Treating Students Fairly: The Economics of Post-School Education Contents

Chapter 10: Student loans and the public accounts

340.There are suggestions that the presentation of student loans in the national accounts was a factor behind three recent policy decisions: the raising of tuition fees in 2012, the setting of the interest rate on post-2012 loans and the sale of the pre-2012 student loan book. This chapter examines the extent to which the national accounting is influencing the financing of higher education and masking its true cost.

341.This chapter is intended to be accessible to the general reader and avoids accounting terminology where possible.

Measuring the value of student loans and cost to the Government

342.Before examining the policy decisions mentioned above, it will be useful to outline how much the Government is currently lending to students and how much this is expected to cost the Government.

Lending to students and expected write-offs

343.For the 2016/17 academic year, the Government issued £13.6 billion of new student loans. When this is added to the student loans issued in previous years which have yet to be paid back (the ‘student loan book’), the Government holds £88.8 billion worth of outstanding student loans (this is referred to as the ‘face value’ of the student loan book).

344.Not all of these loans are going to be repaid, since any amount outstanding after 30 years is written off. The Government acknowledges this outcome and estimates what proportion of the loans issued for the given year will not be repaid. For 2016/17, the Government estimated that £3.9 billion of the £13.6 billion issued will not be repaid.

345.For the full loan book of £88.8 billion, the Government expected that £27.5 billion will not be repaid (around 30 per cent of the face value of the total loan book, a percentage referred to as the ‘impairment rate’341). The Government therefore valued the student loan book at £61.3 billion (this is referred to as the ‘carrying value’ or the ‘face value’342).

Raising the repayment threshold to £25,000

346.The expected loss on student loans will increase as a result of the Government’s decision to raise the repayment threshold on student loans issued after 2012. From April 2018, graduates repay 9 per cent of their salary over £25,000 a year343; the threshold was previously £21,000 a year. This increase will reduce loan repayments.

347.Official figures for 2017/18 are not yet available but the Government said last year that the proportion of the face value of loans issued that will be not be repaid will increase from 30 per cent to between 40 and 45 per cent.344 In work carried out for the Committee, the Institute of Fiscal Studies estimated this would be 48 per cent.345 In other words, around half of the student loans issued by the Government each year will not be paid back.

Cost to the Government of funding higher education through student loans

348.For the 2017/18 academic year, the Institute for Fiscal Studies estimated that the Government would provide around £16 billion of student loans and £750 million of grant funding. Taking into account that around half of these loans will not be paid back, they estimated the cost to the Government of funding the 2017/18 cohort would be £8.4 billion.

349.If the repayment threshold had not been raised, the Institute for Fiscal Studies said the cost of funding the 2017/18 cohort would have been £5.6 billion. Dr Gavan Conlon said London Economics had estimated the long-run taxpayer cost of the change was £2.85 billion a year.346

Effect on the national accounts

350.Because student loans are classified as a ‘loan’ for national accounting purposes, the amount of loans issued each year is not counted as Government expenditure. Only write-offs on loans are counted as expenditure. This means that student loans only affect Government expenditure when the write-offs are made after the loans expire at the end of their 30 year term (student loans are written off 30 years after the April following graduation). In other words, the losses made on students graduating in the summer of 2018 will only be recognised in the national accounts in 2049/50.

Future impact on the deficit (public sector net borrowing)

351.In January 2017, when the impairment rate on student loans was estimated to be around 30 per cent, the Office for Budget Responsibility said that from the mid-2040s, the write-offs on student loans would be worth around 0.3 per cent of GDP every year (0.3 per cent of GDP in 2017/18 is around £6 billion). As the expected impairment rate is now closer to 50 per cent, this percentage will now be higher.

