Education: It's time that we declared war on the class system

A single benchmark is the fairest way to put paid to the old prejudices .

The arrival of an invitation to join a subject benchmarking group looked like a poisoned chalice. It may yet prove to be one, when our group's draft proposals are made public, yet the working sessions have been a pleasant surprise.

The care that has been taken to secure the widest possible representation of the subject's provision might have been expected to produce some dramatic disagreements and confrontations, especially in a subject as notoriously acrimonious as history.

However, in practice, the members seem to have gone out of their way to avoid offending or alarming each other. Most of the problems so far have arisen from almost unanimous reservations about the procedures and priorities of the Quality Assurance Agency - and even here there have eventually been gestures of accommodation from both sides.

Yet if benchmarking (so far) has not proved as corrosive of subject solidarity as some might have feared, less predictable problems are beginning to emerge.

The most important problem concerns final results. Our group (and doubtless those for chemistry and law too) considers that setting standards for final achievement is a fundamental part of benchmarking. And yet we recognise that it is no part of our remit to challenge the hallowed system of classification itself.

That could only be done by all subjects acting together, and in step with universities, too, as degree-awarding bodies. It is perfectly clear from informal discussion, however, that many, perhaps most, of those around the history benchmarking table at least would have no objection to scrapping final classes entirely.

Some of us would actually welcome it. Yet we are in no position positively to recommend it, and so we will end up making complex recommendations appropriate to a system that many of us no longer believe in. In this light, the benchmarking exercise will end up being a colossal missed opportunity.

All of us in higher education have been through the classification system, and most of us have done pretty well out of it. But if that creates a natural disposition in its favour, longer experience of its operation ought surely to raise some doubts. After 30 years examining (internally or externally) in 10 institutions, it certainly leaves me increasingly dissatisfied, and puzzled too at the persistence of a system first designed for 19th century Oxford.

Lumping candidates into a handful of broad classes has always been a crude business. The long-term tendency to sub-divide the classes shows awareness of that.

The division of the second class, or the introduction of the "starred" first, tell us that examiners were at some point convinced that undifferentiated classes were not adequate measures of a meaningful range of ability.

The same tendency is now affecting the upper second class, with periodic calls by external examiners for crackdowns at the lower end, and the clear tendency among increasing numbers of students to regard a 2(ii) as some kind of failure.

When it comes to the third class, there is no longer much doubt. To all intents and purposes it represents a fail, and arguments have been heard in benchmarking for selling the lower threshold for honours at the bottom of the 2(ii). Cogent though these are, this course cannot be recommended because the third class cannot be unilaterally downgraded by any individual subject group.

It is clear that, however much we tinker with classes, each will continue to reflect a huge range of ability and, at the margin, injustice.

That is precisely why so much time is spent at examiners' meetings over borderline cases. And periodic attempts to stem the seemingly inexorable growth of the upper second by more severity at the bottom create their own injustice by penalising students who, by the standards of earlier years, would have achieved a higher class. A further complication in these days when detailed marks are no longer confidential, is that students compare their results and often launch appeals on the basis of achieving lower classes than peers who have apparently similar mark runs.

The whole idea of classifying results in the traditional way, in fact, is creaking. It positively encourages appeals, with all their expense in time and energy. Has the time not come to abandon it entirely?

Already most students receive full transcripts of their results which give a much more subtle picture of their performance than any crude classification.

The problem here is that transcripts alone give no indication of how a student compares with others in his or her group.

Many will see no difficulty with this, but it is not clear that individual transcripts alone will satisfy employers. Anyone who writes references will recognise that employers want to know how applicants stand in relation to their group (the increasing practice of requesting their place in a percentage cohort, incidentally, suggests that even employers find classes no longer enough as a measure of achievement). Would not the answer be to adopt the system that is used by the French grandes ecoles and also by national academic competitions?

In these, results are published in the form of a single ranked list. Now that most marking schemes have become numerical, the means already exists to compile such lists quite easily.

From them, it would be perfectly clear how any candidate had performed relative to the group as a whole. There would be no scope for appeals and no hours spent in meetings agonising over borderline cases.

And the work of benchmarking groups would be so much simplified. They would then be able to concentrate on the really important question of defining a single threshold for admission to honours in their subject.

The predictable objection to abandoning classes is that a class denotes a standard independent of institutions, and is guaranteed by the presence on all boards of examiners of externals.

As an experienced external I think that there is more truth in this than most people are prepared to believe; however, I do not see any prospect of convincing them.

There will always be a widespread view, both within and outside the system, that degrees from some institutions are better (or are worse) than from others.

A single benchmark threshold standard might actually do more to dent these prejudices than persisting with a system of classes whose credibility is becoming increasingly difficult to defend.

The writer of this article is professor of history at the University of Bristol and is a member of the history benchmarking group