Econometrics of Interference
Paper Session
Monday, Jan. 5, 2026 8:00 AM - 10:00 AM (EST)
- Chair: Kirill Borusyak, University of California-Berkeley
Coupling Designs for Randomized Experiments
Abstract
We describe a new family of experimental designs for complex treatments in RCTs, called Coupling Designs. These designs first use baseline covariates to match experimental units into groups. Treatments are then randomly drawn in a correlated fashion to maximize the dispersion of effective treatment within the matched groups. For the randomization step, we take advantage of Monte Carlo integration methods, such as latin hypercube sampling, drawing connections to this literature. Coupling Designs generalize classical stratified randomization, allowing randomization of any treatment distribution within matched groups of any size. For example, we enable matched pairs randomization of continuous treatments, enabling efficient experimentation in new settings. Our analysis shows that the efficiency gain from coupling designs is governed by the product of the dispersion of treatments and match quality. Viewed in this context, stratified randomization is a heuristic that forces perfect dispersion, typically at the cost of poor match quality.The Local Approach to Causal Inference Under Network Interference
Abstract
We propose a new nonparametric modeling framework for causal inference when outcomes depend on how agents are linked in a social or economic network. Such network interference describes a large literature on treatment spillovers, social interactions, social learning, information diffusion, disease and financial contagion, social capital formation, and more. Our approach works by first characterizing how an agent is linked in the network using the configuration of other agents and connections nearby as measured by path distance. The impact of a policy or treatment assignment is then learned by pooling outcome data across similarly configured agents. We demonstrate the approach by deriving finite-sample bounds on the mean-squared error of a k-nearest-neighbor estimator for the average treatment response as well as proposing an asymptotically valid test for the hypothesis of policy irrelevance.Regression Discontinuity Designs Under Interference
Abstract
We extend the continuity-based framework to Regression Discontinuity Designs (RDDs) to identify and estimate causal effects in the presence of interference when units are connected through a network. In this setting, assignmentto an “effective treatment”, which comprises the individual treatment and a summary of the treatment of interfering units (e.g., friends, classmates), is determined by the unit’s score and the scores of other interfering units, leading to
a multiscore RDD with potentially complex, multidimensional boundaries. We characterize these boundaries and derive generalized continuity assumptions to identify the proposed causal estimands, i.e., point and boundary causal effects.
Additionally, we develop a distance-based nonparametric estimator, derive its asymptotic properties under restrictions on the network degree distribution,
and introduce a novel variance estimator that accounts for network correlation. Finally, we apply our methodology to the PROGRESA/Oportunidades dataset to estimate the direct and indirect effects of receiving cash transfers on children’s school attendance.
JEL Classifications
- C21 - Cross-Sectional Models; Spatial Models; Treatment Effect Models; Quantile Regressions