The devil is in the detail: why IPPR is right about loans

Two weeks ago, IPPR’s Commission on the Future of Higher Education published its final report – A Critical Path. There has already been a lively debate on these pages about some of the key findings of the report.

Most recently, Andrew McGettigan posted some detailed technical criticisms about how IPPR’s report treated the higher education funding system. This debate is very important because some of the details are critical to understand as they affect current and future policies. This piece will seek to answer Andrew’s criticisms, and take this debate another step forward.

The case for reforming the funding system

Before responding to Andrew’s detailed comments, it is important to remind readers of the main argument in our report about higher education funding, because I believe it is one that Andrew shares. Our central argument is that the current student support system is flawed and needs reform. We show that the system is storing up long term costs for the government, is an ineffective use of resources and is damaging for some groups of students. We also argue that there are a number of alternatives to the current system – for example lowering tuition fees or introducing a graduate tax – that could be introduced with little additional long term cost to the state. And we model the way that each of these options could be achieved in order to inform policy debates in the run up to the general election. This was done by London Economics, who have one of the most sophisticated models outside of government.

This modelling is only concerned with the long term economic cost of changing the student support system – exploring the aggregate flow of resources between government, graduates and higher education institutions over the lifetime of a student loan. However there are a number of quirks of government accounting that mean each of these options also comes with a bigger short-term impact on government spending. These short-term impacts could have a very important bearing on how a future government tries to reform the student loan system, as they lead directly to political decisions about government spending and deficit reduction in the next parliament. This is where Andrew’s criticisms come in.

How changes to the loan system impact government accounts

Andrew points out that the government sets aside an impairment to cover the estimated losses on student loans, something that is known as the RAB charge. He goes on to argue that IPPR has confused government accounting practices and that if the government is able to reduce the RAB charge ‘any estimated savings on the loans would be reflected in departmental accounts now’. The key implication being that if the government changes the terms of student loans in order to reduce the RAB charge, BIS would immediately be able to spend these resources on other things, such as teaching grants.

Unfortunately it is Andrew that is confused. While the RAB charge does appear as a line in BIS’s budget (DEL) and in its resource outturn, it is treated as ‘non cash’. Due to the unusual nature of the expenditure it is effectively ring-fenced and excluded from BIS’s ‘Total DEL’ and therefore Total Managed Expenditure [see note 1].  Contrary to Andrew’s claim, any long term saving on student loans does not automatically translate into cash that BIS can go and spend on other things in the short term. If BIS alters the repayment terms of student loans in order to save money, it would have to enter into a negotiation with the Treasury about whether it could switch this saving into ordinary grant expenditure [see note 2].

Any reform of student loans therefore hinges on whether the Treasury are prepared to take a long term view of the overall cost to the state, while accepting the short-term impact on the national accounts. In our report we therefore highlight both the long term and short term impacts on government spending, as this is what will be critical to these negotiations. A rational economic logic would assume that the government will only consider the long-term cost, and this forms the main basis of our modelling. However the Treasury may not take an open attitude to any changes that would be detrimental to the national accounts in the short-term, which is why we list these under the ‘disadvantages’ of each scenario.

Ultimately, any reform of the student support system will require politicians to make a decision about what spending they are prepared to sanction in the short-term, even if the long term economic costs are smaller. Unlike Andrew’s blog, IPPR do not try to bury this dilemma in the arcane language of government accounting rules.

Debt or deficit?

Andrew’s second criticism of IPPR’s report is that it ignores the impact that the student loan system has on government debt (PSND), as opposed to the deficit (PSBR). This is an important distinction and one that we make at a number of points in the report (see for example our discussion on p142 of the impact of introducing a new postgraduate loan system). However Andrew is wrong to place such a strong emphasis on debt [see note 3], and as a result he draws the wrong conclusions about the pros and cons of possible reforms to the loan system.

Andrew’s focus on government debt leads him to a different conclusion about the benefits of raising a top rate of interest on student loans. IPPR models a scenario where the government decides to raise the interest rate on loans for high earning graduates, in order to try and recoup more money from them. We point out several pros and cons of this scenario. One of the disadvantages that we list is that the policy may not bring in as much money as the modelling suggests. This is because once the interest rate reaches a very high level, graduates may be incentivised to borrow money more cheaply elsewhere and pay off their entire student debt straight away, in order to avoid paying the high interest rates.

