BetaThis is a new service – your feedback will help us to improve it.
Methodology
Further education outcomes
Published
Last updated
See all notes (2)
Updated to clarify the learner cohort used for progression measures
Updated to note the impact of missing self-assessment data
Background
What are Further Education Outcomes?
The Further Education Outcomes (FEO) present statistics on the employment, earnings and learning outcomes of Further Education (FE) learners. This publication was formerly known as FE outcome-based success measures (OBSM).
The headline measure (the sustained positive destination rate) shows the percentage of learners staying in education, employment (or both) for at least two terms in the year after they achieved their further education qualification.
This publication uses matched administrative data from the Longitudinal Education Outcomes (LEO) dataset, linking together learning data with benefits, employment and earnings data to produce statistics on the employment, earnings and learning outcomes of FE learners.
The FE outcomes (opens in a new tab) were published for the first time as experimental data in 2014, covering outcomes for learners achieving their FE qualifications in academic year 2010/11. The statistics have been released annually since then. This year, data is based on learners achieving an FE qualification during 2021/22 and their sustained activity in the 2022/23 academic year. The current release includes revised estimates for academic years prior to 2021/22 to make use of enhancements made to the data since the last publication.
Why do we publish Further Education Outcomes?
We publish Further Education Outcomes to:
provide clear and comparable information on the success of colleges and further education providers in preparing their students for continued education, apprenticeships, or employment
encourage institutions to ensure their learners receive the necessary support to prepare for and take up education, apprenticeships or employment
provide prospective learners with information on the successes of learners that studied similar aims in progressing onto further learning, or gaining employment
The statistics in the publication are produced using the Longitudinal Educational Outcomes (LEO) dataset. The LEO dataset has been brought together by different government departments and is being used to improve the information available on a range of topics across different policy areas.
The LEO dataset links information about individuals, including:
Personal characteristics such as gender, ethnic group and age.
Education, including schools, colleges and higher education institution attended, courses taken, and qualifications achieved.
Employment and income.
Benefits claimed.
It is created by combining data from the following sources:
The National Pupil Database (NPD) held by the Department for Education (DfE).
Higher Education Statistics Agency (HESA) data on students at UK publicly funded higher education institutions and some Alternative Providers held by DfE.
Individualised Learner Record Data (ILR) on students at further education institutions held by DfE.
Employment data (P45 and P14) held by Her Majesty’s Revenue and Customs (HMRC).
The National Benefit Database, Labour Market System and JUVOS data held by the Department for Work and Pensions (DWP).
The employment data and earnings estimates cover those with P45 and P14 records submitted through the Pay As You Earn (PAYE) system used to collect Income Tax and National Insurance from employment by Her Majesty’s Revenue and Customs (HMRC). These figures have been derived from administrative IT systems that, as with any large-scale recording system, are subject to possible errors with data entry and processing. In addition, the data are primarily collected for the purposes of collecting taxes, so some data cleaning was necessary to improve the quality of any suspect employment records, such that the resulting data looks to provide a good reflection of an individual’s employment and earnings for the year. This data cleaning makes use of algorithms developed by researchers, and uses similar processes as documented in the research in Estimation of the labour market returns to qualifications gained in English further education (opens in a new tab).
Beginning in April 2013, the P45 reporting system was phased out in favour of the Real Time Information (RTI) system, which requires employers to submit information to HMRC each time an employee is paid. This system has now reached full deployment.
RTI offers substantial improvements to the P45 system in terms of data coverage, since employers must now provide information on all their employees if even one employee of the company is paid above the Lower Earnings Limit. The move to RTI means that data coverage is high for the tax years from 2016/17 onwards that are used in this publication.
We can not currently distinguish between part-time and full-time work in the LEO data. This is further discussed in “Calculation of measures - Earning measures”.
As well as employment data for those who pay tax through PAYE, the employment data additionally includes those who pay tax through self-assessment.
Self-assessment forms are completed by a range of people who for example are self-employed, have received income from investments, savings or shares and by people who have complicated tax affairs. A list of people who are required to complete a self-assessment return can be found at www.gov.uk/self-assessment-tax-returns/who-must-send-a-tax-return (opens in a new tab). We receive a self-assessment earnings dataset from HMRC, which contains variables on:
Earnings received through employment (PAYE)
Income from partnership enterprises
Income from sole-trader enterprises
Total earnings for the tax year from the self-assessment form.
Benefit data
Benefit data are taken from the underlying DWP payments systems and are supplemented by the information entered by Jobcentre advisers. The data therefore captures basic information accurately, but non-compulsory fields in either the labour market system or the payment system may be incomplete. Due to the size and technical complexity, these systems are not accessed directly, but at regular intervals, scans are taken that build up a longitudinal picture from repeated snapshots of the data.
Start dates are entered on to the system and are accurate dates of benefit payment, thus provide certain timing and duration of a benefit claim. However, while Job Seeker’s Allowance (JSA) dates have very few discrepancies, due to the way the data are scanned the end dates recorded for other benefits may diverge to some extent from the events they are recording. The potential discrepancy varies from up to two weeks for Employment Support Allowance (ESA) to up to six weeks for Incapacity Benefit (IB).
The Universal Credit Official Statistics dataset has been used to allow those claiming Universal Credit to be included in this publication. This is a relatively new dataset and we are continuing to work closely with colleagues at the Department of Work and Pensions to explore how best to use Universal Credit data in Further Education Outcomes.
The Universal Credit data is retrospectively updated each month when a new caseload snapshot is produced, but the full back series is not updated each time – one month is updated retrospectively and one new month is added.
Education data
The national pupil database
Data from the national pupil database (NPD) were used to calculate education history and to identify learning outcomes.
The NPD is a longitudinal database linking pupil/student characteristics (e.g. age, gender and ethnicity) to school and college learning aims and attainment information for children in schools in England.
