Working Paper: NBER ID: w22363
Authors: Amanda E. Kowalski
Abstract: I examine treatment effect heterogeneity within an experiment to inform external validity. The local average treatment effect (LATE) gives an average treatment effect for compliers. I bound and estimate average treatment effects for always takers and never takers by extending marginal treatment effect methods. I use these methods to separate selection from treatment effect heterogeneity, generalizing the comparison of OLS to LATE. Applying these methods to the Oregon Health Insurance Experiment, I find that the treatment effect of insurance on emergency room utilization decreases from always takers to compliers to never takers. Previous utilization explains a large share of the treatment effect heterogeneity. Extrapolations show that other expansions could increase or decrease utilization.
Keywords: treatment effect heterogeneity; marginal treatment effect; local average treatment effect; health insurance; emergency room utilization
JEL Codes: C1; C9; H4; I13
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| insurance treatment effect of always takers (G52) | insurance treatment effect of compliers (C24) |
| insurance treatment effect of compliers (C24) | insurance treatment effect of never takers (G52) |
| previous ER utilization (I11) | treatment effect heterogeneity among always takers, compliers, and never takers (C22) |
| LATE (Y60) | treatment effect heterogeneity (C21) |
| different policy implementations (D78) | varying impacts on ER utilization (I11) |