The current educational algo-debacle is an exquisitely English cock-up*: a slow-motion train wreck that is the product of 30 years of educational initiatives, reorganisations and adjustments to alleviate the problems generated by previous changes, all piled up on each other without consistent architecture or, needless to say, political consensus. Finally this year Covid nudged it over the cliff of its own contradictions.
Our education system is a perfectly designed generator of grade inflation. Like executive pay under shareholder capitalism, it's an escalator engineered to move in one direction only: up.
This year's events are the culmination of a story that began in 1992 when 38 polytechnics were elevated to university status, nearly doubling the overall estate. Growth has continued ever since: there are now no less than 132 UK universities, with a student body that has expanded to match. Nineteen-seventy’s total of 200,000 students had mushroomed to almost 2m in 2019.
At a stroke, higher education morphed from an elite to a mass education system. Unfortunately, having willed the end, naturally with no diminution of quality, the government neglected to provide the means to bridge the gulf in standards – judged on traditional measures – between the old and the new. Accurately reflecting the gulf in respective resources, it was and in some cases remains large.
Real levelling up would have required a massive injection of resources into the new-borns. Instead, as ever, the government opted for a sleight of hand whose costs would only surface later. Traditionally, to maintain standards new universities underwent an adjustment period during which they administered degrees set by longer-established institutions. By contrast, the post-1992 cohort were granted degree-awarding powers from the start. There was no way a first from an under-resourced new university could be worth the same as a first from a top established one, but at a stroke the difference as made invisible – except to external examiners, who are often still pressured to verify marks that they know are too high or less often too low, depending on where they come from.
Tuition fees did provide universities with extra resources. But they were a two-edged sword. Particularly after 2010, when they jumped to £9,000, they set in motion a programme of marketisation that, as the government intended, turned students from learners into consumers, a process encouraged by the creation of albeit unofficial league tables and increasingly important student satisfaction surveys. By the same token, universities became fierce competitors for their custom. Much of the extra resource was diverted into marketing, facilities and highly paid administration, while students began to argue that shelling out £9,000 a year entitled them to a good degree and the teaching that ensured they got it. Lecturers and their employers had strong incentives to oblige. The casualty: a steady inflation of students’ grades.
As part of the supply chain, schools have naturally been sucked into the upward vortex. They were also subject to strong pressures of their own. Exam boards are competing commercial entities, and schools exploit discreet exam arbitrage between them. Moreover, education was an early testing ground for the New Public Management (NPM), the drive to sharpen up the public sector by subjecting it to private-sector methods and techniques. Unsurprisingly, the regime of targets, inspection, league tables and fierce performance management (‘targets and terror’) had the same dismal effects as in other public services such as health. Particularly harmful were the inducements for heads and teachers to play the numbers game by quietly dropping ‘harder’ subjects, excluding poor performers and ‘teaching to the test’ – a classic illustration of the folly of making professionals accountable to ministers and inspectors rather than those they directly serve. While it is widely accepted that many schools, eg London, have improved, the cost has been high in the shape, again, of grade inflation.
Briefly, consider that the percentage of top ‘A’ passes at A-level had gone up from 12 per cent in 1990 to 26 per cent in 2017, and ‘A’s plus ‘B’s from 27 to 55 per cent. The upward progression in degrees is even more marked. As a New Statesman article put it last year, 'British universities... have increased the number if degrees they award fivefold since 1990, while the proportion of firsts they hand out has quadrupled — from 7% in 1994 to 29% in 2019. For every student who got a first in the early 1990s, nearly 20 do now... The proportion of students getting “good honours” — a first or 2:1 — has leapt from 47% to 79%: at 13 universities more than 90% of students were given at least a 2:1 [in 2018].’ In a perfect self-reinforcing cycle, universities justify this progression by pointing at the schools: it’s not surprising we’re giving more good degrees, they say, because we’re getting better students – just look at the A-level results.
This is the backstory to this year’s school shenanigans, when the creaking system was brought crashing down by the cancellation of GCSEs and A-levels during the lockdown. Without the restraining influence of real marks for real work, the government invented two unreal ones – centrally assessed grades (or CAGS) and a version moderated by the famous algorithm to damp down what it saw as alarming grade inflation. Both measures are barely comprehensible in their complexity (sample: ‘CAGs are not teacher grades or predicted grades, but a centres profile of the most likely grades distributed to students based on the professional views of teachers’). But the circle was unsquarable. While the algorithm did moderate the grades, it could only do so at the price of such manifestly unfair side effects that the government hastily retreated. CAGs, and by extension, grade inflation, on this occasion however justified, rolled on.
So we arrive at a familiar destination. Grade inflation is a symptom of what Ray Ison and Ed Straw, authors of the important new The Hidden Power of Systems Thinking, call a system-determined problem – one that can’t be resolved by first-order change, only by rethinking the system itself. Tinkering with the existing system to make it work better is our old friend doing the wrong thing righter, which ends up making it wronger. And we end up with the worst of both worlds: private-sector market competition moderated by Soviet-style regulation that achieves neither efficiency nor accountability, and whose figures won’t bear the mildest scrutiny. When we most needed a system based on professional trust and respect, we have the reverse, a regime established to assure academic standards that has overseen their almost complete debasement.
This has the potential to be much more than a little local difficulty. Higher and to a lesser extent secondary education, backed up by league tables that conveniently big up their strengths, have long been talked up as one of this country’s strongest international success stories. Covid’s inconvenient intervention suggests a more accurate characterisation might be a house of cards, built on statistical foundations that don’t even come up to O-level standards.
* As a Scottish reader correctly notes, it is increasingly hard to generalise across the component parts of the union in such matters.