I'm afraid I want to write yet more on core/margin, having already covered the issues here, here and here. Managing this system so as to bring the greatest benefit to my employer is likely to be a big part of my job for the next few years, so please excuse me this obsession.
This post is about the AAB exemption. HEFCE have yet to publish their institutional model, but the Higher has provided some roughly indicative data. These exceptionally well-qualified individuals are not widely distributed in the sector, they are concentrated in elite institutions.However even the most elite of the elite don't have a 100% AAB intake. The non-AAB component is made up partly of slightly less well-qualified people (ABBs, say) and partly of very well-qualified people with qualifications that can't be returned within the HESA Qualifications on Entry entity (for instance European students will mostly have foreign qualifications which HESA don't provide codes for). So all institutions will be left with at least a small quota of non-AAB places, and (where it is a very small number) the management of these places will likely become a critical issue for them.
In the first place, the near-miss ABB students will not be evenly distributed across the institution. To make the right number of offers for a given number of places, institutions have to assess how many applicants will accept their offer, and how many of those will meet their conditions. At the whole-institution level these things are quite predictable (maybe less so in 2012!), but at the individual course level random statistical fluctuations will mean that some courses will find themselves oversubscribed, whilst others do not have enough students. An applicant who just misses the AAB offer may therefore still be accepted by the institution if that year for that course there has been a miscalculation and not enough offers have been met, but generally the less well-qualified students will be concentrated in the less popular courses for sufficiently obvious reasons.
If institutions wish to expand - or even maintain - less popular courses this will use up a large proportion of their on-quota places and make it very difficult to manage any cases where they want to take 'near miss' students onto their more popular courses. Alternately, if you want to retain enough flexibility to ensure all the popular courses can be filled (after all, fluctuations in cohorts at the course or department level are irritating), you will need to withhold those on-quota places from your less popular courses. This is why the Russell Group has called for the AAB threshold to be lower for unpopular courses such as Modern Languages, even though it would mean their institutions' quotas would shrink even faster.
A second issue arises about the non-AAB applicants with alternative (e.g. European) qualifications. Exemptions exist in Equalities legislation to allow universities to do things to Overseas students which would normally look like naked racial discrimination (e.g. charging a higher fee to someone just because they don't live in Europe), but as far as I am aware these exemptions don't apply to EU residents. The courts also tend to be very respectful of academic judgement, but only where a judgement can genuinely be shown to be academic. Suppose an elite university gets a significant increase in applications from Europe. If the applicants are strong ones, it may find itself in danger of using up its on-quota places too quickly, making it impossible to recruit ABB-and-below students onto unpopular courses (or even manage numbers on individual popular programmes). Whilst from a HEFCE perspective it could probably close to European and other on-quota applicants whilst remaining open to off-quota AABs, a court might judge that this was racial discrimination. If you previously treated a foreign qualification as equivalent to AAB, but then changed your mind for funding or regulatory reasons, that might not sound like an 'academic judgement' to the judge.
Whilst we are on the equalities issue, I was surprised to see that HEFCE think AAB students are older on average than other students. I would have expected the reverse, as AABs have a very strong propensity to go more or less straight into HE (maybe after a gap year), whilst less well-qualified students may enter the workforce for a while before going on to HE. I suspect this is because of HEFCE's decision to include students with prior HE qualifications in the AAB group. I don't really understand that decision, and I need to talk to HEFCE colleagues to understand why they have taken it.