Distributional robustness of K-class estimators and the PULSE
This work addresses robustness in causal inference for researchers and practitioners, offering an incremental improvement with a new estimator that enhances performance in specific scenarios.
The paper tackles the problem of causal models not being optimal under bounded interventions by proving the K-class estimator's optimality and introducing PULSE, a novel estimator that minimizes prediction error within a confidence region, showing it reduces variability and outperforms others in settings like weak instruments.
While causal models are robust in that they are prediction optimal under arbitrarily strong interventions, they may not be optimal when the interventions are bounded. We prove that the classical K-class estimator satisfies such optimality by establishing a connection between K-class estimators and anchor regression. This connection further motivates a novel estimator in instrumental variable settings that minimizes the mean squared prediction error subject to the constraint that the estimator lies in an asymptotically valid confidence region of the causal coefficient. We call this estimator PULSE (p-uncorrelated least squares estimator), relate it to work on invariance, show that it can be computed efficiently as a data-driven K-class estimator, even though the underlying optimization problem is non-convex, and prove consistency. We evaluate the estimators on real data and perform simulation experiments illustrating that PULSE suffers from less variability. There are several settings including weak instrument settings, where it outperforms other estimators.