Five administrative data sources used in compiling the NPD have been used to determine education history, namely:
Individualised learner record (ILR) covering English colleges, further education (FE) providers and Specialist post-16 institution (SPIs)
School census (SC) covering English schools. This includes state-funded and non-maintained special schools and pupil referral units (PRU)
Awarding body data for independent schools
Alternative provision (AP) census
Higher Education Statistics Agency (HESA) covering United Kingdom higher education institutions and English higher education alternative providers.
The matching of these databases was undertaken at individual level using personal characteristics such as name, date of birth and postcode.
Individualised Learner Record data
The key data source used to identify the learners in scope for FEO is the Individualised Learner Record (ILR) collection run by the Education and Skills Funding Agency (ESFA), which is based on data returns from FE education colleges and training providers, about learners in the system and the learning undertaken by each of them. The data sources section of this methodology document provides more background information about the ILR.
Other personal details fields have high completion rates although there is some use of defaults where information is not known and particular groups such as offender learners have information withheld.
The dates of learning can be assumed accurate to within a week. Key data fields are tied to funding therefore there is a strong incentive for providers to ensure the information returned is accurate.
Higher Education Statistics Agency data
In this release we have included data from alternative providers (opens in a new tab). This was first included in the 2022 release (learners achieving in 2019/20). Alternative providers (APs) are higher education providers who do not receive recurrent funding from the Funding Councils or other public bodies and who are not further education (FE) colleges.
Publication cycle
The publication cycle of the Further Education Outcomes spans four academic years from the beginning of the year in which learning took place to the publication being released.
The first two years allow for learning to take place and for learners to progress onto employment or learning destinations. During the third year data on learners’ destinations becomes available incrementally. It is only once all destination data has been released - almost two years after the end of the academic year in which the learning took place - that work on producing the statistics and publication can begin.
For a detailed breakdown of the publication process including data delivery schedule, refer to Table A below.
Table A - Publication data and production cycle
Time lags
All data used in this process are drawn from administrative sources, which take time to process and collate. Therefore, there are time lags between the reference period and availability of the dataset for analysis.
Benefit data taken from the National Benefits Database contain lags in completeness. At least 3 months is allowed for receipt of backdated claims and data are lagged by around 4 months. Data used in this publication is complete to the end of December 2023 and was released at the end of April 2024.
Employment data have cleaning rules applied, which identify old records when updated with new information. As new information can come through about a job after it has ended this is a source of constant change and historically the data has been considered complete after approximately six months due to retrospection. Some data is retrospectively updated later than six month period, and in these instances will be reflected in the following publication.HMRC started to implement Real Time Information (RTI) in April 2013, which provided more frequent feeds of employment and earnings data and reduced the lag of the P14/45 data used in this publication to 3 months. Self-assessment data covers the whole tax year and becomes available at the beginning of the following tax year, as a result the self-assessment data has a 12 month time lag.
Earnings data were less timely than employment data, and in the past, it has taken up to fifteen months after the end of the tax year for the data to be considered complete. HMRC started to implement Real Time Information (RTI) in April 2013, which provided more frequent feeds of employment and earnings data and reduced the lag of the P14/45 data used in this publication to 3 months.
Individualised Learner Record data are collated from returns by colleges with the provisional data collected to date generally published on a quarterly basis. Returns are not generally complete until up to six months after the end of the academic year, which runs from 1st August to 31st July. This publication uses data that covers the whole academic year, which became available from November 2023.
Higher Education Statistics Authority data are collated from returns by institutions and data for the full academic year are available approximately six months after the end of the academic year.
The Further Education (FE) Outcomes cover all age apprenticeships, and adult (19+) education and training learners that achieved an FE learning aim funded by the Education and Skills Funding Agency (ESFA), and all age Traineeships who completed their aim.
The underlying data files also contain data on 16 to 18 year old education and training learners, but these learners are only reported on in the ‘16 to 18 year olds’ and ‘Industry sections of employment’ sections of the publication.
For each learner their highest and latest learning aim is selected for the purposes of this publication, as such figures are expected to be lower than those presented in the ‘Further education and skills (opens in a new tab)’ (FES) National Statistics publication where each learning aim is reported on (see the FES methodology for further details).
Eligible learning is identified using the Individualised Learning Record (ILR), in particular the achievement status and end date fields to identify learners that achieved.
Adult learners are those that are at least academic age 19 in the year they achieve their qualification, i.e. in the case of learners achieving in 2021/22 this covers learners aged 19 or older on 31 August 2021. Outcomes are reported at a learner basis and learners appear once for each provider where they have achieved a qualification.
The measures do not cover learners funded through the Offenders’ Learning and Skills Service (OLASS).
Learners with multiple learning aims
Outcomes are reported on a learner, rather than a learning aim basis. In the case where a learner achieves multiple aims at the same provider within the academic year, outcomes are reported against their highest-level aim, unless they completed a traineeship in the year. Where a learner achieves two aims at the same level at the same provider, the outcome is reported against the most recently achieved aim. The hierarchy used to select between aims is below.
Highest qualification
Higher Apprenticeship
Level 4 (or higher)
Advanced Apprenticeship
Full Level 3 (including academic qualifications, e.g. A-Levels)
Other Level 3
Intermediate Apprenticeship
Full Level 2 (including academic qualifications, e.g. GCSEs)
Traineeships
Level 2 ESOL
Level 2 English and Maths
Other Level 2
Entry or Level 1 Digital
Entry or Level 1 ESOL
Entry or Level 1 English and Maths
Other Entry or Level 1
Unassigned
Most recently achieved aim (if achieving more than one at the same level)
Where a learner achieves more than one aim at the same level and on the same date at the same provider, outcomes are reported against the aim with the lowest aim sequence number (a unique number used when recording aims in the ILR).
