Working Paper: NBER ID: w29124
Authors: David S. Lee; Justin McCrary; Marcelo J. Moreira; Jack R. Porter
Abstract: In the single-IV model, researchers commonly rely on t-ratio-based inference, even though the literature has quantified its potentially severe large-sample distortions. Building on Stock and Yogo (2005), we introduce the tF critical value function, leading to a standard error adjustment that is a smooth function of the first-stage F-statistic. For one-quarter of specifications in 61 AER papers, corrected standard errors are at least 49 and 136 percent larger than conventional 2SLS standard errors at the 5-percent and 1-percent significance levels, respectively. tF confidence intervals have shorter expected length than those of Anderson and Rubin (1949), whenever both are bounded.
Keywords: instrumental variables; t-ratio inference; statistical inference
JEL Codes: C01; C12; C26; C36
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| first-stage F-statistic (C29) | accuracy of inference in IV estimation (C26) |
| tF adjustment (F16) | corrected standard errors (C20) |
| tF confidence intervals (C29) | expected length compared to Anderson and Rubin (1949) (C29) |
| tF approach (F16) | usable finite confidence intervals for low F-statistic (C51) |
| tF adjustments (F16) | impact on inferences in applied research (C20) |