Why we need to treat statistics about higher apprentices five years on with caution

Oh, look. Another thinktank report that suggests half of all graduates would be better off if they hadn’t bothered with the lecture halls and gotten themself what their lairy uncle would describe as “a trade”.

Well near enough. One of the headline findings of the Centre for Social Justice’s Rewiring Education is that, after five years, higher-level (L4) apprentices can make £12,500 more than people with a “low value degree” and £5,000 more than an average graduate (The Times has the splashier take).

Can they, though?

As is often the case with such statements, there is an element of truth. Some higher level apprentices have better salaries than the average graduates. And to learn more about them we follow the report’s lead in diving into FEO data (it’s like LEO data, but for FE and skills style stuff).

The most recent release of FEO data was in November 2024, covering people who completed their studies during the 2021-22 academic year (or more precisely, achieved their qualification aim before 5 April 2022). If we are looking for apprentices who qualified aged 19-24 five years after they finish we are limited to the 2017-18 cohort.

But I really can’t go much further before I tell you that there were just 1,810 of them.

The majority (1,190) of these L4 apprentices studied in business, administration, and law disciplines. The biggest chunk of this group (590) studied accounting and finance. And the biggest chunk of them (400) were training to be accountants.

Here’s a quick visualisation.

[Full screen]

I don’t think you can get a level 4 apprenticeship in data methods for social sciences, but if you had one I bet you’d immediately spot that these are very low numbers, and that there is quite a lot of variation based on what you study. Some careers pay more than others, and some apprenticeships lead more directly to those careers than others.

And these are issues we are used to coming up against in regular LEO. Here is an ostensibly similar plot from the most recent release of LEO (covering 2016-17 graduates five years on from graduation).

[Full screen]

You’d be tempted to jump straight to looking at salaries by subject area – but before you do check out the grey bars (which on both charts show the number of qualifiers who have their data within the medians and quartiles shown). While the overwhelming majority of level 4 apprentices study business, and a big chunk of them study finance and accounting – the picture for first degrees is very different.

Quite a lot of graduates choose subject areas where there is canonically very little money in associated careers – you know, vanity subjects like (say) nursing, teaching, or social policy that lead to low-paid public sector work. Or arts subjects. Now I like accountants but I feel like on a civilisational level artists are a pretty good thing too.

If we are comparing within the available data an average graduate with an average level four apprenticeship – the latter is more likely to work in financial services. And that’s a world that pays pretty well. So that’s our first clue that doing so is not a fair comparison.

But even so if we have a big population and we are comparing them to a small population we face other problems (this is module two of my data methods for social sciences l4 apprenticeship). What if the members of the small population are weird?

If you were completing a L4 apprenticeship in 2017-18 aged 19-24 you were very weird – it was a very unusual thing for someone in your cohort of learners to be doing. Simply put, at that point apprenticeships at higher education levels were a very new thing. Degree apprenticeships (everything L4 and above) were first introduced in 2015, while the levy by which most of them are funded was introduced only in 2017 (prior to this, employers have to stump up big chunks of the cost of running degree apprenticeships directly).

If you are choosing a young person to fund through training – and your decision (why that person) is backed by your own company’s money you are going to be pretty choosy – and you’d have to be sure that the company would see a benefit. For that reason it is very likely the company involved in the apprenticeship would employ and retain the person in question after qualification – and it is very likely that you had some knowledge of this person beforehand.

Are both these things still true? Difficult to say. I know apprentices that have been told “don’t expect we will definitely employ you afterwards” but that’s anecdote not data. The apprenticeships in the system now would have applied for their apprenticeship rather than been offered it directly. I’d hazard that this, plus the larger numbers involved, suggest they will be less likely to be offered a job at the end.

Indeed, the modern higher apprenticeship (apply via UCAS, have teaching delivered and assessments performed by a higher education provider) feels quite a lot like a degree.