19 December 2005

Interest free student loan costings

“The final costings for the interest free student loans policy justify Labour’s confidence in the policy’s affordability,” Finance Minister Michael Cullen said today.

The operational costs are estimated at $32 million in 2005-06; $218 million in 2006-07; $256 million in 2007-08; $286 million in 2008-09 and $269 million in 2009-10.

“This is is line with, although a little lower, than the Labour Party’s costings and significantly lower than the Treasury scenario released during the election campaign,” Dr Cullen said.

“The figures published today were finalised on 6 December. They are the result of a lot of work by officials in the Ministry of Education, the Ministry of Social Development, Inland Revenue and the Treasury and are higher than the version supplied to the Cabinet on 14 November and referred to by me on 16 November in the House during debate on the interest free loan legislation.

“The earlier estimates assumed a cost on the operating provision of around $200 million a year from 2006-07. The difference between the two sets of costings reflects changes in technical assumptions and refinements to the forecasting model.”

The official costings also provide for an annual cash impact of $157 million in 2005-06; $293 million in 2006-07; $360 million in 2007-08; $367 million in 2008-09 and $308 million in 2009-10. This represents the cost to the government of increased borrowings and slower repayment rates. The volume of voluntary repayments, for example, is expected to reduce over four years to 20 per cent of current levels.

The third component in the costings is the reduction of around $1.5 billion in the value of the student loans portfolio held by the Crown. “This is less than the $2 billion adjustment signalled last month and includes around $500 million from a shift in accounting approach to fair value – a change the government had to make in 2007 anyway and has simply brought forward,” Dr Cullen said.

Attached: 6 December Ministry of Education report.

2 December 2005 < Ministry File Number >
Tertiary Education Report: Costs of No-Interest Student Loans – Appropriations Report

Executive Summary

1. Attached is a report that sets out the costs of the no-interest policy on student loans. Also attached are drafts of letters for you to sign seeking the approval of joint ministers of appropriations to provide for the costs of the implementation of a no-interest policy on student loans.

Background

2. On 14 November 2005, Cabinet delegated to you and to the Associate Minister of Finance (Hon Trevor Mallard), the Minister for Social Development and Employment, and the Minister of Revenue, to make changes to appropriations for the move to a no- interest student loan policy based on any revised assumptions provided by officials so that the impacts may be reflected in the forthcoming Half-Year Economic and Fiscal Update.

3. The attached costing report and the letters to your colleagues seeking their approval are designed to enable you to exercise that delegation.

Structure of the Report

4. This report comprises:

- An explanation of the assumptions that led into the costing;
- An account of the resulting cost figures;
- An explanation of some of the accounting issues involved;
- A statement of the appropriations that are needed for this policy;
- Two appendices that set out the assumptions behind the costing in greater detail;
- Letters to the Associate Minister of Finance (Hon Trevor Mallard), the Minister for Social Development and Employment, and the Minister of Revenue, seeking their agreement to the appropriations.

Linkage to the Half-Year Economic and Fiscal Update

5. The figures in this report are consistent with those submitted for inclusion in the Half Year Economic and Fiscal Update on Monday 5 December 2005.

The Cost of the Policy

6. The costing of this policy has been complicated by accounting issues. The accounting for student loans has shifted from a cash receipts basis to a fair value basis. One of the effects of the change has been that the effective discount to face value of loans shifts from 12.8% to 33.2%. Further work is still being done on the accounting approach – expected to be completed in time for Budget 2006.
7. Three of the main measures of the ‘cost’ of the no-interest student loan policy are:

- Impact on cash;
- Impact on operating provisions;
- Impact in operating balance.

