For a while now we’ve been used to thinking about the national accounts in terms of public sector net debt (PSND).

There’s a couple of common exclusions: if you exclude public sector banks (the bits of bank we still own following the 2008 financial crisis, and the Bank of England) you get what is more precisely called PSND ex. Government debt, expressed in this way, is slightly over 100 per cent of gross domestic product.

Most recent chancellors of the exchequer have set targets (often in the form of “fiscal rules”) based on PSND ex – under the last government it was generally expected that the UK’s aim was to have debt (expressed as PSND ex) on course to fall as a share of national income in a number (usually five, occasionally, the end of the parliament or the wonderfully vague “end of this economic period”).

PSND is not the only way to think about debt. Rachel Reeves, for instance, is reportedly warming to a measure called Public Sector Net Financial Liabilities (PSNFL, or “persnuffle”) which takes into account the financial assets that the government holds, and in this case that includes liquid assets like foreign currency, and illiquid assets like student loans. Here, a loan is an asset because we expect all (or most of it) to be paid back. Moving the fiscal rules away from PSND and towards PSNFL would give Reeves a bit of extra headroom for borrowing and investment.

The case of student loans

Pre-2018 all student loans were also classified as assets in PSND (and thus not really involved in the calculation) – at that point and after some concerted lobbying the Office for National Statistics made the decision to move to what it called a “hybrid treatment” of the loan book as a whole. The portion of the loan expected not to be paid back is counted as a “capital transfer”, the portion that is expected to be paid back is counted as an “asset”.

To make this work you need a way of understanding what size each of these portions might be, and for this reason the Department for Education developed and updated student loan forecasts. So, as of July this year, the magic figure for full time undergraduate plan five loans is that 29 per cent of the total value of loans issued in 2023-24 will not be repaid (of course, many of the loans issued by the government over the years are not plan five).

The upshot of this was that student loans, which used to look very cheap because of what was described as a “fiscal illusion”, suddenly began to look very expensive. Hence the invention of Plan 5 during the May administration – lengthening the term of repayment by a decade to ensure that more money would be counted as an asset (loans that will be repaid) and less counted as a capital transfer (loans that will not be repaid). Plan 5 also removed the interest rate, which counterintuitively ensures that more of the loan capital is repaid.

So, if we targeted PSND as a debt measure, selling the student loan book (as was policy in the 2010s) makes it look like we have less debt. It wouldn’t matter if we sold it for less than it is worth – what was paid for it is taken off the national debt.. It would be a win, despite (on the face of it) making no sense whatsoever.

In contrast, in PSNFL the student loan book is a promise: it is a giant IOU note. You’ll be ahead of me here, but if we handed out more loans we would have a bigger asset.

In both cases you might well wonder about the other end of the loan transaction: the government has to borrow money (via the sale of Gilts) to raise the capital to loan out to students. Under PSND measures and PSNFL this is – as we would expect – debt: it is money borrowed by the government that needs to be repaid.

Future funding and PSNFL

Because student loans represent a hefty chunk of government borrowing (just over £21bn this year in England alone) it could be reasonably suggested that a shift from PSND to PSNFL as a primary debt measure is an end run around the 2018 ONS decision). Rather than counting about 30 per cent of the loan book as actual debt (while ignoring the rest of it) the whole thing would become an asset that has the same value as what was spent.

Thinking like this makes it look easier to raise fee caps (because lending out more money won’t ever show up as debt), easier to return to proper interest rates (it doesn’t matter in debt terms how much of the capital is repaid), and easier to hand out more loans (both).

Both those things would – of course – make PSND look a lot worse. And PSND isn’t going anywhere – you’d better believe economists of all stripes will have an eye on it even if the official government fiscal rules refer to something else.

A graduate tax, however, would always make PSNFL look as bad (borrowing, but no asset) as it would PSND. Additional tax income might eventually show up, but that doesn’t flatter the current debt under either measure.

And giving money to universities directly (for example via the OfS or Research England) is just spending. There is no asset created (no, Exchequer interest doesn’t count here), and no income attached. Universities are assets if you consider the rarely used Public Sector Net Worth (PSNW) – but this presupposes we have the ability to sell bits of them.