Andreas Petrou-Zeniou

I am a second-year PhD student in economics and statistics at MIT. I am interested in econometrics and empirical industrial organization. Currently, I am interested in the econometrics of valued-added measurement. I am working on developing and applying tools from the literatures on multiple testing and inference for set-identified models to quantify uncertainty in settings where researchers use multiple value-added estimates to target and design policy. Formerly, I was an economics and CAAM (concurrent MS) undergraduate at the University of Chicago. My Erdős number is 5.

Please feel free to reach out at apetrouz [at] mit [dot] edu

Working Papers:

Inference After Ranking with Applications to Economic Mobility [slides] (with Azeem Shaikh)

submitted

Abstract: This paper considers the problem of inference after ranking. In our setting, we are interested in any population whose rank according to some random quantity, such as an estimated treatment effect, a measure of value-added, or benefit (net of cost), falls in a pre-specified range of values. As such, this framework generalizes the inference on winners setting previously considered in Andrews, Kitagawa, and McCloskey (2024), in which a winner is understood to be the single population whose rank according to some random quantity is highest. We show that this richer setting accommodates a broad variety of empirically-relevant applications. We develop a two-step method for inference, which we compare to existing methods or their natural generalizations to this setting. We first show the finite-sample validity of this method in a normal location model and then develop asymptotic counterparts to these results by proving uniform validity over a large class of distributions satisfying a weak uniform integrability condition. Importantly, our results permit degeneracy in the covariance matrix of the limiting distribution, which arises naturally in many applications. In an application to the literature on economic mobility, we find that it is difficult to distinguish between high and low-mobility census tracts when correcting for selection. Finally, we demonstrate the practical relevance of our theoretical results through an extensive set of simulations.

Uncertainty Quantification for Compound Decisions with Applications to Ranking and Selection (draft and slides coming soon)

Abstract: The compound decision framework allows for decision making with many, noisy estimates. For example, researchers may seek to guide students to high value-added schools, or audit firms that are likely to engage in discriminatory hiring practices. Empirical Bayes (EB) methods allow researchers to approximate Bayes-optimal decision-making in this framework. However, because these methods rely on estimating a high-dimensional prior, they are subject to sampling uncertainty. This paper considers the problem of quantifying this sampling uncertainty in EB decision rules. To this end, I demonstrate that Bayes-optimal decisions depend on multiple posterior means, making controlling false discoveries difficult. I introduce a notion of uniform coverage of decision-relevant EB estimands that guards against these false discoveries. Leveraging a new connection between this notion of coverage and coverage in partially identified models, I provide a step-down approach to inference which provides tighter confidence sets relative to existing methods. To demonstrate the practical relevance of my framework, I apply my methods to the problem of differentiating between low and high-mobility neighborhoods. My methods uncover large gaps in mobility between census tracts, suggesting that policy-targeting on economic mobility is more credible than previously understood.