TY - JOUR
T1 - Exploiting Discovered Regression Discontinuities to Debias Conditioned-on-observable Estimators
AU - Jakubowski, Benjamin
AU - Somanchi, Sriram
AU - McFowland, Edward
AU - Neill, Daniel B.
N1 - Publisher Copyright:
©2023 Benjamin Jakubowski, Sriram Somanchi, Edward McFowland III, Daniel B. Neill.
PY - 2023
Y1 - 2023
N2 - Regression discontinuity (RD) designs are widely used to estimate causal effects in the absence of a randomized experiment. However, standard approaches to RD analysis face two significant limitations. First, they require a priori knowledge of discontinuities in treatment. Second, they yield doubly-local treatment effect estimates, and fail to provide more general causal effect estimates away from the discontinuity. To address these limitations, we introduce a novel method for automatically detecting RDs at scale, integrating information from multiple discovered discontinuities with an observational estimator, and extrapolating away from discovered, local RDs. We demonstrate the performance of our method on two synthetic datasets, showing improved performance compared to direct use of an observational estimator, direct extrapolation of RD estimates, and existing methods for combining multiple causal effect estimates. Finally, we apply our novel method to estimate spatially heterogeneous treatment effects in the context of a recent economic development problem.
AB - Regression discontinuity (RD) designs are widely used to estimate causal effects in the absence of a randomized experiment. However, standard approaches to RD analysis face two significant limitations. First, they require a priori knowledge of discontinuities in treatment. Second, they yield doubly-local treatment effect estimates, and fail to provide more general causal effect estimates away from the discontinuity. To address these limitations, we introduce a novel method for automatically detecting RDs at scale, integrating information from multiple discovered discontinuities with an observational estimator, and extrapolating away from discovered, local RDs. We demonstrate the performance of our method on two synthetic datasets, showing improved performance compared to direct use of an observational estimator, direct extrapolation of RD estimates, and existing methods for combining multiple causal effect estimates. Finally, we apply our novel method to estimate spatially heterogeneous treatment effects in the context of a recent economic development problem.
KW - causal inference
KW - Gaussian processes
KW - heterogeneous treatment effects
KW - natural experiments
KW - regression discontinuity designs
UR - http://www.scopus.com/inward/record.url?scp=85214400957&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85214400957&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:85214400957
SN - 1532-4435
VL - 24
JO - Journal of Machine Learning Research
JF - Journal of Machine Learning Research
ER -