Future impact on the national debt (public sector net debt)

352.The Office for Budget Responsibility estimated in January 2017 that the student loan book would be worth 11 per cent of GDP in the late-2030s, an increase from around 5 per cent of GDP in 2017/18. They predicted this would fall back to around 9 per cent of GDP by 2066/67. The Department for Education have forecast that the total student loan book will be worth £1.2 trillion in nominal terms (£473 billion in 2018/19 values) by 2049/50. The letter in which the Minister for Universities and Science explained this to the Committee is included in Appendix 6.

2012 funding reforms

353.The difference between Government income and Government expenditure is known as ‘public sector net borrowing’ and is commonly called ‘the deficit’. In June 2010, the coalition government set itself a target of achieving a “cyclically-adjusted current balance” by 2015. To achieve this target it needed to reduce the deficit.

354.The coalition government changed how higher education was funded: rather than a mix of government grant and tuition fees, from 2012 the funding would be provided almost entirely through increase tuition fees, for which larger student loans would be available. Funding higher education through government grant increases government expenditure and the deficit; funding through tuition fees and student loans does not.

355.Table 15 shows the difference between grant funding and loan funding before and after the 2012 reforms.

Table 15: Upfront Government spend on higher education, comparison between 2011/12 and 2012/13 cohorts

Government spend

2011/12 system

2012/13 system

Funding via student loans

£8.4 billion

£15.2 billion

Funding via grants to institutions

£6.4 billion

£2.6 billion

Total upfront spend

£14.8 billion

£17.8 billion

Source: IFS, Options for reducing the interest rate on student loans and introducing maintenance grants: https://www.ifs.org.uk/uploads/BN221.pdf [accessed 24 May 2018]

356.The total upfront spend by the Government was £3 billion higher following the changes but as the spending on grants was £3.8 billion lower, the deficit would have been improved by £3.8 billion.

357.Dr Conlon, when he appeared before the Committee in July 2016, explained the effect that this had on the national accounts:

“the taxpayer is essentially paying through loan write-offs instead of [government grant]. That makes the system a little riskier from the perspective of higher education institutions, but it also kicks the debt down the road towards the next generation. Instead of the Government or the taxpayer paying for higher education now, it will be paid for over the next 30 years, also predominantly by the taxpayer.”347

358.Lord Willetts said that the 2012 changes “took higher education out of public funding, that releases public funding for areas that clearly are in much greater need”.348 Paul Johnson, the Director of the Institute for Fiscal Studies, however said the changes simply “delayed the point in the national accounts at which public funding kicks in.”349 Lord Willetts conceded that the public finances were a factor in the changes: “let’s face it, there were also public expenditure demands.”350

Interest rate on post-2012 loans

359.Interest is charged on post-2012 student loans at RPI plus 3 per cent during study, until the April after the person leaves the course. Following that April, the interest rate used is between RPI plus 0 per cent and RPI plus 3 per cent, depending on a person’s salary.

360.The interest rate on the loans changes every September in time for the forthcoming academic year.351 For 2017/18, RPI was 3.1 per cent for the purpose of calculating the interest rate on student loans. This meant students studying at university, or graduates earning over £41,000, were charged 6.1 per cent interest on their student loan. For 2018/19, the rate of RPI to be used will be 3.3 per cent (meaning the maximum rate of interest charged will increase to 6.3 per cent).352

361.Lord Willetts told us that this interest rate was “brought in to get higher repayments from well-paid graduates.” He accepted that this was not made clear to those graduates.353 At his annual evidence session with the Committee in September 2017, the Chancellor of the Exchequer said:

“It is perfectly well understood that the design intention of the student finance scheme is that there is an element of redistribution in it. Higher earners pay a higher rate of interest on their loans than lower earners. The system is designed such that lower earners will be forgiven the balance of their loan after a certain number of years. It is a very different animal from a loan that one would take from a bank or a building society. Sometimes that is not understood clearly enough.”354

362.We consider in Chapter 8 how the system is not as progressive as the Government claims.

