Imbens and Wager Propose PLRD, a Regression Discontinuity Estimator With Lower Error
lihua_lei_stat · x · 2026-09-09
Aditya Ghosh, Guido Imbens, and Stefan Wager released a new paper, PLRD: Partially Linear Regression Discontinuity Inference (arXiv:2503.09907). Key points:
- RDD is hugely popular in empirical economics, but widely used confidence-interval approaches are often suboptimal in practice.
- PLRD follows the optimal-weights spirit of Armstrong-Kolesar (2018) and Imbens-Wager (2019).
- In simulations calibrated to twelve high-profile RDD applications, PLRD shows substantially lower estimation error, with valid and shorter confidence intervals, plus large-sample guarantees.
- The simulations use Wasserstein GANs to generate synthetic data indistinguishable from the originals, offering a template for credibly evaluating new econometric methods.
More from Research
- Zhejiang University Releases RoboSPA Benchmark to Stress-Test VLA Models on Long-Horizon Tasks — zju · 2026-09-09
- CMU's CoVeR: Training-Free Token Pruning That Preserves Multi-View 3D Reasoning in VLMs — CarnegieMellonU · 2026-09-09
- NVIDIA's ReactVAU: Slow-Fast Decoupled Framework for Real-Time Streaming Video Anomaly Understanding — nvidia · 2026-09-09
- Purdue's SQS Combines Spike-and-Slab Sparsity and GMM Quantization for High-Rate DNN Compression — Purdue · 2026-09-09
- USC Study: LLMs Recognize Unanswerable Questions but Fail to Refuse Due to Routing Misalignment — UniversityofSouthernCalifornia · 2026-09-09
- Mathematician on OpenAI's Navier-Stokes push: picking the right question beats solving it — PTenigma · 2026-09-09