Learners in scope for progression measures
Two progression measures that compare learning achieved or sustained learning outcomes to highest level of prior attainment are calculated for a subset of these learners in scope (namely, ‘achieving at this level for the first time’ and ‘progression for learner overall’). For the figures to be meaningful the department must have a learner’s entire educational history on record. As such, learners have only been included in this analysis when:
the learner’s full academic record is recorded in the NPD and ILR datasets, and
learners were born in or after 1988. This is as far back as the department’s data collection reaches, meaning the department would not hold the full educational history of learners born before this period and could not guarantee full coverage of their prior attainment.
The ‘progression from achieved aim’ measure which captures whether a learner has progressed into a higher level of sustained learning than the qualification achieved in the reporting year does not rely upon data relating to prior attainment, and so this measure is produced for the entire cohort of learners in scope.
Earnings measures
The earnings measures include estimates on the earnings outcomes of learners that have an earnings record on the P14 or self-assessment data (HMRC data), a record of sustained employment on the P45 (HMRC data) and no record of further study at a Higher Education institution in that year. These figures focus on learners that achieved a Full Level 2, Full Level 3 or Level 4+ qualification in academic years 2017/18 to 2021/22 and their observed earnings after training up to the 2022/23 tax year.
They show what learners actually earned post study and include learners who may not have been employed in the same sector in which they achieved their apprenticeship or training.
Benefit learners
Benefit learner status is determined using ILR data matched to DWP’s Customer Information System, rather than through any fields on the ILR itself. The DWP data are seen as more robust, and do not rely on the learner disclosing this information to their training provider. Learners are included in this measure if, on the day their training starts, they are claiming either:
Income Support
Job Seekers Allowance
Universal Credit - Searching for work
Universal Credit - Working with requirements
Universal Credit - Working with no requirements
Universal Credit - Preparing for work
Universal Credit - Planning for work
Employment and Support Allowance - Work Related Activity Group
If a learner achieves two eligible aims in the academic year, and those aims overlap (so that the learner is studying two aims for at least some part of the year) then the learner’s benefit status is measured the day of the earlier start date. For example if a learner achieved two eligible overlapping aims in the year 2014/15, and the highest aim started on 1 December 2014 but the other started on 1 July 2014, then their benefit status would be measured on the earlier date, on 1 July 2014.
Figure 1: Example of when benefit status is measured for learners with multiple overlapping aims.
Destination measures
Sustained employment
The sustained employment measure aims to count the proportion of learners in sustained employment following the achievement of their qualification. Employment destinations are produced by matching ILR data to HMRC tax records.
The definition of sustained employment is consistent with the definition used for 16-19 accountability. This looks at employment activity in the six month October to March period following the end of the academic year in which the learning aim took place. For 2021/22 achievers to be counted as in sustained employment:
A learner must have completed a self-assessed return for tax year 2022/23, or
A learner must be in paid PAYE employment in five out of the six months between October 2022 and March 2023.
A learner needs to be in paid PAYE employment for at least one day in a month for that month to be counted.
If a learner is employed in the five months between October 2022 and February 2023, but not in March 2023, then they must also be employed in April 2023.
Figure 2: The eight possible scenarios that would lead to a learner being classified as in sustained employment.
The measure allows for a one-month pause in PAYE employment to reflect that there may be more volatility in initial employment post learning. Where the pause is in March, activity in April is checked to see if it is a short pause or a more substantial break.
Sustained employment only
The sustained employment only measure reports on learners in ‘sustained employment’ excluding those that are also in ‘sustained learning’.
Sustained self-employment
If a learner is found in HMRC self-assessment data in the tax year following the achievement of their qualification they are flagged as self-employed and are subsequently counted in the sustained employment measure.
Sustained further/higher education learning
The ‘sustained further education learning’, and ‘sustained higher education learning’ measures aim to count the proportion of learners in sustained FE or HE learning, regardless of economic activity, following the achievement of their qualification. Learning destinations are produced by matching ILR data to ILR data (the following year) and Higher Education Statistics Authority (HESA) data, respectively.
The definition of sustained learning for both measures is consistent with the definition used for 16-19 accountability. This looks at learning activity in the six-month October to March period following the end of the academic year in which the learning aim took place. For 2021/22 achievers to be counted as in sustained learning:
A learner must be in further education training (sustained further education) in each of the six months between October 2022 and March 2023.
A learner must be in higher education training (sustained higher education) in each of the six months between October 2022 and March 2023.
A learner needs to be in learning for at least one day in a month for that month to be counted.
For this release the definition of sustained further education has been updated to include community learning aims, and therefore the proportion of sustained further education destinations may be slightly higher that previous releases.
Figure 3: The single scenario that would lead to a learner being classified as in sustained learning.
Sustained apprenticeship
The ‘sustained apprenticeship’ measures aims to count the proportion of learners who went on to study a sustained apprenticeship, following the achievement of their course. Learning destinations are produced by matching ILR data to ILR data (the following year).
The definition of sustained apprenticeships is consistent with the definition used for the 16 to 18 destination measures publication. This looks at apprenticeship activity in the year after the end of the academic year in which the learning aim took place. For 2021/22 achievers to be counted as in a sustained apprenticeship:
A learner must have had any 6 months consecutive apprenticeship learning.
A learner needs to be in learning for at least one day in a month for that month to be counted.
Figure 4: The single scenario that would lead to a learner being classified as in a sustained apprenticeship.
Sustained learning
The ‘sustainedlearning’ measure takes into account the ‘sustained further education’, ‘sustained higher education’, and ‘sustained apprenticeship’ measures. If a learner has met the criteria for one or more of these measures they will be included in the ‘sustained learning’ measure.
Sustained learning only
The sustained learning only measure reports on learners in ‘sustained learning’ excluding those that are also in sustained employment.
Sustained employment and learning
The sustained employment and learning measure reports on learners that were counted as being in both ‘sustained employment’, and ‘sustained learning’.
Sustained positive destination
The ‘sustained positive destination’ measure aims to count the proportion of learners with a sustained positive outcome, either into learning or employment (or both). For 2021/22 achievers to be counted as having a sustained positive destination, a learner must either:
Have a sustained positive employment outcome, or
Have a sustained positive learning outcome (including sustained apprenticeship), or
Be engaged in either learning (further education or higher education) or paid employment in each of the six months between October 2022 and March 2023.
