Generalised Transportability via Causal Abstractions

2026-08-16

AI Summary

This study addresses the challenges of cross-population transfer in causal inference and quantitative evaluation under target-data scarcity. Grounded in causal abstraction theory, we formulate transfer as mechanism alignment and introduce a model-level transfer perspective that employs distributionally robust optimization for approximate transfer. Crucially, this framework provides certified intervals to quantify errors even without precise mappings or target data. Experiments on synthetic and real-world ecological datasets demonstrate that our method accurately covers true interventional query values, effectively resolving bound estimation difficulties in non-transferable scenarios. Consequently, this work enables reliable model-level generalization and facilitates rigorous assessment of approximate causal transfer where conventional evaluation is infeasible.

Abstract

Transporting a causal conclusion from a source study population to a target one is a fundamental problem in causal inference. The theory of transportability provides a criterion for when this is possible: given experimental data from the source and observational data from the target, it determines whether a target query is identifiable and does so completely; i.e. if the query can be transported, the criterion finds the exact formula. However, it works one query at a time and returns an expression rather than the value itself. It is also silent in two practically important regimes: when the query is not transportable and when no target data exist at all. To tackle both, we take a model-level perspective grounded in Causal Abstraction theory. Source and target share variables, graph, and interventions, differing only at a known set of mechanisms, which makes transportability a special case of same-level abstraction. Thus, instead of asking whether one query transports, we ask whether a single map aligns the source and target across their interventional behaviour. We characterise when such a map exists in both the Markovian and semi-Markovian settings; when it does, every target query transports at once. Our main contribution lies in the approximate case. When no exact map exists, the best approximate one still yields certified query intervals, recasting abstraction error as a quantitative notion of approximate transportability. We formulate model-level transport as distributionally robust optimisation over mechanism and environment perturbations of the unseen target and derive certificates for both challenging regimes: bounds for non-transportable queries, and guarantees under target-agnostic settings. We evaluate our framework on synthetic Markovian and semi-Markovian benchmarks and a real ecological dataset, and we show that the certified intervals bracket the true interventional query.

Problem

• Causal Transportability
• Causal Abstraction
• Approximate Transportability
• Target-agnostic Settings
• Distributionally Robust Optimization

Innovation

• Causal Abstraction
• Generalised Transportability
• Distributionally Robust Optimisation
• Approximate Transportability
• Certified Intervals