An extension to employment coding means that, for the first time, the government can cleanly separate out “key worker” roles or “STEM-MH” roles from other employment.
This seems a little close to an extension of the “low quality course” debate for comfort. And it’s all to do with a coding frame that already chucks up anomalous data.
There’s a hole in my SOC
If you’ve been digging into the detail of the Graduate Outcomes data, or trying to engage with the “highly skilled employment” arm of the “low quality courses” debate, you’ll know that the Standard Occupational Classification (SOC, specifically SOC2020) are a big deal.
It’s a way of fitting people’s job role into what for our purposes is a hierarchy – is what they do a “professional occupation” (major group 2) and thus “highly skilled”, or is it an “administrative and secretarial occupation” and thus only requiring a medium level of skill.
Though in higher education data we usually deal with only the nine “major groups”, these are aggregates of three additional levels of data. For example:
- Major group 3 is “Associate professional occupations”
- This contains “Culture, media, and sport occupations” (34)
- In which we find “Artistic, literary and media occupations” (341)
- And, at the highest level of resolution, ”Authors, writers, and translators” (3412)
That last one is still surprisingly broad – so there has been a demand from some SOC users for an additional level, the Sub Unit Groups (SUGs).
We call it madness
The Office for National Statistics, which owns this classification, has recently published a new extended SOC framework as a spreadsheet that contains all of these new SUGs. So, occupations within 3412 (Authors, writers, and translators) can now be coded to a range of SUGs:
- 3412/01 – Authors
- 3412/02 – Bloggers
- 3412/03 – Copywriters
- 3412/04 – Literary editors
- 3412/05 – Poets
- 3412/06 – Script writers
- 3412/07 – Technical writers
- 3412/08 – Translators and interpreters
- 3412/99 – Authors, writers and translators not elsewhere classified.
It’s difficult to argue with this – a technical writer clearly requires a different skill set, and a different qualification route, to a blogger? I mean, would a blogger have to wade into huge amounts of technical detail on coding specifications? Really?
You’ll recall – if not see Peter Brant’s superb explanation elsewhere on the site – that SOCs (and the designation of some large and diverse groups of jobs as “highly skilled” or “graduate level and some not) can be disturbingly arbitrary.
The skill level of each job is defined by “an approximation based on the time needed for a person to become fully competent in the performance of the tasks associated with a job” – which can include formal qualifications, work-based training, or experience.
And that there’s also another way of splitting these occupations – some roles are “graduate jobs”, others are not. A graduate job is defined as:
those occupations identified as graduate jobs by academics at the University of Warwick based on their assessment of whether or not people in those occupations “normally require knowledge and skills developed on a three-year university degree to enable them to perform the associated tasks competently.
There’s some – but not a complete – match between “highly skilled” jobs and “graduate jobs”, but the overlap isn’t complete.
A clean pair of heels
Previously, some poor sap had to figure out whether “Authors, writers, and translators” (as a group) is a highly skilled (or a graduate level) job by looking at a range of interview and demographic data to figure out what these people do, how long it takes them to do it, and whether most people doing the job have a degree.
Bringing in a new coding level makes it easier to add more nuance to this process – a “technical writer” may be a highly skilled graduate role, a “blogger”… not so much. For our purposes in higher education this makes it easier to make one of those value judgements about the usefulness of a course: she’s a technical writer and has thus made good use of her degree, he’s a blogger and has not.
One of the main drivers for the development of this course was a need to identify science, technology, engineering and mathematics, plus medicine and health (STEM+MH) occupations from non-STEM+MH. Given that the user engagement meeting participants included the Office for Students, DfE and HESA it is not hard to draw a line between this coding change and the development of policy around which degree courses should be supported and which they would rather not exist.
Toe to toe
Apart from perhaps census data at national or regional levels, we’ll never see data published at this level of resolution. The closest we generally get to see in HE (Graduate Outcomes) is the top level groups, with the top three of these used as a rough proxy for both highly skilled jobs and graduate jobs.
The introduction of these new SUGs allows us to see other binaries – STEM+MH or non-STEM+MH, for example – using a marker. So we may be told that a particular course at a particular provider gets only 12 per cent of graduates into STEM+MH roles while down the road the number is 60 per cent for broadly similar provision. We won’t (easily) be able to drill down to see what is going on, but you had better believe that policy decisions will be made (and funding will flow) based on these findings.
And all this relies on the exact words a graduate uses to describe their job being matched to a coding frame, and another assessment as to whether a role on this framework is STEM-MH or not.
So what’s going to happen? Assuming that we are not going to dedicate lecture theatre time to coaching final year students in describing their jobs in ways that helps make the course look better, and assuming that professional bodies don’t get into campaigning that “their” jobs should be classified in a particular way, this feels like another instance where a descriptive qualitative coding frame does not have the rigidity needed to support policy decisions that are based primarily in prejudice.