In the history of Australian higher education policy Brendan Nelson, the Liberal minister for education from 2001 to 2006, is perhaps under-rated. Several student funding structural changes he legislated 20 years ago are still in place. These include:

• Student contributions set by universities up to a legislated maximum and going to universities (previously HECS was a fixed government charge);
• A per full-time equivalent student Commonwealth contribution based on subject field of education (previously universities received an overall operating grant, which although informed by an early 1990s costing exercise did not directly tie money paid to discipline-level enrolments);
• Commonwealth-university funding agreements as a method of allocating student funding to institutions, which made funding arrangements more transparent (but also turned into a backdoor instrument of policy and regulation that bypasses Parliament);
• Through FEE-HELP, extension of student loans to full-fee undergraduates and students in private higher education institutions (the more limited Postgraduate Education Loan Scheme, PELS, was already supporting university full-fee postgraduates).

The 2003 Cabinet papers

The annual National Archives release of 20-year-old Cabinet papers, with the 2003 papers released earlier this week, gives us a look behind the scenes as Nelson’s reform package was developed and debated. Three digitised Cabinet documents record proposals and decisions, but not the Cabinet discussion. Sometimes, however, Cabinet thinking can be inferred from requests for further work and contextual material in the submissions.

This post focuses on changes to income contingent student loans.

The loan scheme that did not make it through Cabinet

In the first digitised Cabinet submission, dated 18 February 2003, three new income-contingent loan schemes are suggested: F-HELP, an extension of loans to full-fee students, OS-HELP for study abroad, and E-HELP.

E-HELP was to cover educational costs for full-time students in Commonwealth-supported places, offering $2,000 a year up to a maximum of $5,000 overall.

By 15 April 2003, the date of the next digitised Cabinet submission, F-HELP had been renamed FEE-HELP and along with OS-HELP was approved. E-HELP, although included in the explanatory papers provided to ministers, was omitted from the Cabinet decision.

No reasons for E-HELP’s demise are given, but the submissions offer clues. The overall cost of student loans gets significant discussion, and in this context E-HELP was perhaps one loan scheme too many.

But why did E-HELP get cut and not one of the other two? Perhaps the parallels with the Student Financial Supplement Scheme (SFSS), which let students trade in $1 of their income support for a loan of $2, was on the minds of ministers. It is mentioned in the 18 February submission. Separately, the government was considering abolishing the SFSS, and announced its decision to do so in April 2003.

The SFSS, versions of which had been in place since 1993, was often criticised for putting students into unnecessary debt and for its low repayment rates. Even more now than in 2003, the SFSS is a case study of a loan scheme that attracts high-risk borrowers.

Although no new SFSS loans were made after 31 December 2003 the Guardian reported last year that 140,000 SFSS borrowers still owed $2 billion. According to 2020-21 ATO statistics only 74,203 SFSS debtors filed a tax return, and of these just 15,368 earned enough to make a repayment ($40.3 million in total).

On my calculations, the SFSS lent approximately $2.6 billion between 1993 and 2003, so the outstanding balance is about three-quarters of the original debt. Debt indexation causes the outstanding SFSS balance to increase despite some repayments being made.

Perhaps to the Cabinet E-HELP looked too similar to the SFSS – a loan scheme that will appeal to students with poor repayment prospects. But the SFSS experience was not fresh in the minds of subsequent policymakers. The current Start-up Loan, established by a later Liberal government to replace a start-up grant, does not require income support trade-offs but otherwise has SFSS parallels. The Universities Accord interim report mentioned an E-HELP-type suggestion, which I hope did not make it to their final 47 recommendations.

For OS-HELP, a background document in the 15 April submission suggests that its borrowers would have a good chance of repaying. Unlike the other loan schemes, OS-HELP was not to be (and isn’t) an entitlement scheme open to everyone meeting its criteria. Lending was to be capped each year. This fixed amount was to be (and is) allocated to universities, who then decide which students should get it. The submission suggests that rationing OS-HELP makes it more likely to go to ‘meritorious’ students. Students had to complete at least one year of study to be eligible, screening out students at high risk of attrition, and to have study remaining after returning, to reduce the chance they will stay overseas and avoid repaying their debt.

Debt held by actual or potential overseas debtors

The issue of HECS debtors living overseas and not repaying was on the minds of policymakers in 2003. At that point, however, they did not think requiring them to repay was practical (it was later introduced, starting with the 2016-17 tax year).

Instead, the government resorted to clumsy visa proxy indicators of flight risk. The rules on New Zealand citizens and permanent residents receiving HECS loans had been tightened earlier in the Howard era, with the 2003 Cabinet submissions proposing their complete removal from loan eligibility, other than for permanent humanitarian visa holders.

