Dear Roger

Scottish Labour Market Statistics for Young People

I am writing to you following our recent compliance check of the Labour Market Statistics for Young People (16 to 24 year olds) published by the Scottish Government, against the Code of Practice for Statistics.

Labour market statistics are key economic indicators used by a wide range of users and subject to high user interest. The introduction of a Young Person’s Guarantee in Scotland, means there is also particular interest in rates of unemployment amongst young people. Estimates of youth unemployment in Scotland are produced using data from both the Annual Population Survey (APS) and the Labour Force Survey (LFS). Whilst the APS estimates are published with a greater time lag, the larger sample size means that these statistics are considered more robust and reliable estimates of youth unemployment in Scotland than those produced using the LFS.

There has been some confusion in the past as to which of the two data sources is the most appropriate to use, with LFS estimates sometimes being quoted inappropriately in place of more-reliable APS figures. In acknowledgement of this, Scottish Government decided to publish a separate bulletin for youth unemployment statistics using APS data, to reduce the likelihood of the data being misused. The new bulletin provides clarity for users on the appropriate use of the data, and is a positive response to concerns we raised in 2019.

Our review found a range of positive features that demonstrate the trustworthiness, quality and value of the new bulletin:

  • The context behind the new release is clearly set out for users on the first page of the bulletin – with the relative strengths and weaknesses of the APS and the LFS figures well-explained. The new format makes the distinction between the two data sources clear to support appropriate use of the data.
  • The bulletin is easy to follow and brings out the headline figures and messages clearly for users. There is a good balance between statistical commentary and the accompanying graphs.
  • The section on ‘Further information’ provides useful information on the data sources and methods. Concepts such as ‘confidence intervals’ and ‘statistical significance’ are summarised in simple terms, which are accessible for users with a wide range of experience and expertise. The page on ‘Concepts and Definitions’ is particularly helpful for less-expert users, as it provides clear definitions of the key terms used in the bulletin in one place.
  • We welcome the gender breakdowns that have been added to the latest release. We are also pleased to hear of plans to introduce further breakdowns by local authority area in upcoming releases – in response to user demand for more-granular statistics. This increased granularity of the statistics will enhance the value to those who use them.

We also identified some areas where further improvements could be made to aspects of the statistics:

  • The statistics team should consider how the statistics can be better presented to improve users’ understanding of how the labour market is changing over time. Currently, statistics on the unemployed, employed and economically inactive are presented largely in isolation, under separate sections in the bulletin. Explaining the relationship between these different groups better would tell a more holistic story of youth unemployment rates in Scotland. This was also a recommendation of a compliance check we conducted last year of Scottish Government’s monthly labour market trends release.
  • More should be done to better reflect uncertainty in the bulletin and help users understand the precision of the estimates. Statements such as ‘the unemployment rate for young people in Scotland was X per cent’ present the figures as absolute, which risks misleading users. We also highlighted this as an area for improvement in last year’s labour market compliance check. We are pleased to hear from the statistics team that it is developing its approach to communicating uncertainty in the release we reviewed last year, and would encourage it to adopt a similar approach for the youth unemployment bulletin.
  • We were reassured by the statistics team that robust quality assurance is carried out on the data, both by Scottish Government and the Office for National Statistics (who collect the APS data). However, there is currently no information given on quality assurance processes in the bulletin itself. The statistics team should consider how they can give users an indication of the quality of the data in the bulletin, or signpost to other relevant QA information to enhance user confidence and understanding.
  • We welcome the fact that the statistics team has engaged with policy colleagues who use the statistics, to make them aware of the new release and ensure that the data are being used appropriately. The team should also consider how it can widen its approach to user engagement, to understand the needs of other potential users outside of government, and establish whether any further changes to the statistics would add value to those users.

We appreciate your team’s cooperation with us on this review. Our labour market and welfare domain team will continue to engage with your team over the coming months to discuss progress.

Yours sincerely

Mark Pont

Assessment Programme Lead