8. The report suggests that these three measures are as set out in the following table.

$ million increase / (decrease)
2005/06 2006/07 2007/08 2008/09 2009/10
Annual cash impact 157 293 360 367 308
Total impact on operating provision 32 218 256 286 269
Total impact on operating balance 1,649 369 416 442 415

9. The operating provision is higher than earlier versions of this costing has implied. The reasons for this change are:

o review and quality assurance of the model, revealing some areas where it was considered that the model did not represent the new policy appropriately – leading, in most cases, to increases in costs;

o review of information leading to refinement of the behavioural assumptions – a slight revision downwards;

o alignment of Ministry of Education forecast figures with forecast figures supplied to BEFU 2005 by Inland Revenue;

10. The alignment of Ministry of Education and Inland Revenue forecasts meant that the revenue baseline against which the cost impacts of the policy change was measured was shifted up. The effect of this was to increase the costs from that baseline.

Recommended Action

We recommend that the Minister for Tertiary Education:

1. note the contents of the appropriations report for no-interest student loans;

2. note that on 14 November 2005 Cabinet authorised the Minister for Tertiary Education, the Associate Minister of Finance (Hon Trevor Mallard), the Minister for Social Development and Employment, and the Minister of Revenue to make changes to appropriations for the move to an interest free student loan policy (CAB Min (05) 38/3 refers);

3. note that the operating provision impacts of the switch to a no-interest student loan policy are:
$ million 2005/06 2006/07 2007/08 2008/09 2009/10
Total impact on operating provision 32 218 256 286 269

4. note that this cost differs from that advised in the Cabinet paper of 14 November 2005, where the operating provision impact in 2008/09 was stated to be $202.22 million;

5. sign and forward the enclosed letters to the Associate Minister of Finance (Hon Trevor Mallard), the Minister for Social Development and Employment, and the Minister of Revenue ask them to agree to:

1. approve the establishment of “Student Loan Fair Value Recognition” as a new Other expense to be incurred by the Crown appropriation in Vote Revenue from 2005/06;

2. agree that the Vote Social Development “Bad Debt Provisions” appropriation scope statement be:
write-downs of Crown debt administered by MSD due to debt write offs or debt provisions resulting from the need to value debt in accordance with generally accepted accounting practice

3. agree to increase funding for the move to an interest free student loan policy as set out below:

All figures are $m, GST-inclusive where applicable
2005/06 2006/07 2007/08 2008/09 2009/10 & Outyears
Operating Balance Impact 1,648.620 369.451 415.617 442.117 415.156
Debt Impact 156.785 293.466 359.718 366.993 307.712

4. approve the following changes to appropriations to put into effect the decisions in 3 above:

$ million increase / (decrease)
Vote Revenue 2005/06 2006/07 2007/08 2008/09 2009/10 & Outyears
Other Expenses to be Incurred by the Crown:
Student Loan Fair Value Recognition
1,479.435 - - - -

Vote Social Development
Capital Expenditure:
Student Loans 80.534 199.238 266.689 292.912 358.001

Other Expenses to be incurred by the Crown:
Bad Debt Provisions 284.071 418.731 457.393 483.363 504.972

Total Operating 1,763.506 418.731 457.393 483.363 504.972
Total Capital 80.534 199.238 266.689 292.912 358.001

5. agree that the amounts approved under recommendation 4 be included in the 2005/06 Supplementary Estimates and that, in the interim, these expenses be met from Imprest Supply.

Hon. Dr Michael Cullen
Minister for Tertiary Education

Tertiary Education Report: Costs of No-Interest Student Loans – Appropriations Report
Introduction

1. The government agreed on 7 November 2005 to introduce a no-interest policy for student loans [CBC Min (05) 15/1-4 refers] and on 8 November introduced legislation to give effect to that policy.

2. On 14 November 2005 Cabinet authorised the Minister for Tertiary Education, the Associate Minister of Finance (Hon Trevor Mallard), the Minister for Social Development and Employment, and the Minister of Revenue, to make changes to appropriations for the move to an interest free student loan policy based on any revised assumptions provided by officials so that the impacts may be reflected in the forthcoming Half-Year Economic and Fiscal Update.

3. Cabinet also noted that ‘further work is still needed to finalise the financial implications of the ‘no interest’ on student loans policy’. [CAB Min (05) 38/3 refers].

