Correction to New Keynesian Policy Counterfactual Method
A new technical note identifies a calibration error in a widely used semistructural methodology for predicting policy outcomes.
A technical correction to the semistructural methodology used in New Keynesian modeling shows that a high-profile example of policy counterfactuals was not identified at its printed calibration. The original work allowed multiple structures to satisfy every condition of its identification step, meaning the model could not uniquely determine the coefficients required for its predictions.
This failure stems from a redundancy in the identifying restrictions. Two of the six restrictions coincided on equilibrium equations, leaving eleven conditions to determine twelve coefficients. This gap creates a mathematical ambiguity where the reduced form data cannot isolate a single structural truth, potentially undermining the reliability of the resulting policy simulations.
To fix this, the note derives a rank condition that is computable from the reduced form and the restrictions alone. This provides a concrete diagnostic tool for researchers to verify if a model is actually identified before running counterfactuals. A one-clause amendment to the original theorem restores the theoretical conclusion, while a correction to the counterfactual display fixes a misprint in the reported results.
For firms and institutions that rely on New Keynesian frameworks to forecast the impact of central bank pivots or fiscal shifts, this correction highlights the fragility of semistructural inference. If the identification step is not rigorously verified via the rank condition, the resulting policy projections may be based on one of many possible structures rather than the actual economic mechanism.
This creates a need for a broader audit of structural macro models used in institutional forecasting. When models are used to justify specific policy shifts, the inability to uniquely identify coefficients means the "counterfactual" is an estimate with an invisible margin of error.
Market participants should watch for a shift toward more transparent identification checks in macroeconomic research. The ability to compute these conditions directly from reduced-form data makes it possible to stress-test the validity of policy predictions before they are baked into strategic planning.