Skip to content

Interview with Real Media – parts 1, 2 & 3

3 x 10min films on students loans and related topics.

Critical Education

Earlier in January, I saw down with Real Media for a couple of hours to talk about higher education reform and student loans.

Real Media have now released their first edited video. It’s ten minutes long and from somewhere in the middle of that two hours. I’m not sure how much more they are planning to release at this stage but I’ll put updates here.

The discussion in this segment focuses on how English HE is moving towards a funding system based on creditworthiness (of individuals and institutions). More detail on my theory of ‘financialisation’ can be found in the final chapters of The Great University Gamble.

Update – 6 Feb 2017

Part 2 of the interview is now available. This 10-min segment covers debt: student loan & commercial debt and why comparisons with the US are mostly misleading.

Part 3 of the interview – which appeared on 7 Feb…

View original post 30 more words

HESA data – HNCs & HNDs

HESA has today published data on enrolments and qualifications achieved for students at alternative providers designated for student support.

I have a particular interest in the data for students registered on Pearson’s HNC and HND qualifications, which were the focus of a National Audit Office investigation in 2014.

Today’s data shows that nearly 15,000 full-time students were registered on an HNC or HND in 2015/16 – 8200 of them were first years. This compares to 22,605 enrolments in the previous year when the split between 1st and 2nd years was 9500:13 100.

So numbers continue to fall since the high point of 2013/14, when roughly 40,000 such students were in receipt of loan funding.

In terms of qualifications, 4180 were achieved by full-time students in 2015/16. Unfortunately – for reasons unknown – HESA does not separate HNCs from HNDS here. An HNC is level 4; and HND level 5. It is important to be able to assess how many FT first years are achieving HNCs and also how many students go into the second year, aim for an HND but only achieve an HNC.

The best possible scenario from the HESA data is that 4180 HNDs were awarded from a pool of roughly 6 800 candidates. This though is unlikely.

The data needs to be improved but perhaps can go back to the 2014/15 data in light of the new figures. Last year, 4300 qualifications were achieved.

Since we can make a reasonable assumption that the 9500 FT first years in 2014/15 translated into 6800 second years, there is a pool of 2700 students who might have achieved HNCs in 2014/15. This would mean that a ‘rough achievement rate’ for 2014/15 was in the region of 27 per cent: 4300 qualifications and 15,800 students leaving study.

Obviously a bit rough and ready, but the HESA data isn’t satisfactory on this score. Just a reminder: each full-time student could be accessing £6000 per year in tuition fees and c. £10,000 in maintenance support.

Sale of loans – media coverage a success for government

The government will be pleased with the way the press covered its announcement about selling student loans yesterday.

Firstly, it was keen to impress on borrowers that there would be no changes to their terms and conditions and that SLC and HMRC will continue to collect payments as they do now. In regard to the latter point, this sale is therefore unlike the sales of mortgage-style loans in the late nineties and 2013, when the outstanding balances were sold to specialists in debt collection. Mortgage style loans were collected solely by SLC through direct debit and the government declared a profit on the deal, because the purchaser decided it would pay more than the fair value of the loans in the government accounts – because the purchaser believed they could do a better job of collecting outstanding payments and crosschecking the information people put on their loan deferment forms with other financial declarations.

The NUS and others have been targeting a potential sell-off by warning borrowers that the purchasers might have the power to change terms and conditions. As I’ve said on here recently, that’s not the case. Legislation prevents third parties from doing that even if they purchase loans outright (and we’re now talking about a securitisation not outright sale). And government will avoid future changes by fixing terms in contracts – it is otherwise hard to get purchasers to play ball: they want as much uncertainty removed as possible so that they can work on a price.

Here’s the key point – the repayment threshold for pre-2012 loans is now permanently indexed to RPI to enable a price to be agreed. The preparations for the sale of pre-2012 loans have improved the situation for borrowers – a higher threshold means lower repayments.

So why does it look like the sale of loans is a bad deal? Because it’s probably a bad deal for the public finances. Which leads to the second point.

The government knows it is likely to make a loss on the loan sales. A profit or loss – which will crystallise in various measures of the ‘deficit’ – is determined by the price it receives compared to the fair value of the loans (in the accounts).  Every major outlet – including the FT – has taken the government’s lead and written about the proceeds likely to be lower than face value.

This isn’t news – we know loans aren’t worth the face value. That’s what all the coverage on non-repayment is about – the government thinks it loses 25p on every new £1 loaned, so the loans can’t be worth their face value.

The issue is the fair value. And as I said in the previous post, whether the government makes a loss or profit is not the same as whether a ‘value for money’ test is passed. The government has a skewed VfM test that values present money higher than the fair value does – that means it will determine a lower value for the same projected cash stream.

