rdrobust

RDROBUST

The rdrobust package implements the statistical inference and graphical procedures for Regression Discontinuity designs employing local polynomial and partitioning methods. It provides point estimators, confidence intervals estimators, bandwidth selectors, automatic RD plots, and many other features.

This work was supported in part by the National Science Foundation through grants SES-1357561, SES-1459931, SES-1947805 and SES-2019432.

Authors

Sebastian Calonico ([e-mail])

Matias D. Cattaneo ([e-mail])

Max H. Farrell ([e-mail])

Ricardo Masini ([e-mail])

Rocio Titiunik ([e-mail])

Website

https://rdpackages.github.io/rdrobust

Major Upgrades

This package was first released in Spring 2014, and had two major upgrades in Fall 2016 and in Winter 2020.

• Fall 2016 new features include: (i) major speed improvements; (ii) covariate-adjusted bandwidth selection, point estimation, and robust inference; (iii) cluster-robust bandwidth selection, point estimation, and robust inference; (iv) weighted global polynomial fits and pointwise confidence bands for RD plots; and (v) several new bandwidths selectors (e.g., different bandwidths for control and treatment groups, coverage error optimal bandwidths, and optimal bandwidths for fuzzy designs).

• Winter 2020 new features include: (i) discrete running variable checks and adjustments; (ii) bandwidth selection adjustments for too few mass points in and/or overshooting of the support of the running variable; (iii) RD Plots with additional covariates plotted at their mean (previously the package set additional covariates at zero); (iv) automatic removal of co-linear additional covariates; (v) turn on/off standardization of variables (which may lead to small numerical/rounding discrepancies with prior versions); and (vi) rdplot output using ggplot2 in R.

Installation

To install/update use pip

pip install rdrobust

Usage

from rdrobust import rdrobust, rdbwselect, rdplot

• Replication: rdrobust illustration, rdplot illustration, senate data.

Dependencies

• numpy
• pandas
• scipy.linalg
• sklearn.linear_model
• plotnine

References

For overviews and introductions, see rdpackages website.

Software and Implementation

• Calonico, Cattaneo and Titiunik (2014): Robust Data-Driven Inference in the Regression-Discontinuity Design.
  Stata Journal 14(4): 909-946.

• Calonico, Cattaneo and Titiunik (2015): rdrobust: An R Package for Robust Nonparametric Inference in Regression-Discontinuity Designs.
  R Journal 7(1): 38-51.

• Calonico, Cattaneo, Farrell and Titiunik (2017): rdrobust: Software for Regression Discontinuity Designs.
  Stata Journal 17(2): 372-404.

Technical and Methodological

• Calonico, Cattaneo and Titiunik (2014): Robust Nonparametric Confidence Intervals for Regression-Discontinuity Designs.
  Econometrica 82(6): 2295-2326.
  Supplemental Appendix.

• Calonico, Cattaneo and Titiunik (2015): Optimal Data-Driven Regression Discontinuity Plots.
  Journal of the American Statistical Association 110(512): 1753-1769.
  Supplemental Appendix.

• Calonico, Cattaneo and Farrell (2018): On the Effect of Bias Estimation on Coverage Accuracy in Nonparametric Inference.
  Journal of the American Statistical Association 113(522): 767-779.
  Supplemental Appendix.

• Calonico, Cattaneo, Farrell and Titiunik (2019): Regression Discontinuity Designs Using Covariates.
  Review of Economics and Statistics 101(3): 442-451.
  Supplemental Appendix.

• Calonico, Cattaneo and Farrell (2020): Optimal Bandwidth Choice for Robust Bias Corrected Inference in Regression Discontinuity Designs.
  Econometrics Journal 23(2): 192-210.
  Supplemental Appendix.

• Calonico, Cattaneo and Farrell (2021): Coverage Error Optimal Confidence Intervals for Local Polynomial Regression.
  Working paper.
  Supplemental Appendix.