Under the final scenario, learners may only ‘switch’ between learning and employment once. For example, if they are in learning for 2 months, then employment for 4 months they are counted as having a sustained positive destination. However if they are in learning for 2 months, then employment for 2 months, then learning 2 months, they are not counted as having a sustained positive destination.
Figure 5: The twelve possible scenarios that would lead to a learner being classified as having a sustained positive destination
Not recorded as a sustained positive destination
The ‘not recorded as a sustained positive destination’ measure has been provided to give an overview of the full cohort, including those learners that did not go onto a sustained positive destination. This measure encompasses three distinct groups:
Destination not sustained – learner had a positive destination, but it did not span the whole period necessary to be counted as sustained
Destination not sustained and in receipt of benefits– learner had a positive destination but it did not span the whole period necessary to be counted as sustained, and the learner was also in receipt of benefits during the destination reference period (see ‘Benefit learners’ section above for a breakdown of included benefits).
On benefits only
This measure indicates that a learner had no evidence of a positive destination but was in receipt of benefits during the destination reference period (see ‘Benefit learners’ section for a breakdown of included benefits).
Any learning (sustained and non-sustained)
The ‘any learning’ measure aims to count the proportion of learners who are in any learning, regardless of economic activity, following the achievement of their course. This uses the same data sources as the sustained learning measure, but looks across the whole of the following academic year to be counted as in learning:
A learner must be in learning (in either further education or higher education) in any of the twelve months between August 2022 and July 2023.
A learner only needs to be in learning for a single day in a month for the learning to be counted.
Figure 6: The scenario that would lead to a learner being classified as in learning.
No activity captured in data
The ‘no activity captured in data’ measure has also been provided to give an overview of the full cohort, and includes those learners who do not appear to have any learning or economic activity in the destination reference period.
Progression measures
A series of experimental progression measures are now included in the release, which use a learner’s educational history to compare the level of learning achieved in the reporting year (2021/22) with i) the highest level of learning achieved prior to this year and ii) the level of learning in their destination year (2022/23), for those who progress to sustained learning.
Two of these measures are only applicable to a subset of the learners included in the FE outcome-based success measures publication, as for the figures to be meaningful the department must have a learner’s entire educational history on record. As such, learners have only been included in the ‘achieving at this level for the first time’ and ‘progression for the learner overall’ measures when:
the learner’s full academic record is recorded in the NPD and ILR datasets, and
learners were born in or after 1988. This is as far back as the department’s data collection reaches, meaning the department would not hold the full educational history of learners born before this period and could not guarantee full coverage of their prior attainment.
Sustained progression from achieved aim
The ‘sustained progression from achieved aim’ measure compares the level of the highest and latest achieved aim in the reporting year, to the highest level of sustained learning the learner is studying in the destination year.
As a result, this measure is calculated as a proportion of learners who went on to a sustained learning destination following achieving their learning aim in the reporting year.
To be counted as ‘progressing from achieved aim’, the level of the learning aim a learner has gone on to study in the destination year must be higher than the aim they achieved in the reporting year.
For example, if a learner achieved a Full Level 2 learning aim in the academic year 2021/22, then they would be counted in this measure if they were studying for a Level 3 or higher aim in 2022/23.
For the purposes of this measure, Full Level 2, and Full Level 3 are treated as higher than Level 2 and Level 3 respectively.
Achieving at this level for the first time
This measure denotes whether the aim the learner achieved in the reporting year was the learner’s first time achieving at that level. As discussed above, it is calculated for younger learners born in or since 1988 only, for whom we hold a full educational record.
It is calculated by comparing the level of the sustained learning aim in the reporting year, to the highest level the learner has achieved prior to the beginning of the reporting year.
When the level of learning in the reporting year is higher than the learner’s previous highest level of attainment, they are counted as ‘achieving this level for the first time’.
For example, if a learner had previously achieved a Full Level 2 qualification, and in the academic year 2021/22 (reporting year) they were studying a Level 3 learning aim, they would be counted as ‘achieving this level for the first time’. Conversely, if the learner was studying a Level 2 or lower aim in 2021/22 they would not be classed as having achieved this level for the first time.
Sustained progression for learner overall
The ‘sustained progression for learner overall’ measure compares the level of learning of the aim the learner is studying for in the destination year to the highest level the learner has achieved prior to the destination year.
As a result, this measure is only calculated for learners who went on to a sustained learning destination following achieving their learning aim in the reporting year.
To be counted in the ‘sustained progression for learner overall’ measure, the level of the sustained learning aim a learner has gone on to study in the destination year must be higher than the highest level of learning they have achieved over the course of their educational history.
For example, if a learner achieved a Full Level 2 learning aim before the beginning of the 2021/22 academic year, then they would be counted in this measure if the learning aim they were studying for in 2021/22 was Level 3 or higher.
Table B: Comparison of progression measures
Earnings measures
This report presents the median annualised earnings of learners. The median is calculated by ranking all learners’ annualised earnings and taking the value at which half of learners fall above and half fall below. In addition to the median, the annualised earnings for the top 25% (or upper quartile) andbottom 25% (or lower quartile) are also presented to help users understand more about how earnings are distributed.
In the case of all earnings measures, the estimates only include that achieved a Full Level 2, Full Level 3 or Level 4+ qualification and have an earnings record on the P14 or self-assessment data (HMRC data), a record of sustained employment on the Real Time Information submitted to HMRC and no record of further study at a Higher Education institution.
Annualised earnings are calculated for learners that started or left employment part way through the tax year by adjusting their recorded earnings to the equivalent earnings had they been employed for the entire tax year. The PAYE records from HMRC do not include reliable information on the hours worked in employment so it is not possible to accurately distinguish between learners in full time and part time employment. Therefore, part time earnings are not adjusted to the full-time equivalent amount.
