The chances of a clear-headed decision on the merits of post-qualification admissions to higher education receded further this week, as politics intruded on an already fraught debate. Research on predicted grades that should have clarified some of the key issues has been lost in a welter of recrimination over the misinterpretation of its findings by the Department for Education and Skills. Now all the attention is focused on alleged social engineering rather than on the establishment of a system that is fair to all candidates. Since ministers made such play of the supposedly disproportionate impact on applicants from poor homes and state schools, they have only themselves to blame for the hay being made by Opposition spokesmen who want the consultation relaunched.
Someone at the DfES should have seen that the Oxford research suggested the opposite. But why was this the kernel of the argument in the first place? Is it any more unjust for a working-class student to miss out on a degree place because of inaccurate predictions than for a middle-class student to do so? There are legitimate hopes, expressed in Steven Schwartz's 2004 report on admissions, that PQA might help more students from poor homes win places at elite universities. The knowledge that they had achieved top grades might be expected to bolster those who presently lack the confidence to aim high. That is not social engineering. But the idea that the whole system might change to benefit particular types of applicant, however deserving, was asking for trouble.
The row has distracted attention from the real conclusions of the research, which may be seen as undermining the case for PQA. In particular, the important second section of the report - on the impact of inaccurate predictions on the making of offers - has received almost no attention. The fact that fewer than half of all grade predictions are correct may be a poor reflection on schools and colleges, but it matters only if it affects admissions. The report finds little evidence that it does. Although the researchers require more information on the actual offers made by universities to be certain, they found "only weak and negative evidence" of overprediction increasing applicants' chances of success.
This may be because most inaccurate predictions are out by only one grade, and the lower grades are most susceptible to error, according to the research. Oversubscribed courses, which inevitably produce most admissions controversy, tend to deal in A grades, which are the most reliably predicted. A post-qualification system would still be fairer, but the report demonstrates the need for more evidence on the numbers affected adversely before expensive upheaval can be justified.