363.The high interest rate may have been influenced by the fact that interest on student loans is recorded as income by the Government as it accrues. The Office for Budget Responsibility estimated that the accrued interest on the student loan book will be £3 billion in 2017/18. This is forecast to rise to £7.5 billion by 2022/23 and will carry on increasing as the proportion of post-2012 loans in the loan book increases (this is because the interest rate on pre-2012 loans is much lower: 1.25 per cent for 2017/18). Table 16 shows the forecast for accrued interest on student loans alongside the forecast for public sector net borrowing (the deficit); by 2021/22 the accrued interest will be a substantial proportion of public sector net borrowing.

Table 16: Office for Budget Responsibility’s forecasts for accrued interest on student loan and public sector net borrowing, 2017/18 to 2021/22

2017/18

2018/19

2019/20

2020/21

2021/22

Forecast accrued interest

£3.2 billion

£4.7 billion

£5.6 billion

£6.7 billion

£7.5 billion

Forecast public sector net borrowing

£45.2 billion

£37.1 billion

£28.7 billion

£26.0 billion

£21.4 billion

Source: Office for Budget Responsibility, Economic and Fiscal Outlook, Cm 9572, March 2018: http://cdn.obr.uk/EFO-MaRch_2018.pdf [accessed 24 May 2018]

364.As with the changes to funding, the high interest rate makes it easier for the Government to achieve its target to reduce the deficit. The present Government has a target to reduce the deficit to below 2 per cent of GDP by 2020/21.355

365.The Office for Budget Responsibility said in the March 2018 Economic and Fiscal Outlook that “much of the accrued interest will eventually be written off rather than repaid, so the National Accounts methodology for measuring interest does not reflect fiscal reality.” They described the recording of the accrued interest as income as a “fiscal illusion”.356

366.Dr McGettigan described the accounting as “unexpected, if not bizarre, because we know that the interest accruing is very unlikely to be repaid, so we are scoring the receivable as income before we have necessarily received it, and we anticipate that we may never even receive it.”357

367.Paul Johnson said he would not be surprised if these flattering effects on the public finances were part of the motive for the Government’s choice of rate, “but I really can’t comment on it.”358 Lord Willetts said that “I was in the room at the time. I can tell you that there was pressure from our coalition colleagues to make this as progressive as possible.”359

Selling the student loan book

368.In the 2013 Autumn Statement the coalition Government announced plans to sell the pre-2012 student loan book. The sale of the first tranche of loans was concluded in December 2017. Loans with a face value of £3.5 billion were sold for £1.7 billion, with £1.8 billion (51 per cent of the face value) written off. The Government plans to sell off £12 billion of loans over the next five years.

369.Why was the Government prepared to make a sale at such a large loss? It prefers cash today over a larger sum of cash tomorrow. This preference is a result of another of the Government’s targets: to reduce the national debt. The present Government is aiming to see the national debt (referred to as ‘public sector net debt’) falling as a percentage of GDP by 2020/21.

370.When student loans are issued, the value of the loan is added to the national debt. When repayments on the loans are made, the national debt is reduced accordingly. If the loans are sold, the value of the sale is taken off the national debt. As repayments of the loans takes place over 30 years, the national debt is reduced much more quickly by selling the loans, even at a substantial loss.

371.For the Government to sell assets such as student loans, the sale must pass HM Treasury’s value for money test. The test is skewed to reflect the Government’s preference for cash today: a sale can pass this test despite it taking place at a price far below the value the Government itself places on the asset.360

Effect of the sale on the national accounts

372.The sale of the loans means effectively that the Government has brought forward the write-offs on the loans and recognised them today. But a sale means that the write-offs are never recognised in the deficit (if the Government had held onto the loans, any write-offs would have been recognised at the end of the loan term). In a February 2018 report, the House of Commons Treasury Committee drew attention to the size of this difference if that level of loss remains the same for the rest of the planned sales:

“If the rate of losses on these sales is maintained, billions of pounds of student loan losses will be crystallised without having any impact on the deficit. Its inclusion would increase the deficit as forecast by the Office for Budget Responsibility (OBR) by 13 per cent, from £45.5 billion to £51 billion.”361