The government recently lowered the discount rate used in the fair value calculation of loans (from RPI + 2.2 to RPI + 0.7). It hasn’t done the same for the ‘VfM’ discount rate which starts at RPI + 3.5 (1 percentage point is then deducted for catastrophe risk and a risk element specific to loans is added back in).

This ‘social time preference’ rate (the ‘Green Book’ rate) hasn’t been changed for a decade and supposedly captures society’s preference for cash today over future cash.  In essence it captures a short-termism in government thinking: ‘the polite expression for the rapacity’ (Harrod) with which present reason views the future.

In student loan sale coverage, there’s a great deal of difference between following the press release prompts to observe that “of course, the price won’t match the face value” and asking whether the loans have been sold for less than the government says they are worth in the accounts.

A late aside: even if the government is successful at this point and manages to raise £12bn over the next 4 years without losing too much, we’re still a long way from where we are meant to be.

The original plan was to sell new loans as they are issued. This protracted retrospective sale is a fallback option – the main goal is still to get there. It’s easy to see why: that £12bn over 4 years will be dwarfed by this year’s £15bn+ of new loans. New issuance is set to pass £20bn before the end of parliament.

Government put loan sale hopes in securitisation

This morning, Jo Johnson set a written statement before parliament in which he confirmed that the government is ‘starting the process to sell part of the English student loan book’. (A sale will not affect borrowers from the rest of the UK).

Although the intention to sell has been announced several times in recent years, what today marks is the start of a formal process through which government tries to woo purchasers with a specific set of products. It may now take months before a final decision on a sale is made and any proceeds are announced. This decision is subject to a value for money test being passed, but this test can be passed if even the sale makes a significant loss for government (more on this below).

What is most significant about today’s announcement is that the government has shifted its position regarding the sale structure.

In 2014, when Vince Cable (then Secretary of State for Business), vetoed a sale, the government was proposing to sell whole cohorts of graduates  to pension funds and insurance companies.

Now, the government is still looking at the ‘pre-2012’ loans where repayment started between 2002 and 2006. These loans will form the backbone for the first sale in a planned four-year programme. But it is no longer looking for buyers to take on whole cohorts – it is instead looking to securitise those loans.

This means the government is looking to create and sell a set of tradeable notes (‘securities’) based on the rights to income steams which will be backed by the future repayments associated with those student loan accounts. As I understand it, the sale will offer purchasers five different types of note, which will vary by seniority, maturity and coupon (interest rate).

The maturity of the note indicates how many years it is in effect for: on a ‘vanilla’ bond, the purchaser gets their principal back at maturity and receives bi-annual coupon payments every year.

The seniority of the notes indicates who gets paid first and where different risks lie. The notes are backed by student loan repayments – these will vary each year.  A senior note gets paid first and is therefore less risky, but the payments will probably be fixed along the lines of the ‘vanilla’ bond outlined above. The holders of the most ‘junior’, or ‘equity’, tranches gets aid last and therefore take more risk – but they get what’s left and are potentially rewarded with higher repayments in good years.

When the government has looked at securitisation in the past, buyers have been keen on senior tranches, but less so on the junior tranches.

The problem for government is that if the junior tranches don’t get sold, then it may have to keep hold of those risks. In which case, it is unlikely that the securitisation would have achieved the main objectives of a sale and the Office for National Statistics might have to conclude that the loans were still on the government’s balance sheet. Securitisation is a very flexible – and therefore attractive – approach to selling financial assets, but it normally requires a special purpose vehicle to maintain ownership of the loans (only the rights to income from the loans have been sold to purchasers). If the SPV is still owned by government (and in many cases the SPV will buy the equity tranches), then it won’t have achieved two explicit loan sale aims:

  • de-risking the government balance sheet;
  • removing the loan accounts from Public Sector Net Debt.

It’s also worth noting here that the government has also abandoned the ‘synthetic hedge’ designed to deal with concerns over the interest rate terms on pre-2012 loans (interest accrues at the lower of RPI or bank base rates + 1%). This may also affect buyer appetite for the tranches with lower seniority.

Update (3pm): the ONS has confirmed to me that it will only make a final decision on classification after a sale is completed. An ONS spokesperson said: “ONS has considered a policy proposal from Her Majesty’s Treasury on the sale of the student loan book. However, we will not provide a final decision on the sale of the loan book for National Accounts purposes until the sale itself has happened.”



A sale will likely make significant losses for government

At this stage, £4bn of loan accounts are involved. This is the face value of those loans – take all the outstanding balances of borrowers who graduated between 2001 and 2005 and sum them together. As of the end of July 2016, the face value of pre-2012 loans came to £44.5bn.