Where there are high levels of part time employment within a group of learners, the median annualised earnings will be lower as a result. This is the case for sector subject areas like child development and wellbeing where many of the employment opportunities are part time. It is important to note that the number of people in part time employment may be as much due to the preferred working pattern of the learners as what is being offered by employers.
For the first time, the earnings estimates include self-employment income recorded through the self-assessment tax system in addition to earnings from paid employment collected by the PAYE system. Self-assessment tax returns to not contain dates of self-employment, therefore it is not possible to annualise self-employment earnings in the same way that PAYE earnings are annualised. It is assumed that the earnings reported in self-assessment tax return relate to a spell of self-employment covering at least the whole tax year.
Where a learner has income from both sustained employment paid through PAYE and though self-employment, the earnings used for this graduate is the sum of their annualised PAYE earnings and their raw earnings from self-employment. It should be noted that a learner with a PAYE records (that does not reach the ‘sustained’ criteria) and a self-employment earnings record will be counted as being in ‘sustained employment’ but we do not include their earnings in the earnings calculation. This is to avoid the risk of annualising PAYE data that could be based on a very short earnings spell.
The full definition of how the destination measures are created is found in ‘Calculation of measures’, but this section provides a broad overview and illustrates how the measures relate to each other.
Destination
Definition
1
Sustained positive destination
Proportion of learners progressing into sustained employment, learning, or both in the academic year after achieving their qualification. [A combination of measures 3, 9 and 10]
2
Sustained employment
Proportion of learners in sustained employment in the academic year after achieving their qualification.
3
Sustained employment only
Proportion of learners in ‘sustained employment’ [2] excluding those that are also in ‘sustained learning’ [8].
4
Sustained self-employment
Proportion of learners found in HMRC self-assessment data in the academic year after achieving their qualification. A subset of those in ‘sustained employment’ [2].
5
Sustained further education learning
Proportion of learners in sustained FE learning, regardless of economic activity, in the academic year after achieving their qualification.
6
Sustained higher education learning
Proportion of learners in sustained HE learning, regardless of economic activity, in the academic year after achieving their qualification.
7
Sustained apprenticeship
Proportion of learners who went on to study a sustained apprenticeship, in the academic year after achieving their qualification.
8
Sustained learning
Proportion of learners who fall into one or more out of the 'sustained further education’ [5], ‘sustained higher education’ [6], or ‘sustained apprenticeship’ [7] measures.
9
Sustained learning only
Proportion of learners in ‘sustained learning’ [8] excluding those that are also in sustained employment [2].
10
Sustained employment and learning
Proportion of learners that were counted as being in both ‘sustained employment’[2], and ‘sustained learning’ [8].
11
Not recorded as a sustained positive destination
Provided to give an overview of the full cohort - including those without a sustained destination [12 and 13] and those on benefits only [14]
12
Destination not sustained
Proportion of learners who had a positive destination, but it did not span the whole period necessary to be counted as sustained
13
Destination not sustained and in receipt of benefits
Proportion of learners who had a positive destination but it did not span the whole period necessary to be counted as sustained, and the learner was also in receipt of benefits during the destination reference period
14
On benefits only
Proportion of learners who had no evidence of a positive destination but were in receipt of benefits during the destination reference period
15
Any learning
Proportion of learners who are in any learning (sustained and non-sustained), regardless of economic activity, following the achievement of their course.
16
No activity captured in data
Proportion of learners who do not appear to have any learning or economic activity in the destination reference period.
Progression measures
Progression
Definition
Sustained progression from achieved aim
Whether the level of sustained learning in the destination year is above the level of the highest and latest achieved aim in the reporting year.
Achieving at this level for first time
Whether the aim the learner achieved in the reporting year was the learner’s first time achieving at that level.
Sustained progression for learner overall
Whether the level of sustained learning in the destination year is above the highest level the learner had achieved prior to the destination year.
Age breakdowns use the learner's age on 31st August of the academic year in which the learning aim took place (i.e. the base year).
Some of the breakdowns in this release only cover young learners (for whom we have a full education historical record). Details on the reason for this can be found in ‘Learners in scope for the measures - Learners in scope for the progression measures’.
Benefit learner
A learner who has a benefit claim that overlaps the start date of their learning aim. See ‘Learners in scope for the measures - Benefit learners’ for more detail.
Provider type
FE learning can be undertaken via a range of FE providers: General FE Colleges (including Tertiary), Sixth Form Colleges, Schools, Special Colleges (Agricultural and Horticultural Colleges, and Art and Design Colleges), Specialist Colleges, Private Sector Public Funded and Other Public Funded (i.e. LA's and HE).
Provision
In this publication, FE provision is divided into Apprenticeships, Traineeships and Education & Training. Results are also presented separately for Community Learning.
Apprenticeships
Apprenticeships are paid jobs that incorporate on-the-job and off-the-job training leading to nationally recognised qualifications. As an employee, apprentices earn as they learn and gain practical skills in the workplace.
There are currently two types of apprenticeships: ‘frameworks’ and ‘standards’.
An apprenticeship ‘framework’ typically contains the following separately certified elements:
A knowledge-based element (the theoretical knowledge underpinning a job in a certain occupation and industry, typically certified via a technical certificate).
A competence-based element (the ability to discharge the functions of a certain occupation, typically certified via work-based assessed national vocational qualifications (NVQs)).
As part of these reforms, new employer-led apprenticeship ‘standards’ were introduced in 2014 and were initially known as ‘trailblazers’.
Apprenticeship standards are designed with employers to help ensure apprentices have the skills businesses need, and outline the skills, knowledge and behaviours (KSBs) required to carry out a certain job role. All apprentices must take an independent assessment at the end of their training to demonstrate the KSBs set out in the occupational standard.
New standards continue to be made available and have been replacing the older apprenticeship frameworks in recent years. All new apprenticeship starts will be on standards by the beginning of the 2020/21 academic year.
