Principled estimation of regression discontinuity designs

TL;DR

This paper introduces a principled method for estimating regression discontinuity designs that improves precision by integrating adaptive LASSO, reducing bias from covariate selection, especially in small samples.

Contribution

It develops a new approach combining adaptive LASSO with RDD estimation, implemented in the adaptiveRDD package, to enhance treatment effect precision.

Findings

Significantly improves treatment effect precision in small samples

Reduces bias caused by covariate selection

Effective in datasets with fewer than 200 observations

Abstract

Regression discontinuity designs are frequently used to estimate the causal effect of election outcomes and policy interventions. In these contexts, treatment effects are typically estimated with covariates included to improve efficiency. While including covariates improves precision asymptotically, in practice, treatment effects are estimated with a small number of observations, resulting in considerable fluctuations in treatment effect magnitude and precision depending upon the covariates chosen. This practice thus incentivizes researchers to select covariates which maximize treatment effect statistical significance rather than precision. Here, I propose a principled approach for estimating RDDs which provides a means of improving precision with covariates while minimizing adverse incentives. This is accomplished by integrating the adaptive LASSO, a machine learning method, into RDD estimation using an R package developed for this purpose, adaptiveRDD. Using simulations, I show that this method significantly improves treatment effect precision, particularly when estimating treatment effects with fewer than 200 observations.