4. This note responds to those decisions and seeks ministers’ approval of the necessary changes to appropriations.

5. The paper:

o reviews the key assumptions used in costing this policy;

o gives the latest estimate of the costs of the policy;

o discusses the accounting issues that impact on the costing;

o identifies the size of the appropriation needed to implement this policy.

The paper is in two parts. Part A deals with the assumptions behind the modelling and Part B with the costs. Appendices set out illustrative material on the modelling, additional detail of costing assumptions and some further information on expected borrowing.

6. The analysis was prepared by the Ministry of Education but with input from and information supplied by the Treasury, Inland Revenue and the Ministry of Social Development.

7. Appendices set out the assumptions and the costing in greater detail.

PART A: MODELLING THE NEW POLICY

The Behavioural Analysis Used in the Costing

8. The modelling of the no-interest policy for student loans is built on an understanding of how students will behave when faced with no-interest loans. The analysis took account of how an economically rational student might behave, and then considered other factors which may impact on this behaviour – both borrowing behaviour and repayment behaviour.

9. Taking account of all factors – including the time value of money – the no-interest Student Loan Scheme is likely to be the most attractive option for many or most students seeking to finance their studies.

10. It is clear, however, that borrowers don’t always act as economically rational agents. Many factors will moderate the tendency to borrow, despite the financial advantages of doing so. Those factors include:

o inertia;

o debt aversion;

o transaction costs – the effort needed to get a loan may outweigh the benefits if the amount of benefit is small and alternative, attractive sources of finance are available;

o conscientious objection – such as moral, religious or ethical objections to acceptance of subsidies;

o lack of awareness or understanding;

o non-financial impediments – such as the intention to travel overseas post-study (when interest would apply) or the wish to maintain a ‘cleaner’ credit record.

11. The modelling also took account of the observed behavioural responses to the introduction in 2000 of the no-interest-while studying policy.

12. The effects on loan take-up and borrowing are set out in the following tables.

Table 1: Forecast changes in the number of borrowers – no-interest policy compared with BEFU 2005

2004 2005 2006 2007 2008 2009 2010
Number of Borrowers Actual Actual* forecast forecast forecast forecast forecast
Forecast under existing policy (BEFU 2005) 157,041 153,274 155,679 158,113 160,852 163,576 165,766
Forecast under no-interest policy 157,041 153,274 175,749 191,403 200,342 205,583 209,482
Increase 20,070 33,290 39,490 42,007 43,716
Percentage increase on BEFU 2005 13% 21% 25% 26% 26%

Table 2: Forecast changes in the amount of borrowing – no-interest policy compared with BEFU 2005

2004 2005 2006 2007 2008 2009 2010
Amount Borrowed [$M] Actual Actual* forecast forecast forecast forecast forecast
Forecast under existing policy (BEFU 2005) 975.6 971.7 1,011.1 1,055.9 1,104.6 1,155.1 1,203.5
Forecast under no-interest policy 975.6 971.7 1,155.6 1,303.6 1,401.3 1,468.0 1,532.8
Increase 144 248 297 313 329
Percentage increase on BEFU 2005 14% 23% 27% 27% 27%

Table 3: Forecast changes in the average borrowing – no-interest policy compared with BEFU 2005

2004 2005 2006 2007 2008 2009 2010
Average Amount Borrowed [$] Actual Actual* forecast forecast forecast forecast forecast
Forecast under existing policy (BEFU 2005) 6,213 6,339 6,495 6,678 6,867 7,062 7,261
Forecast under no-interest policy 6,213 6,339 6,575 6,811 6,995 7,141 7,317
Increase 80 133 128 79 57
Percentage increase on BEFU 2005 1% 2% 2% 1% 1%

13. Points to note about the behavioural response set out above and included in the costing are:

o the extent of response assumed to occur under the no-interest policy is less than observed under the 2000 no-interest-while-studying policy; and

o the extent of response is less than that assumed in the costing of this policy completed on 28 October, which informed the Cabinet papers of 7 November and 14 November;

Repayments

14. The original costing assumed that the level of voluntary repayments would fall straightaway to one eighth of the current level. This revised costing has the level of voluntary repayments reducing over a four year period to 20% of the current level. Table 4 below sets out the phasing of the reduction.

