Uncertain Neighbors: Bayesian Propensity Score Matching For Causal Inference

TL;DR

This paper compares traditional and Bayesian propensity score matching methods, demonstrating that the Bayesian approach better utilizes information and reduces arbitrary decisions, with evaluations through simulations and a case study on land reform in Colombia.

Contribution

It introduces a Bayesian propensity score matching method and empirically compares its performance to standard methods in various scenarios.

Findings

Bayesian matching makes less arbitrary decisions about data inclusion.

Bayesian approach utilizes available information more effectively.

Simulation and case study show improved matching performance.

Abstract

We compare the performance of standard nearest-neighbor propensity score matching with that of an analogous Bayesian propensity score matching procedure. We show that the Bayesian approach makes better use of available information, as it makes less arbitrary decisions about which observations to drop and which ones to keep in the matched sample. We conduct a simulation study to evaluate the performance of standard and Bayesian nearest-neighbor matching when matching is done with and without replacement. We then use both methods to replicate a recent study about the impact of land reform on guerrilla activity in Colombia.