Dear Sir Andrew 

Response to OSR’s review of population estimates and projections

Thank you for your correspondence of 28 July 2021 concerning the response to our review of the population projections and population estimates produced by the Office for National Statistics (ONS).    

As you are aware, ONS published its response to our review on 29 July which took the form of a letter, blog and a more detailed article setting out how it plans to address the findings of our review. We welcome the response and its focus on improving methods, enhancing communication and active engagement.   

We are encouraged by the steps ONS has already taken to engage with users, including the briefing session and the preliminary meeting ahead of the publication of the 2020 mid-year estimates that you refer to in your letter, which the Office for Statistics Regulation (OSR) had made CPRE aware of on 23 June. We will continue to encourage ONS to keep user engagement at the centre of its plans for addressing our recommendations.  

I have today written to Jonathan Athow, Deputy National Statistician, to welcome the response and to also urge ONS to make some quick wins in developing guidance and presentation of uncertainty ahead of the Census 2021 data being available. We are pleased to see ONS’s plans to review how existing work using administrative sources can provide insights into the student population ahead of being able to review Census data. I have encouraged ONS to publish regular updates on the emerging findings from this work.  

As you have highlighted in your letter, OSR has previously made a number of interventions around migration data. Most notably, in August 2019 we took the decision to remove National Statistics designation of the Migration Statistics Quarterly Report, including long-term international migration estimates. This sent a strong signal to the analytical community and the public more widely about our view of the trustworthiness, quality and value of these statistics. It is not the role of OSR to determine the methodology and data sources that are used by statistics producers.   

Our review of population estimates and projections produced by ONS referenced the historic challenge of estimating international outward migration, due to there being few and only partial data sources which do not provide a complete picture. This challenge has been exacerbated by the International Passenger Survey being paused in March 2020 as a result of the coronavirus pandemic. It has been positive to see ONS take a collaborative approach to tackling the challenges in measuring migration, population and the labour market during the pandemic and we welcome its commitment to continue to take a joined-up approach to addressing the opportunities to improve methods that were highlighted in our review.  

With regards to our ability to be more decisive around our findings for Coventry, we are not in a position to state what the ‘true’ value is. Our review found that the population estimates for some smaller cities with large student populations, such as Coventry, did seem to be inconsistent with local evidence. It is the role of ONS to develop the estimates, and our job to assess how reliable those estimates are. It is not, and cannot, be the job of the regulator to propose variant or competing estimates.  

You have asked whether OSR will ask ONS to withdraw its current population and household projections for Coventry and make a new and informed set of projections, using the 2021 Census and drawing on local data sources. ONS has stated its intentions to replace the existing statistics with the rebased set of population estimates, based on new intelligence from Census 2021, once they are available. It is not normal practice for individual data points to be withdrawn where there is no improved alternative. We understand that ONS will be working with local authorities to compare provisional Census data with local insight, as part of its quality assurance process.  

OSR’s focus moving forward is to support ONS to deliver its plans to address our recommendations to improve the public value and quality of population statistics. We would be very happy to arrange a meeting with you to discuss our remit and our role in monitoring these developments.  

Yours sincerely  

Ed Humpherson 

Director General for Regulation