There is an excellent German word – verschlimmbesserung – that roughly translates as “trying to fix things but making them worse”. It would serve as a suitable epitaph for the government’s increasingly panicked attempts to award exam grades in the absence of any actual exams.
Last summer, the Department for Education (DfE) tried an algorithm tying school’s results to the previous year’s performances. Unfortunately, it did it in such a ham-fisted and rigid way that it was forced into a last-minute U-turn as multiple examples of manifest unfairness came to light.
The lesson officials seem to have learnt from this is not that their algorithm was stupid but that all algorithms – which are just sets of rules – are bad. So this year they decided to eschew rules entirely and opt for a laissez-faire, anything-goes approach.
Gavin Williamson argued in the Commons yesterday (25 February) that this year’s model would provide “fairness and consistency”, but it would be hard to design a system less likely to provide consistency if you tried.
[See also: Gavin Williamson: The dunce of Westminster]
Schools will be responsible for deciding grades and can base them on pretty much whatever they want. They can use mini-tests provided by exam boards on parts of the course; or old exam papers; or write their own tests. These tests can be done under exam conditions or in class or at home. If they can’t or don’t want to use new tests, they can look at old mock exams, unfinished coursework, or examples taken from classwork. Even with the best will in the world it’s impossible to be consistent in assessment if the work you’re assessing is completely different.
Moreover, the algorithm ban means there will be very limited moderation by exam boards except in extreme cases, so unless teachers have engaged in absolutely egregious manipulation, the school’s grades will stand.
This puts teachers in an almost impossible position. They will know – as anyone can see – that this is a system wide open to abuse. Teaching is a profession with a very high level of moral purpose and integrity, but if you think there’s a risk that even a small number of schools will exploit the system, and you know that they won’t get caught, then sticking to your principles could harm your students’ life chances.
It’s a rare example of a perfect prisoner’s dilemma in public policy. The best outcome requires all schools to do the right thing. But if even some schools “defect” and take advantage of the system, the ones that act with integrity lose out. Or rather, their students do.
On top of this pressure, schools will have to deal with parents campaigning on their children’s behalf (we all know the type of parent who’ll do this) and an appeals process that will pit their students against them, and not an exam board as would be the case in a normal year.
These factors, combined, will inevitably lead to significant grade inflation. Last year, when the DfE reverted from the algorithm to school grades, there was a 13 per cent increase in A grades awarded at A-level. And schools set those grades under the assumption there would be strong external moderation. Without that threat this year, it’s hard to see how this won’t go a fair bit higher.
To be clear: this won’t require schools to break any rules, but simply to apply a generous reading of the guidance. We won’t see E grade students given As, but borderline cases, or cases that can be justified as borderline under this very vague system, will be given the benefit of the doubt.
You may be thinking, “so what if more pupils, who’ve lived through a pretty miserable year, get high grades?” I think it matters for three reasons.
First, even though more students overall will get higher grades, these will not be evenly distributed. Some schools will be more generous in their interpretation of the extremely accommodating guidance. Their pupils will have an unfair advantage. Moreover, we know that teacher assessment – even when the system is well-designed – is prone to stereotype biases around gender, race, and class.
Second, it’s unfair on teachers, who will have to engage in an enormous amount of work to produce portfolios of evidence to justify grades; worry over tricky cases while trying to apply impossibly inconsistent rules; and potentially end up in conflict with students and families; only to then get attacked by the tabloids for inevitable grade inflation in August. This government will have no compunction over teachers getting all the criticism this year if things go wrong. Blame avoidance is pretty much its sole motivation at this stage.
Finally, by risking very high levels of grade inflation, it puts the whole future of the system in danger. What happens next year when (fingers crossed) normal exams are possible? Do they have to be recalibrated to incorporate the grade inflation so as to avoid unfairness on that year group? If so, pre- and post-pandemic grades will not be comparable for employers, which is hardly fair on those in their early twenties who have seen Covid-19 hammer their job opportunities already. In its desperation to get out of a hole this year, the government is creating an even bigger problem for the future.
[See also: An urgent action plan to mitigate the damage of school closures]
What’s so frustrating is that all of this was avoidable. After last summer’s fiasco it was obvious that the same problems would arise this year should Covid peak over winter. And yet here we are in late February, two months after exams were cancelled, with a plan that’s little better than a shoulder shrug. Had the DfE reacted promptly we could have had a coherent plan B with shorter standardised assessments given through the year. Even at this stage the government could have accepted precise grades were impossible and focused instead on ensuring all students progressed to a suitable destination, by supporting schools, colleges, and universities to work together.
Instead, we have the laziest possible solution. Leave it up to schools, let them take the blame – and sod the consequences for them, their students, or the future.
Content from our partners
Related
