Dear all


I am writing to you following our recent review of the key Labour Market statistical reports for the devolved nations against the Code of Practice for Statistics. The statistics published by the Welsh Government, Scottish Government and Northern Ireland Statistics and Research Agency (NISRA) that we reviewed are:

The statistics have been considered as part of a wider review of labour market statistics, along with our assessment of UK employment and jobs statistics produced by the Office for National Statistics (ONS). I am pleased to confirm that all three labour market reports should continue to be designated as National Statistics.

Labour market statistics are key economic indicators which are used by a wide range of users and are subject to high user interest. This review focussed mainly on the quality and public value of the data, statistics and supporting information. We recognise that the outputs we have reviewed differ between the three countries in terms of purpose, each team’s access to the underlying data and the time window available to produce them. These factors have been considered as part of our recommendations.

In reviewing the labour market reports, we found examples of clear supporting methodology information, effective sign-posting and presentation of uncertainty, which we detail separately for each country later in this letter. We have identified some common areas for improvement across the three producer teams, which also correspond to areas for improvements we’ve highlighted in our assessment report of ONS’s employment and jobs statistics. The recommendations in this letter build on those we have made to ONS and we encourage all four countries to continue to work together to ensure that labour market statistics across the UK continue to provide the necessary insights. In order to improve the quality and public value of these statistics, the teams should:

  • Consider how the statistics can be better presented to help improve users’ understanding of how the labour market is changing over time. We found some examples where the key labour market measures are defined but the relationship between these groups of people (for example, the unemployed and the economically inactive) could be more clearly explained. We encourage the three producer teams to also work with ONS to develop a way to understand the flows of people into, out of and within the labour market.
  • Build on existing collaboration between all the producer teams, including ONS, to enhance the coherence of labour market statistics. We found strong evidence of effective cross-producer collaboration through regular meetings and steering groups. However, discrepancies between the Labour Force Survey (LFS) and the Annual Population Survey (APS) data currently present issues with coherence of data sources. Greater collaboration could support a consistent approach in presenting data from the LFS or APS respectively and in turn, lead to a better read-across between the different countries’ statistics. This will require leadership and coordination from ONS and is highlighted in our assessment report of employment and jobs statistics (para 2.5). It could also prove an effective part of finding a solution to address the concerns raised by Scottish Government and Welsh Government about the future funding for APS which is explained in the ONS assessment report (para 1.9).

Welsh Government

  • Key Economic Statistics is well presented and the narrative provides relevant context to the statistics. The section on ‘Key quality information’ is appropriately detailed and provides useful information on data sources and methods. The bulletin includes links to supporting documentation and StatsWales data tables throughout. This could be further improved by signposting relevant sections from the ‘Key quality information’ within the main bulk of the bulletin to help aid understanding.
  • The statistics team has presented confidence intervals for the LFS estimates which provides some context for the level of uncertainty associated with the data. However, these are relatively inaccessible, and the language used in the narrative presents the latest figures as absolute, for example “The employment rate in Wales was x%”. This is particularly important when comparing data across the four countries, where estimated differences are not always statistically significant. Welsh Government should improve the way uncertainty is reflected in the narrative, following the lead of ONS as recommended in our assessment report, for example referring to the latest figures as estimates.
  • We were pleased to hear from the statistics team about its plans to potentially introduce a new bulletin covering protected characteristics in the labour market, which is an area of interest identified in its 2012 user consultation. We encourage Welsh Government to keep published statistical development plans up to date and to ensure users are aware of progress being made against these developments.
  • The statistics team told us that some of the main users of Key Economic Statistics go straight to the data tables to find the information they require and not the bulletin by default. We encourage Welsh Government to find out how its users engage with the various statistical outputs to ensure they remain relevant to users.

Scottish Government

  • The Labour Market Trends bulletin is easy to follow and we welcome the improvements that have been made to the presentation of chart headings and footnotes. The bulletin signposts to the new quarterly youth APS publication, which was previously included in the monthly LFS bulletin, as well as a number of ONS pages relating to the LFS. To improve clarity of the statistics, Scottish Government should look to expand on the methodology information within the bulletin itself.
  • The statistics team told us that the process for producing the monthly bulletin has largely been automated to ensure the statistics can be published at the same time as the ONS release. As a result, the narrative in the bulletin focuses on the latest figures and the change on the previous quarter or year. We would encourage the statistics team to consider how to bring out more insight from the statistics to improve their public value.
  • We welcome the work Scottish Government and ONS are doing to ensure uncertainty is properly reflected in the bulletin, as part of our recommendation in the assessment of employment and jobs statistics, to help users understand the precision of estimates. For example, the statistics team should avoid presenting figures as absolute in the headline
    infographic such as “x% of people aged 16 to 64 were in employment” and instead refer to the latest figures as estimates.
  • We were pleased to hear from the statistics team about its ongoing user engagement and its plans for developing alternative products for accessing the data to complement the ScotGov open data platform and to meet a range of users’ accessibility needs. The statistics team also told us that its economic statistics development plan is being updated to cover a wider range of economic statistics than in previous years. We would encourage Scottish Government to increase the visibility of its developments by publishing updates and outcomes of user engagement to highlight the good work they are doing in this area and to keep users informed of their plans and progress.

Northern Ireland Statistics and Research Agency

  • NISRA’s Northern Ireland Labour Market report is engaging and comprehensive. The narrative is proportionate to the statistics and the ‘Further information’ section of the report is thorough in addressing the strengths, limitations and comparability of the data.
  • The presentation of uncertainty in the bulletin and supporting materials is effective, for example including statistical significance and confidence intervals of estimates, and we are pleased to hear the report is being used as a case study for presenting uncertainty by the Government Statistical Service’s Good Practice Team. To improve this further, NISRA should ensure comparisons between Northern Ireland and the UK also take into the account the level of uncertainty for the estimates.
  • The team carried out a user consultation of labour market statistics in 2019 and has published its planned developments in response. We would encourage NISRA to seek feedback on its progress against the developments and continue to collate feedback on its various statistical outputs.

We appreciate each of the teams’ willingness to engage with us in this review as well as the wider assessment process with ONS. We wish to thank them for taking on board our recommendations. Our labour market and welfare domain team will continue to engage with your teams over the coming months to discuss progress.

I am copying this letter to Melanie Brown (Welsh Government), Gayle Mackie (Scottish Government) and Cathryn Blair (NISRA), the lead statisticians.

Yours sincerely
Mark Pont
Assessment Programme Lead

Related Links:

Assessment Report: UK employment and jobs statistics (March 2020)

Assessment of the UK employment and jobs statistics (March 2020)