Select Committee on Education and Skills Minutes of Evidence

Supplementary memorandum from Dr Nicholas Barr[28] (SS11)


  1.  In 2002-03, the Student Loans Company will lend £2,500 million. The cost of the interest subsidy (some 35 per cent of total lending) is around £800 million.

  2.  If interest rates are held to the rate of inflation, the resulting subsidy does not benefit a single student, but only successful professionals in mid career. The argument is worth spelling out:

    —  Interest subsidies do not help students (graduates make repayments, not students).

    —  They do not help low-earning graduates, since unpaid debt is eventually forgiven.

    —  They do not help high-earning graduates early in their careers—with income-contingent loans, monthly repayments depend only on earnings; thus interest rates have no effect on the size of monthly repayments, but only on the duration of the loan.

    —  The only people they help are higher-earning graduates in mid career, whose loan repayments are switched off earlier because of the interest subsidies than would be the case without them.

  3.  Thus the question is whether we want to spend £800 million on helping wealthy people become wealthier, or whether we would rather use it to promote access through targeted student support, and improved quality through more resources to education institutions.

  4.  Note that this line of argument goes a long way to solving the political problems of raising the interest rate on student loans to the government's cost of borrowing (not "commercial rates") discussed in the parallel note by Iain Crawford.


  5.  The objective of the NUS is "targeting money effectively". There is complete agreement about that objective.

  6.  However, the NUS puts forward the following argument (paragraphs 188 and 213 of their oral evidence) against abolishing interest subsidies. With income-contingent repayments, higher earners repay more quickly and lower earners more slowly. Thus lower earners pay more interest in total than higher earners. Thus raising the interest rate on student loans harms lower earners. There are three answers.

  7.  Answer 1. As a proposition in logic the statement is true. But the basis of the argument is that in order to protect the poor you subsidise the commodity for everyone. This is the postwar argument for food subsidies: food subsidies did, indeed, help the poor, but they helped the rich much more. The same resources could do much more to help the poor if spent in a way that did not leak out to the rich.

  8.  Answer 2. The whole point of income-contingent repayments is that they tailor monthly repayments to ability to pay by automatically extending loan duration for lower earners. Someone who buys a house for £100,000 will pay more interest if she opts for a 25 year mortgage with lower monthly repayments than a 15 year mortgage with higher monthly repayments. Income-contingent repayments do the same thing automatically.

  9.  It is, of course, entirely possible to solve the NUS's "problem" by having mortgage repayments (eg equal annual repayments for (say) 10 years, regardless of income); thus loan duration, and hence interest payments, are the same for all students with a given size of loan. That cure, however, is vastly worse than the disease.

  10.  Answer 3. There is a range of mechanisms (my main written evidence, paragraphs 58-62) for helping low earners which are much more effective than blanket interest subsidies:

    —  Income-contingent repayments automatically project low earners (the argument above).

    —  Grants for students from poor backgrounds mean that such students have no loan, or a smaller loan.

    —  Targeted interest subsidies based on current income. The attack on blanket interest subsidies is not an attack on targeted interest subsidies. It would, as examples, be possible to pay an interest subsidy to people who are unemployed or looking after young children. Such targeted interest subsidies are both possible (as currently in New Zealand) and desirable.

    —  Conditional subsidies. It would also be possible to pay interest subsidies to people with low current earnings, but to claw them back if they subsequently become high earners. Design work for the Hungarian Government established that such an arrangement presents no administrative problems.

  11.  Any or all of the mechanisms in the previous paragraph help low earners much more effectively than blanket interest subsidies.


  12.  Make the cost of interest subsidies transparent. Expenditure on interest subsidies is currently buried in the overall cost of government borrowing, and hence does not appear in the education budget. The most minimal reform is to make spending on interest subsidies an explicit line item in the education budget. A major advantage is to make the high cost of interest subsidies visible. In addition, figures on expenditure on student support, by correctly including the cost of interest subsidies, would be more accurate and more readily comparable with earlier years.

  13.  Abolish the blanket interest subsidy for all the reasons given above.

  14.  Ensure that the education budget benefits from its abolition. This is the issue discussed in main evidence, paragraphs 38-41. The core of the issue is that the figure of £800 million referred to above is the present value of the savings from moving to the government's cost of borrowing, but the savings in cash-flow terms are small in the early years. To ensure that the withdrawal of blanket interest subsidies leads to an immediate increase in spending on access and quality, a deal will be needed between the Treasury and DfES to accelerate the cash flow. As noted in paragraph 39, one way to finance that deal would be to sell a further tranche of student debt, yielding up to £2 billion.


  15.  Some analytical equivalences. The following are logically equivalent:

    —  An income-contingent loan.

    —  A grant plus an income-related graduate contribution.

    —  A grant paid for out of taxation, except that the tax (a) is paid only by people who have been to university, and (b) does not go on for ever.

  Though analytically identical, the three representations clearly have major differences in terms of perceptions. The companion document by Iain Crawford suggests a way of addressing the perception problem.

  16.  The balance between loans and grants should be variable, not an either-or proposition:

    —  Middle class students would be entitled to an income-contingent loan sufficient to cover all living costs and tuition fees, thus making higher education free at the point of use (the Cubie advance).

    —  A student from a poor background would receive a grant. Depending on the size of the grant, he or she might also need to take out a loan—but a smaller loan.

    —  A student from a particularly vulnerable background might receive a full grant, and would therefore not need to take out a loan at all.

  17.  Tuition fees.

    —  All the previous discussion about loans can be decoupled from the fees issue.

    —  There is a strong case for a move towards flexible fees (which can be higher or lower than the present tuition fee).

    —  However, deregulation of fees should be phased, not "big bang".

Dr Nicholas Barr

May 2002

28   Department of Economics, London School of Economics and Political Science, Houghton Street, London WC2A 2AE, UK. Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 19 June 2002