Education and Training is mainly classroom-based adult further education that is not classed as an apprenticeship or community learning. It can also include distance learning or e-learning. Contrary to the Education & Training statistics in the ‘FE and Skills’ National Statistics, it excludes traineeships and offender learning.
Traineeships
Traineeships were introduced in the 2013/14 academic year to provide young people with essential work preparation, English, maths and work experience to secure an apprenticeship or other work, and can last up to 6 months in duration. From August 2014, traineeships were available to young people aged 16-24 and prior to this only to young people aged 16-23.
Traineeships support social mobility by providing training for young people who need to develop their skills and experience to enter the labour market. They are aimed at those young people who are motivated to work but lack the necessary skills and work experience to gain an apprenticeship or other job.
Community learning funds a wide range of non-formal courses, from personal development through to older people’s learning, IT courses, employability skills, family learning and activities to promote civic engagement and community development. Courses may be offered by local authorities, colleges, and voluntary and community groups, and include activity targeted at deprived areas and disadvantaged groups.
Level of learning
In this release, data is aggregated by the level of learning in a broad and more detailed measure, as presented below. Other Level 2 aims refers to aims which are not Full Level 2, and excludes Essential skills. These level definitions have been aligned with the FE & Skills publication.
Provision Type
Level of Learning Group
Level of Learning
Education and Training
Essential Skills
Entry Level Digital
Entry Level English & Maths
Entry Level ESOL
Level 1 Digital
Level 1 English & Maths
Level 1 ESOL
Level 2 English & Maths
Level 2 ESOL
Below Level 2 (excluding Essential Skills)
Entry Level
Level 1
Level 2 (excluding Essential Skills)
Other Level 2
Full Level 2
Level 3
Other Level 3
Full Level 3
Level 4 and Level 5
Level 4
Level 5
Level 6
Level 6
Other
Other
Apprenticeship
Intermediate apprenticeship
Intermediate apprenticeship
Advanced apprenticeship
Advanced apprenticeship
Higher (Level 4) apprenticeship
Higher (Level 4) apprenticeship
Higher (Level 5) apprenticeship
Higher (Level 5) apprenticeship
Higher (Level 6) apprenticeship
Higher (Level 6) apprenticeship
Higher (Level 7+) apprenticeship
Higher (Level 7+) apprenticeship
Traineeship
Traineeship
Traineeship
Essential Skills
Essential skills, formerly known as ‘Basic Skills’ and ‘Skills for Life’ are designed to give learners the necessary basic reading, writing, maths, digital and communication skills for everyday life, to operating effectively in work and/or succeeding on other training courses. Having basic digital proficiency and achieving a level 2 qualification in both English and maths is often required for further study, training and skilled employment.
Essential skills cover GCSEs, Functional Skills, Adult Basic Skills Certificates, including English for Speakers of Other Languages (ESOL) Certificates, and Qualifications and Credit Framework Certificates and Awards in English Maths and Digital.
Qualification / aim
The qualification data have been aggregated by the qualification level and either the standard/framework title (for Apprenticeships) or the aim title as recorded on the Learning Aims Database (opens in a new tab).
Geography
This information is collected in the ILR and is based on the postcode of the learner. For example, Local Authority District breakdowns contain data on learners residing in the Local Authority District.
Mayoral Combined Authority funding
This data is based an annual data return submitted to DfE by Mayoral Combined Authorities and the Greater London Authority and is used to identify aims funded using devolved budgets.
The ILR student records are matched to DWP’s Customer Information System (CIS)[1] using an established matching algorithm based on the following personal characteristics: National Insurance Number (NINO), forename, surname, date of birth, postcode and sex.
Some of these characteristics are simplified to make the matching process less time-intensive and allow more matches, for instance if a surname misspelt in one of the datasets. Only the first initial of the forename is used, the surname is encoded using an English sound-based algorithm called SOUNDEX[2] , and for most matches only the sector of the postcode is used.
All records accessed for analysis are anonymous so that individuals cannot be identified. The personal identifying records used in the actual matching process are accessed under strict security controls.
There are five match processes carried out, ranging from the highest quality and most likely to be accurate (Green) to the lowest quality and most likely to be a false match (Red-Amber). Table B shows the criteria for each match type.
Once the ILR records have been matched to the CIS the corresponding tax and benefits records for that individual can then be linked to their ILR record.
Table C: Criteria for each type of match
[1] The CIS is a computer system used by the Department for Work and Pensions to store basic identifying information about customers and provides information on all individuals who have ever had a national insurance number
[2] SAS function that turns a surname into a code representing what it sounds like, which allows some flexibility for different spellings. For example Wilson=Willson
Coverage and robustness of the data
The measures are calculated using administrative data sources already held by the Government, placing no additional burden on providers, individuals or employers to collect new information. Learner records are linked to DWP and HMRC data to observe benefit and employment activity, and to other education datasets to observe learning activity before and after training. Around 99% of learners are matched to DWP or HMRC data, rising to almost 100% for apprenticeships, and so provide representative coverage of activity for FE learners.
It should be noted that the match rate would never be expected to reach 100% for a number of reasons including inaccurate recording of personal information in the datasets used in the matching exercise, and the fact that LEO data on learners aged 80 or above is not included.
Table A: Match rates to LEO dataset by year
Academic Year
Provision
Number of learners[1]
Match rate to LEO dataset
2021/22
All learners
785,120
98.7%
Apprenticeship
113,640
99.9%
Education & Training
648,790
98.5%
Traineeship
12,700
99.0%
2020/21
All learners
798,790
99.1%
Apprenticeship
138,850
99.9%
Education & Training
647,920
99.0%
Traineeship
12,020
98.9%
2019/20
All learners
799,080
99.0%
Apprenticeship
145,520
99.9%
Education & Training
643,680
98.8%
Traineeship
9,870
99.0%
2018/19
All learners
1,004,480
99.0%
Apprenticeship
185,010
99.9%
Education & Training
807,070
98.8%
Traineeship
12,400
99.2%
2017/18
All learners
1,085,390
98.9%
Apprenticeship
275,970
99.9%
Education & Training
795,330
98.6%
Traineeship
14,100
99.4%
[1] Number of learners may differ to the figures provided in the Further education and skills National Statistics publication as the Further Education Outcomes reports only a learner's highest and latest aim.
