Working Paper: NBER ID: w9497
Authors: James Heckman; Salvador Navarro-Lozano
Abstract: This paper investigates four topics. (1) It examines the different roles played by the propensity score (probability of selection) in matching, instrumental variable and control functions methods. (2) It contrasts the roles of exclusion restrictions in matching and selection models. (3) It characterizes the sensitivity of matching to the choice of conditioning variables and demonstrates the greater robustness of control function methods to misspecification of the conditioning variables. (4) It demonstrates the problem of choosing the conditioning variables in matching and the failure of conventional model selection criteria when candidate conditioning variables are not exogenous.
Keywords: No keywords provided
JEL Codes: C31
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Conventional matching methods do not appropriately handle excluded variables (C52) | Biased estimates (C51) |
| Control functions can utilize exclusion restrictions (C24) | Mitigate bias from excluded variables (C24) |
| Sensitivity of matching methods to conditioning variables (C52) | Significant bias (D91) |
| Control function methods are more robust to misspecification of conditioning variables (C51) | More reliable estimate of treatment effects (C22) |
| Failure of conventional model selection criteria in matching (C52) | Incorrect inferences (C20) |