24.7 Steps for Sharp RD

  1. Graph the data by computing the average value of the outcome variable over a set of bins (large enough to see a smooth graph, and small enough to make the jump around the cutoff clear).

  2. Run regression on both sides of the cutoff to get the treatment effect

  3. Robustness checks:

    1. Assess possible jumps in other variables around the cutoff

    2. Hypothesis testing for bunching

    3. Placebo tests

    4. Varying bandwidth