Government ministers have been expressing increasing concern about the possibility that university tuition fees will be much higher than they expected. Most recently, David Willetts, the universities and science minister, stated that institutions charging £9,000 a year may look "silly". What is behind this?
The problem for the government is that it has said that it expects the new fee arrangements to deliver major savings in public expenditure, but that assertion is based on assumptions that are looking somewhat heroic.
First, its calculation is founded on pretty optimistic notions about how much students will earn (and therefore the speed at which they will repay, and the amount).
Second, the loans that it will have to provide will be based on the fees that universities charge. It has recently become apparent that the government's calculations have been based on the assumption that £7,500 will be the average fee. Few universities are even considering charging less than that. Yet if £7,500 is to be the average, there must be roughly as many below as above.
Why wouldn't a university charge as much as is permitted? We saw with the current fee regime how universities gravitated towards the maximum. As there is no sign yet of a market and demand is likely to outstrip supply for the foreseeable future, market disciplines do not exist.
The problem is that creating a market would require allowing universities to recruit as many students as they wish - and that will be well-nigh impossible if the government is to control expenditure.
The cost to the state could be really serious. Presumably the deal done with the Treasury is based around a £7,500 average, so if the reality is above that, something has to give.
But what? Theoretically, this is not this government's problem. The immediate public expenditure cost of the new arrangement is far higher than at present, because each student will be entitled to a loan of up to £9,000 instead of £3,000. But because of the rules of public accounting, the money that the government borrows in order to supply students with repayable loans does not count as public borrowing.
At the moment, the markets seem to be satisfied. And of course, the unplanned cost to the taxpayer will be downstream, when the money does not come back as forecast. It will be somebody else's problem - some other politician and some other taxpayer.
Nevertheless, assuming that the government is serious that it will not wear the cost of an average fee of more than £7,500 a year, what can it do to ensure that its expenditure plans are not breached?
Its first option is to impose rules on what universities can charge to ensure that £7,500 will be the average. The Office for Fair Access is the only game in town at the moment to do this, which is bizarre given that Offa was created largely to ensure that the more prestigious universities admit a fair share of poorer students.
It has not been much concerned with Leeds Metropolitan University, for example, but more interested in the University of Leeds. It would take some pretty gerrymandered rules to use this mechanism to stop Leeds Met from charging as much as Leeds, and indeed there are doubts about whether such a mechanism would be legal.
The criterion most often mentioned is dropout rates. That would affect Leeds Met more than Leeds; it would also be outrageous and presumably have half the coalition up in arms. It would have a direct impact on widening participation and provide a disincentive to universities recruiting students least likely to complete, who tend to hail from disadvantaged backgrounds.
The second option is to cut student numbers even further, enabling the government to reduce the amount it will lend under the new regime. But of course, that too would have a negative impact on participation.
The third is to make students pay more. There are any number of measures that the government could take to achieve this - for example, a higher interest rate or a lower effective threshold for repayment.
And the fourth is to cut even further than has been planned the residual grant that the government intends to pay to universities, or to slash other higher education expenditure, such as research funding.
We are in something of a mess at present. The government is reliant on market disciplines to stop prices spiralling out of control, but it lacks the means for creating a market, and so far there is no sign of any other plausible mechanism for moderating fees. The pressure is on to find some basis, however unsatisfactory, for keeping its costs down.