Recursive

Recursive <: AbstractIdentificationMethod

General-purpose identification method based on recursive, or conditional ignorability, assumptions.

This method assumes that, conditional on variables ordered before, a given variable (often interpreted as a treatment or shock) is as good as random. This enables causal interpretation of the identified shock as structural.

In macroeconomic applications, this is typically operationalized through a Cholesky decomposition in structural VARs (SVARs), but the method is not limited to this setting. It applies to any model in which recursive ordering (conditioning on other variables) can be used to justify identification.

Conditioning variables are those that are ordered before the treatment variable in the provided dataset.