The same students at the same providers?

Did the increase in students placed during the 2021 cycle do anything for widening access? David Kernohan keeps it in proportion

David Kernohan is Acting Editor of Wonkhe

Has anything actually changed?

One of the enduring concerns around the 2021 UCAS cycle was that the upward shift in level three (A levels and the like) results meant that “popular” traditionally selective providers ended up taking on more of the kinds of student than they would generally admit anyway.

Your hardworking but not academically outstanding archetypal 18 year old from a nice home who – previously – would have missed their offer at Bristol and taken up an (excellent, I hasten to add) insurance place at UWE, held the same initial offer as they would have in previous years. The difference was that in 2021 they got the results Bristol needed, and ended up living in Clifton rather than Frenchay. Or so the story goes.

A level day jump in the air

In the data we would expect to see this as an uptick in the proportion POLAR4 quintile 5 placed 18 year olds. While POLAR is area based rather than individually based, we don’t get the detailed background information at provider level from UCAS that we do from HESA – so it remains our best proxy.

Looking at a ranked plot of the differences in the proportion of placed 18 year olds from POLAR4 quintile 5 between 2020 to 2021, we see that the biggest increase in placements among this type of 18 year olds was not at Bristol – indeed the proportion at Bristol fell very slightly. Indeed, the proportion rose at UWE – again very slightly.

[Full screen]

Keeping things in proportion

But hang on, I hear you cry. That slight increase at UWE represented 195 extra placed applicants from quintile 5. But the slight proportional dip at Bristol actually meant an extra 500 placed quintile 5 applicants. It’s counterintuitive, but this just confirms that the news is good. For 500 extra Q5 students at Bristol to be a drop, there must have been substantial recruitment outside of the usual suspects.

Indeed so, as we can see on this time series of POLAR proportions:

[Full screen]

The big proportional rise in the University of Bristol between 2020 and 2021 was in POLAR4 quintile 2 – a group traditionally less likely to attend a very selective provider like Bristol. The numerical rise was similar between groups (apart from a slight fall in quintile 1, down from a record high in 2020).

Compare the University of the West of England. A smallish proportional rise in quintile 1 and in quintile five placed 18 year old applicant, proportional falls elsewhere – but a rise in numbers overall.

In citing proportions or absolute values of placed students as evidence of success in providers reaching underserved areas we miss out a fair bit of context. To understand these changes we need to see the provider in context.

Tariff effects

I used UWE and Bristol, above, as a kind of lazy regional shorthand for provider types (mint and pink would, of course, be another way of describing this). There is a mindset that suggests that we could divide the higher education sector into:

  • Providers that serve groups that would not otherwise access higher education, and that have low tariffs for entry
  • Providers with high tariffs for entry, that serve primarily people who would traditionally have gone to universities (traditionally selective).

But does this hold up in 2021? Or in any other year?

Because we don’t get provider tariff data from UCAS (yet…), I’ve instead used data from the most recent Guardian University Guide. The team that prepares that dataset buys the information in, and there are documented issues about using this with Scottish providers (because of the strange way tariff points treat Scottish higher qualifications) and the subject mix can have a huge impact (a big medical school in a small provider pushes the tariff up quite a lot). This is, to be clear, the actual qualifications that entrants hold – not the level at which offers are made (wouldn’t it be great to have that data?).

[Full screen]

A scatter plot showing provider tariff against the proportion of POLAR4 quintile 1 (left) and POLAR4 quintile 5 (right) demonstrates that for 2021 (and every other year, use the slider!) this distinction still largely holds. Hats off in recent years to Nottingham Trent University and Manchester Metropolitan University for bucking the trend in one direction, and to the University of Leeds and the University of Manchester in the other – but if you are bringing in larger numbers of quintile 1 acceptances you are still likely to have a lower tariff and a lower number of quintile five acceptances.

Is this a problem? Well yes, in a sense that social mixing is a big part of university life and it would be best to see students have peers from widely diverse backgrounds if we want to get the maximum benefit out of this. And yes, in that the provider you attended holds a disproportionate sway in the graduate job market (though, as we have seen, not quite enough to erase differences by background whatever the Department for Education likes to imagine).

So – yes – in conclusion. Driving up numbers overall will necessarily expand access, but from where we currently sit there is some danger we are creating a kind of binary divide.

Leave a Reply