Participation of young people in education and the labour market: methodology note

The statistics published in our annual release ‘participation of young people in education and the labour market’ provide information on the learning activities and labour market status of young people aged 16 to 24 in Wales at the end of the calendar year. They are derived from a number of different sources which are described below. 

Each annual release provides final data for previous years and provisional estimates for the latest year. Provisional estimates use the best data currently available for each aspect of participation. These include some final data, some provisional data and some modelling depending on the different sources of data. Estimates for previous years take into account final data not available at the time when they were originally published.

The Participation of young people in education and the labour market release published in March 2025 included revisions to the estimates of young people not in education, employment or training (NEET), from 2012 to 2022: 

1. The population data used in the compilation of the figures has been revised from 2012 to incorporate rebased mid-year population estimates using Census 2021 data (Office for National Statistics (ONS)).
2. The methodology used to produce the Higher Education Statistics Agency (HESA) data has been updated from the 2022/23 academic year, impacting NEET estimates from 2022 onwards. The impacts of this methodology change on the estimates are small. Further information is available in the Higher Education section below.

The release also provides the definitive source for estimates of the proportion of young people who are NEET in Wales. Further information on this and the other sources of NEET statistics in Wales are available in our guide to understanding the different sources of statistics on young people not in education, employment or training.

School pupil numbers

Information is collected in January each year from the Pupil Level Annual School Census (PLASC). Due to the level of coronavirus (COVID-19) cases in January 2022, the 2022 census date was delayed to 15 February 2022. School closures between December 2020 and March 2021 due to the coronavirus (COVID-19) pandemic meant that the 2021 census date was delayed to 20 April 2021.

Ages are as at 31 August prior to the start of the academic year.  The census only requests a combined figure for those aged 19 & over - these figures are included for 19 year olds only. PLASC includes all maintained and independent schools. 

Any changes to the provisional PLASC data will be reflected in the final participation estimates in next year’s release. 

Higher education

Information regarding those at higher education providers (HEPs) is collected by the Higher Education Statistics Agency (HESA). This collection also included those studying further education courses at higher education providers up until 2016/17. From 2016/17, it also includes those studying at higher education level at further education colleges in Wales.

Prior to 2022/23, data was deduplicated based on a student’s unique identifier, such that only one record was retained per unique identifier. From 2022/23, data was deduplicated based on the combination of the student’s unique identifier and higher education provider, as students are no longer required to retain their unique identifier if studying at different providers. This means that students studying at more than one provider may be counted more than once.

Sex unknown and other have been aggregated due to known data quality issues associated with the sex identifier field in the HESA data.

For academic years prior to 2022/23, the analysis in this release is based on the 1st December population (DecPop). From 2022/23, the 1st December population is no longer directly captured in the HESA student record. Instead, from 2022/23, activity on the 1st December is determined by student course sessions identified as active or writing up at 1st December (SesActDec). These two methods differ somewhat due to changes in the structure of the HESA record. However, tests for the 2022/23 academic year using the base student population (the standard registration population) showed that using the SesActDec resulted in less than 1% fewer students than would have been expected if the DecPop were available for 2022/23, suggesting that there can be continuity between the two methods for time series purposes.

Ages are as at 31 August prior to start of academic year. Data relate to the total number of students, though students studying at different providers simultaneously may be counted more than once.

Provisional year end estimates are modelled using the Higher Education Students Early Statistics Survey (HESES), collected by the Higher Education Funding Council for Wales (Medr), which provides an early indication of the number of HE students studying in the latest academic year. These provisional estimates will be revised with final HESA data in next year’s release.

We also use figures for Welsh domiciled students studying at the Open University. These are "whole year" figures, i.e. they include enrolments throughout the year. The reference point for age is 1 January. Data for the latest academic year is provisional and will be revised with final data in next year’s release.

Further education and work-based learning

The Welsh Government collects information on enrolments at further education institutions (FEIs) and work based learning (WBL) via the Lifelong Learning Wales Record (LLWR).  

The analysis in this release is based on enrolments as at the week of 1 December. Data for the latest academic year is still under validation so the provisional estimates are based on an early freeze of the data. Where a learner has activities in more than one provision type, allocation to a single type has been made in the order: WBL at FEIs, HE at FEIs, other at FEIs, WBL at other training providers. Learners with both WBL activities at other training providers and learning activities at FEIs will be included under the activity at the FEI. These early estimates will be revised with final data in next year’s release. 

Population

The ONS calculates mid-year estimates of resident population by single year of age as at 30 June. For this release, end of year populations have been derived from the latest mid-year estimates and national population projections. These estimates have been adjusted to age as at 31 August.

The estimates in the latest release have been revised from 2012 to include rebased mid-year population estimates based on Census 2021, published by the ONS in November 2023.

Economic activity

Economic activity is estimated from survey data using proportions derived from the Annual Population Survey (APS), which is a household survey carried out by the ONS. The APS combines the boosted samples of the Labour Force Survey (LFS) and provides rolling four quarter data. These latest data cover the calendar year. 
Some of the estimates are based on small samples and therefore may have a large margin of error. As a result changes in trends in labour market participation need to be interpreted with a degree of caution, as changes could be attributed to sampling effects as well as real effects. Currently it is not possible to distinguish between these impacts. 

The APS has seen a fall in sample sizes over recent years and has not been reweighted to latest population estimates, resulting in increased volatility and uncertainty in estimates produced from the survey. It is still appropriate to use the statistics in the participation of young people not in education and the labour market release, however users should be aware of the increased volatility and uncertainty of the estimates.

The APS is used to estimate the following proportions: 

• Labour market status of those in full-time education;
• Labour market status of those in part-time education;
• Full-time and part-time employment of those in work-based learning, who are employed;
• Employer sponsored ‘off-the-job’ training for those in employment.
  Note that the employment status of work-based learners is derived from the LLWR rather than the APS.

These proportions are then applied to the numbers known to be in education, work-based learning and the total population to derive estimates of participation by education and employment. For work-based learners, the labour market status at the start of the learning programme collected via the LLWR is used with the addition of some APS data to estimate the proportions in full-time and part-time employment.  

This document below provides further detail on the methodology used to estimate participation by education and economic activity.