Working Papers

AI and the Extended Workday: Productivity, Contracting Efficiency, and Distribution of Rents (with Wei Jiang, Junyoung Park, and Rachel Xiao)

Abstract: This study investigates how occupational AI exposure impacts employment at the intensive margin, i.e., the length of workdays and the allocation of time between work and leisure. Drawing on individual-level time diary data from 2004–2023, we find that higher AI exposure—whether stemming from the ChatGPT shock or broader AI evolution—is associated with longer work hours and reduced leisure time, primarily due to AI complementing human labor rather than replacing it. This effect is particularly pronounced in contexts where AI significantly enhances marginal productivity and monitoring efficiency. It is further amplified in competitive labor and product markets, where workers have limited bargaining power to retain the benefits of productivity gains, which are often captured by consumers or firms instead. The findings question the expectation that technological advancements alleviate human labor burdens, revealing instead a paradox where such progresses compromise work-life balance.

Venture Capital Investment and Employee Effort: Evidence from Open-Source Projects (with Rachel Xiao)

Abstract: we study the effect of venture capital investment in incentivizing effort at work. We find that after peers receive venture capital investment, computer programmers reduce their participation in open-source projects unrelated to their paid jobs, indicating that they put more effort into work. We rule out alternative explanations and find that the effect is more pronounced when the programmer is in closer proximity with her peers, and when her personal projects have low synergy with those owned by her employers. The affected programmers are less likely to leave their current employers, and the startups experience improved funding and exit opportunities.

Up or Out: Markup, Growth and Startup Valuation (with Hongyu Shan and John Lynch)

Abstract: This study develops a dynamic equilibrium model where entrepreneurs trade off between markup and growth. The model indicates that startups optimally lower markup to increase the growth of customer base which, in equilibrium, is associated with higher valuations. This predicted outcome is more pronounced in contested markets. We construct two empirical tests to elaborate on the model implications. Using the daily number of Twitter followers as a proxy of the startup’s customer base, we show that a higher level of customer growth is associated with higher valuation, and this relationship is stronger in a fast-evolving market. In a selected sample where we observe the pricing of products, we show that startups strategically lower prices, leading to a higher likelihood and greater amount of CV financing in the subsequent six months.

Work-in-Progress

Employee Network and Venture Capital Financing (with Mo Shen and Jack (Jie) He)

Publications

TAXI! Do Mutual Funds Pursue and Exploit Information on Local Companies? (with David Cicero, Andy Puckett, Albert Wang), Journal of Financial and Quantitative Analysis 59 (2024), 3340 - 3375.

The Role of Market Sentiment in Asset Allocations and Stock Returns (with Jitka Hilliard and Arun Narayanasamy), Journal of Behavioral Finance 21 (2019), 423-441.

Opportunity Zones: Do Tax Benefits Go to the Most Distressed Communities (with James Barth and Yanfei Sun)? Journal of Financial Economic Policy 13 (2020), 301-316.

Gramm-Leach-Bliley Act: Creating a New Bank for a New Millenium (with James Barth), Encyclopedia of Finance (3rd Edition), 529-540 (invited contribution).

Natural Disaster Impacts on U.S. Banks (with James Barth, Stephen Miller, and Yanfei Sun), American Business Review 25 (2020), 452-487