Comparing entry rates to university: the (data) gap that could easily be reduced

The First Minister was interviewed yesterday by Andrew Neil. One journalist reported that “Sturgeon says difference in university access rates for poorer pupils between Scot & Eng down to different sets of figures.”

I didn’t see the interview, but this sounds plausible.  The Scottish Government has become increasingly confident and robust over the past year in dismissing UCAS figures as a basis for making cross-country comparisons in entry rates to university for young people from disadvantaged backgrounds. UCAS figures show 19.5% of 18 year olds from most disadvantaged 20% of areas in England entering university, and 10.9% for Scotland. It’s worth noting that this gap is not uniform across young people: it disappears as levels of disadvantage fall.  For those at the other end of scale, there’s hardly any cross-country difference in university entry rates at 18.

There are two possible problems being invoked by the FM here. One is a technical point about different ways of measuring area disadvantage. For England UCAS uses a measure called POLAR3, which looks particularly at education disadvantage: for Scotland, the figure above uses a different measure called SIMD. However, I haven’t seen an argument from the Scottish Government that specific differences between POLAR and SIMD explain the gap above.

It’s much more likely this argument is about the coverage of UCAS data.  UCAS doesn’t include all the 18 year olds who go to college and from there to university, which is a much more common route here than south of the border. The Scottish Government leans on this point increasingly hard, and it is an absolutely true observation about the coverage of the UCAS figures.

There’s a substantial argument to be had about how analogous college-to-university entry is with direct entry. In particular, slightly under half of those who do it (in total, at all ages) get full credit (“advanced standing”) for their time at college. The rest mostly go back to square one and into the first year of a degree.  A few get partial credit (“advanced progression”), such as doing two years at college and then starting in second year at university.  Anything other than full credit means repeating one or two years, depending if students have done an HNC or an HND. That means five or six years of full-time study to get an undergraduate degree which would have taken a direct entrant four years. That in turn carries extra living costs, and means later labour market entry, putting these entrants at a relative disadvantage.

These cases also raise a reasonable question whether their English analogues are 18 year old direct entrants, or those south of the border who also took a couple of years doing something else before starting their degree course, and who also are not in the age 18 UCAS figures there. Of course, we shouldn’t only be interested in school leavers.  But there are particular reasons to monitor how far access to starting a degree programme straight from school is socially biased.

Does the SG have the data on how many people entered college at 18, from the most disadvantaged 20% of areas (SIMD20), and how many of those got full credit? This information isn’t published, but other material suggests it could be calculated.

The SFC publishes how many college-to-university movers got full credit (Table 22 here): it was 3,999 in 2014-15.  The SFC notes that after 2014-15 numbers will have been rising: but it appears not very sharply. The “national ambition” for 2016-17 for those with full credit is 4,100.  Elsewhere it has published that 71% of all movers in 2013-14 were under 25, and 23% of them were from SIMD20 (para 10 here). So it feels like it should at  least be possible to isolate the number who moved at an age consistent with having entered college at 18, were from SIMD20 and got particular amounts of credit.

Without those particular figures being public, there is still a way to test how likely it is that the UCAS gap at 18 is closed by college-to-university movers who get full credit, from the figures we do have.

UCAS reported 1,345 SIMD20 age 18 Scottish entrants in 2016. To make that 10.9% into something like the 19.5% in England would require that figure to be a little over 1,000 higher.

1,000 19 and 20 year olds from SIMD20 areas getting full credit would equal around one-quarter of all those getting full credit from all ages and all backgrounds.  That feels high.

If I assume that those aged 19 or 20 make up one-third of all people at all ages moving with full credit, and that they are half as likely again to be SIMD20 as the entire college-to-university population (which isn’t an obvious assumption, given disadvantage tends to delay engagement with higher education), I can get this group to 11.5% of the full credit group (33% x (23% x1.5)). That only half closes the gap, and these to me feel like quite hopeful assumptions.

So it’s not immediately obvious from what we do know that “UCAS leaves out colleges” is necessarily a powerful rebuttal to concerns about how Scotland compares to the wider UK in giving young people from disadvantaged backgrounds access to university-level study on equal terms.

Including those who don’t get full credit would come close to closing the gap: although that would still be using the same optimisitic assumptions.  But in any case, as argued above, it’s open to question whether those who had to go back to the start of the process after a year or two at college should be included in any comparison of age 18 university entrants, even if they went straight to college at 18.

It’s frustrating that while the Scottish Government has been keen to stress what’s missing from the UCAS data, it hasn’t moved with equal speed to use the information it holds to fill that gap, even with a few footnotes and caveats. At the moment, “ignore this data, it doesn’t tell you the whole story, but we aren’t saying what whole story is” is being too easily played as a get out of jail free card.   It’s time to move from general dismissal to the actual numbers.