Members in attendance

• Mr Grant Fitzner (ONS) (Chair)
• Professor Ian Crawford
• Professor Bert Balk
• Mr Rupert de Vincent-Humphreys

ONS contributors

• Mr Stephen Burgess (ONS)
• Mr Chris Payne (ONS)
• Mr James McGregor (ONS)

Secretariat

• Mr James Wilkins (ONS)

ONS attendees

• Ms Abi Casey (ONS)
• Mr Peter Bailey (ONS)
• Mr Ian Boreham (ONS)

Apologies

• Professor Rebecca Killick
• Mr Peter Levell
• Dr Martin Weale
• Professor Paul Smith
• Dr Jens Mehrhoff

1. Introduction, apologies, and actions

1. Mr Fitzner opened the meeting and passed on apologies from members unable to attend.

2. Experimental methodology for regional CPI for Northern Ireland

1. Mr McGregor presented a paper circulated with Panel members on producing experimental consumer price inflation for Northern Ireland (NI), with the work in collaboration with NI’s Department for Economy (DfE), Consumer Council for NI (CCNI) and Northern Ireland Statistics and Research Agency (NISRA). The aim of the paper was to present the methods chosen for creating a regional CPI for NI and get the Panel’s input on the methods chosen.
2. Mr McGregor presented the experimental approach to Panel members, outlining the way in which the price collection has been boosted, how regional weights were constructed and how we can identify low quality component regional indices. Mr McGregor asked for feedback on each of these areas and for the Panels thoughts on the next steps for the project.
3. On the experiment regional weighting methodology, a Panel member shared the concern that the NI CPI may drift from the truth, and the effect a sudden rebasing could have on the series. The Panel member highlighted research signposted in the paper, that considered the use of a three-year rolling average. The Panel member suggested if three-year rolling averages were too volatile as the research suggests, a five-year rolling average could be investigated and possibly used. Mr Payne highlighted the possibility that the rolling average could smooth out longer-term effects but agreed the idea should investigated to balance trade-offs.
4. A Panel member suggested testing the assumption that changes over time in NI are similar to the UK. The approach recommended by the Panel member would be to take a region with a large enough sample to be confident in, then reduce the sample size to NI levels, next apply the same methodology to this larger region, and finally compare the regional indices.
5. A Panel member queried whether the difference in the dynamics of the UK and NI energy markets were driving the differences seen within the weights. The Panel member questioned whether the pattern presented was consistent across time due to the volatility of energy prices, which lends support to the use of rolling averages. The Panel member recommended exploring the time variation, to investigate the co-movements between the energy market and differences in consumption patterns.
6. A Panel member questioned by how much the NI as a region differs from the UK or from the Republic of Ireland. If different dynamics of NI are more like one region or the other, it may provide practical solutions for constructing a NI CPI. Ms Casey responded that this was being investigated with the collaborative partners.
7. Mr Fitzner questioned whether the answer to the problems presented would be to boost the sample size of the LCF to improve the consumer expenditure data, Mr McGregor agreed. Mr Fitzner stated it would be good to know what a full four country level LCF survey would look like in terms of sample size and the additional costs in advance of any review of financial surveys. Mr Payne stated some of the complexities in doing so.
8. A Panel member asked whether the differences in the absolute sterling expenditure values had been looked into, rather than comparing shares of the total at divisional level. Mr McGregor highlighted these haven’t been considered but could be looked into.
9. A Panel member highlighted that the time period considered when comparing weights was across 2022 when inflation was unusually high. They highlighted that the dimension is important, so it may be helpful to see the time series of NI and the rest of the UK to identify any similar patterns, rather than intrinsic differences in dynamics. Mr McGregor explained that the next step is to bring this time series data up to date.
10. A Panel member questioned what the response rate to the LCF survey in NI was, and if previous research done by ONS could help improve the response rate in NI to help with consumer expenditure sample sizes. Mr Payne provided context to current LCF rates, and challenges with NI LCF responses. Mr Payne agreed it would be interesting to determine NI specific LCF response rates. Mr Burgess highlighted the ONS drive to improve LCF survey response rates more generally.
11. A Panel member noted the large difference in the shape of the distributions before and after the boost to price collections in NI. Mr McGregor agreed and that the boosted sample seemed to reduce some of the volatility which is to be expected. Mr Payne added that the sample sizes are still relatively small, and that there may still be seeing some small sample size issues.
12. Mr Fitzner asked what published outputs and estimated could be expected. Mr McGregor stated the target of an October publication, where the content will be decided as Panel members feedback is taken on. Mr Fitzner highlighted the potential usefulness of seeing methodological papers for future reference along with published estimates.

3. Proposal to delay the implementation of COICOP18 in Consumer Prices Inflation

1. Mr Burgess introduced a paper circulated with Panel members and gave thanks to Mr Jenkins as author of the paper. The purpose of the paper was to propose the ONS’s provisional decision to delay the implementation of COICOP18 for the UK, subject to conversations with key Stakeholders.
2. Mr Burgess explained the background of the Classification of Individual Consumption by Purpose (COICOP) classification, which underpins household final consumption expenditure data. Mr Burgess explained that in 2018, a significant revision was made to the COICOP classification by the UN Statistics Division which involved an additional division and restructuring throughout the system.
3. On COICOP18, Mr Burgess stated countries which produce a Harmonised Index of Consumer Prices (HICP) are expected by Eurostat to implement the new classification in early 2026 into consumer price inflation measures. Mr Burgess stated that the ONS had planned to meet this timetable to be internationally compliant and follow best practice. The implementation of COICOP18 is complex as there are a number of updates across the ONS that would be needed. Due to organisational prioritisation, this project was put on hold. Mr Burgess highlighted that the rushed implementation of COICOP18 could pose risk to regular Prices production, and to current Prices transformation projects.
4. Mr Burgess presented the ONS preferred option to delay the implementation of COICOP18 in order for it to be completed comprehensively across the ONS. The provisional implementation date in consumer price inflation would be early 2028. Mr Burgess explained this would mean less risk for regular CPI production but may make research projects that compare the UK with other countries at the class or divisional level more challenging. Mr Burgess emphasised that headline CPI would not be affected.
5. To help negate potential reputational risk, Mr Burgess presented an alternative option which would involve modelling and the partition of existing numbers to calculate. However, Mr Burgess highlighted this would not be the preferred approach due to operational complications which lead to more potential risk and work for the division.
6. Mr Burgess explained that it would be possible to implement COICOP18 into the Purchasing Power Parities (PPP) which are produced as part of an OECD led programme. Several factors would mean there is less risk involved with updating the COICOP structure into PPP.
7. Panel members were happy to support the ONS preferred approach.
8. Mr Fitzner highlighted that the decision taken reflected the current ONS resource landscape and emphasised the aim to build back resilience and capacity within teams to address issues such as the one presented.

4. Publication status of papers

1. The NI CPI paper is to be published alongside the minutes.
2. The paper on the proposal to delay the implementation of COICOP18 will be published once a formal decision has been made following conversations with key stakeholders.

5. Any other business and date of next meeting

1. The next meeting will be held on Friday 27 September 2024 following a joint Panel session with the Stakeholder Panel.
2. A Panel member requested that the Technical Panel be included in conversations on seasonally adjusted CPI. Mr Fitzner agreed.
3. Mr Payne requested a meeting action for Panel members not in attendance to provide comments on the NI CPI agenda item through correspondence. Mr Fitzner agreed.

Action:

Panel members not in attendance to provide comments on the NI CPI agenda item through correspondence.