While the recent IFS report attempted to model graduates’ potential lifetime earnings – and how prior attainment and subject of study help or detract from these – the Department for Education’s inhouse data team has been focusing on earnings in-year.
Building on the Longitudinal Educational Outcomes (LEO) data – familiar to the sector as a mash-up of HESA student records, HMRC taxpayer profiles, and DWP information on interactions with the welfare system – a new release of comparative annualised earnings outcomes (CAEO) by post 18 pathway builds in more detail on GCSE level attainment.
Don’t have a CAEO, man
DfE splits the 2012-13 GCSE cohort into quintiles based on their best 8 GCSE passes, as follows:
- Quintile 1 (lowest attainment): an average of four Cs and one E
- Quintile 2: an average of eight Cs
- Quintile 3: an average of four Bs and four Cs
- Quintile 4: eight Bs
- Quintile 5 (highest attainment): four A* and four As
We can then look at LEO-style median earnings aged about 27 (in 2023-24) alongside upper and lower quartiles, and the 10th and 90th percentile, and destinations (in terms of employment and/or further study) for each group. All of this is separated out by highest level of study, and – for level four and above – into three main subject groups (STEM, LEM [law, economics, business and management], and other).
Before we get into the interpretation, here’s what the data looks like:
You can choose which levels of qualification to view together: anything up to level three (the comparator), level 4 or 5 apprenticeships, other level 4 or 5 qualifications, and level 6 – the latter being split by subject group. The chart at the top shows the median (solid dot) alongside other parts of the salary range for each GCSE attainment quintile – the chart at the bottom shows the proportion of young people in a sustained destination. There are separate data sets for male and female.
The law of small numbers
First up, you’ll note by default I’ve not shown you apprenticeships or other level four and five qualifications (don’t worry, you can add them in via the filter if you want) – my reason for this is that in the mid-teens there were very few young people who went to to this kind of study after level 3. I’m always suspicious of small numbers in LEO-derived data, because small groups of people that do unusual things are likely to be weird in other ways, and therefore not representative of what would happen if a larger population suddenly got interested in foundation degrees.
The comparison that will make headlines is that the median young person with four Cs and one E in 2012-13 earned £24,500 aged 27 if they just did anything below level 4, and £23,900 if they did a level six qualification. That £600 difference will drive an entire industry of commentary, so it is worth digging into.
So, how many people with about four Cs and an E at GCSE were there in 2012-13? We are not told in the data (I’m going to go ahead and ask DfE if we can get that in future). But for now, I can tell you that 81.6 per cent of all state pupils in that year got five or more GCSEs at grades A* to C. So we would have to conclude there were not many.
Early specialisation
The lower end of this attainment scale is likely to contain nearly all pupils at special schools (just 3 per cent of that cohort hit that target, which actually makes me quite angry – only 44 per cent were even entered for five GCSEs), and others likely to be receiving specialist support within the mainstream. Special educational needs cover a diverse range of pupils – some flourish in higher education where their abilities often become apparent with support, others are unlikely to ever enter the job market. And everything in between.
The earnings presented in the CAEO release refer only to those who were in sustained employment. It doesn’t include anyone in further study at all, and the figure doesn’t control for the number of people who are not earning. What you will spot from my bar chart is that 43 percent of people in quintile 1 who didn’t go beyond level three had no sustained destination aged 27 – compared to 24 per cent who got to level 6.
So, if your GCSEs were a mess would we rather have a 57 per cent chance of earning a median salary of £24,500, or a 76 per cent chance of earning a median salary of £23,800? In answering this question, imagine that your school career has moved in and out of mainstream education, your parents have had to fight to get you an EHCP and a school place, and you have lately developed a passion and aptitude for film studies. Should you go to university on that basis, if that is what you want to do? Yes.
And for everyone else, are you likely to experience a financial detriment because of a decision to study at university? Yes, possibly – but it is unlikely, and could be based on a whole bunch of other things that happen before you hit 30. Off the top of my head, you could get married, start a family, and work part time (LEO still can’t understand part time by the way). You can figure out your own way to net all this out against the wider social and societal benefits.
Quintiles. Quintiles everywhere
We get those prior attainment quintiles now in grown-up LEO too. Here’s a plot showing the way more detailed subjects of study and prior attainment has an impact on median salary.
YAG is “years after graduation, you can choose from 1,3,5 and 10 with each choice giving you a range of academic years to select via the filter. There is separate data for male and female graduates, or you can look at them in total. Each coloured dot is a subject area, with the height showing median salary and the size showing the number of people who graduated in that subject in that year.
What strikes me is that, for most subjects, quintile five (very good GCSEs) earn more than other quintiles. What I’m reckoning here is that this might mean that better employment opportunities to those on the academic superstar pathway (high social capital, no history of disadvantage, support from parents through GCSE and A level, selective university), and as luck would happen I can test this.
This plot shows industry of employment (SIC, at a very fine level of granularity) showing median earnings (axis), numbers of students (via the dots), and prior attainment quintile (the lanes). You can filter by subject of study, year after graduation (we just see the latest available year for each by default), and SIC section (to make the chart a bit less crowded.
What’s lovely here is that the very best paid graduates a year after graduation are around 25 securities traders in quintile 1, who are pulling down a median salary of £95,500. I’ve no idea what they studied, or how this affected their career, but fair play to them.
A rummage around this chart will show you any number of anomalies, and a huge range of careers for given subjects for every level of prior attainment.
Traditional LEO
In case you thought I’d forgotten, here’s the traditional real terms equivalent charts, by years after graduation, level of study, and subject area. The upward trend in salaries overall since the pandemic continues, with particular boosts from public sector works who are beginning to receive pay rises and scale uplifted at last.
And the “by provider” version will be with us very soon, don’t worry.