Table 4: Projected voluntary repayments as a proportion of the current level

2005 2006 2007 2008 2009 2010
Share of current level 90% 50% 35% 20% 20% 20%
Estimated value of voluntary repayments ($m) 134 99 77 66 76

15. When the model was reviewed as part of the current costing and taking account of information supplied by Inland Revenue, it was considered that the estimate of voluntary repayments made in the model was too low. The voluntary/compulsory split in the model has since been adjusted to correct for this. That correction would have a negligible impact on the operating provision.

Payments by Those Overseas

16. Currently, around 6% of all those with loans are declared as resident overseas. That is an undercount of those overseas as significant numbers of borrowers live overseas without ever telling Inland Revenue. It is estimated that the number of borrowers actually overseas is at least twice the number declared overseas. Around 20% of the total aggregate Student Loan Scheme balance is estimated to be held overseas.

17. The penalty payments amnesty agreed to by Cabinet on 7 November and provided for in the current bill is expected to have the effect of encouraging significant numbers - the modelling has assumed 7,000 - currently overseas to declare themselves as non-resident. That declaration will provide additional interest revenue as such people now receive substantial interest write-offs.

18. In the longer term, the better policing of the requirement to declare residency through data-matching will provide an assurance that a high proportion of those who are not resident in New Zealand declare themselves as overseas. It is assumed that 28,000 additional borrowers would be declared overseas over five years.

The Effects of the Policy on Overall Debt Levels

19. This new forecast assumes that the number of borrowers would rise in 2007 by 21% on the BEFU 2005 forecast under the existing policy. In 2010, the number of borrowers be up by nearly 44,000 or 26% on the current policy. Borrowing would rise by similar levels.

20. The face value of the aggregate Student Loan Scheme balance would rise, but less sharply. The forecast for the value of aggregate balance on 30 June 2010 under the existing policy was $10,813 million. Under the new policy that would increase to $11,299 million, an increase of 4.5%.

21. Fuller information on the basis of the assumptions and an explanation of the analysis leading to those assumptions are set out in Appendix One to this paper. Appendix Two summarises the main assumptions.

PART B: COSTS OF THE NEW POLICY

Introduction

22. The costs of the new no-interest student loan policy derive from three main factors.

o the effects of the behavioural assumptions discussed in Part A of this paper – the amount of extra borrowing that is forecast to occur and the effects of the new policy on repayment behaviour;

o the change in accounting policy to value student loan on a ‘fair value’ basis, rather than on the basis of expected cash receipts;

o the change in the fair value of the scheme that results from the fact that interest is no longer to be charged.

23. The modelling that has led to the costing set out below is particularly complex. Therefore, the figures have been subject to ongoing review and refinement and hence, have changed several times. In addition, the accounting treatment of the changes is still being discussed with auditors and external advisors and hence, should be treated as provisional at this stage. It is expected that the outstanding issues will be resolved for Budget 2006.

24. This part of the paper describes the main features of the costing.

The Shift in the Basis for Accounting for Student Loans – Fair Value vs Cash Receipts

25. The costing of the new policy is complicated by the shift in the accounting basis. Under the current accounting policy, for each dollar of student loan borrowing, 12.8 cents is not expected to be recovered - hence the student loan principal is valued at 87.2 cents in the dollar. This has been called a Doubtful Debt Provision (DDP) and means that the accounting is done on a cash receipts basis.

26. The current cash receipts accounting approach does not fairly reflect the costs of a no-interest loan scheme. In addition, in 2007, the government will have to comply with a new international financial reporting standard, IAS39. While considerable work is still needed to confirm the approach required under the new standard, this new standard will be applied in 2005/06. Under the new approach, the loans will be initially recognised at fair value, which reflects the amount that a willing buyer would pay for the stock of debt and takes into account the expected timing of repayments. Once a loan is brought onto the books, it will be subsequently measured at amortised cost, which includes recognising any imputed interest income. .