Andrew argues that this view is ‘misinformed’. He thinks the Treasury would stand to benefit from a graduate repaying their loan early as it will help to pay down the stock of debt. But he misses the point of raising the top rate of interest. If the government raised the top rate of interest to 4.5 per cent it would mean many high earning graduates would actually ‘overpay’ the total value of their loan. We estimate that once discount rates have been taken into account, the highest earners would overpay by around 20 per cent, helping to recuperate £679m more for the Treasury from graduates. It would be far better for the Treasury if high earning graduates took this route. If they followed Andrew’s logic, and opted to repay their loan in a lump sum, they would deprive the Treasury of these additional payments. While it would help reduce the debt stock in the short term (which is Andrew’s focus), it would actually prevent the Treasury from making additional gains in the long term. This is why raising the top rate of interest may not actually be such a good idea. Incidentally, it may also fall foul of legal regulations about whether such payments could be classed as a loan.

Back to the big picture

Given Andrew made a number of very technical claims about the accuracy of IPPR’s report, it has been necessary to descend into a detailed discussion of government accounting practices. Ultimately, however, I think we agree on the bigger picture: that the current student support system is not fair or sustainable and that it will need to be reformed in the future. IPPR provided a series of costed alternatives to the current system and set out the trade-offs that will face policymakers when they come to choose between them. Our intention was to inform debates about reforming student finance in the run up to the general election. In that, at least, we seem to have succeeded.


  1. While Andrew is right that the impairment (RAB charge) appears as a line in BIS’s DEL and total outturn, he has mistakenly assumed that it can be treated like any other line of grant expenditure. The ‘noncash’ nature of the as can be seen from this spreadsheet – note that the total figure for knowledge and innovation excludes the line for student loans. Neither does the impairment count directly towards Total Managed Expenditure, as explained on page 15 of this note from the House of Commons library.
  2. The RAB charge is treated differently from other lines of expenditure on the BIS accounts because it ultimately has a very different impact on the National Accounts. The RAB charge that appears on BIS’s DEL is essentially a device put in place by the Treasury to help them contain long term costs. Generally speaking, if BIS want to lower the RAB charge and use that to cover grant expenditure they will have to get Treasury approval, because it could have an impact on the National Accounts in the short-term. On the other hand, if the RAB charge increases, BIS can use savings in grant expenditure to compensate for that, as the short-term impact on the national accounts would not be a problem for the Treasury. That is why on page 125 of our report we say that if the RAB charge increases it will add pressure on BIS to cut current expenditure in other areas, but conversely why on page 134 we show that if BIS reduces the RAB charge it cannot automatically assume that this long term saving to the state can be used to fund increased grant expenditure in the short-term. It would have to agree this with the treasury first.
  3. Andrew argues that the Treasury is more concerned about the overall stock of debt that results from student loans (the impact on PSND). Of course this is important, as having high levels of government debt will increase interest payments and ultimately the government will want to see PSND fall. But having high levels of debt is not such a big problem if the government is confident that it will be able to service this debt. In the case of student loans it can be confident of doing this because it knows it will receive graduate repayments in the future. To use an everyday analogy, it doesn’t matter if a household takes on a substantial loan for a mortgage, provided it is able to pay the money back further down the line. This is why the government is more concerned about the proportion of loans that will ultimately go unpaid – as this represents the long term cost of student loans to the state. This is the figure that impacts the deficit and that the Treasury pays most attention to. None of this is to say that debt levels are not important – it is just a question of emphasis.

3 responses to “The devil is in the detail: why IPPR is right about loans

  1. Thanks to Jonathan for the reply. I don’t think it responds to my substantial points that the IPPR report only tells part of the story because it largely ignores the impact of the loans on public debt and because it confuses accounting conventions governing departmental budgets (accruals) and national accounts (cash in, cash out).

    But I will make two points of clarification here:

    1.Departmental Budgets
    “Contrary to Andrew’s claim, any long term saving on student loans does not automatically translate into cash that BIS can go and spend on other things in the short term.”

    I do not make this claim. Each year for new loan issues, BIS is required to set aside a ringfenced ‘impairment’ in the departmental expenditure budget. A reduction here would not ‘automatically translate into cash’, the budget holder would be required to make the case that they should see the benefit from the additional resource. It is much a political negotiation as anything. Liabilities have to be accounted for somewhere, but a budget holder who reduces a liability in their own budget should be in a position to make such a case.

    The key point from my previous post is that in an accruals system the impairment is recorded in budgets at time of issue.

    I agree that the impairment is non-cash from a national accounts perspective and indeed this was explained in my original post.

    2. Debt

    “ It would be far better for the Treasury if high earning graduates took this route. If they followed Andrew’s logic, and opted to repay their loan in a lump sum, they would deprive the Treasury of these additional payments. “

    This is not my logic – as can be seen from this post for the Guardian.

    My basic point was that the IPPR fails to grasp the Treasury position and hence the political contours of student loans today by concentrating on the ‘deficit’, rather than the ‘debt’

Leave a Reply