Local Randomization Methods for RD Designs

The rdlocrand package implements estimation, inference, and graphical procedures for Regression Discontinuity (RD) designs using local randomization methods.

• rdrandinf: randomization inference and confidence intervals under local randomization.
• rdwinselect: data-driven window selection around the cutoff using covariate balance tests.
• rdsensitivity: sensitivity analysis of randomization p-values and confidence intervals over treatment-effect and window grids.
• rdrbounds: Rosenbaum bounds for sensitivity to unobserved confounders under local randomization.

Python Implementation

To install/update in Python type:

pip install rdlocrand

• Help: PyPI repository.

• Replication: py-script, senate data.

R Implementation

To install/update in R type:

install.packages('rdlocrand')

• Help: R Manual, CRAN repository.

• Replication: R-script, senate data.

Stata Implementation

To install/update in Stata type:

net install rdlocrand, from(https://raw.githubusercontent.com/rdpackages/rdlocrand/main/stata) replace

• Help: rdrandinf, rdwinselect, rdsensitivity, rdrbounds.

• Replication: do-file, senate data.

References

For overviews and introductions, see rdpackages website. Source code is available at https://github.com/rdpackages/rdlocrand.

Software and Implementation

• Cattaneo, Titiunik and Vazquez-Bare (2016): Inference in Regression Discontinuity Designs under Local Randomization.
  Stata Journal 16(2): 331-367.

Technical and Methodological

• Cattaneo, Frandsen and Titiunik (2015): Randomization Inference in the Regression Discontinuity Design: An Application to Party Advantages in the U.S. Senate.
  Journal of Causal Inference 3(1): 1-24.

• Cattaneo, Titiunik and Vazquez-Bare (2017): Comparing Inference Approaches for RD Designs: A Reexamination of the Effect of Head Start on Child Mortality.
  Journal of Policy Analysis and Management 36(3): 643-681.

Funding

This work was supported in part by the National Science Foundation through grants SES-1357561.