Two Essays on Online Platforms

Department of Decision Sciences and Managerial Economics


The prevalence of online platforms, from Uber Eats to Amazon, and from Coursera to Wikipedia, play a vital role in our daily behavior and lifestyle change. The platforms sometimes have spillover effects on other markets (e.g., labor market), which may realize important social values. In this thesis, we present two essays regarding online platforms.

In the first essay, we explore whether personalized information is always necessary for recommendation systems. We discuss whether the neighborhood-based collaborative filtering (CF) recommendation based on personalized information input indeed always performs better than the recommendations generated by other algorithms (e.g., globally-popular recommendation algorithm). We conduct a field experiment on an established retailing platform to compare the influence of globally-popular recommendations and neighborhood-based CF recommendations on consumer behaviors in January 2021. The results are surprising that there is no difference in recommendation-to-purchase conversion between the two algorithms. We then classify the users, and decrypt the effect of various recommendation systems on various consumers’ psychology trade-offs when making decisions in different shopping stages. Our findings suggest that personalized information may not always improve the performance of recommendation system. Platform designers should carefully develop the algorithm considering the distribution of different types of users on their platforms.

In the second essay, we investigate the spillover effects of online food delivery platforms on female labor force participation. Female labor force participation is often explained by factors such as schooling, wage gap, fertility, etc. We identify how technology-induced time savings from household chores led to increased female labor force participation in South Korea. Using a leads-and-lags difference-in-differences model, we find that the entry of an online food delivery platform significantly increased the female employment rate in the next three years. We estimate the positive externality generated by the online food delivery platform: this new technology-induced female employment accounts for 0.27% of South Korea’s GDP, or 17 times the revenue of the online food delivery platform.