Working Papers

Modern recommender systems, which rely on large-scale machine learning (ML) models to predict the next item a consumer will engage with, often lack generalizability and understanding of consumer behaviors due to their black-box nature. To address…

In this overview piece, I review the literature and identify important directions for how developing societies can break out of such ethnic growth traps and instead leverage the gains that can often be had from ethnic diversity. To do this…

What can we learn from the high-frequency responses in bond and currency markets to the recent tariff announcement about the status of the U.S. dollar as the global reserve currency? The dollar depreciated by 3.4% after April 4 in spite of rising…

Geoeconomic pressure — the use of existing economic relationships by governments to achieve geopolitical or economic ends — has become a prominent feature of global power dynamics. This paper introduces a methodology using large language models (…

We develop a model of disclosure timing in which the firm can commit to disclose its revenues at a fixed date (a time-based disclosure policy) or after a certain number of transactions have occurred (a news arrival-based disclosure policy). We…

Modern recommender systems use machine learning (ML) models to predict consumer preferences based on consumption history. Although these “black-box” models achieve impressive predictive performance, they often suffer from a lack of transparency…

For several decades, researchers and practitioners have wondered whether a
“hot hand” exists in domains with repeated, human-controlled trials. Using a comprehensive play-by-play dataset from the game show “Jeopardy!”, we demonstrate that…

Deferred prosecution agreements (DPAs) are now a standard tool used by prosecutors to punish corporate crime. Under a DPA, the defendant escapes prosecution by living up to the terms of the contract. However, if the prosecutor declares a breach,…

In this paper, we argue that the New York City congestion pricing plan whose implementation was paused in the summer of 2024 had a major shortcoming: as designed, it would have had a much more severe impact on the drivers of personal vehicles…

Workers who lose their jobs during recessions face strikingly large and persistent declines in their future earnings. Using individual-level administrative data from the United States, this paper shows that an important driver of these costs is…

We consider a retailer running a switchback experiment for the price of a single product, with infinite supply. In each period, the seller chooses a price from a set of predefined prices that consist of a reference price and a few discounted…

We analyze the effects of veto players when the set of available policies isn’t exogenously fixed, but rather determined by policy developers who work to craft new high-quality proposals. If veto players are moderate, there is active competition…

This paper investigates decision-making in A/B experiments for online platforms and marketplaces. In such settings, due to constraints on inventory, A/B experiments typically lead to biased estimators because of interference; this phenomenon has…

description description

In recent years, U.S. public companies have increasingly begun to voluntarily disclose official workforce diversity data (i.e., EEO-1 reports), which they previously only confidentially filed with the U.S. Equal Employment Opportunity Commission…

Algorithms frequently assist, rather than replace, human decision-makers. However, the design and analysis of algorithms often focus on predicting outcomes and do not explicitly model their effect on human decisions. This discrepancy between the…

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…

We examine the labor market consequences of the 2020 Regulation S-K requiring human capital disclosure in 10K filings. Using large-sample job-level data, we observe that public firms subject to the regulation increase their disclosure of…

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…

We consider the problem of using observational bandit feedback data from multiple heterogeneous data sources to learn a personalized decision policy that robustly generalizes across diverse target settings. To achieve this, we propose a minimax…