Between Models
Counterfactual Frontiers across Causal-Response Experiments
Research question
How far must a structural model’s causal response move before its monetary counterfactual changes sign?
Abstract
When do several structural models support the same policy conclusion? I compare the response distance needed to represent every declared family with the distance to the first policy-sign reversal. A positive gap gives a range in which family coverage and sign agreement coexist. The framework separates agreement on response fibers, the effect of additional evidence, and uncertainty in estimated responses. In a monetary application, the baseline four-family support has a positive gap. A lag-augmented local-projection refit gives negative gaps under three covariance specifications. Common-support comparisons trace the roles of response weighting and shock-path profiling. The positive target crossings are small, and a calendar re-estimation exercise produces gap values on both sides of zero. Each comparison is conditional on its response protocol, evaluated structural domain, and policy target.
In one sentence
Models can agree near the data and still disagree on the sign of a policy counterfactual once the estimated causal response moves far enough.