Scotland's Labour Market Insights: February 2025

Data and Methodology

Further to the Labour Market Trends publication that was released on 21 January 2025, this publication provides more detailed trends and analysis for topics of interest from a wider range of labour market data sources for Scotland.

Figure 15: Measures and data sources presented in this publication

Infographic showing measures and data sources included in this publication for people in work, people not in work and people moving into work.

Note that the majority of participants receiving employment support were unemployed or inactive before joining No One Left Behind, however a small proportion were employed when starting.

The statistics in this release are Official Statistics or Official Statistics in Development. Information on the data sources is given in the Data sources section.

Where can labour market data for Scotland be accessed

The data contained in this release can be obtained from the following sources:

Table 2: Table of data sources and where the data can be accessed

Further labour market information for Scotland from the ONS Annual Population Survey is also published on Nomis

Data sources

ONS have published a Comparison of labour market data sources methodology document which compares data sources and discusses some of the main differences.

ONS Labour Force Survey

The Labour Force Survey is a survey of UK households collected and published by the Office for Nationals Statistics. Information are obtained from a sample survey and are therefore subject to some error. LFS estimates are currently classed as official statistics in development until further review.

ONS Annual Population Survey

The APS combines results from the ONS Labour Force Survey (LFS) with the English, Welsh and Scottish LFS boosts. This provides a larger annual sample of households. Compared with the quarterly LFS, the annual data is more robust.

The population totals used for the latest APS estimates use projected growth rates from Real Time Information data for UK, EU and non-EU populations based on 2021 patterns. The total population used for the APS therefore does not take into account any changes in migration, birth rates, death rates, and so on, since June 2021, so level estimates may be under- or over-estimating the true values and should be used with caution. Estimates of rates will, however, be robust.

ONS Annual Population Survey (APS) estimates have not been reweighted to the new population estimates used for the LFS. Consequently, all APS estimates remain weighted to the previous population totals, which will be inconsistent with those used for ONS Labour Force Survey (LFS) in the latest periods.

ONS have recently conducted analysis to assess the impact of falling sample sizes over recent years on the quality of APS estimates. ONS state that although the APS estimates are robust at National and headline regional level, there are concerns with the quality of estimates for smaller groups of the population, for example local authority level estimates.

This assessment of the APS estimates alongside the fact that APS estimates have not been reweighted to new population estimates has led ONS, in agreement with the Office for Statistics Regulation (OSR), to temporarily suspend the accreditation of all APS-based outputs. Further details on this can be found in the exchange between ONS and OSR:

Michael Keoghan to Siobhan Tuohy-Smith: Request to suspend APS accreditation – Office for Statistics Regulation (statisticsauthority.gov.uk)

Therefore, APS estimates are currently classed as official statistics in development until further review.

The longer-term solution remains the replacement of the Labour Force Survey with the Transformed Labour Force Survey (TLFS).

Labour Market Transformation

ONS are transforming the LFS. They are publishing Labour market transformation articles providing updates on the transformation of labour market statistics.

ONS also welcome any feedback on this latest update and their plans. Please email them at labour.market.transformation@ons.gov.uk to tell them what you think.

Scottish Governent are stakeholders in the transformation of the labour force survey. To provide any feedback on labour market statistics for Scotland, please feedback to LMStats@gov.scot

ONS Annual Survey of Hours and Earnings

The Annual Survey of Hours and Earnings (ASHE) is the official source of UK earnings estimates. Data are sourced from the Annual Survey of Hours and Earnings, first released on 29 October 2024 by the Office for National Statistics (ONS). The latest estimates relate to the pay period that includes 17 April 2024.

The Annual Survey of Hours and Earnings (ASHE) is the official source of UK earnings estimates.  ASHE data is published annually and is based on a 1% sample of the Pay As You Earn (PAYE) system. The ONS have released three statistical bulletins covering Employee earnings in the UK, Gender Pay Gap in the UK and Low and high pay in the UK.

