What does double robustness actually guarantee?

Question: Consider an augmented inverse-probability-weighted estimator of an average treatment effect. Under consistency, unconfoundedness and positivity, which statement best describes its double-robustness property? A) It is consistent if either the propensity-score model or the outcome-regression model is consistently estimated B) It is unbiased in every finite sample whenever either nuisance model is estimated using machine learning C) Its confidence interval has correct coverage if either nuisance model is correct, regardless of the variance estimator D) Its variance is lower than both pure regression adjustment and inverse-probability weighting whenever one model is correct Correct: A Explanation: Double robustness concerns consistency of the point estimator when at least one of two nuisance components is correctly specified or consistently estimated. It does not automatically imply finite-sample unbiasedness, minimum variance, or valid inference under every variance-estimation procedure. Topic: advanced ML / causal inference / doubly robust estimation