More important to note is that the face value of loans is not what the government thinks they are worth – the fair or carrying value. The fair value is a present value estimate of what the loan accounts are likely to generate in future repayments.  The pre-2012 loans had a fair value of £33.7bn in the 2015/16 BIS accounts. (An illustrated explanation of the difference between face and fair value can be found towards the end of this video.  If I offer you an IOU for £100 in ten years’ time, what would you be prepared to pay? The face value is £100 but that’s not what it’s worth).

The key point that the government’s announcement obscures is that a value for money test can be passed even if the loans are sold for significantly less than the fair value – ie what the government thinks the loans are worth.

This is because the value for money test uses a higher discount rate to that used in the  fair value calculation. A higher discount rate values the same projected cash stream lower because a higher discount rate gives a higher value to today’s cash.

The fair value calculation uses RPI  plus 0.7, while the VfM test uses RPI plus 2.5 (at a minimum).  Let’s say RPI is 3 per cent, then the fair value discount rate is 3.72% (1.03 * 1.007), while the VfM discount rate is 5.58% (1.03 * 1.025). For the former, £1.037 next year is worth £1 today; the VfM test will sell £1.056 of next year’s repayments to get that £1: an additional 2p of next year’s repayments.

Compound over several years of repayments and that is a signficant loss that would be tolerated.

By selling the loans, the government may improve the headline public finance statistics (it says today that there is a ‘good prospect’ of achieving value for money), but its misplaced valuation of cash today over holding the loan assets will lead at best to a presentational gain, not a fiscal or economic one.

Interview with Real Media – parts 1, 2 & 3

Earlier in January, I saw down with Real Media for a couple of hours to talk about higher education reform and student loans.

Real Media have now released their first edited video. It’s ten minutes long and from somewhere in the middle of that two hours. I’m not sure how much more they are planning to release at this stage but I’ll put updates here.

The discussion in this segment focuses on how English HE is moving towards a funding system based on creditworthiness (of individuals and institutions). More detail on my theory of ‘financialisation’ can be found in the final chapters of The Great University Gamble.

Update – 6 Feb 2017

Part 2 of the interview is now available. This 10-min segment covers debt: student loan & commercial debt and why comparisons with the US are mostly misleading.

Part 3 of the interview – which appeared on 7 Feb – concentrated on why the government is trying to sell student loans and how to value student loans – including an explanation of the difference between face and fair value.


Why undergraduate tuition fees are not prices when backed by SLC loans

My contribution to Tuesday’s event at Kingston raised an issue to which several audience and panel members returned.

I made the claim that the undergraduate tuition fee is not a price.

What I meant by this is that 90 per cent of eligible students will take out a loan to fully cover tuition fees. The cost of the degree is then determined by the loan repayments made. The government subsidises Student Loans Company loans. This subsidy is largely unpredictable for individuals and can vary wildly – someone who never earns more than the repayment threshold pays nothing and so is fully subsidised; very high earners see no subsidy at all.

Repayments are therefore vague and uncertain, but primarily determined by future income not by graduating debt.

According to neoclassical economics, this means the tuition fee cannot signal as a price should in a perfectly competitive market. The undergraduate course fee cannot give consumers or producers information about relative quality or opportunity.

Only those who go on to be very high earners will see any difference in repayments between a course that charges £7500 per year and one that charges £9 000, or £9 250.   So it does not make sense for universities to diverge from the maximum fee allowable. A potential student would rightly reason that, while they see little to no difference in cost from a higher fee, they will be better off in an institution enjoying more resource per student.  (A postgraduate or international student is in a very different situation – for them, the fee is a price unless they have borrowed on a similar income contingent repayment deal).

As long as there is a subsidy in the loan scheme, the tuition fee cannot be ‘the single best indicator of relative quality’ (as the Browne review put it) and there is no alternative pricing signal.

A neoclassical economics website explains the centrality of the price signal function for markets to co-ordinate activity efficiently:

“[P]rices must reflect costs and benefits. . If people do not take account of substantial costs of their actions, they will act in inappropriate ways. They will either engage in too much of an action (if the ignored effects are costs imposed on third parties) or too little of it (if the ignored effects are benefits enjoyed by third parties).”

The problem in HE is then perceived to be that people are making the wrong choices, because the tuition fee cannot signal the difference between – for the sake of argument – economics and creative arts or economics at different universities. And the costs of a poor decision fall on government / public money (as the backer) when graduates fail to earn enough to repay the equivalent of what was lent to them.

This is a serious problem facing efforts to create a market in English HE. Back in October Theresa May announced an interventionist role for the state where markets are not seen to be working.

That’s why where markets are dysfunctional, we should be prepared to intervene.

Where companies are exploiting the failures of the market in which they operate, where consumer choice is inhibited by deliberately complex pricing structures, we must set the market right.

In undergraduate study there isn’t even a price! So something drastic needs to happen!