373.The accrued interest on the loans that were sold was nevertheless counted as income while they were held by the Government. The Government had likely received little of the interest payments on these loans before the sale. These interest payments, if ever made by the borrowers, will now be paid to the purchaser of the loans. However, the Government will not make a deduction in the national accounts to reflect the fact that the accrued interest it has already counted as income will never be received. As the interest rate is low on pre-2012 loans, the accrued interest will not be a large figure. Should the Government pursue sales of post-2012 loans, which have a higher rate of interest and lead to the Government recording substantial amounts of accrued interest as income, the distorting effect on the national accounts could be substantial.

374.The Chancellor of the Exchequer explained the Government’s thinking to us in September 2017:

“It is the Government’s intention, where they find that they hold assets on the public balance sheet for which there is no policy or strategic reason, to realise those assets and thus reduce public sector debt, helping us to achieve our debt targets and/or create capacity to do other things in line with policy priorities.”362

375.When we put this explanation to Paul Johnson, he said that “it was hard to know why in any rational world it would give you more space … There is a short-term benefit to the measured financial debt, but that is an illusion created by the slightly odd way in which we look at the public finances.” He concluded that most policy changes in student loans “move in the same helpful direction of short-term public finances at the expense of the long-term public finances.”363

376.Recent changes to higher education financing have been motivated mainly by the desire to lower the deficit.

377.The decision to switch almost all higher education funding to tuition fees hides the true cost of public spending on higher education. When the change was made in 2012, the upfront spend by the Government on higher education increased by £3 billion but as the vast majority of funding was provided through loans rather than grants, the deficit figure was improved by £3.8 billion. Write-offs on student loans will be included in the deficit only when the loans expire in 30 years; if the loans are sold before that point, the write-offs never hit the deficit.

378.The high rate of interest on student loans creates the illusion that Government borrowing is lower than it actually is. It was presented as a progressive measure but in reality, the motivation appears to have been the flattering effect that accrued interest on those loans will have on the deficit.

379.Future governments will have to adjust spending plans to recognise historic student loan losses: in today’s money, that would mean the 2047/48 government having to find an extra £8.4 billion to cover expected losses on the 2017/18 student loans. Alternatively, a future government may attempt to abandon the use of public sector net borrowing as a measure of the strength of the economy. It is unacceptable to expect future taxpayers to bear the brunt for funding today’s students.

380.The total upfront spend by the Government was £3 billion higher following the changes but as the spending on grants was £3.8 billion lower, the deficit would have been improved by £3.8 billion.

Different ways of accounting for student loans

381.The public accounts will increasingly present an inaccurate indication of the strength of the economy. Dr McGettigan said student loans will have “an increasingly distorting effect on those statistics.” That is particularly the case once the write-offs begin to show in the public accounts. This section will consider other ways in which student loans could be recorded in the public accounts.

382.Dr McGettigan said the problem was with the use of statistics by the Government:

“It is perhaps not so much a question of reforming national accounts as taking a step back and realising the presentational force of the headline statistics. The fiscal mandate is what the Government present to the country as, “Judge our macroeconomic and fiscal competence on hitting these targets”. The issue is the statistics in the targetry rather than whether you should re-categorise student loans as income or expenditure in different ways.”

383.The Office for National Statistics said it was their responsibility to compile the deficit (public sector net borrowing) and the national debt (public sector net debt) in a way that ensures that “the underlying economic reality of transactions and financial instruments is followed when deciding whether or not a particular transaction should impact [the statistics] … ONS reaches these decisions in compliance with the international (UN) statistical guidance for National Accounts.”364

384.They said that “it might be conceivable” to consider student loans in a different way. For example as a series of transfers out of government, at the point of loan issuance, and into government when interest and capital repayments are made. This would mean that all cash paid when issuing the loans would be recorded as government spending, affecting the deficit. They were, however, “firmly of the view” that the economic nature of student loans closely matches the definition of a loan in the international accounting rules used by the government.

