Student loans: how much are they costing?
"Everybody's now saying the government isn't going to get the money back anyway so what's the point".
David Dimbleby, BBC Question Time, 10 April 2014
"You're not comparing apples with apples here David".
Sajid Javid MP, BBC Question Time, 10 April 2014
Last week, an Institute for Fiscal Studies report estimated that 73% of graduates will not earn enough to pay back their loans in full - a figure briefly mistaken as an upward estimate of the government's 45% on last week's Question Time.
The figures are referring to separate things, one referring to the public cost of student loans and the other to the long term costs of the loans to graduates. As forecasts, if the economy or the graduates act differently to how the models predict then the reality could be quite different.
Government: Public subsidy to loans is increasing
When the cap on university fees was increased from £3,000 to £9,000 a year, it was intended to shift more of the costs of higher education funding from the central budget onto individuals.
But, the government have always said that a proportion of the money loaned to students is not expected to be repaid. This is known as the Resource and Accounting Budget (RAB) charge, estimates of which are produced by the government and are factored into public expenditure as a permanent cost to the taxpayer.
In March, the government estimated that the RAB now stands at around 45% - up from an estimate of 40% in December 2013, and from 30% in February 2012. This means 45% of the amount spent on loans in any given year will be accounted for as permanent expenditure in the annual budget.
There are a number of factors affecting this estimate. Not all graduates will earn over £21,000 - the minimum earnings threshold for repayments - or they may take a while to reach that level. Not all will pay the full level of interest on the loan either, which depends on the level of earnings. This means that some will not finish repaying the loan within 30 years, after which the remaining balance is written off. People who develop a long-term disability will also stop repaying, and some may die.
Some of the money will also be 'lost' because of inflation, as writer and researcher Andrew McGettigan explains in his blog.
The National Audit Office and the House of Commons Public Accounts Committee have criticised the government's modelling used to predict repayments, finding actual repayments to be less than those forecast. For example, they say in 2011-12 £1.28 bn was forecast to be repaid, but £1.17 bn was collected.
The Government have said that they continue to review the modelling they use and that "by its nature an estimate is subject to change as it is highly dependent on macroeconomic circumstances, and the growth of graduate earnings over the next 30 years".
IFS: 73% of graduates will not repay the full loan
The Institute for Fiscal Studies' report used their own projections of future graduate earnings.
They predicted that most graduates will continue paying back their student loan until their early 50s - because the salary at which graduates start repaying their loan has increased to £21,000, and because of the higher debt that students will graduate with.
This could mean that, although students now have 30 years in which to repay the loan (up from 25 years under the old system and after which the amount is 'written off'), the majority - 73% - still will not have repaid the full amount before the 30 years is up.
They told us they're publishing a further report in the next few weeks which will look at how the government model differs to their own so we'll update this piece then.
A lower estimate of 60% of students under the post-2012 funding system expected to have some or all of their loan written off was recently published by the government, based on its own modelling. They told us the reason for the difference to the IFS' estimates was due to the models making very different assumptions in their forecasts.
The IFS has also now published their second report, which discusses some of the differences in the modelling. It also makes its own estimate of the 'loan subsidy' but notes that the report's estimate and the government's RAB charge are not directly comparable.