No-harm calibration for generalized Oaxaca–Blinder estimators
- 8 June 2023
- journal article
- research article
- Published by Oxford University Press (OUP) in Biometrika
- Vol. 111 (1) , 331-338
Abstract
Summary: In randomized experiments, adjusting for observed features when estimating treatment effects has been proposed as a way to improve asymptotic efficiency. However, among parametric methods, only linear regression has been proven to form an estimate of the average treatment effect that is asymptotically no less efficient than the treated-minus-control difference in means regardless of the true data generating process. Randomized treatment assignment provides this do-no-harm property, with neither truth of a linear model nor a generative model for the outcomes being required. We present a general calibration method that confers the same no-harm property onto estimators leveraging a broad class of nonlinear models. This recovers the usual regression-adjusted estimator when ordinary least squares is used, and further provides noninferior treatment effect estimators using methods such as logistic and Poisson regression. The resulting estimators are noninferior to both the difference-in-means estimator and to treatment effect estimators that have not undergone calibration. We show that our estimator is asymptotically equivalent to an inverse-probability-weighted estimator using a logit link with predicted potential outcomes as covariates. In a simulation study, we demonstrate that common nonlinear estimators without our calibration procedure may perform markedly worse than both the calibrated estimator and the unadjusted difference in means.Keywords
This publication has 18 references indexed in Scilit:
- Agnostic notes on regression adjustments to experimental data: Reexamining Freedman’s critiqueThe Annals of Applied Statistics, 2013
- Improved double-robust estimation in missing data and causal inference modelsBiometrika, 2012
- A local maximal inequality under uniform entropyElectronic Journal of Statistics, 2011
- Bounded, efficient and doubly robust estimation with inverse weightingBiometrika, 2010
- Simple, Efficient Estimators of Treatment Effects in Randomized Trials Using Generalized Linear Models to Leverage Baseline VariablesThe International Journal of Biostatistics, 2010
- On regression adjustments to experimental dataAdvances in Applied Mathematics, 2008
- Efficient Estimation of Average Treatment Effects Using the Estimated Propensity ScoreEconometrica, 2003
- On the Role of the Propensity Score in Efficient Semiparametric Estimation of Average Treatment EffectsEconometrica, 1998
- Male-Female Wage Differentials in Urban Labor MarketsInternational Economic Review, 1973
- Wage Discrimination: Reduced Form and Structural EstimatesThe Journal of Human Resources, 1973