The Code of Practice for Statistics requires us to take reasonable steps to ensure that our published or disseminated statistics protect confidentiality. Where appropriate we apply disclosure control to protect confidentiality.
Data has been suppressed:
For destination measures where the row count for number of matched learners is less than 5.
For earnings measures where the row count is less than 11.
For rates with a numerator less than 3, or denominator less than 6.
The following symbols have been used in this publication:
( 0 ) zero
( c ) small number suppressed to preserve confidentiality or for accountability reasons
( z ) data is unavailable
Data on the number of learnings or earnings has been rounded to the nearest 10, and destination rates are rounded to the nearest percent.
The Further Education (FE) Skills Measure is a type of value-added indicator designed to assess if a provider’s progression rate is higher or lower than expected in relation to learner progression to sustained employment and/or higher learning. It has been designed to account for as many factors as possible that may impact learner progression, that are outside of the provider’s control. The output is a provider “score” which can be interpreted as a change in the expected progression rate that is driven either by the provider or some factor that it has not been able to account for.
Coverage
This measure looks at the progression of learners in FE aged 16+ who achieve a qualification in a General FE College, Sixth Form College, Designated Institution, and local authority FE provider in receipt of £1m or more annually from DfE or devolved funding. This means learners in Independent Training Providers, Higher Education Institutions offering FE provision, 16-19 academies and school sixth forms are excluded from this measure.
All qualifications taught in these providers are in-scope, such as:
Technical/vocational
Academic (i.e. GCSEs, A-levels)
Apprenticeships
Below level 2 qualifications (including ESOL)
Community learning
A small number of qualifications (around 1%) are removed from the dataset if they do not map to a qualification type and do not have a level assigned.
Data
The same data as the Further Education Outcomes publication is used. This is individual level data for learners and includes a flag for those that progress to different outcomes (e.g. sustained employment or learning), as well as other learner characteristics (e.g. benefit status).
GCSE prior attainment, FSM, and SEN status are also included in the Skills Measure dataset. This comes from the National Pupil Database (NPD) which is linked to the Individualised Learner Record (ILR) and Higher Education Statistics Authority (HESA) data to detail students’ attainment throughout their educational history. The resulting dataset includes attainment records and pupil characteristics from schools. The cohort with these records is limited to learners born in 1988 or later.
The Index of Multiple Deprivation deciles, local authority unemployment rates, and Urban/Rural indicators are also used.
Method
Logistic regression modelling
The outcome variable is learner progression to sustained employment and/or higher learning, in the year after achievement. This is a binary outcome of 1 if a learner progresses to this destination or 0 if not.
The aim is to determine how far a provider may impact learner progression, or get as close to this as possible. To do this, logistic regression modelling is used to account for different learner and contextual factors that may impact the likelihood of a learner’s progression. The output of the logistic regression modelling includes predicted probabilities of the likelihood of the progression outcome occurring, after considering the relationship that the outcome has with different independent variables.
The outcomes of learners aged 16-18 are modelled separately as there are differences in the patterns of their progression compared to learners aged 19+. The data is limited for learners born before 1988 as these learners do not have linked school records in LEO, so any data from the National Pupil Database (e.g. GCSEs, FSM, SEN) is missing for them. For this reason, learners born in 1988 or later are modelled separately. For the latest output looking at achievers from academic year 2021/22, this means modelling learners aged 19-35 and 36+ separately.
The outcomes of progression to sustained employment and progression to sustained higher learning are also modelled separately, as the independent variables will have different relationships with both outcomes. To avoid double counting when aggregating the predicted probabilities at the provider level, a decision is made on how to categorise those who progress to both sustained outcomes (i.e. to employment and higher learning). Those learners who progress to both are categorised as having progressed to sustained higher learning and not sustained employment (flag=1 for the dependent variable for sustained higher learning and 0 for sustained employment). This means that in total, six different regressions are run.
Independent variables
The independent variables in the models are based on both sector knowledge and the literature. The variables differ slightly depending on the age of the learners, either due to data limitations (e.g. no GCSE or school data for learners born after 1987) or because it is unlikely to improve the model (e.g. the 16-18 regression does not include the interaction between age and qualification type, because it is unlikely that the impact of qualification type on progression will be different depending on if a learner is aged 16, 17, or 18).
Two interaction terms are included in the modelling: Qualification Type X Age and Qualification Type X Benefit status. These are included because it is likely that the impact of age and benefit status on progression varies depending on the qualification type that is studied. For the benefit status interaction, this means that while there is a significant link between being a benefit learner and progression, the size and direction of this relationship (e.g. positive or negative association) could vary depending on the qualification that the learner completes.
