I am an economist trained in empirical industrial organization. I study how data and AI technologies generate useful information in imperfect markets, and the mechanisms through which they create and distribute economic value.
I focus on applications in insurance and digital platforms as well as the design of pricing, incentives, disclosure, and matching mechanisms. I specialize in field experiments, structural modeling, and independent collaboration with firms.
My research brings a design and entrepreneurial approach to empirical IO. I consider the tools in empirical IO to be a bridge between economic theory and business and policy applications, catalyzed by the proliferation of data and a growing appreciation for market imperfections.
How to Prevent Traffic Accidents? Moral Hazard, Inattention, and Behavioral Data
Road accidents killed 1.4 mil people in 2019. Can novel behavioral data and better auto insurance design curb risky driving? Can we do so without increasing the punishment of accidents? Focusing on phone use while driving, we combine observational and experimental evidence to show that risky behaviors are insensitive to changes in risk exposure, as inattention precludes moral hazard. But a simple behavioral nudge with mild added incentives led to large and persistent behavioral improvement. Our structural model shows that risky phone use behaviors are inelastic to incentives, but behavioral nudge can catalyze its effect by substantially raising perceived incentives. In order to recover the private and social optima, a combined approach is needed based on counterfactual simulations..
(first author | w/ Thomas Yu)
Firms in many markets directly elicit large amounts of data from consumers. The data is used to mitigate information problems, to gain competitive advantage, and to extract rent from consumers. We develop an equilibrium framework to trade-off these countervailing forces, and use it to study the creation of detailed driving data in U.S. auto insurance through a voluntary monitoring program. Surprisingly, the incentive effect from the data collection process, not the data itself, is the main source of social value creation. Despite large switching inertia, the product market is still competitive enough so that forcing the firm to make the data public hurts short-term consumer welfare by discouraging data creation.
(first author | w/ Shosh Vasserman)
Digital Training and the Online Growth Barrier: Experimental Evidence with Two Million E-Commerce Businesses
We conduct a large experiment where we randomize access to a digital training program among e-commerce entrants. We document high growth barriers in the competition for consumer attention, and that the training program boosts entrant traffic by closing the knowledge gap on practical operational skills. But the training data are far more valuable as it can identify more high-quality entrants. We develop a consideration-set based equilibrium model to capture the welfare benefit of reallocating consumer attention to these higher-quality entrants. Counterfactual analyses show that, at social optima, the platform should expand digital training and value entrant traffic significantly more when ranking firms and assigning traffic.
(Zhengyun Sun's job market paper)
What kind of firms collect what types of data? Can such data facilitate growth? If so, by influencing what strategies? We analyze the adoption and the effect of analytics tools by firms on an e-commerce platform. We then conduct a high-stake experiment among non-adopting stores and found evidence of information friction despite low take-up.