This blog has been contributed by Rob Cuthbert, an independent consultant at Practical Academics and an Emeritus Professor of Higher Education Management. This piece follows on from his previous HEPI blog, ‘A-Levels 2020: what students and parents need to know’.
The horror show that is A-levels in 2020 descended to complete farce over the weekend, but no one is laughing. There are tens of thousands of ruined dreams after A-levels results day, because for five months the Government and Ofqual have been too secretive, made bad choices, refused to listen to constructive criticism, tried to tough it out and then made the wrong concessions too late. The risk in trying to summarise the position on Sunday is that it will no doubt be very different on Monday, after major changes on Saturday. But let us try.
Faced with mass public protest, Scotland u-turned on the Highers awards announced on 6 August, with First Minister Nicola Sturgeon admitting on Monday (10 August) that they had got it wrong, and Scots education minister John Swinney the next day announcing that they would abandon their algorithm and use only Centre-Assessed Grades (CAGs). Whitehall’s Department for Education went into full panic mode in relation to England. Secretary of State Gavin Williamson had nailed his colours to the this-algorithm-is-robust-and-fair mast so many times that he would not follow Scotland’s lead, and there was no sensible alternative. So he went for something that wasn’t sensible – the announcement late on Tuesday night (11 August) that students could use mock grades under certain circumstances instead of the algorithm’s grades.
For thousands of students who had taken mocks, this sounded like blessed relief. Not only could they now make an individual appeal (something previously ruled out), they knew it would succeed. But that was late Tuesday night. By Wednesday morning Ofqual, Schools Minister Nick Gibb and Universities Minister Michelle Donelan were doing their best to dilute the message, saying only that mocks might form part of the grounds for an appeal and even suggesting that not many appeals were expected. Schools and colleges, who had only that day received their students’ grades with shock and horror, pointed out the huge variability and complete lack of standardisation of mocks even within one school, let alone across the whole sector.
Williamson stood firm on his ‘triple lock’ – mocks or algorithm grades or Autumn exams. It was presented as a solution for all, when it was nothing of the sort. He had already announced that Ofqual would issue guidance on how the new appeals system would work, and Ofqual understandably said they would need a few days to work out how to operationalise the process, with a statement promised ‘next week’. Unfortunately, a few days were not available.
A-level results day was Thursday. Universities had as usual had prior notice of grades (on the preceding Friday) to enable them to hit the ground running on Thursday with offers and Clearing. The announcement on mocks-based appeals came much too late for them to change their starting position on Thursday. UCAS had already agreed to extend their deadline for decisions from 31 August to 7 September, and frantic statements from the Department for Education asked universities to hold places open for students whose grades depended on appeals.
Universities, with no prior notice of these changes, reacted variously – however sympathetic, they were facing pressures of their own. There is always tricky arithmetic for universities in deciding which students should have offers confirmed and who, even with a confirmed place, may not come. This year added hugely to the problems, which now included complete guesswork about how many international students might eventually want to enrol, even if they were able to travel to the UK, and the impact of a temporary Government-imposed set of student number controls. Earlier in the year, the Government had through the Office for Students also more-or-less outlawed unconditional offers to students, which would have reduced some of the uncertainty for students and universities.
In practice, what this meant was that many courses and even some whole universities rapidly filled all or almost all their places. The most selective universities and the most selective courses were most likely to be full; stories abounded of brilliant students with Oxbridge offers or would-be medical students being told they could not take up their place, irrespective of the outcome of any appeal. The worst effects seem to be in the most popular subjects with the largest cohorts – Mathematics, Chemistry, Biology and Psychology – ironic indeed for a Government so determined to promote STEM (Science, Technology, Engineering and Mathematics) subjects and talk down the humanities and creative arts.
Universities were doing their best to help, with initiatives like Leicester’s splendid declaration that it would accept mock grades regardless of any appeal. The Department for Education intimated that student number controls would be eased, but the overall cap on numbers would stay. It seemed unlikely that any university would actually receive the swingeing financial penalties previously threatened for exceeding their controlled targets.
Meanwhile, the A-levels story was headline news and stayed on the front pages despite suspiciously-timed announcements about quarantine for holidaymakers in France, seemingly designed to knock A-levels out of the public consciousness. The Ofqual algorithm had led to CAGs which would have created unacceptably high levels of grade inflation, with Ofqual not quite blaming teachers, but Government spin-doctors trying to make the charge stick. There was too much evidence of good professional judgment and integrity, so Government tried to blame the universities, with no more success.
By the end of Thursday, attention had switched to the finally-published Ofqual algorithm and their lengthy Interim Report. The storm of criticism of the approach became ever more expert as the statisticians weighed in. Imperial professors Guy Nason and George Constantinides expressed their polite but damning reservations in comments on Twitter which were retweeted hundreds if not thousands of times. It emerged that Ofqual had refused help from the Royal Statistical Society (RSS) some months earlier, unless the helpers agreed to sign draconian non-disclosure agreements which the RSS deemed completely inappropriate.
