1: \begin{abstract}\singlespacing

2: Consider estimating the G-formula for the counterfactual mean outcome under a given treatment regime in a

3: longitudinal study. \cite{Bang&Robins05}

4: provided an estimator for this parameter based on a sequential regression formulation. This approach is doubly robust in that

5: it is consistent if either all outcome regressions or all treatment mechanisms are consistently

6: estimated. 

7: We define a stronger notion of double robustness, termed sequential double robustness, for estimators of the longitudinal G-formula. The definition emerges naturally from a more general definition of sequential double robustness for outcome regression estimators. An outcome regression estimator is sequentially doubly robust (SDR) if, at each subsequent time point, either the outcome regression or the treatment mechanism is consistently estimated. This form of robustness is exactly what one would anticipate is attainable by studying the remainder term of a first-order expansion of the G-formula parameter. We introduce a novel SDR estimator, whose development involves a novel infinite-dimensional extension of targeted minimum loss-based estimation. These new developments have the potential to dramatically improve both the robustness of estimators of the marginal G-formula and the rate of convergence of outcome regression estimators, with improved rates translating into dramatically improved finite sample behavior.\vspace{.2in}

8: \end{abstract}