If there is no price signal, complex reforms have to be put in place to replicate its function or the market is no more efficient as an allocator of resources than an alternative form of organisation such as central planning. And there would be no point in engaging in extensive reforms to university funding.

The Teaching Excellence Framework is best understood then as an effort to create a proxy signal to indicate quality to potential students and rival institutions through its gold, silver and bronze badges. (In contrast to this consumer intervention, the Research Excellence Framework is a public procurement tendering exercise where future research is funded on the basis of past results).

As a concluding aside, if ‘neoliberalism’ is to be more than a shibbloleth when discussing HE reforms, then its more substantive meanings are to be found by understanding why markets are desired and what impediments there are to their implementation. This then makes the contours of future struggle clear – only by eroding the subsidy in student loans can tuition fees (as they rise and differentiate) become the prices they are so obviously meant to be.


To this I should add the obvious corollary, when people don’t understand how student loans work they are likely to massively inflate the cost and avoid HE as a poor purchase.

Think tank garbage

How mandatory repayments on new student loans are calculated isn’t news. 9 per cent of your gross annual earnings over £21 000 is demanded as a contribution to your graduate debt.

If you earn £41 000 at present (on post-2012 loans) you will face monthly payments of £150 amounting to £1800 per year. (£41 000 minus £21 000 is £20 000, and 9% of £20 000 is £1800).

I tell people to bear that in mind when think tanks and financial companies are scaring students and graduates with tales of total repayments clocking in at around £100,000 over a lifetime (the 30 to 35 years that they are in effect).

What sort of sums would you have to be earning today to make the average annual contibution needed to hit those levels of repayment? That’s a useful question to keep in mind when reading news coverage of loans; having a toy model of this kind in your head helps avoid basic misunderstandings.

Unfortunately, Intergenerational Foundation thinks you can base an analysis of English student loan policy on such simplified calculations.

Even worse, journalists have fallen for it. The BBC runs with the headline that someone earning £41 000 would repay £54 000. That’s taking the £1800 above and multiplying by 30 to cover the 30 years of repayments.

The problems with such an approach should be obvious. It assumes that the individual has exactly the same salary and faces the same repayment threshold for 30 years, but is prevented from clearing their debt because of interest accruing. (The addition of interest to a toy model will distort the analysis from the off without the addition of other real-world features such as earnings growth. Here, a measure of inflation has been used to determine interest, but plays no part in earnings or other features.)

IF attempts one realistic salary pathway calculation. Here a graduate ‘earns £22,000 for 3 years, then £25,000 for five years, then £30,000 for five years, £35,000 for ten years and £41,000 for seven years’.  Their graduate ends up making £31 000 in repayments, but this is on the assumption that the repayment threshold is frozen at £21 000 throughout, but with RPI running at 2.8% for thirty years for the purposes of calculating interest.

There are two key points to make in relation to those results:

  • The value of that final £41 000 salary would be the equivalent of around £16 000 today if the RPI used is a guide. But IF nowhere discusses how the value of money changes, leading the reader to the mistaken thought that the value of total cash repayments (£31 000) is similar to the amount loaned (£28 000 of tuition fee loans) despite that first sum and the repayments being years apart.
  • The repayments would not have been altered by the addition of maintenance loan debt. Here IF removes a ‘generous’ feature from the government-backed scheme in an effort to present it as worse than it is. I can’t see any legitimate reason for only discussing tuition fee loans and indeed alternative loan companies use the same trick to make their loans seem like a better deal than those from the SLC.

So the IF modelling is very ropey and their findings aren’t really open to sensible interpretation. The problems though run deeper. From a public policy angle, the BBC focused on claims that the taxpayer will have to pick up the tab for write-offs.

“At present the taxpayer picks up the tab for unpaid loans after 30 years, allowing graduates to walk away from tens of thousands of pounds of debt and interest charges.”
But the cost to government is not – as IF assumes – the sum of nominal loan balances written off after 30 years. It’s the difference between initial loan outlay and the value of repayments received (and that means introducing the time value of money).The government currently issues loans thinking it will only get back the equivalent of 75p for every £1 that goes out the door. (IF cites the old 45% figures not realising how much has changed in the last two years).
I could go on.There are many mistakes in the report. I’ll just point out one more:

Student loans are cancelled 30 years after “eligibility to repay”.This means that loans will have a term exceeding thirty years where eligibility to repay (i.e.reaching 21,000 salary) is delayed.
Completely wrong. “Eligibility to repay” is determined by when the student left the course – not the repayment threshould. For those finishing in the summer, the clock starts in the following April regardless of earnings levels. Someone becomes eligible to repay and then the  mandatory repayment is calculated in relation to the threshold: the life of student loans is not extended by periods when borrowers are earning under the threshold.

(Declaration – I used to be on IF’s advisory board and wrote a report for them on student loans in 2012, but resigned a few years ago).