The independent variables included in the model are in the following table:
16-18 Employment
19-35 Employment
36+ Employment
16-18 Higher learning
19-35 Higher learning
36+ Higher learning
Sex
Sex
Sex
Sex
Sex
Sex
Age
Age
Age
Age
Age
Age
LLDD
LLDD
LLDD
LLDD
LLDD
LLDD
Level
Level
Level
Level
Level
Level
% FSM in provider
% FSM in provider
% FSM in provider
% FSM in provider
% FSM in provider
% FSM in provider
Learner IMD decile
Learner IMD decile
Learner IMD decile
Learner IMD decile
Learner IMD decile
Learner IMD decile
Benefit status
Benefit status
Benefit status
Benefit status
Benefit status
Benefit status
Qualification type
Qualification type
Qualification type
Qualification type
Qualification type
Qualification type
Benefit learner * Qualification type (interaction)
Benefit learner * Qualification type (interaction)
Benefit learner * Qualification type (interaction)
Benefit learner * Qualification type (interaction)
Benefit learner * Qualification type (interaction)
Benefit learner * Qualification type (interaction)
Learner urban/rural classification
Learner urban/rural classification
Learner urban/rural classification
Learner LA unemployment
Learner LA unemployment
Learner LA unemployment
GCSE group
GCSE group
GCSE group
GCSE group
SEN ever
SEN ever
SEN ever
SEN ever
FSM ever
FSM ever
FSM ever
FSM ever
Highest FE qualification in last 5 years
Highest FE qualification in last 5 years
Highest FE qualification in last 5 years
Highest FE qualification in last 5 years
Benefit learner in last 5 years
Benefit learner in last 5 years
Benefit learner in last 5 years
Benefit learner in last 5 years
Age * Qualification type (interaction)
Age * Qualification type (interaction)
Age * Qualification type (interaction)
Age * Qualification type (interaction)
Creating the Skills Measure
The regression models are run in RStudio using the glm function. The outputs of the model include a predicted probability of progression for each learner that are extracted using the predict(type=”response”) function. The probabilities are extracted for each of the six models, and the average is taken for each provider, within each regression. These expected progression rates are aggregated into one weighted average expected rate of progression to sustained employment and/or higher learning. There is a minimum threshold of 100 learners within each regression. This means that any expected progression rates created using fewer than 100 learners are not included. This number was chosen after seeing results that showed that the extreme outliers in the distribution of provider scores had fewer than 100 learners.
A simple example for one provider is below:
Regression
Expected progression
No. of learners in regression
Weight
Weighted expected progression
Overall expected progression
16-18 employment
60%
80
0 (under 100 so excluded)
0
23%+
47%+
3%+
7%=
80%
19-35 employment
70%
100
100/300 = 0.33
70%*0.33 = 23%
36+ employment
70%
200
200/300 = 0.67
70%*0.67 = 47%
16-18 higher learning
30%
80 (under 100 so excluded)
0 (under 100 so excluded)
0
19-35 higher learning
10%
100
100/300 = 0.33
10%*0.33= 3%
36+ higher learning
10%
200
200/300 = 0.67
10%*0.67= 7%
The expected progression rate is compared to the providers actual progression rate, and the difference between these is the final output for the Skills Measure – a score determining if a provider has a progression rate that is higher or lower than expected. This score may be driven by something the provider is doing differently, as the modelling aims to account for factors that influence progression that are outside of the providers control, however the score may also be driven by something unrelated to the provider that it has not been possible to capture in the modelling.
Most providers perform as expected, with scores close to 0. There are some providers on either side of the distribution of scores with very high or low scores. A case-by-case investigation would be needed to determine what is driving these scores.
The historical data prior to academic year 2021/22 have been revised as part of this publication, taking into account:
improvements made in the matching and processing of the administrative data sources, and
revisions to historic employment and earnings figures that may have changed retrospectively due to the addition of newer data
Revisions to methodology
A summary of methodological changes to FE Outcomes and timeline is included below:
Publication date
Cohort
Detail
November 2024
21/22 into 22/23
Expanded underlying data files to also include 16 to 18 year old education and training learners, although most reporting still excludes these from the reporting cohort. Added destination measures for learners moving into T level foundation year or T level aims. New data filters added to identify aims funded by Mayoral Combined Authorities, T level aims and aims funded by the Adult Skills Fund.
November 2023
20/21 into 21/22
Limited data to a five year rolling period. Updated benefits in scope for defining benefit learner. Expanded definition of cohort previously known as ‘young learners’ to include any learner born in or since 1988 rather than under 25 year olds only. Addition of community learning as a distinct destination. Update to definition of Essential Skills (formerly Basic Skills) to include new digital skills qualifications.
November 2022
19/20 into 20/21
Updated self-assessment data for 19/20 which now includes all records. New HE AP provider data and updated records for 2014/15. Change in definition to Level 2 and Level 3 classifications. New benefits groups. Addition of community learning in sustained further education destinations. Addition of CJRS data.
December 2021
18/19 into 19/20
‘Not recorded as a sustained positive destination’, ‘on benefits only’ and ‘no activity captured in data’ destination measures were updated to only be calculated for learners matched to LEO data. 'Progression from achieved aim' measure amended to include only sustained progression and to be calculated for all learners, rather than just young learners. ‘Progression for learner overall’ measure amended to only be calculated for young learners whose subsequent learning was sustained. Removed analysis on ‘Outcomes by learner category’.
November 2020
17/18 into 18/19
Addition of new headline measures ‘sustained employment only’, ‘sustained learning only’ and ‘sustained employment and learning’. Apprenticeship standards and frameworks split out. Correction to calculation of level of further education destinations.
October 2019
16/17 into 17/18
Introduction of experimental ‘progression measures’ based on the subset of young learners with full educational record. Reclassification of a number of FL2 and FL3 to L2 and L3 impacts volumes of learners in these categories in Education & Training from 2016/17 onwards (see below for more detail).
October 2018
15/16 into 16/17
October 2017
14/15 into 15/16
Measures produced using LEO dataset for the first time. Introduction of ‘Outcomes by learner category’ analysis.
Full level 2 and Full level 3 methodology in 2016/17
In 2016/17 the number of qualifications classed as Full Level 2 and Full Level 3 were reclassified by the ESFA for the 19-23 entitlement, and to align with the 16-19 offer and recommendations in the Wolf Review of Vocational Qualifications (opens in a new tab).
The methodology change involved a number of Level 2 and Level 3 vocational qualifications no longer being classed as Full Level 2 or Full Level 3 for funding purposes, now being reclassified to Level 2 and Level 3, respectively.
Therefore, the number of learning aims (qualifications) designated as ‘full’ for 2016/17 onwards has decreased. The new methodology aligns more closely with the 16 to 19 Performance Tables in terms of the qualifications included.
