Working Papers

We consider the problem of learning personalized decision policies from observational bandit feedback data across multiple heterogeneous data sources. In our approach, we introduce a novel regret analysis that establishes finite-sample upper…

Qini curves have emerged as an attractive and popular approach for evaluating the benefit of data-driven targeting rules for treatment allocation. We propose a generalization of the Qini curve to multiple costly treatment arms that quantifies the…

One thread of empirical work in social science focuses on decomposing group differences in outcomes into unexplained components and components explained by observable factors. In this paper, we study gender wage decompositions, which require…

The rise of marketplaces for goods and services has led to changes in the mechanisms used to ensure high quality. We analyze this phenomenon in the Uber market, where the system of pre-screening that prevailed in the taxi industry has been…

Has the United States economy become less competitive in recent decades? One might think so based on a body of research that has rapidly become influential for antitrust policy. We explain that the empirical evidence relating to concentration…

Many empirical studies of labor market questions rely on estimating relatively simple predictive models using small, carefully constructed longitudinal survey datasets based on hand-engineered features. Large Language Models (LLMs), trained on…

We formulate the problem of constructing multiple simultaneously valid confidence intervals (CIs) as estimating a high probability upper bound on the maximum error for a class/set of estimate-estimand-error tuples, and refer to this as the error…

Advances in artificial intelligence (A.I.) are a double-edged sword. On the one hand, they may increase economic growth as A.I. augments our ability to innovate. On the other hand, many experts worry that these advances entail existential risk:…

When choosing whether and how much to donate, potential donors often observe a set of default donation amounts known as an “ask string.” In an experiment with more than 400,000 PayPal users, we replace a relatively unused donation amount ($75) on…

This study investigates the labor market value of credentials obtained from Massive Open Online Courses (MOOCs) and shared on business networking platforms. We conducted a randomized experiment involving more than 800,000 learners, primarily from…

We evaluate a contest-based intervention intended to increase the usage of an educational app that helps children in India learn to read English. The evaluation included approximately 10,000 children, of whom about half were randomly selected to…

We consider the role of parental influence on the industry choice of entrepreneurs and the success of their ventures. Almost 75% of male entrepreneurs start a firm in an industry that is the same or closely related to their fathers’ industry of…

Gross Domestic Product (GDP), though often used as a proxy for economic well-being, was not designed for this purpose and is inadequate. In our current research, we are building new metrics for measuring changes in consumer well-being that…

We conducted two experiments on PayPal’s Give at Checkout feature to learn about the effect of 1) information about charity outcomes on donations, and 2) exposure to these point-of-sale microgiving requests on subsequent giving. In this “…

Many proposed surface water transfers undergo a series of regulatory reviews designed to mitigate hydrological and economic externalities. While these reviews help limit externalities, they impose substantial transaction costs that also limit…

In many applications, e.g. in healthcare and e-commerce, the goal of a contextual bandit may be to learn an optimal treatment assignment policy at the end of the experiment. That is, to minimize simple regret. However, this objective remains…

In many settings, interventions may be more effective for some individuals than others, so that targeting interventions may be beneficial. We analyze the value of targeting in the context of a large-scale field experiment with over 53,000 college…

We use equity returns to construct a time-varying measure of the interest rate that we call the zero-beta rate: the expected return of a stock portfolio orthogonal to the stochastic discount factor. The zero-beta rate is high and volatile. In…

ICT is increasingly used to deliver customized information in developing countries. We examine whether individually targeting the timing of automated voice calls meaningfully increases engagement in an agricultural advisory service. We define,…

A large literature in labor economics seeks to decompose observed gender wage gaps (GWGs) into different sources, including portions explained by cross-gender differences in education, occupation, and experience. This paper provides new methods…