Research
I study the economics of data and its implications for platform strategy and regulation.
Job Market Paper
Opening the Black Box: A Statistical Theory of the Value of Data
Abstract
This paper develops a theory of the value of data for prediction. A platform predicts a target variable for a target individual. To improve its prediction, the platform can collect covariates on that individual and training observations from similar individuals, where each observation contains the target variable and a possibly different set of covariates. The model delivers two main results. First, covariates exhibit economies of scope: one covariate becomes more valuable when the platform observes others. Second, covariates and observations are complements in small datasets but can become substitutes in large datasets. I then apply the model to competition for covariates between two ecosystems, showing that economies of scope can lead to asymmetries in data accumulation and cause tipping.
Working Papers
The Great Moderation and the Rise of Markups: Demand Volatility and Tacit Collusion
Abstract
Business cycle stability has traditionally been considered beneficial for the economy. This paper identifies a cost: stable demand makes tacit collusion easier to sustain. In a repeated-oligopoly model, incentives to undercut rivals are greatest when demand peaks, so a decline in volatility increases the maximum sustainable collusive markup---most strongly in concentrated markets, with an elasticity equal to the inverse demand elasticity. The paper tests these predictions on a US state--sector panel covering the Great Moderation---the sustained decline in US macroeconomic volatility that began in the early 1980s---using a shift--share instrument for volatility based on the staggered deregulation of interstate banking. A 1% decline in volatility raises markups by 0.2%; the effect is absent in the least concentrated state--sectors and grows stronger and turns significant in the most concentrated quartile. Tacit collusion is consistent with a monopoly markup of 1.317, close to observed markups. The Great Moderation's decline in volatility accounts for up to two-thirds of the rise in markups between 1980 and 1997. More generally, the paper highlights a trade-off between macroeconomic stabilization and competition: merger guidelines should become stricter when demand stabilizes.