We investigate the potential for Large Language Models (LLMs) to enhance scientific practice within experimentation by identifying key areas, directions, and implications. First, we discuss how these models can improve experimental design, including improving the elicitation wording, coding...

The extant literature predicts market returns with "simple" models that use only a few parameters. Contrary to conventional wisdom, we theoretically prove that simple models severely understate return predictability compared to "complex" models in which the number of parameters exceeds the...

We theoretically characterize the behavior of machine learning asset pricing models. We prove that expected out-of-sample model performance--in terms of SDF Sharpe ratio and test asset pricing errors--is improving in model parameterization (or "complexity"). Our empirical findings verify the...

We introduce artificial intelligence pricing theory (AIPT). In contrast with the APT's foundational assumption of a low dimensional factor structure in returns, the AIPT conjectures that returns are driven by a large number of factors. We first verify this conjecture empirically and show that...

Whether police misconduct can be prevented depends partly on whether it can be predicted. We show police misconduct is partially predictable and that estimated misconduct risk is not simply an artifact of measurement error or a proxy for officer activity. We also show many officers at risk of...

What do machines learn, and why? To answer these questions we import models of human cognition into machine learning. We propose two ways of modeling machine learners based on this join: feasibility-based and cost-based machine learning. We evaluate and estimate our models using a deep learning...

Income is simultaneously one of the most important variables used by economists and the variable most likely to be missing due to item non-response. While observations that are missing income responses are often dropped from analyses, such treatment is usually inappropriate. More appropriate...

This paper proposes a new way of displaying and analyzing macroeconomic time series to form recession forecasts. The proposed data displays contain the last three years of each expansion. These allow observers to see for themselves what is different about the last year before recession. Based on...

Using a new comprehensive survey of adults in large U.S. media markets we show that minority and low-skill individuals, who are heavily exposed to shocks to the local economy, typically have stronger preferences for and stronger ex- posure to local news than high-skill and white individuals. At...

We study the decentralization of liquor policies in the Post-Prohibition Era, which is the most famous natural experiment ever conducted with respect to policy decentralization in the U.S. Our empirical analysis exploits a unique feature of this policy change, namely that we observe votes of...