Academics are seeking access to 20 years' worth of student loan information that could be used by ministers to stop some universities from charging top-end fees.
Neil Shephard, professor of economics at the University of Oxford, is in negotiations with the Student Loans Company about gaining access to the records to aid his research into higher education funding models.
If permission is granted, the data would be used to look at the rate of loan repayment among graduates according to the institution they attended and the subject they studied.
In theory, this information could then be used to estimate the cost to the taxpayer - in terms of the amount of debt that graduates fail to repay - of lending money to students at particular universities and on specific courses.
This would allow the government to introduce a policy that stops individual universities from charging high tuition fees if they have relatively poor graduate employment records and their students draw large loan subsidies as a result.
Although Professor Shephard's work is independent of the government, David Willetts, the universities and science minister, is known to be interested in the policy implications of being able to estimate loan subsidies for individual universities and courses.
Writing in Times Higher Education this week, Mr Willetts says the percentage of student loan costs borne by the taxpayer - known as the Resource Accounting and Budgeting (RAB) charge - will be "at the core of university financing for many years".
"I expect that, in the future, as the data accrue, the policy debate will be about the RAB charge for individual institutions," he writes.
Economists have argued that under the finance system effective from 2012-13, universities have no incentive to charge fees of less than £9,000 because the taxpayer, rather than the university, carries the financial risk of students not repaying their loans.
If some of the risk could be shifted on to universities, a key benefit would be that student numbers could be expanded without additional cost to the Treasury, creating a truly competitive market in the process.
A proposal to charge universities a levy on high tuition fees was included in last year's Browne Review, but the government rejected the idea, partly because, Mr Willetts says, it was "indiscriminate and did not reflect the actual Exchequer risk from lending to students at specific universities".
The main sticking point on releasing the SLC data is the danger that it might be possible to link graduate repayments to specific individuals.
It is thought that a plan is being worked on whereby the information would be grouped into wider subject "bands" to prevent individuals from being identified.
Professor Shephard said the SLC had "entirely reasonable" confidentiality concerns, but he was in discussions with officials to resolve the problems.
An SLC spokeswoman said it was "working to provide (the information) to him once the necessary legal frameworks are in place to ensure all graduate data are anonymised".
Although prior to 2006 student loans were offered for living costs rather than fees, the wider system has been in place since 1990, and Professor Shephard said this meant that the data would provide a rich source of information on graduate repayment.
"If you can improve the level of research in this area then it creates better policy discussions," Professor Shephard said.
Along with Nicholas Barr, professor of public economics at the London School of Economics, the Oxford scholar has proposed a policy whereby institutions are charged an "insurance premium" to cover the cost of subsidising loans once fees rise above £7,000.
Professor Shephard said that the data from the SLC could also be used to improve general information for students on the projected income returns from attending different universities and studying specific subjects, but that this was not his primary motive for the research.
A paper published last month by the CentreForum think tank suggests that a true market could be created in higher education by allocating places using historical data on graduate employment and loan repayments, forcing universities to bid at realistic fee levels.