385.Eurostat said that the UK’s recording of student loans in the public accounts was in line with the rules prescribed by the applicable European accounting rules in relation to the accounting of “standard loans”.365

386.They described two different ways of treating student loans issued by governments, which depend on the amount repaid:

(1)“When student loans are provided by government (or on behalf of government) and are expected to be largely repaid, a loan asset is recognised in GFS/EDP data. When debt cancellations occur, these impact net lending at the time of debt cancellation.

(2)When student loans are provided by government and are mostly not repaid, a capital transfer impacting net lending should be recognised at the time the loan is granted. Any recoveries should impact net lending at the time of recovery.”366

387.With the Government expecting just under half of loans to be repaid at present, student loans in England may be getting close to the second scenario. The House of Commons Treasury Committee recommended:

“Loans that are intended to be written off are, in substance, a partially repayable grant rather than a loan. The ONS should re-examine its classification of student loans as financial assets—which they are in legal form—and consider whether a portion of the loan should, in substance, be classed as a grant.”367

The Office for National Statistics announced subsequently that “to consider the treatment of such financial assets and the accounting issues they raise, we have begun work with international agencies and other National Statistical Offices.”368

388.Most student loans will not be repaid in full: some will be paid in full, some not at all, and a lot only partially repaid. The expected write-offs should be shown in the deficit when the loan is issued. The true cost of funding higher education would then be immediately apparent. This would allow for a better discussion as to where funding in the higher education system should be allocated.

Cost of our proposed changes to student loans and maintenance and the effect on the national accounts

389.We asked the Institute for Fiscal Studies to calculate the cost of our proposals to reduce the interest rate on student loans and restore maintenance grants.

Cost of reducing the interest rate

390.Table 17 compares the cost of funding the 2017/18 cohort through the current system, with an interest rate of RPI plus 0–3 per cent, with the cost of our proposed system, with an interest rate equal to the 10-year gilt rate (currently around 1.5 per cent). The change does not affect the deficit but it increases the proportion of the 2017/18 loans which would not be repaid.

Table 17: Cost of reducing the interest rate on student loans to the 10-year gilt rate

Present system (RPI plus 0–3%)

Proposed system (the 10-year gilt rate)

Upfront funding via student loans

£16 billion

£16 billion

Upfront funding via grants (counted in the deficit)

£0.7 billion

£0.7 billion

Value of student loans in 2017 prices which will not be repaid

£8.4 billion

£10.4 billion

Source: Institute for Fiscal Studies (see Appendix 7). The simulation by the Institute for Fiscal Studies assumed a two per cent nominal interest rate in the long-run (the Institute for Fiscal Studies were asked originally to model the results of lowering the interest rate to CPI).

391.The lower interest rate means that a smaller proportion of the £16 billion issued is ever repaid. This is because higher earning graduates are making fewer interest payments on their loans, which under the present system are helping subsidise the losses made on loans issued to lower-earning graduates.

Cost of restoring maintenance grants

392.Table 18 compares the cost of restoring maintenance grants of £3,500 a year to students with household incomes of less than £25,000, tapered between £25,000 and £45,000, with the cost of supporting students through maintenance loans under the present system (for the 2017/18 cohort).

Table 18: Cost of restoring maintenance grants

Present system (no maintenance grants)

Proposed system (£3,500 grants per year, tapered between £25,000 and £45,000)

Upfront funding via student loans

£16 billion

£16 billion

Upfront funding via grants (counted in the deficit)

£0.7 billion

£2.4 billion

Value of student loans in 2017 prices which will not be repaid

£8.4 billion

£8.8 billion

Source: IFS, Options for reducing the interest rate on student loans and introducing maintenance grants: https://www.ifs.org.uk/uploads/BN221.pdf [accessed 24 May 2018] and and Institute for Fiscal Studies (see Appendix 7).

