Rigid-Central Inflation in the UK: Replication, Tests and Two Extensions

The latest reading of inflation in central sectors with rigid prices (RC) can be reconstructed from public data (3.29% for August 2026 against the published 3.3%), and its gap to actual headline inflation is almost fully explained by item coverage and the aggregation method. However, the Bank of England's empirical claim that RC better reflects the output gap is not confirmed with public gap measures, and RC adds no confirmed predictive power in out-of-sample forecasts of future headline inflation. For a main composite measure of labour-market pressure, RC's in-sample same-quarter R² point estimate is higher than those of the baseline comparison indices (services, core, headline and CPI84) and of composition control indices, and nominal additional information remains in regressions that control for services. The stand-alone fit is, however, highly sensitive to whether 2022–23 is included, and in the current direct four-quarter-ahead forecasting specification adding RC to services yields no realised forecasting gain. For energy pass-through, point estimates of the pipeline increase are positive in all 18 lag specifications examined, but their size is sensitive to specification and sample, and the interval for the baseline specification includes zero. The quantitative paths are conditional on the model and on the assumed energy price path.

The record contains the paper (PDF) and the replication code (rc_study.py with a README). All results can be reproduced from public data: ONS MM23, QNA and LMS; FRED; and the RC weight file published by the Bank of England (August 2026 CPI vintage). Not peer reviewed.

JEL classification: C43, E31, E37, E52