Two Layers of Instability in Causal Estimation
There is a precise sense in which drawing causal inferences from observational data is hard, even when identifiability is assumed. In particular, Robins and Ritov (1997) and Robins et al. (2003) showed that causal effects can be discontinuous as a function of the data distribution: two arbitrarily close data distributions might correspond to different causal effects. This is a fact independent of the choice of estimator; however, not all estimators are equally unstable. Our contribution is to surface a second layer of instability that depends on the choice of estimator. We show that many standard point estimates can be read as point summaries of multimodal distributions over the space of structural causal models. As such, estimators can jump discontinuously in the data distribution. This defines a taxonomy of estimators that admits a decision-theoretic reading: stability depends on whether the implicit loss function an estimator optimizes is aligned with the causal effect itself. Specifically, inverse propensity weighted estimators and regression estimators are examples of discontinuous summaries, while explicit posterior means and medians are shown to be continuous.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Scalable Counterfactual Risk Estimation for Rare Events in Longitudinal Data
Estimating the causal effect of time-varying treatments on survival outcomes in large observational studies is computationally demanding, particularly when outcomes are rare. While g-formula-based methods such as the ite…
Causal InferenceLearning high-dimensional causal effect
The scarcity of high-dimensional causal inference datasets restricts the exploration of complex deep models. In this work, we propose a method to generate a synthetic causal dataset that is high-dimensional. The syntheti…
Causal InferenceVocal Bursts Intensity PredictionMechanistic Interpretability as Statistical Estimation: A Variance Analysis
Mechanistic Interpretability (MI) aims to reverse-engineer model behaviors by identifying functional sub-networks. Yet, the scientific validity of these findings depends on their stability. In this work, we argue that ci…
The Dynamics of Instability
We study a model in which two players with opposing interests try to alter a status quo through instability-generating actions. We show that instability can be used to secure longer-term durable changes, even if it is co…
Combining T-learning and DR-learning: a framework for oracle-efficient estimation of causal contrasts
We introduce efficient plug-in (EP) learning, a novel framework for the estimation of heterogeneous causal contrasts, such as the conditional average treatment effect and conditional relative risk. The EP-learning framew…