Members in attendance

Stakeholder Panel

• Dame Kate Barker (Chair)
• Jonathan Camfield (Lane Clark & Peacock)
• Richard Gibson (Barnett Waddingham)
• Simon Kirby (Bank of England)
• Jill Leyland (Royal Statistical Society)
• Tara Murphy (HMT)
• Ellie Price (HMT)
• Ian Rowson (Independent Policy Analyst)
• James Smith (Resolution Foundation)
• Geoff Tily (TUC)

Technical Panel

• Ian Crawford (University of Oxford)
• Rupert de Vincent-Humphreys (Bank of England)
• Peter Levell (Institute for Fiscal Studies)
• Jens Mehrhoff (German Bundesbank)

Secretariat

• Rifah Abdullah
• David Beckett
• Richard Harris
• Aimee North

ONS presenters

• Ian Boreham
• Mike Hardie
• Emily Hopson

ONS attendees

• Stephen Burgess
• Abi Casey
• Tanya Flower
• Chris Payne

1. Introduction and apologies

1. Dame Kate Barker (KB) opened the meeting and passed on apologies from members unable to attend.

2. Update on scanner data, local collection impact analysis and Northern Ireland rents impact analysis

1. Mike Hardie (MH) presented an overview of the changes the ONS hope to make as part of the transformation project such as inclusion of scanner data, move off legacy systems including methods changes, implementation of consumption segments and new Northern Ireland (NI) rents data. MH explained that while it will be feasible to go live with the grocery scanner data in March 2025, there will only be time to conduct a one-month parallel run and the ONS feel that this is an unacceptable level of risk given the complexity of the systems and data. The project would be ready for April 2025 however due to the fixed basket approach, the ONS are only able to incorporate changes in March each year, marginally missing this deadline equates to a whole year’s delay. Therefore, these data can only be included in March 2026. The ONS requested feedback from the panels regarding the delay in scanner data implementation but proceeding with all other changes and production of supplementary analysis of scanner data in the interim.
2. Panel members fully supported the delay of implementation of scanner data, agreeing with ONS that they would prefer to know the data has been through a much more thorough quality assurance process.
3. Aside from scanner data, ONS were still planning to go ahead with five important methods changes in March 2025, which Emily Hopson (EH) presented impact analysis for:
   1. moving locally collected price data into a new system (the Cloud platform);
   2. moving from a two-month to a one-month lag for imputing prices for non-comparable products;
   3. changing the aggregation structure of parts of the CPI and RPI calculation process to incorporate “consumption segments”, to prepare for introducing grocery scanner data in the future;
   4. where prices are missing in the middle of a year and need to be imputed, doing this using CPI methods rather than RPI methods;
   5. changing the imputation process for seasonal items by imputing the base prices rather than carrying them forward.
4. These impacts were compared with a synthetic baseline which was very similar to the existing published CPI and RPI data. For changes (1), (2), (3) and (5), the impact was very small.
5. The impact for (4), moving from RPI to CPI base price imputation, was more significant. A panel member noted the impact on the wedge between CPI and RPI, and another panel member pointed out this could be perceived to have a permanent effect on the rate of CPI inflation for a given set of economic conditions, which would almost certainly be of interest to users, including financial market participants and the MPC. The panel highlighted that this change might need additional explaining to users. KB suggested reaching out to a smaller group of panel members to collaborate on communications.
6. Tanya Flower (TF) explained why the change in base price imputation method was expected to affect CPI growth rates. The RPI index generally has higher monthly price changes and that by using the CPI index for base price imputation for CPI and CPIH, we would see a higher base price resulting in a consistently lower price change overall. TF highlighted to the panel that the data analysed were from a volatile period (last 5 years) resulting in a larger difference than what we might expect ordinarily. Another panel member asked whether there was any evidence of the larger effects from the 2021-22 period being concentrated in a small number of products.
7. Ian Boreham (IB) presented an update on the NI rents transformation from September 2024. IB explained the key differences between the Price Index of Private Rents (PIPR) and Index of Private Housing Rental Prices (IPHRP) data and why additional methodological considerations were needed when handling the NI data. Additionally, there is a data delivery lag of 2 months therefore advice is sought from the panel regarding what would be the most appropriate nowcasting method. Previously, GB growth rates were used to project the missing months, but they were found not to be a good predictor of NI data. The suggestion instead is to use the previous monthly relationship in the NI data to nowcast and any differences are expected to be extremely small. IB asked the panel if they are happy with the proposed nowcasting approach and if they had any concerns.
8. A panel member queried whether any analysis had been done to assess if there are any identifiable bias in the nowcasting approach that could be potentially corrected. IB confirmed that this was investigated and there is no discernible bias in the data. No other concerns or questions were raised and it was concluded that the panels are happy with the proposed changes.
9. A panel member requested more information around the cloud platforming security and highlighted the need for transparency. They suggested publishing ONS’ cloud service strategy to get ahead of any issues that might occur in the future. MH explained the layers of quality assurance the cloud platform system has undergone, including contracting specialists to assess the system’s robustness.
10. Another panel member raised concern around ONS’s reliance on an ongoing supply of scanner data. This was reassured by TF who also mentioned that this is partly mitigated by progress on the data acquisition side to recruit more retailers.
11. Regarding publishing the scanner data analysis as a supplementary publication, a panel member raised concern around the timetable proposed and the possible market implications. Another panel member highlights whether this might cause the public to lose trust in the headline figure regardless of the experimental label on the supplementary analysis.

Action:

MH to reach out a small group of panel members to discuss comms on the impact of base price imputation change on CPI and CPIH.

Action:

MH to check whether ONS already published this information.

3. Publication status of papers

1. The slides will be published at the same time as the minutes following official press release from the ONS regarding these changes on the 23^rd of January 2025.

4. Any other business

1. The Chair thanked Panel members for their contributions to the meeting and the ONS for their presentations.
2. The Chair asks panel members to send any further comments on Jill Leyland’s paper to her directly please.