Economics · Yonsei University

Wonjin Seo

I am an M.A. student in Economics at Yonsei University. I study how firms respond to competition, information, and technological change. My research spans industrial organization, international trade, and econometrics.

Wonjin Seo

Research

Working papers

How Does Trade Competition with a Technological Leader Reshape Domestic Innovation? Evidence from South Korea

Sole-authored Draft

Abstract

I study how import competition from a technological leader reshapes innovation in a catching-up economy. South Korea’s abolition of restrictions on Japanese imports generated a Japan-specific import surge and reduced patenting in exposed technologies, especially where Japan was initially further ahead. Broadly diversified incumbents reallocated patenting away from exposed technologies, whereas narrower incumbents experienced substantial overall contraction. Broad firms redirected patenting toward technologically related fields with greater technological opportunities. Import competition can therefore reshape innovation through selective technological retreat, discouraging research in affected fields while redirecting inventive activity toward alternative technological opportunities.

Premium or Penalty? Consumer Inference from the Sponsored Label

Sole-authored Draft

Abstract

The same sponsored label can make a listing less attractive to consumers in some markets and more attractive in others. I develop a model in which sponsored placement serves both to signal quality and to provide visibility to sellers passed over by the organic algorithm. When both forces operate, consumers favor organic listings under a precise algorithm but sponsored listings under a noisy one. Algorithm noise also shapes the quality distribution of listings shown to consumers: greater noise improves the sponsored pool but shifts the overall quality distribution downward in the sense of first-order stochastic dominance. When choosing algorithm noise, the platform has an incentive to monetize the sponsored label’s informational value. This incentive has policy consequences: mandatory sponsorship disclosure can backfire by inducing greater algorithm noise.

Work in progress

Identification and Estimation of Discrete Choice Models with a Fixed Latent Information Structure

with Paul S. Koh

Summary

Studies identification and estimation in single-agent and multi-agent discrete choice models when the information structure is unknown but fixed across markets.

The Production of Artificial Intelligence: Evidence from Korean AI Firms

with Paul S. Koh and Devesh Raval

Summary

Uses linked firm-level microdata from the Korean Artificial Intelligence Industry Survey to study how labor, R&D, knowledge stock, computing resources, and cloud use shape AI production and productivity.