393.As maintenance is paid as grants rather than loans, Government spending (which counts towards the deficit) is increased by £1.7 billion. The removal of maintenance loans decreases the total borrowing of students. Those students have a lower level of borrowing on which interest is charged. The Government therefore receives fewer repayments and the value of loans which will not be repaid for the 2017/18 cohort increases by £400 million (in today’s prices).

Combined cost of the proposals

394.Finally, the Institute for Fiscal Studies calculated the combined cost of these proposals.

Table 19: Cost of lowering interest rates and restoring maintenance grants

Present system

Proposed system

Upfront funding via student loans

£16 billion

£16 billion

Upfront funding via grants (counted in the deficit)

£0.7 billion

£2.4 billion

Value of student loans in 2017 prices which will not be repaid

£8.4 billion

£10.8 billion

Source: Institute for Fiscal Studies (see Appendix 7)

395.The combined effect of the proposals on the national accounts is to increase Government spending (and therefore the deficit) by £1.7 billion, and to increase the amount of the 2017/18 loans that will not be repaid by £2.4 billion (in today’s prices).


341 The impairment rate is referred to in the Department for Education’s accounts as the ‘Resource and Accounting Budgeting’, or ‘RAB’, charge.

342 The Department for Education says that the carrying value of the loan book is a reasonable approximation of the fair value, “in the absence of an active market, readily observable market trends or similar arm’s length transactions.”

343 The threshold will rise in line with average earnings for subsequent years.

344 Q 52 (James Bowler)

345 See Appendix 7.

346 Q 27 (Dr Gavan Conlon)

347 Q 1 (Dr Gavan Conlon)

348 Q 6 (Lord Willetts)

349 Q 6 (Paul Johnson)

350 Q 3 (Lord Willetts)

351 The rate of RPI used is RPI for the year up the previous March.

352 Office for National Statistics, ‘Consumer price inflation tables’, 23 May 2018: https://www.ons.gov.uk/economy/inflationandpriceindices/datasets/consumerpriceinflation [accessed 25 May 2018]. RPI was 3.3 per cent in March 2018.

353 7 (Lord Willetts)

354 Oral evidence taken before the Economic Affairs Committee on 12 September 2017 (Session 2017–19), Q 4 (The Rt Hon Philip Hammond MP)

355 The latest Charter for Budget Responsibility requires the “structural deficit (cyclically adjusted public sector net borrowing) to lie below 2 per cent of GDP by 2020/21”.

356 Office for Budget Responsibility, Economic and Fiscal Outlook, Cm 9572, March 2018: http://cdn.obr.uk/EFO-MaRch_2018.pdf [accessed 24 May 2018]

357 Q 26 (Dr Andrew McGettigan)

358 Q 9 (Paul Johnson)

359 Q 9 (Lord Willetts)

360 The sale passed HM Treasury’s value for money test because of the discrepancy between the discount rate used to work out the fair value of the future cash flow from the loans (RPI plus 0.7 per cent) and the discount rate used to value the asset for the value for money test (RPI plus 3.5 per cent). The future cash flows from the student loans are valued at a lower rate under the higher discount rate, and therefore a sale today will look more attractive.

361 Treasury Select Committee, Student Loans (Seventh Report, Session 2017–18, HC 478)

362 Q 1 (The Rt Hon Philip Hammond MP)

363 Q 10 (Paul Johnson)

364 Written evidence from the Office for National Statistics (HFV0103)

365 European Commission, ‘European System of Accounts: ESA 2010’ (4 December 2013): http://ec.europa.eu/eurostat/web/products-manuals-and-guidelines/-/KS-02-13-269 [accessed 24 May 2018]

366 Written evidence from Eurostat (HFV0104)

367 Treasury Select Committee, Student Loans (Seventh Report, Session 2017–18, HC 478)

368 Office for National Statistics, ‘Public sector finances, UK: March 2018’, 24 April 2018: https://www.ons.gov.uk/releases/ukpublicsectorfinancesmar2018 [accessed 30 May 2018]




© Parliamentary copyright 2018