Research

Job Market Paper

Poverty Targeting with Imperfect Information

Is it enough to target the households that look poorest?

Some households look especially poor because their incomes are underestimated. I develop a targeting method that accounts for this uncertainty. In simulations, it achieves the same poverty reduction as standard targeting with 6.7 percent less spending.

Abstract

How should antipoverty programs allocate transfers when household income is known only through noisy predictions? I formulate this as a statistical decision problem in which a policymaker chooses nonnegative transfers within a fixed budget to minimize squared deviations of post-transfer income from the poverty line. The standard plug-in rule, which treats predictions as exact, is inadmissible. I develop a nonparametric empirical Bayes rule that allocates the budget using posterior mean poverty gaps in place of estimated gaps. Its Bayes regret is bounded by the mean squared difference between its posterior mean gaps and the oracle's, so the budget and nonnegativity constraints do not slow its convergence to the oracle. The approach extends to an objective that penalizes only remaining poverty and to fixed benefit menus, with weaker guarantees. In simulations using household surveys from nine African countries, the empirical Bayes rule reaches about 1.8 times as many poor people as plug-in OLS with the same budget and achieves the same average poverty-gap reduction with 6.7 percent less spending.

arXiv

Working Papers

When and How to Pilot: Design Rules for Two-Wave Experiments

How much should a small pilot change the experiment that follows?

A pilot can help researchers choose a more efficient split between treatment and control, but a small pilot can also mislead. I develop a method that guards against overreacting to limited evidence and approaches the optimal split as the pilot grows.

Abstract

Experimenters often run pilots, but how much a small pilot should shape the main-wave design has no settled answer. This paper shows how noisy pilot evidence should guide treatment assignment probabilities in two-wave experiments. Two canonical rules mark the extremes. Balanced assignment guards against worst cases but ignores evidence that one arm is noisier. Feasible Neyman allocation adapts, but with a finite pilot it can overreact to noise, producing arbitrarily large precision losses. I propose a Conditional Minimax Regret (CMR) rule that minimizes worst-case regret over a finite-sample confidence set for the treatment and control variances. CMR retains balance’s worst-case protection with high probability, converges to the Neyman allocation as the pilot grows, and attains the minimax-regret rate up to constants. It extends to multi-arm and stratified designs, and simulations calibrated to four field experiments show it avoids feasible Neyman’s severe small-pilot losses while matching its large-pilot gains.

arXiv Software

Two-Way Effects Models: A Nonparametric Empirical Bayes Approach

with Cole Davis

How should we estimate worker and firm effects when the two are sorted?

Wage decompositions split pay into worker and firm components, and existing methods assume the two are unrelated. We develop a method that lets the distribution of worker effects depend on the firm, so estimates reflect the sorting that occurs.

Abstract

Researchers use AKM models to decompose wages into worker effects and firm wage premia. The resulting fixed-effect estimates are often noisy because firm effects are identified by worker mobility, and conventional empirical Bayes approaches rely on parametric prior assumptions. We develop a nonparametric empirical Bayes framework in which the distribution of worker effects varies with latent firm effects. This allows worker and firm effects to be sorted, and it induces dependence among workers attached to the same reference firm. We propose a feasible estimation procedure, characterize the resulting shrinkage rules, and define an oracle benchmark that respects the hierarchical structure of AKM effects. Simulations show mean squared error close to an oracle benchmark and improvements over i.i.d.-based methods.

Draft on request

Publications

Birds of a Feather Collude Together: Subnational Alignment and Corruption

with Leopoldo Fergusson, Arturo Harker, and Carlos Molina · Forthcoming, American Political Science Review

Does partisan alignment across levels of government facilitate corruption?

Using close elections in Colombia, we find that partisan alignment between mayors and governors increases the fabrication of “ghost” students to obtain education transfers. Actual enrollment and student performance do not improve.

Abstract

We examine how subnational partisan alignment influences corruption in clientelistic environments, focusing on the fabrication of “ghost” students to inflate education transfers to local governments in Colombia. Using a Regression Discontinuity Design, we find that partisan alignment between municipal mayors and departmental governors increases ghost students by 0.3 standard deviations, without improving genuine enrollment or student performance. Alignment also leads to more discretionary hiring, patronage-based outsourcing, and increased electoral fraud risk. The effects are strongest in municipalities with weaker institutions and entrenched clientelism. Alignment also raises the likelihood that mayors’ relatives are appointed to departmental posts and governors’ relatives to municipal posts, consistent with reciprocal patronage. These findings support the view that resource diversion benefits politicians with few benefits for local constituencies. Aligned politicians also experience better future electoral prospects, suggesting a breakdown in accountability. Our results highlight how clientelistic networks distort public service delivery, reinforcing the persistence of political corruption.

PDF

Selected Presentations

2026

Midwest Econometrics Group Annual Meeting, Cincinnati

Poverty Targeting with Imperfect Information
Accepted for presentation · October 9–10, 2026

2025

World Congress of the Econometric Society, Seoul

Poverty Targeting with Imperfect Information

2025

Advances with Field Experiments (AFE), Chicago

When and How to Pilot: Design Rules for Two-Wave Experiments