Those London Economics/Nuffield Foundation fee models in a nutshell

A welcome cross-UK comparison of fee and maintenance arrangements that also looks at exchequer spending and provider income per student

David Kernohan is Deputy Editor of Wonkhe

It’s always a delight to dive back into the fine fee modelling work of Gavan Conlon and Maike Halterbeck.

I say “back” as these are variations on the same models London Economics (LE) have been using for a while – if you’ve been to a Wonkhe event on funding, or if you remember our 2019 coverage of the Augar review, you’ll be familiar with the basic approach that underpins today’s release.

The LE team has examined the cost – to students, and to the taxpayer or exchequer – of each of the four funding systems in the UK. This fact alone makes this documentation (four slide decks – one for each of England, Scotland, Wales, and Northern Ireland – plus a short comparative overview) worth engaging with. In each nation of the United Kingdom, the detail of funding policy is very poorly understood – with the “national debate”, such as it is, dominated by the sticker price.

It’s worth a quick summary:

Northern Ireland

Northern Ireland domiciled students studying in NI pay a maximum yearly tuition fee currently set at £4,710 – this increases with inflation over the course. If they wish to study elsewhere in the UK, it costs £9,250 – in each case there are fee loans available.

Students get a means-tested maintenance grant of up to £3,475 and a means-tested loan of up to £6,776, each year.

Loans are repaid at 9 per cent of earnings over £22,105, are written off after 25 years, and interest rates are equivalent to RPI.

The exchequer contribution – the amount paid by the taxpayer, rather than recouped from loans – per student per year is £4,810, most of which is teaching grants at £3,030.


Scotland domiciled students studying in Scotland do not pay fees – notional annual tuition fees (£1,820 a year) are covered by non-repayable grants from the Scottish Government. If students study elsewhere in the UK, the fee is £9,250 backed by fee loans.

In Scotland the mean-tested annual maintenance grant is up to £2,000 and the means-tested loan has a £7,000 maximum, irrespective of where students live. (There’s no mention here of the “special support loan” coming next year – £2,400 for all full-time students)

Scottish graduates pay back 9 per cent of earnings above £27,660 (uprated by RPI each year), and the remainder is written off after 30 years. Interest is levied at RPI.

The contribution of the exchequer per student per year is £9,130, mostly via the teaching grant.


For Welsh students studying in Wales annual tuition fees are £9,000. These, and the £9,250 payable by Welsh students studying elsewhere, are covered by fee loans.

Welsh students can access up to £11,720 a year in maintenance support (for students living away from home outside of london. A means test controls the split between grants (a maximum of £8,100 a year, a minimum of £1,000) and loans covering the remainder.

Welsh graduates pay 9 per cent of earnings over £27,295 (this should go up by RPI each year but is frozen until 2025-26), with the remainder written off after 30 years. Real interest rates (up to RPI plus 3 per cent) are charged – and there is an initial write off of £1,500 from maintenance loan totals at the point the first payment is made.

The average exchequer contribution per student, per year, is £3,780 – mostly from maintenance funding (£3,800)


The maximum fee for English domiciled students is £9,250 – this is payable wherever in the UK students study and is backed by a fee loan.

English domiciled students can access a means-tested maintenance loan of up to £9,978 a year (for students living away from home studying outside London).

Graduates repay 9 per cent of their earnings above £25,000 (this should rise with RPI, but is frozen until 2027-28, with loans written off after 40 years. There is no real interest payable on borrowing (RPI is used).

The exchequer contribution is around £1,630 per undergraduate student, per year – about 16 per cent of the total cost of provision.

Making changes

Broadly speaking, the impact of changes are counter-intuitive when you consider socio-economic equality – stuff that those who engaged on Plan 5 will be aware of, but equally valid UK wide.

  • Lowering fee levels has a positive impact on the most advantaged graduates, who will save money by paying off loans quicker. It has almost no impact on the most disadvantaged graduates (who would never have paid off their full loans anyway, but now pay more).
  • Conversely, raising fees affects everyone, though disadvantaged students will repay for longer (and thus repay more) than the better off.
  • Scrapping real interest rates likewise provides a disproportionate benefit to better off graduates (who will pay less), with no impact on repayment for the rest.
  • Lengthening the repayment term before write off disproportionately affects less well off graduates, who will pay more.

The differing amounts of government contribution per student are also counterintuitive – Scottish universities get a greater exchequer contribution per student than the other three systems, though the overall income per student is lower than elsewhere.

These are all things to bear in mind when discussing the complex mess that is the fee system in any UK higher education sector – though political considerations (the near totemic tuition fee pledge in Scotland, the new pressures on Northern Ireland’s government to harmonise with an Irish Sinn Fein pledge to scrap fees entirely, should the latter come to power) generally have the upper hand.

3 responses to “Those London Economics/Nuffield Foundation fee models in a nutshell

  1. The problem with all this as the authors themselves highlight if you can be bothered to read all the annexes is that the cost to the exchequer of the English system is based on a frankly laughable estimate of the RAB charge (ie the cost to government of making the loans). Even with the plan 5 changes the idea that on average the cost is only 4% is back to magic money tree levels of credibility.

  2. The difference between the DfE modelling and the London Economics modelling is that the DfE model is massively more pessimistic about graduate earnings. The IFS student loan model reaches similar conclusions as the London Economics model – see this very useful calculator which allows you to change the parameters including the discount rate

    I’m not clear what is driving this huge difference in earnings forecasts between DfE and other forecasters?

    The question about discount rates is entirely valid. The current discount rate used by the DfE of RPI minus 1.3 per cent is roughly equal to CPI – in other words, a real-terms payment of £500 in 2060 is valued the same as a £500 payment in 2024. This is set on the basis as that this is the current cost of government borrowing. When reforms were first introduced in 2011 the RAB charge was instead calculated on the basis of a discount rate of RPI plus 3.5 per cent (which the IFS model says would equate to a RAB of 54%) based on social time preference as per the HMT Green Book guidance on cost-benefit analysis

Leave a Reply