Class 14: Regression Discontinuity, Part II

Today we’ll apply RD to real data using Medicare Advantage star ratings, and learn to validate the design.

Objectives

• Implement an RD strategy with real data (MA star ratings)
• Define “manipulation” of the running variable and test for it
• Define what we mean by “covariate balance” and why it matters for RD
• Estimate RD effects across multiple thresholds

Activities

• Implement an RD strategy with Medicare Advantage star ratings
• Test for manipulation of the running variable
• Test for covariate balance
• Estimate and compare effects across star rating cutoffs