Average Controlled and Average Natural Micro Direct Effects in Summary Causal Graphs
This work addresses challenges in causal inference for epidemiological and other real-world contexts where non-linear relationships and hidden confounding complicate effect estimation, offering incremental theoretical advances in non-parametric methods.
The paper tackles the problem of identifying average controlled and natural micro direct effects in causal systems using summary causal graphs, which are abstractions for complex dynamic systems with cycles and omitted temporal information, and provides sufficient conditions for identifiability even with hidden confounding, with necessary conditions shown in specific no-hidden-confounding settings.
In this paper, we investigate the identifiability of average controlled direct effects and average natural direct effects in causal systems represented by summary causal graphs, which are abstractions of full causal graphs, often used in dynamic systems where cycles and omitted temporal information complicate causal inference. Unlike in the traditional linear setting, where direct effects are typically easier to identify and estimate, non-parametric direct effects, which are crucial for handling real-world complexities, particularly in epidemiological contexts where relationships between variables (e.g, genetic, environmental, and behavioral factors) are often non-linear, are much harder to define and identify. In particular, we give sufficient conditions for identifying average controlled micro direct effect and average natural micro direct effect from summary causal graphs in the presence of hidden confounding. Furthermore, we show that the conditions given for the average controlled micro direct effect become also necessary in the setting where there is no hidden confounding and where we are only interested in identifiability by adjustment.