From 2021, ONS have moved from Standard Occupation Classifications (SOC) 2010 to SOC 2020 for their occupation coding. This means earnings estimates for April 2021 based on SOC 2020 represent a break in the ASHE time series. Therefore, estimates pre- and post-2021 are not directly comparable.

In their October 2024 release, ONS have implemented improvements to the methods used for processing data returns to help address differences between ASHE and other sources of earnings data. Due to these methodological changes, data for 2023 and 2024 might not be directly comparable to data for 2022 and earlier years. For further details on the methodological changes implemented by ONS, please see the data sources and quality section of the ONS publication.

Other sources

This publication also contains HMRC PAYE Real Time Information (RTI) on median monthly earnings and payrolled employees. These are classed as Official Statistics in Development.

HMRC RTI

This release covers people paid through the HMRC’s Pay As You Earn (PAYE) system where their pay is reported through the Real Time Information (RTI) system.

Information presented in this release are experimental monthly estimates of the number of payrolled employees and their median earnings. It includes UK and geographical regions (NUTS1) early estimates of payrolled employment and median pay for the most recent month.

Statistics in this release are based on people who are employed in at least one job paid through HMRC’s PAYE system and the monthly estimates reflect the average for each day of the calendar month.

The publication and background information can be accessed on the ONS website.

Textkernel online job adverts

Textkernel is the source of the ONS online job adverts data. Textkernel online job advert data is collected using comprehensive web-scraping software which downloads job advert information from approximately 90,000 job boards and recruitment pages. The scraped data includes job titles, descriptions, posting dates and expiration dates. These describe location, salary, seniority, skill requirements, home/office working, and more. Textkernel perform some proprietary data cleaning to identify duplicate job adverts, which ONS have removed in this release. Duplication can occur when the same job is posted on multiple job boards, or when multiple recruiters advertise the job at the same time. Textkernel remove some adverts when they are low quality, e.g. if they are missing information, and remove adverts for jobs not based in the UK. Textkernel-based statistics are official statistics in development.

Reliability of estimates

Estimates from the ONS Annual Population Survey for October 2004 to September 2005 through to October 2023 to September 2024 are presented. For the latest time period, the sample size is around 8,900 households in Scotland.

The ONS Labour Force Survey (LFS) and Annual Population Survey (APS) are sample surveys. As such, these estimates are subject to an associated sampling error that decreases as the sample size increases. It is the nature of sampling variability that the smaller a group is the (proportionately) less precise an estimate is. The Office for National Statistics (ONS) have published a detailed guidance note on this. Shading is one method used by ONS to indicate where estimates should be used with caution.

Employees who were furloughed between April 2020 and the end of September 2021 were classed as employed, but temporarily away from work. This is consistent with labour market definitions outlined by ONS.

Confidence Intervals

Confidence limits can be used to assess the range of values that the true value lies between. 95 per cent confidence intervals for rates are included in all tables and charts.

What does the 95 per cent confidence limit mean?

If, for example, we have an APS estimate and confidence limit of 63 per cent ± 0.27, this means that 19 times out of 20 we would expect the true rate to lie between 62.73 per cent and 63.27 per cent. Only in exceptional circumstances (1 in 20 times) would we expect the true rate to be outside the confidence interval around the APS estimate. Thus the smaller the confidence limits, the more reliable the estimate is.

The confidence limits use a design factor of 1, which may not be likely in some cases but given the lack of further information an average design factor of 1 is assumed to be reasonable. Further information on estimating confidence intervals can be found in the LFS user guidance.

Statistical Significance

Statistical significance is based on 95 per cent Confidence Intervals. Statistical significance means that the change was large enough that it is unlikely to have resulted only from the variable nature of the sample.

Quality Assurance

Annual Population Survey data is collected and produced by the ONS.

When producing estimates for this publication, Scottish Government statisticians conduct in-depth quality assurance.

These checks include:

• analysis of the sample size obtained in the collection process
• production of estimates from the microdata using statistical software and relevant coding
• cross checking of coding between team members

Further checks relate to:

• crosschecking historical time series data with previously published results
• benchmarking the results against other relevant data sources