About

Avi Feller is an associate professor at the Goldman School of Public Policy at the University of California, Berkeley, where he works at the intersection of public policy, data science, and statistics. His methodological research centers on learning more from social policy evaluations, while his applied research focuses on collaborating with governments to use data for designing, implementing, and evaluating policies. Feller has a background that includes serving as Special Assistant to the Director at the White House Office of Management and Budget and working at the Center on Budget and Policy Priorities. He holds a Ph.D. in Statistics from Harvard University, an M.Sc. in Applied Statistics from the University of Oxford as a Rhodes Scholar, and a B.A. in Political Science and Applied Mathematics from Yale University.

Research topics

• Artificial Intelligence
• Computer Science
• Mathematics
• Econometrics
• Statistics

Selected publications

• On Nonasymptotic Confidence Intervals for Treatment Effects in Randomized Experiments

  ArXiv.org · 2026-01-16

  articleOpen access

  We study nonasymptotic (finite-sample) confidence intervals for treatment effects in randomized experiments. In the existing literature, the effective sample sizes of nonasymptotic confidence intervals tend to be looser than the corresponding central-limit-theorem-based confidence intervals by a factor depending on the square root of the propensity score. We show that this performance gap can be closed, designing nonasymptotic confidence intervals that have the same effective sample size as their asymptotic counterparts. Our approach involves systematic exploitation of negative dependence or variance adaptivity (or both). We also show that the nonasymptotic rates that we achieve are unimprovable in an information-theoretic sense.

  PublisherOA PDF

• Abortion Bans and Maternal, Pregnancy-Related, and Pregnancy-Associated Mortality in 14 US States, 2016-2023: Estimated Impacts Amid Substantial Measurement Challenges.

  PubMed · 2026-06-01

  articleOpen access

  . 2026;116(6):819-828. https://doi.org/10.2105/AJPH.2026.308465).

  PublisherOA PDFDOI

• Authors’ reply to the Discussion of ‘Augmented balancing weights as linear regression' by Bruns-Smith et al

  Journal of the Royal Statistical Society Series B (Statistical Methodology) · 2026-04-16

  article
  PublisherDOI

• Discussion of "Matrix Completion When Missing Is Not at Random and Its Applications in Causal Panel Data Models"

  Open MIND · 2026-02-24

  preprintSenior author

  Choi and Yuan (2025) propose a novel approach to applying matrix completion to the problem of estimating causal effects in panel data. The key insight is that even in the presence of structured patterns of missing data -- i.e. selection into treatment -- matrix completion can be effective if the number of treated observations is small relative to the number of control observations. We applaud the authors for their insightful and interesting paper. We discuss this proposal from two complementary perspectives. First, we situate their proposal as an example of a "split-apply-combine" strategy that underlies many modern panel data estimators, including difference-in-differences and synthetic control approaches. Second, we discuss the issue of the statistical "last mile problem" -- the gap between theory and practice -- and offer suggestions on how to partially address it. We conclude by considering the challenges of estimating the impacts of public policies using panel data and apply the approach to a study on the effect of right to carry laws on violent crime.

  DOI

• Methodological considerations for investigating the impact of abortion restrictions on outcomes using aggregate panel data

  American Journal of Epidemiology · 2026-02-12 · 1 citations

  article

  Dramatic changes in the US abortion policy landscape have led to growing interest in studying the health and social impacts of abortion bans. Many studies of population-level impacts necessarily rely on panel designs using aggregate state-level data to strengthen causal inference, yet such analyses risk pitfalls if they apply generic evaluation frameworks that overlook the complexity of the US abortion context and relevant outcomes. This commentary provides practical guidance for researchers engaged in panel studies of abortion policy, as well as for peer reviewers who may be less familiar with the methodological and substantive considerations in this area. Drawing from recent work, we highlight abortion-specific challenges that require attention, including time-varying confounding and violation of parallel trends, COVID-era disruptions, data suppression, spillover effects, and subgroup heterogeneity. We further recommend assessing sensitivity to including Texas, given its earlier implementation of abortion restrictions and potential outsized influence on results. Ultimately, we emphasize that rigorous evaluation of abortion policies requires thoughtful study design, context-specific considerations, and collaboration between methodologists and subject-matter experts.

  PublisherDOI

• Discussion of "Matrix Completion When Missing Is Not at Random and Its Applications in Causal Panel Data Models"

  ArXiv.org · 2026-02-24

  articleOpen accessSenior author

  Choi and Yuan (2025) propose a novel approach to applying matrix completion to the problem of estimating causal effects in panel data. The key insight is that even in the presence of structured patterns of missing data -- i.e. selection into treatment -- matrix completion can be effective if the number of treated observations is small relative to the number of control observations. We applaud the authors for their insightful and interesting paper. We discuss this proposal from two complementary perspectives. First, we situate their proposal as an example of a "split-apply-combine" strategy that underlies many modern panel data estimators, including difference-in-differences and synthetic control approaches. Second, we discuss the issue of the statistical "last mile problem" -- the gap between theory and practice -- and offer suggestions on how to partially address it. We conclude by considering the challenges of estimating the impacts of public policies using panel data and apply the approach to a study on the effect of right to carry laws on violent crime.

