Drop-out rates for first degree students in England are apparently rising.Before this induces panic there are some important points to make.

First, though we now have participation rates comparable to other countries, we are still better than most at getting students through.

Second, if you open the doors wider, enrolling more less well-prepared students, drop-out rates must rise unless standards are lowered to accommodate them. This we have made strenuous efforts to avoid.

Third, drop-out rates cannot yet be attributed to fees. The available figures pre-date their introduction. They may be linked to the gradual replacement of grants by loans. But most likely they are the predictable consequence of moving from an elite to a mass system.

ÖйúAƬ

That said, many will see a one in four drop-out rate as failure and waste.Some will argue that admission should be more selective. But if it were, the least privileged students from the worst schools would be first to be excluded. Welcome policies aimed at wider participation would be vitiated.

Better instead to give priority to improving schools so that more students reach ÖйúAƬ well prepared. This the government is addressing with vigour. As part of this, reform of the terminal school exams, which form the basis for further study, is long overdue. This should provide a broader, more consistent base for students enrolling in ÖйúAƬ.

ÖйúAƬ

Reform of A level has been advocated now for 30 years. Unfortunately the government appears to be nervous and we are to have a staged change. This risks public confusion and defensive lobbying by schools which specialise in getting their pupils through traditional A levels. It would be better to go in one jump to a baccalaureat-type of qualification that could be clearly explained and understood.

On ÖйúAƬ's part, better interim qualifications are needed so that those who drop out have something to show for their efforts and somewhere to start again later. Institutions have been slow to do this. The Dearing committee had a go at sorting out levels and qualifications but that attempt was flawed because it was tangled up with a desire to ensure that all institutions' qualifications were of comparable standard. This is cloud cuckoo land. Everyone knows, though not everyone will say so,that output standards vary between institutions.

It follows from this that there should be more honesty in acknowledging that we already have a highly differentiated system. A group of increasingly valued and confident community colleges is emerging offering post-16 courses plus sub-degree ÖйúAƬ courses and some first degrees. These colleges open up opportunities locally to students unfamiliar with ÖйúAƬ. They need to be latched into universities so that high-flyers can take their studies on as far as their talents allow.

Equally clear is the emergence of a group of research dominated universities better funded than others and achieving not only high research ratings but high teaching assessment scores as well (pages 6 and 7).

ÖйúAƬ

In the middle is a large, varied group of institutions where the role is more mixed. Some will be drivers of regional consortia. A few may find themselves under increasingly critical scrutiny: the Thames Valley syndrome. Morale in many is low as the awful truth dawns that a longed-for Labour government has not brought an end to per capita cuts and miserable salaries. A number still need to rethink what they are for.

More clarity in this area would help students make wiser choices and improve their chances of staying the course. The welcome development of sophisticated performance indicators that encourage difference should help that process.