Did higher-level apprenticeships replace the traditional kind, or did apprenticeships just shrink?
The apprenticeship levy took effect in April 2017. Since then the mix of starts has shifted sharply toward higher levels — but the share shifted more than the numbers grew.
In 2024/25, 18.58% of apprenticeship starts were Intermediate (level 2), 41.61% Advanced (level 3), and 39.81% Higher (level 4+).
In 2016/17, the last full year before the levy, 52.67% of starts were Intermediate, 39.94% Advanced, and 7.39% Higher.
The level mix has shifted sharply
39.81%
Higher-level share
2024/25, up from 7.39% in 2016/17
18.58%
Intermediate share
2024/25, down from 52.67% in 2016/17
17.1%
Degree-level (L6/7) share
2024/25, up from 14.8% the year before
Apprenticeship starts by level, England
Higher-level starts have taken a growing share of the total since the levy, while the Intermediate share has more than halved.
View data table
| Year | Intermediate (level 2) | Advanced (level 3) | Higher (level 4+) |
|---|---|---|---|
| 2006 | 69.09 | 30.86 | 0.05 |
| 2007 | 67.53 | 32.43 | 0.04 |
| 2008 | 66.07 | 33.89 | 0.08 |
| 2009 | 68.11 | 31.36 | 0.54 |
| 2010 | 65.86 | 33.66 | 0.48 |
| 2011 | 63.2 | 36.09 | 0.71 |
| 2012 | 57.39 | 40.71 | 1.92 |
| 2013 | 65.05 | 32.86 | 2.09 |
| 2014 | 59.67 | 36.36 | 3.95 |
| 2015 | 57.2 | 37.47 | 5.33 |
| 2016 | 52.67 | 39.94 | 7.39 |
| 2017 | 42.95 | 44.24 | 12.81 |
| 2018 | 36.5 | 44.42 | 19.08 |
| 2019 | 30.76 | 43.67 | 25.57 |
| 2020 | 26.18 | 43.08 | 30.74 |
| 2021 | 26.21 | 43.33 | 30.46 |
| 2022 | 22.63 | 43.88 | 33.5 |
| 2023 | 20.86 | 43.15 | 35.99 |
| 2024 | 18.58 | 41.61 | 39.81 |
England · Source: DfE, Apprenticeships — historical series: starts, participation by age, level, SSA · as of 2024/25 (latest full academic year)
Level 6/7 (degree-level) starts reached 60,350 in 2024/25, 17.1% of all starts, up from 14.8% the year before.
Share vs volume: the mix shift is not the whole story
Between 2016/17 and 2024/25, Intermediate starts fell -74.8% and Advanced fell -25.6% — both in absolute numbers, not only as a share. Higher starts grew 284.8%, but total starts still fell -28.6%.
Apprenticeship starts by level, England — absolute numbers
Intermediate starts fell by around three-quarters in raw numbers, not only as a share of a shrinking total — and total starts remain below where they were before the levy.
View data table
| Year | Intermediate (level 2) | Advanced (level 3) | Higher (level 4+) |
|---|---|---|---|
| 2006 | 127400 | 56900 | 100 |
| 2007 | 151800 | 72900 | 100 |
| 2008 | 158500 | 81300 | 200 |
| 2009 | 190500 | 87700 | 1500 |
| 2010 | 301100 | 153900 | 2200 |
| 2011 | 329000 | 187900 | 3700 |
| 2012 | 292800 | 207700 | 9800 |
| 2013 | 286500 | 144700 | 9200 |
| 2014 | 298280 | 181760 | 19770 |
| 2015 | 291330 | 190870 | 27160 |
| 2016 | 260650 | 197660 | 36570 |
| 2017 | 161390 | 166220 | 48150 |
| 2018 | 143590 | 174730 | 75060 |
| 2019 | 99220 | 140840 | 82460 |
| 2020 | 84150 | 138490 | 98810 |
| 2021 | 91520 | 151310 | 106360 |
| 2022 | 76280 | 147930 | 112930 |
| 2023 | 70840 | 146520 | 122230 |
| 2024 | 65680 | 147090 | 140730 |
England · Source: DfE, Apprenticeships — historical series: starts, participation by age, level, SSA · as of 2024/25 (latest full academic year)
“Replaced” is a share claim unless both charts agree
Higher-level starts grew both in share and in absolute numbers, so “higher-level apprenticeships grew” is well supported. But Intermediate starts did not simply lose ground to Higher — they collapsed in raw numbers too, and total starts across all levels are still below the last full pre-levy year. The honest summary is that apprenticeships shrank overall and what remained shifted upward, not that higher-level apprenticeships replaced lower-level ones on a like-for-like basis.
Level mix by year, England
View as table
| Academic year | Intermediate | Advanced | Higher | Total starts |
|---|---|---|---|---|
| 2016/17 | 52.67% (260,650) | 39.94% (197,660) | 7.39% (36,570) | 494,880 |
| 2024/25 | 18.58% (65,680) | 41.61% (147,090) | 39.81% (140,730) | 353,500 |
Full 19-year series (2006/07–2024/25) is in the chart above; this table shows the two years the levy comparison uses.
Who starts an apprenticeship
Most apprentices are not school leavers: 51.3% of 2024/25 starts were aged 25 or over.
Business, Administration and Law was the largest sector for apprenticeship starts in 2024/25, with 103,890 starts.
Starts, not people
Every figure counts APPRENTICESHIP STARTS, not people — one person starting two apprenticeships in a year (or in different years) is counted each time. This is DfE's own published unit, not a site calculation.
What this does not show
Why the levy changed the mix
DfE’s own release commentary treats the levy as one contextual factor among several, including the ongoing transition from apprenticeship frameworks to standards over the same period. This page shows what changed, not a controlled test of why.
Whether higher-level apprenticeships pay better
Published earnings-by-level data (DfE’s Further Education Outcomes release) does not show a clean higher-level-means-higher-pay pattern — Level 6 sits below Level 5 one year after achievement. A future pass on this page should carry that comparison directly rather than assert a monotonic relationship the data does not support.
Whether apprentices finish what they start
Completion rates are published separately by level and cluster fairly tightly across levels (roughly 55–61%, save Level 6’s outlier), which does not support a claim that higher-level apprenticeships are dramatically better completed.
Last verified 2026-08-31. Data: DfE, Apprenticeships — historical series: starts, participation by age, level, SSA.
Related pages
- In this sectionFrom classroom to workThe rest of this chapter — what school leads to.
- In this sectionThe university expansionThe other major post-16 route this chapter covers, over the same recent decades.
- Leaves this sectionEmploymentWhere the labour-market half of this evidence comes from.
- In this sectionEducationThe other chapters, and how this section is built.
Scope and definitions
England, not the UK
Education is devolved. Every figure on this page is England only. Scotland, Wales and Northern Ireland operate different frameworks with different legal definitions, and their figures are not directly comparable.
Population
These figures cover schools and staff in England as recorded in the annual censuses: the school census for pupils and schools, and the school workforce census for staff. They count institutions and employees, not children with any particular characteristic, and they are not the 0–25 population used in the SEND chapter.
The levy year
The apprenticeship levy took effect 6 April 2017 (Finance Act 2016 Part 6), so 2016/17 is the last full pre-levy academic year and 2017/18 the first full post-levy year. Every before/after comparison on this page uses those two years.