What is the Social Mobility Index measuring?
Jim is an Associate Editor at Wonkhe
Tags
The exercise, devised and compiled by London South Bank and published by HEPI, says that it compares the performance of individual English higher education providers by measuring the “social distance travelled” by graduates from each institution – as well as the proportion of graduates so transported.
It combines access, continuation and outcomes measures for undergraduates for all modes of study (except, this year, apprenticeships).
The problem is that while a university might be great and getting disadvantaged students in and getting them to the second year on the one hand, and getting them into a well paid job on the other, they may not be the same students.
In other words, what if the people who got in via the “poor door” are not the people who end up in the plum jobs? Medics tend to top up salary performance, for example – but few are looking at their med students as a rich source of examples of students from low-income backgrounds.
Don’t just take my word for it. When the IFS and the Sutton Trust had a run at a similar exercise a few years ago – albeit using different variables and slightly different weightings – it specifically highlighted the “considerable variation” in mobility rates across subjects within the same institution – noting that many institutions had courses (it meant subjects) both in the top and bottom 10 percent in terms of mobility:
This suggests policy might be more appropriately focused at specific combinations of universities and subjects rather than on universities as a whole.
Better would be an index that surfaced a subject-level multiplication of “inputs” and “outputs”, or even one that worked at individual student level. But outside of the impressive top few in the SMI tables (where even playing the averages couldn’t account for impressive numbers), further down the exercise becomes meaningless once the provider is large.
Over the past year or so I’ve been handling the admin on a death of a family member, and at various points I’ve had to have someone in one of the “professions” watch me sign something. Each time I’ve briefly wondered which provider they got their education in, and what sort of background they were from.
If we want society – and those “top jobs” – to be more equal, we’ll need measures that test how well we’re doing on that. The danger is that measures of this sort distort the self-perception and avoid tackling the need to regulate and improve both access and participation at least at subject level.
There are certainly a few peculiarities with the index.
Nick Hillman of HEPI comments as follows:
The fact that some relatively new and less prestigious institutions beat Oxbridge reminds us of the different contributions made by different institutions.
The ranking doesn’t, however, demonstrate this clearly. Two “relatively new and less prestigious institutions” (Oxford Brookes and Anglia Ruskin), for example, rank much less highly than Oxford and Cambridge. Other similar comparisons could be cited. If the ranking in split into two halves, older Universities and newer Universities are to be found in both halves but there appear to be proportionately more relatively old and selective institutions in the top half.
Consequently, drawing conclusions is difficult. Perhaps the ranking indicates some selective Universities are doing better than others in recruiting and supporting disadvantaged students. Perhaps it also indicates that some recruiting Universities may be offering low value added courses. But perhaps the ranking methodology itself just raises more questions than it answers.
You may have missed the purpose of the meaningful word ‘some’ in my quote.
I’d be delighted to chat with you about all this further if you’d like to get in touch. (As you’ve used a pseudonym, it’s impossible for me to contact you directly.)
The point of indices like this one is to get a debate going and to recognise that different league tables measure different things, most of which are important. So thanks for engaging with it.
Thanks for the feedback, Jim. One of my main aims when I first published the SMI in 2021 was to generate discussion around how we measure social mobility and also to move the discussion beyond stereotypical view that only certain institutions ‘do social mobility’.
The basis was to consider the size of the intake from lower socioeconomic groups and levels of continuation and outcomes for these same groups. In short, the idea that both moving many people a fair distance or fewer people a greater distance should be celebrated. It also gives institutions the chance to look at their own peer group and, where appropriate, share good practice.
I’ve always acknowledged that the model has limitations and finding some way to account for subject mix is certainly an improvement I would to make in the future. I’m not entirely in agreement with your argument that the SMI doesn’t account for ‘if the people who got in via the “poor door” are not the people who end up in plum jobs’ however, because the outcome measures of the Index do, in fact, only look at those students from IMD 1 and 2.