OeNB Freitagseminar with Johannes Binswanger

A Unifying Framework for Expectation Distortions in DSGE Models

OeNB Freitagsseminar with Johannes Binswanger, University of St. Gallen

Abstract
While various specifications of expectation distortions have been proposed for DSGE models, their typically non-nested nature makes it difficult to assess their relative empirical merit. We propose a unified framework that resolves this challenge and enables a simultaneous “horse race” between competing behavioral theories. Our approach models distortions as “phantom signals” – perceived but unrealized news about fundamentals. This maintains structural rationality – assuming agents understand the economy's structural equations – while nesting multiple linear distortions within a single estimation. We demonstrate the framework's identification power by focusing on diagnostic expectations and over-persistence bias, two prominent mechanisms selected because their similar tendency to generate over-reaction makes them particularly prone to observational equivalence. Applying this framework to a medium-scale New Keynesian model (Smets and Wouters, 2007), we successfully disentangle belief biases from objective shock persistence. For policy institutions, our framework offers a disciplined mechanism to systematically screen the diverse menu of behavioral specifications, selecting only those distortions that genuinely improve model performance.

Date
Friday, 30 January 2026 | Start: 11:00 AM | End: 12:30 PM                    

Venue
The event is planned both, online via Webex and onsite at the Oesterreichische Nationalbank, Otto-Wagner-Platz 3, 1090 Vienna