59. A grant is offered only below a score threshold. When might comparing applicants just on either side estimate its effect?
HardBusiness cases
Case
Show model answer
A regression-discontinuity design can compare outcomes near the cutoff if potential outcomes vary smoothly there and applicants cannot manipulate their position. Check whether other policies also change at the same threshold.
Investigate score bunching, baseline balance and sensitivity to bandwidth and specification. The result is local to applicants near the cutoff; it does not automatically describe the effect for all applicants or establish that the entire score range is comparable.
What interviewers listen for
- Check continuity and manipulation risks
- Examine cutoff-specific robustness
- Keep the conclusion local