Working Paper: NBER ID: w24833
Authors: Guido Imbens; Konrad Menzel
Abstract: The bootstrap, introduced by Efron (1982), has become a very popular method for estimating variances and constructing confidence intervals. A key insight is that one can approximate the properties of estimators by using the empirical distribution function of the sample as an approximation for the true distribution function. This approach views the uncertainty in the estimator as coming exclusively from sampling uncertainty. We argue that for causal estimands the uncertainty arises entirely, or partially, from a different source, corresponding to the stochastic nature of the treatment received. We develop a bootstrap procedure that accounts for this uncertainty, and compare its properties to that of the classical bootstrap.
Keywords: Potential Outcomes; Causality; Randomization; Inference; Bootstrap; Copula; Partial Identification
JEL Codes: C01; C31
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| classical bootstrap method (C59) | inadequate for causal estimands (C51) |
| causal bootstrap procedure (C59) | more accurate estimates of average treatment effect (C22) |
| causal bootstrap (C59) | nuanced understanding of variance of estimator (C51) |
| causal bootstrap (C59) | tighter confidence intervals (C46) |
| causal bootstrap (C59) | more reliable statistical inference (C12) |
| design uncertainty (D80) | inadequate causal estimates (C20) |