There has been an interesting debate about the IPPR’s report ‘A Critical Path‘ on this blog and elsewhere over recent days. My problems with the IPPR report relate to the lack of care taken with respect to loan accounting. I hope that this post will layout these concerns clearly.
Higher Education policy around loans is only partially understood by cashflow models. The debt used to finance the loans in the first place and its impact on the national balance sheet and departmental accounts also needs to be understood, especially when the headline statistics are used to present macroeconomic competence to the public. That is, the issues are not merely about the economic long-run but the political present: think of the Coalition’s targets for the debt and the deficit.
Unfortunately the IPPR makes two significant errors here:
Firstly, IPPR has been less than clear in distinguishing the conventions used in the National Accounts – which records transactions when they occur (‘cash in, cash out’) – and those for the departmental accounts, which are run on an accruals basis. The latter record obligations when they are incurred. This distinction is vital for understanding student loans which have lifetimes of over 30 years.
The problem of non-repayment is not ‘hidden’ until write-offs occur as IPPR claim (pp. 120-4). An impairment to cover estimated losses – the ‘RAB charge’ – is created and ringfenced within departmental expenditure now. Any significant upwards revisions to those estimates must be reflected in additional budget being assigned to build up the impairment accordingly.
These movements are not reflected in national accounts because this money has not yet been spent (no ‘cash out’). This is recognised in discussion of the fourth ‘weakness’ on page 125 but not carried through consistently: ‘In order to compensate for this increased long-term cost, the Treasury will require BIS to make an equivalent saving in its current expenditure. A higher RAB charge will therefore add more pressure on BIS to cut essential current spending in the short term, in areas such as science and research.’ For instance, in the discussion of specific scenarios involving loans, we find the following:
‘There is a high upfront cost to the government, as it would need to immediately increase teaching grants for HEIs. Under this scenario, the Treasury would have to fund an additional £1.67 billion in HEFCE teaching grants. This would be the upfront annual cost of lowering tuition fees to £6,000 (although the overall cost of this policy to the government would be less than that in the long term, as it would make savings in the future as a result of the reduced cost of student loans). [page 134]
Precisely confusing national account conventions with departmental. Any estimated savings on the loans would be reflected in departmental accounts now.
Secondly, IPPR recognises that from a national accounts perspective the ‘impairment’ does not figure in Total Managed Expenditure (because the money in the impairment has not yet been spent) and so is not included in the ‘deficit’ (write-offs will appear here when they occur).
However IPPR misunderstands what the significance of this is, because they forget about the second headline statistic, debt.
The normal logic of deficit reduction sees less pressure on borrowing from expenditure, which lessens the pressure to accumulate additional public sector net debt (PSND). The rate of growth of the latter is normally slowed down by reductions in annual borrowing requirement used to cover the deficit.
Student loans work differently. Borrowing is needed to finance new annual outlay and this is only paid down by the receipt of graduate repayments.
PSND is currently the key to loans, not the deficit, as the normal link between the two is broken for ‘policy lending’.
In order to understand PSND, a balance sheet measure, we need to understand how assets and liabilities figure in the calculation.
IPPR confuse the resulting assets and liabilities in the following passage:
‘When the government borrows money to issue in the form of student loans, it is able to treat this as an asset that doesn’t count against the deficit’ [p. 139]
The borrowing [‘this’] is not an asset. The borrowing is a liability – money the government owes – the loan book created is an asset – money owed to the government. Two accounting events occur. Importantly, from the government’s point of view, the positive side of the equation, the asset does not appear in PSND as it is classed as an illiquid asset. That is, the impact on PSND of loans is worse than you might think, because the positive (the asset created) is not netted against the negative (liability). Only the latter counts.
In sum, although the move from grants to loans reduced the deficit, the new funding regime has increased PSND in the short to medium term. This means that, for example, the first scenario IPPR propose and discuss is misinformed when it concludes: ‘Charging a high real rate of interest increases the likelihood that wealthy graduates will opt to repay their debt as a lump sum, in order to avoid having to pay the interest. This means the Treasury would not benefit from increased repayments for this group.’ [p. 132].
Since early repayments pay down the debt faster, the Treasury would very much like to encourage this – a higher interest rate is one of the best ways to achieve this (cf. a discount incentive).
Since debt is ignored in general (both public and private incidentally), the risks the Treasury associates with the growing portfolio of student loans simply escape the report as does the logic of a sale to third parties, and its intimate relation to an expanding loan book.