Efficient Estimation of Data Combination Models by the Method of Auxiliary-to-Study Tilting (AST)
We propose a locally efficient, doubly robust, estimator for a class of semiparametric data combination problems. A leading estimand in this class is the average treatment effect on the treated (ATT). Data combination problems are related to, but distinct from, the class of missing data problems analyzed by Robins, Rotnitzky and Zhao (1994) (of which the Average Treatment Effect (ATE) estimand is a special case). Our procedure may be used to efficiently estimate, among other objects, the ATT, the two-sample instrumental variables model (TSIV), counterfactual distributions, and poverty maps. In an empirical application we use our procedure to characterize residual Black-White wage inequality after flexibly controlling for 'pre-market' differences in measured cognitive achievement as in Neal and Johnson (1996). We find that residual Black-White inequality is negligible at lower and higher quantiles of the Black wage distribution, but substantial at middle quantiles.
You may purchase this paper on-line in .pdf format from SSRN.com ($5) for electronic delivery.
An online appendix is available for this publication.