Williamson “blaming Ofqual”. Yet it has performed the role it was set up to do-by ministers

Sanctuary Buildings, home of the Department for Education. Surely, the buck has to stop here (or with the Prime Minister), rather than with officials.

Gavin Williamson, the Education Secretary, is reportedly “seeking to blame Ofqual for exams debacle”, as he performed a humiliating u-turn over the grading of A-levels and GCSEs.

This means, of course, that students’ teacher assessment grades will now be accepted if they are higher than those which would have been awarded by a controversial algorithm overseen by the regulator, Ofqual.

Williamson was reported as having said that he had only become aware “over the Saturday and Sunday” of the scale of the problems with the Ofqual algorithm.

He was reported in the Guardian saying: “Over the weekend it became apparent to me, with evidence that Ofqual…and external experts had provided, that there were real concerns about what…[grades] a large number of students were getting… and whether or not that was a proper and fair reflection of their efforts.

“We… consistently asked a large number of challenging questions about the system. Its robustness and its fairness. We’d been consistently reassured about that. Over the weekend, Ofqual released some of the algorithm to the public and actually shared that quite broadly and obviously we saw a number of what I would call just outliers that didn’t make sense.”

He also said that he had been “consistently reassured” by senior Ofqual officials that the algorithm was fair, the Telegraph reported.

This begs the question, clearly, as to how searching those “challenging questions” had been, given that Williamson has said that his department only grasped the scale of the problem days after the results had been released to students.

While experts advise, it is ministers’ job to ask those questions, to learn from them and then to take responsibility if there is a public policy failure.

This is how modern democratic systems are supposed to function. Yet this reported attempt to pass off blame seems to carry echoes of the ongoing controversy over the government’s claim to be “following the science” with respect to the coronavirus crisis as a whole, with concerns that such statements are an attempt to pass off blame for failings to scientific advisers.

There is another point to make with regard to Ofqual’s responsibility, or not. While it is impossible to know at this stage whether Ofqual was warning ministers in advance of difficulties, and what questions it was asked, a more basic point is that the regulator appears to have been following the requirements of the way it was set up, in acting the way it did.

For Ofqual was founded with one of its defining missions being to keep national exam results roughly constant, from year to year. This, as I wrote last week, it achieved through the “comparable outcomes” policy, which aimed to stop annual increases in average student grades.

Ofqual was founded with such a mission through the Apprenticeships, Skills, Children and Learning Act, which was passed in the dying days of New Labour, under Ed Balls as Education Secretary, in 2009.

This listed, as Ofqual’s first objectives, that it should ensure “that regulated qualifications give a reliable indication of [students’] knowledge, skills and understanding,” and that they “indicate a consistent level of attainment (including over time) between comparable regulated qualifications”. The section on Ofqual was amended by the coalition under Michael Gove in 2011, although these words did not change.

Now, you can argue whether these two objectives might, in reality, be in conflict this year, as it could certainly be argued that, in cases where the algorithm had produced “outlier” results for students, then it was not “reliable”.

But the second objective above seems very clear, in requiring Ofqual basically to deliver results which do not change much year on year.

Yet it emerged in public last month, when the regulator held a seminar on results, that, if teacher assessments were to be used as the basis for grades, then national results would jump markedly. It thus reacted as its second objective essentially says it should, and used the algorithm in a bid to stop that happening.

All this, then, seems predictable weeks, if not months, in advance. Given that we are in a unique year, a more foresighted government might have sought to pass legislation temporarily removing the requirement on Ofqual to hold results broadly constant, given the situation where no exams were taken.

But it did not do so. So, instead, we are in this situation where it has, in the end, happened anyway. And yet, Williamson has seemingly sought to blame the regulator, for a situation which seems to have been at least in large part a consequence of the way that successive governments have required it to operate.