These measures were implemented, although later changes allowed some long-term NZ citizen Australian residents to get a loan and, more recently, the government made it easier for NZ residents of Australia to become Australian citizens.

Limits on debt

The large amount of student debt held by some people is now a major issue, for them and for HELP’s finances. In 2003, policymakers combined their proposals for new loan categories with steps to control debt.

Borrowers using the new FEE-HELP scheme – with significant potential to escalate debt due to supporting fees with no price caps – were given a lifetime FEE-HELP borrowing limit of $50,000.

Annual lending for OS-HELP, as mentioned, was set by the government.

For HECS-HELP no direct cap on borrowing was set, but a de facto one was proposed with a 5-year Student Learning Entitlement on Commonwealth supported places. The eventual SLE was seven years. In practice, this meant that HECS-HELP limits were driven by student contribution levels (the SLE was abolished in the Rudd-Gillard era but restored by the Morrison government).

Since 2020, combined borrowing limits apply for all tuition loans.

The 15 April Cabinet papers include a proposal to reduce the $50,000 limit by $2,500 for each year once a person turns 50. With reduced potential workforce years in which to repay older borrowers are more likely to produce bad debt. The Department of Education pushed back on this idea on multiple grounds: breaching anti-discrimination law, undermining lifelong learning, administrative complexity, and low savings given small student numbers in the 50+ age bracket. This proposal did not proceed.

Partial cost recovery for student loans

Right from the start of income contingent loans in the late 1980s the money departments, Treasury and Finance, have wanted to reduce their cost. A favourite policy has been charging real interest, above the CPI indexation that has always applied.

The politics of real interest, however, have never been good. The 18 February Cabinet submission tries to head this off by suggesting a loan fee of 30% on the new HELP schemes. The advantage of this over real interest, according to the submission, is that it fixes the cost of borrowing and reduces the accumulation of debt during periods of low repayments.

Nelson had trouble convincing his Cabinet colleagues on this point. Although at the 15 April 2003 meeting they approved FEE-HELP and OS-HELP, the loan cost issue was not resolved. Instead, the 30% proposal and two other ideas were to be considered further. The first alternative was a CPI + 3.5% interest rate for the first 10 years, before reverting to CPI only. The second was a 15% loan fee, suggesting that some people in the room saw the political difficulties of both the other options.

The policy as announced was a victory for Treasury and Finance, CPI + 3.5% real interest for 10 years. But the parliament was harder to convince than the Cabinet. In a 27 November 2003 Cabinet document on Senate negotiations Nelson is encouraged to keep trying for the 3.5%, but to consider options such as a lower interest rate or (I think, the wording is unclear) a split system that let people who would have been eligible for PELS remain with its system (i.e. no loan fee and CPI-only indexation) while other students paid CPI plus 3.5% interest.

The final result, however, was postgraduates remaining on CPI-only indexation while undergraduates were charged a 20% loan fee plus CPI indexation. Remarkably, this distinction between undergraduates and postgraduates, which reflects Cabinet and Senate negotiating trade-offs in late 2003 rather than any policy principle, is still in place.

Repayment thresholds

Cabinet’s decisions don’t reveal their reaction to a recommended first HELP repayment threshold increase to $30,000 from 2005-06, up from the 2002-03 level of $24,365. The supporting papers note an estimated increase of 0.5% in the level of HECS-HELP doubtful debt as a result.

Possibly Cabinet accepted that the 1996 Budget decision to reduce the threshold from $28,495 in 1996-97 to $20,701 in 1997-98 affected living standards for lower-income debtors too much and caused more political grief than it was worth. And as Cabinet was planning to let universities increase most student contributions by up to 30% (25% in the end) they needed some positive news for students.

Similarly, the digitised decisions include no specific reference to the proposed introduction of further repayment thresholds on higher incomes, up to a maximum of 8% of income (the highest rate was 6% at the time).

Although the government was already planning to increase the first repayment threshold its level was a point of contention in the Senate. The 27 November Cabinet decision authorised Nelson to increase the first repayment threshold to $33,150 for 2005-06. But the government gave further ground. I don’t have any internal documents, but transitional legislation lifted the threshold to $35,000 for 2004-05, while the principal new funding legislation increased it again to $36,184 for 2005-06.

Conclusion

The 2003 Cabinet papers contain much more material on student loans than previously released submissions; on both how new loan functions could drive change in the higher education sector and the potential problems of high debt. Changes to the first repayment threshold suggest an understanding that repaying HELP debt on a low income can cause problems, but the digitised papers don’t show any deeper analysis, such as how student loan repayments interact (or would interact) with the broader tax and welfare systems. That’s something I hope to see in the Universities Accord final report.