The 318-page Ofqual Interim Report was taken apart piece by piece. Even things we thought we knew turned out to be wrong. One example: ‘large’ cohorts – any with more than 15 students – turned out not to mean 15 students this year. Size was defined as the ‘harmonic mean’ of the cohort size this year and the total numbers in the cohorts for the last three years. If there had been 120 students in the last three years, an apparently ‘medium-sized’ cohort of 8 this year would be defined by the algorithm as ‘large’. That one detail alone had led to widespread mystification among schools and journalists about how the method worked. What the algorithm called a C might have been a borderline B/C or C/D, but it might also have been anywhere between A and E, depending on the school’s cohort and history. That is why some students saw their CAGs plummeting by two or even three grades, in too many cases.
In response to the wave of protests, the Department made a further concession, announcing on 15 August that appeals would be free of charge and there would be a ‘gold command’ task force chaired by Schools Minister Nick Gibb to make sure appeals would be handled quickly.
Under extreme pressure, Ofqual put out its advice on the amended appeals process by early afternoon on Saturday, ahead of the date it had promised. That introduced two new possibilities.
- First, students who had not taken mocks could still appeal on the basis of some other non-mock assessment held under controlled conditions.
- Secondly, CAGs might also be used in deciding the outcome of the appeal – if the CAG was lower than the mock grade, then the CAG would be used instead.
But this, of course, suggested that the CAG was a more reliable basis for judgment. While this was almost certainly true – CAGs, unlike mocks, rested on a great deal of careful and collective teacher judgment – it nevertheless still failed to allow for the range of school behaviours, where many schools had probably decided to propose no more than the maximum likely to be allowable as A*, say, even though they might have more students potentially able to achieve that top grade.
The implications of the advice were still being debated and worked out in social media when very late on Saturday evening Ofqual withdrew its advice, saying that the Ofqual board would review it and another statement would follow ‘in due course’. So just when we thought it could not get any worse, it did. Speculation on Sunday centred on the suspicion that it was the mention of CAGs that might have caused the Department for Education to tell Ofqual to change tack, mostly because of a report in The Sunday Telegraph by Camilla Turner, often very well-briefed on new government education initiatives. This was the position at midday on Sunday.
It did not have to be like this. The likely problems were obvious from the moment that Ofqual had announced it would require detailed rankings, and had been predicted from an early stage. The Education Select Committee had called on 11 July for action by Ofqual which would have improved the situation, but Ofqual had in particular refused to publish their algorithm immediately, as the Committee had asked. Ofqual had already refused offers of expert help by setting unacceptable conditions.
Policymakers in England repeatedly doubled down on supporting the supposed ‘integrity’ of the national system, at the expense of the many thousands of individual students experiencing terrible injustice, overlaid by what appeared to be class-based privileging of the already-advantaged. However this is in many ways an equal-opportunities disgrace, because many individual unfairnesses are built in for every kind of school. Every badly-treated student has the same problem: they have been judged not by their own performance, but by the record of previous students at their school. They have been forced to compete with their own schoolmates, within each subject, for the limited number of top grades available as determined by the algorithm. And for many it is already too late for them to get the place they deserve and would probably have achieved if they could only have been judged on their own performance, not other people’s.
Ofqual explained in their 318-page Interim Report why they went for the ‘meso’ level solution rather than a completely national or completely individual level approach. That was probably the right choice, and it could have been made to work. But now it has failed, and the only acceptable solution is something like this:
- abandon the algorithm and award grades based on CAGs, as in Scotland;
- abandon completely the student number controls on higher education institutions and the financial penalties for over-recruitment;
- consider additional targeted funding for universities with capacity constraints so that offers for downgraded students might still be available this year; and
- consider some kind of compensation, whether financial or non-financial, for all downgraded students unable to go to the university of their choice this year, if only to head off the expected deluge of legal action from the Good Law Project and others.
In 2016 Nick Hillman made a key point about the difference between individual and collective social justice in his discussion of grammar schools:
The difference between those who place the interests of society above the interests of individuals and those who put individuals above society goes to the heart of politics.
In reality, each person – not just each political party – is a mix of the individualistic and the communitarian. … a good manager simultaneously cares about the health of their organisation and its individual members of staff. … MPs oscillate between acting in the interests of society as a whole and acting in the interests of individual citizens. It is the difference between fixing the problems raised by someone at a constituency surgery and being a Minister, frontbench spokesperson or Select Committee member
Most big political questions can only be answered by balancing the needs of individuals against society.
For this year’s A-levels, government and Ofqual have got that balance horribly wrong. GCSE results are announced on Thursday 20 August, with Ofqual admitting that their algorithm is less reliable for GCSE than for A-levels. Up to two million centre-assessed grades might be changed. Will the government get the balance right this time?