Published November 14, 2026 | Version v1

Exploratory Tests of the Correction Freeze Problem in a Household Panel: Routes 2 and 3 on the New York Fed Survey of Consumer Expectations

  • 1. ROR icon Phenikaa University
  • 2. Faculty of Nursing, Widya Mandala Surabaya Catholic University

Description

The Correction Freeze Problem (CFP) describes four routes by which beliefs can fail to correct while the world changes. Earlier exploratory tests of Routes 2 and 3 used national aggregates. This note takes both routes to individual respondents, using the public microdata of the New York Fed Survey of Consumer Expectations from June 2013 to January 2026. The rules of each test were written into a script, checked on simulated data, and fingerprinted before the test was run. Route 3 predicts that respondents who are less sure of their own judgment lean more on an outside reference. Taking the Federal Reserve’s 2% target as that reference and the spread of each respondent’s own probability forecast as self-uncertainty, the prediction fails clearly. Less certain respondents tracked inflation’s distance from 2% more closely, not less, in all three predictions and in every robustness check. Route 2 predicts that checking stays flat while the stakes look ordinary and rises once they look serious. Taking a revision of the inflation forecast as the visible outcome of checking, the prediction is supported. Among respondents, the chance of revising did not rise with inflation below 4% and rose by about 2.4 percentage points per point of inflation above it. The verdict holds at all thresholds and in all subgroups tested. It rests, however, on one episode, May 2021 to June 2023, and cannot be separated from a rise in news coverage that affected everyone at once. Route 2 has now passed three times, while the readings of Route 3 tested so far have failed four times.

Files

CFP_SCE_Routes2and3.pdf

Files (381.9 kB)

Name Size
md5:f7ce4c63ac56d5f828012890dfe14d3f
381.9 kB Preview Download