixed effects estimation. Throughout, the emphasis is on combining economic theory with scalable computational methods to extract insights from large datasets.