Showing papers by "David S. Lee published in 2021"
TL;DR: This article introduced the tF procedure, which provides F-dependent adjusted t-ratio critical values, and re-examine 57 AER papers and find that corrected inference causes half of the initially presumed statistically significant results to be insignificant.
Abstract: In the single IV model, current practice relies on the first-stage F exceeding some threshold (e.g., 10) as a criterion for trusting t-ratio inferences, even though this yields an anti-conservative test. We show that a true 5 percent test instead requires an F greater than 104.7. Maintaining 10 as a threshold requires replacing the critical value 1.96 with 3.43. We re-examine 57 AER papers and find that corrected inference causes half of the initially presumed statistically significant results to be insignificant. We introduce a more powerful test, the tF procedure, which provides F-dependent adjusted t-ratio critical values.
116 citations
Posted Content•
TL;DR: This article introduced the tF critical value function, leading to a minimized standard error adjustment factor that is a smooth function of the first-stage F-statistic, and applied the correction to a sample of 61 AER papers.
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 the approach for correcting inference of Stock and Yogo (2005), we introduce the tF critical value function, leading to a minimized standard error adjustment factor that is a smooth function of the first-stage F-statistic. Applying the correction to a sample of 61 AER papers leads to a 25 percent increase in standard errors, on average. tF confidence intervals have shorter expected length than those of Anderson and Rubin (1949), whenever both are bounded intervals.
28 citations
TL;DR: In this paper, the deadweight loss associated with possible reforms is defined and its measurement typically requires specifying a fixed set of rules that must be satisfied to aid analytical tractability, which is difficult to obtain.
Abstract: Central to the welfare analysis of income transfer programs is the deadweight loss associated with possible reforms. To aid analytical tractability, its measurement typically requires specifying a ...
10 citations
TL;DR: Treatment effect estimates in regression discontinuity (RD) designs are often sensitive to the choice of bandwidth and polynomial order, the two important ingredients of widely used local regressio...
Abstract: Treatment effect estimates in regression discontinuity (RD) designs are often sensitive to the choice of bandwidth and polynomial order, the two important ingredients of widely used local regressio...
4 citations
Posted Content•
TL;DR: This article introduced the tF critical value function, leading to a minimized standard error adjustment factor that is a smooth function of the first-stage F-statistic, and applied the correction to a sample of 61 AER papers.
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 the approach for correcting inference of Stock and Yogo (2005), we introduce the tF critical value function, leading to a minimized standard error adjustment factor that is a smooth function of the first-stage F-statistic. Applying the correction to a sample of 61 AER papers leads to a 25 percent increase in standard errors, on average. tF confidence intervals have shorter expected length than those of Anderson and Rubin (1949), whenever both are bounded intervals.
Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.
1 citations