  PublisherOA PDF

• Deconfounding Scores and Representation Learning for Causal Effect Estimation with Weak Overlap

  arXiv (Cornell University) · 2026-04-01

  preprintOpen accessSenior author

  Overlap, also known as positivity, is a key condition for causal treatment effect estimation. Many popular estimators suffer from high variance and become brittle when features differ strongly across treatment groups. This is especially challenging in high dimensions: the curse of dimensionality can make overlap implausible. To address this, we propose a class of feature representations called deconfounding scores, which preserve both identification and the target of estimation; the classical propensity and prognostic scores are two special cases. We characterize the problem of finding a representation with better overlap as minimizing an overlap divergence under a deconfounding score constraint. We then derive closed-form expressions for a class of deconfounding scores under a broad family of generalized linear models with Gaussian features and show that prognostic scores are overlap-optimal within this class. We conduct extensive experiments to assess this behavior empirically.

  PublisherDOI

• On Nonasymptotic Confidence Intervals for Treatment Effects in Randomized Experiments

  arXiv (Cornell University) · 2026-01-16

  preprintOpen access

  We study nonasymptotic (finite-sample) confidence intervals for treatment effects in randomized experiments. In the existing literature, the effective sample sizes of nonasymptotic confidence intervals tend to be looser than the corresponding central-limit-theorem-based confidence intervals by a factor depending on the square root of the propensity score. We show that this performance gap can be closed, designing nonasymptotic confidence intervals that have the same effective sample size as their asymptotic counterparts. Our approach involves systematic exploitation of negative dependence or variance adaptivity (or both). We also show that the nonasymptotic rates that we achieve are unimprovable in an information-theoretic sense.

  PublisherDOI

• A Weighting Framework for Clusters as Confounders in Observational Studies

  Open MIND · 2026-02-04

  preprint

  When units in observational studies are clustered in groups, such as students in schools or patients in hospitals, researchers often address confounding by adjusting for cluster-level covariates or cluster membership. In this paper, we develop a unified weighting framework that clarifies how different estimation methods control two distinct sources of imbalance: global balance (differences between treated and control units across clusters) and local balance (differences within clusters). We show that inverse propensity score weighting (IPW) with a random effects propensity score model -- the current standard in the literature -- targets only global balance and constant level shifts across clusters, but imposes no constraints on local balance. We then present two approaches that target both forms of balance. First, hierarchical balancing weights directly control global and local balance through a constrained optimization problem. Second, building on the recently proposed Generalized Mundlak approach, we develop a novel Mundlak balancing weights estimator that adjusts for cluster-level sufficient statistics rather than cluster indicators; this approach can accommodate small clusters where all units are treated or untreated. Critically, these approaches rest on different assumptions: hierarchical balancing weights require only that treatment is ignorable given covariates and cluster membership, while Mundlak methods additionally require an exponential family structure. We then compare these methods in a simulation study and in two applications in education and health services research that exhibit very different cluster structures.

  DOI

• A Weighting Framework for Clusters as Confounders in Observational Studies

  arXiv (Cornell University) · 2026-02-04

  articleOpen access

  When units in observational studies are clustered in groups, such as students in schools or patients in hospitals, researchers often address confounding by adjusting for cluster-level covariates or cluster membership. In this paper, we develop a unified weighting framework that clarifies how different estimation methods control two distinct sources of imbalance: global balance (differences between treated and control units across clusters) and local balance (differences within clusters). We show that inverse propensity score weighting (IPW) with a random effects propensity score model -- the current standard in the literature -- targets only global balance and constant level shifts across clusters, but imposes no constraints on local balance. We then present two approaches that target both forms of balance. First, hierarchical balancing weights directly control global and local balance through a constrained optimization problem. Second, building on the recently proposed Generalized Mundlak approach, we develop a novel Mundlak balancing weights estimator that adjusts for cluster-level sufficient statistics rather than cluster indicators; this approach can accommodate small clusters where all units are treated or untreated. Critically, these approaches rest on different assumptions: hierarchical balancing weights require only that treatment is ignorable given covariates and cluster membership, while Mundlak methods additionally require an exponential family structure. We then compare these methods in a simulation study and in two applications in education and health services research that exhibit very different cluster structures.

  PublisherOA PDF

Frequent coauthors

• Eli Ben‐Michael

  62 shared

• Luke Miratrix

  46 shared

• Jesse Rothstein

  University of California, Berkeley

  31 shared

• Todd Grindal

  SRI International

  24 shared

• Guillaume Basse

  Citadel

  19 shared

• Peng Ding

  University of California, Berkeley

  18 shared

• Elizabeth A. Stuart

  Johns Hopkins University

  18 shared

• Noah Haber

  Center for Open Science

  13 shared