Please Note: Third parties issued press releases about this analysis on 17 November 2010. We will be revising this document shortly and have detailed how and why here.

Last week, the Minister of State for Universities and Science, David Willetts, announced the Government's proposals for higher education funding in England, in response to last month's publication of the Browne Review. IFS released some initial reaction to these proposals last week. Here we quantify the main implications of that announcement. The key differences compared with Lord Browne's proposals are:

• Tighter means testing of grants, which saves the taxpayer money;
• A fee cap of £9,000 but with no levy on fees above £6,000. The levy was intended to recoup the cost to the Government of providing larger loans to cover these fees; without it, the cost to the taxpayer is significantly higher than under Lord Browne's recommendations;
• Increasing the interest rate for higher earning graduates and reducing it for lower earning graduates, via the proposed interest rate taper. We calculate that this will actually cost the taxpayer money compared to the system proposed by Lord Browne;
• Making the maintenance loan system more complex without any significant savings to the taxpayer compared to Lord Browne's proposals.

Who are the winners and losers from David Willetts's announcement?

Compared to the system proposed by Lord Browne, universities wanting to charge between £7,000 and £9,000 a year are the main winners, while the richest half of graduates would gain slightly. Universities gain because the absence of a levy enables them to keep 100% of any additional fee income above the basic £6,000 level, and graduates gain because the tapered interest rate - a sliding scale between 0% real at earnings of £21,000, and 3% real at £41,000 - provides an overall subsidy, relative to Lord Browne's proposal, to middle- and high-earning graduates.

The main loser of the Government's proposed system is the taxpayer: the reduction in maintenance grants is more than outweighed by the cost of not imposing a levy. At higher fee levels, this becomes an increasingly important factor: if all universities charged £9,000 a year, we calculate that total the taxpayer burden of higher education would only be slightly lower than it is at the moment (by around £770 per graduate) costing the government potentially billions of pounds of savings compared to Lord Browne's proposals.

Table 1 summarises the balance of contributions to the cost of higher education under the current system, the Browne Review recommendations and the Government's proposals. The figures presented are total amounts per graduate over the course of their degree.

Table 1. Balance of contributions to higher education under current, Browne, and Government systems (Click here for a larger table)

What are the implications for graduates?

The Government's own analysis of its proposals suggests that most graduates - those in the top eight deciles of lifetime earnings - would pay back more than under the Browne Review proposals, and that the top 30% would actually pay back more than they borrowed.

On these points, our conclusions differ from the Government's. We find that the poorest 40% of graduates would be unaffected. As shown in Figure 1 (using a £7,000 fee), we calculate that most graduates, particularly those between the sixth and eighth deciles of lifetime earnings, would be better off under the Government's proposal and that no decile group would be worse off. We also calculate that at most 1.0-1.65% of graduates would pay back the full value of their debt. Only the very highest earners are likely to pay "over the odds" for their degree; the top decile as a whole would pay, on average, around 95% of their debt (compared with around 97% under the Browne Review proposals).

Figure 1. Graduate repayments under current system, Browne proposals and Government proposals (£7,000 fee)

This discrepancy arises because of differences in the assumed levels of annual earnings across the distribution. As a result of differences in underlying data, the Government's analysis over-estimates annual earnings at the top of the distribution. Our profiles of lifetime earnings imply average annual earnings of £60,000 in the top decile over the period during which loans are repaid (and higher earnings thereafter as graduates' progress through their careers). As a result, the Government's analysis over-estimates the number of graduates at the top of the distribution who would earn enough to face the full 3% real interest rate while they are making repayments. In fact, a significant number of graduates in the top half of the distribution could face a lower average interest rate than under the Browne proposals.

The £41,000 threshold for the interest rate taper effectively provides a subsidy to high earning graduates and will penalise only a very small number of high-fliers at the very top of the distribution. Since most graduates are unlikely to breach this level early on in their careers while they are making loan repayments, the Government could both save money and extract revenue more revenue from the high-earners by opting for a lower threshold. If it were set at £31,000 a year, for example, we estimate that the richest quarter of graduates would all pay back slightly more than under these proposals (while other graduates would be unaffected), thereby costing the taxpayer less.