Publications

Innovations bring many benefits to society, but they can also bring harm. We study the problem of a regulator deciding whether to approve an innovation where information about the impact of the innovation is held within the firms that are…

In this Closer Look, we take a deep dive into a sample of companies that failed their say-on-pay votes and the changes they implemented the subsequent year. We find many failed votes are in response to large, one-time special awards.…

In summer 2024, Stanford Graduate School of Business, the Hoover Institution Working Group on Corporate Governance, and the Arthur and Toni Rembe Rock Center for Corporate Governance at Stanford University hired Prolific to conduct a nationwide…

In this paper, I share a few insights on our remarkable field to explore what creates a beautiful life. I conceptualize that the idea of a “beautiful life” is defined not by transient moments of happiness, but by a profound sense of meaning and…

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 develop a model of the global financial cycle with one key ingredient: the international demand for safe dollar assets. The model matches patterns of dollar borrowing and currency mismatch, the U.S. external balance sheet, exorbitant privilege…

We provide the first large-sample evidence on third-party verification of firms’ environmental and social metrics in ESG reports (ESG assurance) in the United States. Focusing on the S&P 500 from 2010–2020, we document a striking increase in…

We combine newly collected election data with records of public denials of the results of the 2020 election to estimate the degree to which election-denying Republican candidates for senator, governor, secretary of state, and attorney general…

Eleven large language models (LLMs) were assessed using 40 bespoke false-belief tasks, considered a gold standard in testing theory of mind (ToM) in humans. Each task included a false-belief scenario, three closely matched true-belief control…

There are a number of available methods for selecting whom to prioritize for treatment, including ones based on treatment effect estimation, risk scoring, and hand-crafted rules. We propose rank-weighted average treatment effect (RATE) metrics as…

Cognitive differences can catalyze social learning through the process of one-to-one social influence. Yet the learning benefits of exposure to the ideas of cognitively dissimilar others often fail to materialize. Why do cognitive differences…

Carefully standardized facial images of 591 participants were taken in the laboratory while controlling for self-presentation, facial expression, head orientation, and image properties. They were presented to human raters and a facial recognition…

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 develop a new approach for estimating average treatment effects in observational studies with unobserved group-level heterogeneity. We consider a general model with group-level unconfoundedness and provide conditions under which aggregate…

In recent decades, economic activity has become increasingly concentrated in major global metropolises. Yet, the rise of working from home threatens this dominance of cities. Using multiple high-frequency datasets on spending, commuting,…

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…

ChatGPT-4 and 600 human raters evaluated 226 public figures’ personalities using the Ten-Item Personality Inventory. The correlation between ChatGPT-4 and aggregate human ratings ranged from r = 0.76 to 0.87, outperforming the models…

We study the identification and estimation of long-term treatment effects by combining short-term experimental data and long-term observational data subject to unobserved confounding. This problem arises often when concerned with long-term…

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…