FMQs update: +10% = 1.4% (probably, not -1.7%)

Update 17:15 12 June:  Further digging in the newest UCAS numbers suggests that steps have already been taken in those to take into account the “UHI effect”, so that the further adjustment below isn’t needed to get a like-for-like effect. That’s because UCAS also excluded  category called “RPA” from the new data, which produces an adjustment which turns out to deal with the UHI dip almost exactly.

In that case, “all age” entrants for SIMD Q1 Scots in Scotland do after all see a rise, but of 1.4% (55/4020) not 10%. That still relies on all of any increase reported for SRUC being real, not administrative (see below). Non-18 years old go up by a more positive 5.4%, rather than 0.9%.

But the age 18 figures still drop by 7.5%. The analysis below did not make any further adjustment for those and used the new UCAS figures as presented.

The total like-for-like rise across all Scottish domiciles of all backgrounds in Scotland shown in the new UCAS figures is 2% (595/27,770).

However, if these figures do include all of the increase of 445 for SRUC,  which more than doubled its reported entrants from 335 to 780, then that would be doing most of the work here: in that case, the rest of the sector would have seen a like-for-like increase of around 0.5% in entrants.

Greater transparency and detail about issues affecting comparability over time when these numbers are published in the first place, and in any subsequent analyses, would really help here.

Meanwhile, I’ve corrected my initial mis-reading. Over to the SG to correct its rather larger one …

Original post below

The post linked here drew attention to a wrong reading of UCAS data by the First Minister at last week’s FMQs.  It explained why the increase in the numbers going to university in Scotland from the most deprived areas between 2014 and 2015 was likely to be well below the 10% quoted.  Some further data released by UCAS yesterday, plus a more detailed reading of the data already released, makes it possible to be more specific.  There was in fact a slight fall in the numbers.

The problem with the 10% is that, as both UCAS and the SG have both previously flagged up,  it includes the effect of things which have nothing to do with changes in actual student numbers. One is the switching of initial teacher training in Scotland into the main UCAS scheme (it was not counted in these numbers before); the other is the apparent one-off omission by unnamed institutions (it turns out to be due mainly to one) to record most of those admitted later in the process, in 2014 only.

UCAS have just put out some supplementary information which shows how the acceptance figures look if initial undergraduate teacher training (UTT) is removed from them, for each SIMD (deprivation)  quintile – giving the true underlying pattern, and incidentally also putting the figures on the same basis as those for the other UK nations, which continue to use a separate scheme for UTT.

Separately, a careful look at the detailed information provided for individual institutions reveals that the institution whose figures dipped spectacularly in 2014 was the University of the Highlands and Islands. Again, it is now possible to show how that affected each SIMD quintile (see table below).

Putting all this together confirms that these effects had little impact on 18 year olds in SIMD1 (the most deprived 20%). For them, a like-for-like comparison shows there was a 7.5% fall in the numbers entering via UCAS, a small change from the 6.9%  fall the original UCAS figures showed.

This is five times the general fall in the age group  (-1.6%) and really shouldn’t be quickly dismissed with references to other age groups, college entry etc –  on any reading, this is a significant fall in the number of 18 year olds from the most deprived  areas who were able to take advantage of the most direct route to university between 2014 and 2015. It needs noticing and understanding, not downplaying.

Moreover, as predicted, there is a very significant effect when the “all age” figures are looked at. The 10% increase becomes a 1.7% like-for-like drop.

UCAS has previously suggested that there was a small increase in total main scheme acceptances for Scots, even after allowing for the two effects above (which they estimated accounted for four-fifths of acceptances). This seems to be explained by the 150 more Scots who went to other parts of the UK, plus two Scottish institutions not included in the latest numbers. Both of these saw some increase in Scots admitted: Glasgow School of Art (195, up from 185) and particularly Scotland’s Rural College (780, up from 335 – although there was only an increase of 50 in the College’s SFC funded places between the two years, so it is possible most of this increase reflects another administrative change, rather than an actual growth in numbers – that needs more explanation).

As the table below shows, there is a little  bit of better news if only those not aged 18 are considered. There was a slight like-for-like rise in the number  of non-18 year old SIMD1 students entering through UCAS, of just under 1%.  But it wasn’t enough to off-set the larger fall in 18 year olds.

Technical note

The table below shows the calculations.

Scotland: SIMD 1 (most deprived 20%), excluding UTT

2014 and 2015 figures for age 18 and “all ages” from here. Change  and “not 18” calculated from these.

UHI data correction estimated from university-level report here , Table P.25

UHI: Scottish domiciled, all placed applicants, all ages, all quintiles

UHI: Scottish domiciled, all placed applicants, Quintile 1

The simplest assumption to make is that the 2014 figure was under-declared by 145 (180 could be used, based on a comparison with 2013, but that year is unusually high). However, this number will include some new UTT cases as well (probably around 20, based on looking at the total increase in UTT at UHI of 210 – see here – and working out that around 10% of the UTT intake is SIMD1, by comparing the figures with and without UTT).  To avoid double-counting these, 145 has been reduced to 125. (None of this appears to affect 18 year olds.)

This means the increase of 55 shown in the  new figures needs reducing further, by 125,before the figures are like-for-like, giving a reduction of 70.

By excluding the UTT numbers, any real growth in those numbers is excluded too, so there may have been some genuine increase in SIMD1 due to that, but I can’t see a reliable way to identify that from the available numbers and there’s no obvious evidence at first sight of any substantial growth in UTT.

Conversely, the University of the West of Scotland also saw a noticeable dip in 2014, but it was smaller (3680/3255/3905) and the pattern for Q1 is different (900/940/1065), so none of the UWS Q1 increase between 2014 and 2015 has been treated here as due to changes in recording practice – but it is possible some of it may be.