Working Papers

The allocation of decision authority by a principal to either a human agent or an artificial intelligence (AI) is examined. The principal trades off an AI’s more aligned choice with the need to motivate the human agent to expend effort in…

We develop an equilibrium theory of business cycles driven by spikes in risk premiums that depress business demand for capital and labor. Aggregate shocks increase firms’ uninsurable idiosyncratic risk and raise risk premiums. We show that risk…

For many machine learning algorithms, two main assumptions are required to guarantee performance. One is that the test data are drawn from the same distribution as the training data, and the other is that the model is correctly specified. In real…

This paper provides a positive identification result for first-price procurement models with asymmetric bidders, statistically dependent private signals,

and interdependent costs. When bidders are risk neutral, the model’s payoff-relevant…

A transplant can improve a patient’s life while saving several hundreds of thousands of dollars in healthcare expenditures. Organs from deceased donors, like many other scarce public resources (e.g. public housing, child-care, publicly funded…

We develop a valuation model for venture capital-backed companies and apply it to 135 US unicorns, that is, private companies with reported valuations above $1 billion. We value unicorns using financial terms from legal filings and find that…

Eliminating firms’ access to tax havens can have unintended consequences for their domestic economic activity. We study a policy that limited profit shifting by U.S. multinationals and show it raised the tax cost of domestic investment. Firms…

This paper studies families’ preferences for peers in the school and the implications of those preferences for the distribution of academic outcomes. I develop an equilibrium model of school competition and student sorting under social…

A common challenge in estimating the long-term impacts of treatments (e.g., job training programs) is that the outcomes of interest (e.g., lifetime earnings) are observed with a long delay. We address this problem by combining several short-term…

We study repeated games in which players learn the unknown state of the world by observing a sequence of noisy private signals. We find that for generic signal distributions, the folk theorem obtains using ex-post equilibria. In our…

We document three new facts about entrepreneurship. First, a majority of male entrepreneurs start a firm in the same or a closely related industry as their fathers’ industry of employment. Second, this tendency is correlated with intelligence:…

We develop a Loan Portfolio Risk (LPR) variable that measures time-varying volatility in default risk for a portfolio of bank loans.  An Equity-to-LPR ratio (ELPR) is incrementally important in predicting bank…

Data is nonrival: a person’s location history, medical records, and driving data can be used by any number of firms simultaneously. Nonrivalry leads to increasing returns and implies an important role for market structure and property rights. Who…

Many learning algorithms require categorical data to be transformed into real vectors before it can be used as input. Often, categorical variables are encoded as one-hot (or dummy) vectors. However, this mode of representation can be…

This paper studies the potential small and large scale effects of a policy designed to produce more informed consumers in the market for primary education. We develop and test a personalized information provision intervention that targets…

We examine Kreps’ (2019) conjecture that optimal expected utility in the classic Black–Scholes–Merton (BSM) economy is the limit of optimal expected utility for a sequence of discrete-time economies that “approach” the BSM economy in a…

We prove a fundamental result concerning the connection between discrete-time models of financial markets and the celebrated Black–Scholes–Merton continuous-time model in which “markets are complete.” Specifically, we prove that if (a) the…

This paper proposes a method for estimating consumer preferences among discrete choices, where the consumer chooses at most one product in a category, but selects from multiple categories in parallel. The consumer’s utility is additive in the…

Random forests are a powerful method for non-parametric regression, but are limited in their ability to fit smooth signals, and can show poor predictive performance in the presence of strong, smooth effects. Taking the perspective of random…