27. Under the current student loan policy, the fair value of the principal of each student loan is 81.6 cents in the dollar .

28. Under the new no-interest policy, the fair value is expected to fall to 66.8% in 2005/06 .

29. Further work is being undertaken to clarify the impact of the new financial reporting standard, which requires loans to be initially valued at fair value and then subsequently tested for impairment at each reporting date. The standard is particularly complicated and is not clear when applied to long term debt. External advice is being sought. While the fair value impact is known with reasonable certainty, the impairment aspect is not and may require a further adjustment to appropriations which, if required, will be sought as part of the Budget Economic and Fiscal Update. Thus the financial approach and consequently, the figures, should be treated as indicative and will be updated as part of Budget 2006.

30. Earlier advice to ministers separated the revaluation impacts into those due to the accounting changes and those due to the change to the no-interest policy. The impact of the accounting policy change was treated as a remeasurement, not a revaluation. The Auditor-General has now, however, ruled that it is necessary to make an appropriation to cover both sources of the revaluation.

31. The fair value impacts of this change - i.e. the revaluation change due to the shift to a no-interest policy - should not impact the operating provisions. This approach is consistent with how the costs for 2005/06 are treated and with the treatment of other policy-related costs, i.e. the costs incurred in each year are considered against operating provisions, but not the fair value components of those costs (such as the time value of money).

32. Of the write-down of the value of $1,479 million, around $455 million is due to the application of the accounting policy change to the existing stock of debt at 31 October 2005.

Table 5: No-interest student loan policy: Revaluation impacts
Revaluation impacts ($ millions)
Book value at 1 July 2005 6,464.5
Fair value at 1 July 2005 6,121.3
Face value of gross student loan debt at 31 October 2005 7,730.1
Current provisions -1,084.6
Net student loan debt - book value at 31 October 2005 under current accounting treatment 6,645.5
Fair value of the current scheme 6,190.3
Revaluation due to change in accounting policy 455.2
Fair value at 31 October 2005 under no-interest policy 5,166.1
Fair value write-down due to moving to no-interest policy 1,024.2
Total revaluation impact at 31 October 2005 1,479.4

The Overall Position

33. The tables and comments below summarise the overall budget impacts of the new policy.

34. The net effect of the new policy on the cash position of the Student Loan Scheme in 2005/06 is $157 million relative to BEFU 2005, resulting from additional borrowings of $81 million and lower repayments of $76 million.

Table 6: No-interest student loan policy: Cash impact

$ million increase / (decrease)
2005/06 2006/07 2007/08 2008/09 2009/10
Current borrowings 1,017.00 1,062.00 1,111.00 1,163.00 1,163.00
New borrowings 1,097.53 1,261.24 1,377.69 1,455.91 1,521.00
Change in borrowings 80.53 199.24 266.69 292.91 358.00
Current repayments 655.00 735.00 819.00 909.00 909.00
New repayments 578.75 640.77 725.97 834.92 959.29
Change in repayments -76.25 -94.23 -93.03 -74.08 -50.29
Annual cash impact 156.78 293.47 359.72 366.99 307.71

35. This cash impact is the total impact on gross debt. Part of this impact on gross debt will be managed through the operating provisions as set out in Table 7. The impact on the operating provisions is defined as the change in income from net interest charged under the new scheme compared to the old scheme, discounted for capital write-offs.

Table 7: No-interest student loan policy: Operating provision impact

$ million 2005/06 2006/07 2007/08 2008/09 2009/10
Decrease in net interest
Current policy 303 345 391 437 437
New policy 289 137 150 162 173
Impact 14 208 241 275 264
Decrease in capital write-offs
Current policy 15 17 18 19 19
New policy 15 16 17 19 18
Impact -1 -1 -1 0 -1
Provision
Current policy 83 84 85 86 86
New policy at old rates 101 95 100 98 92
Impact 18 11 15 12 6
Total impact on operating provision 32 218 256 286 269

36. The current costing has produced results that differ from the 28 October costing considered as part of the two Cabinet papers on the no-interest policy. The main reasons for the changes are:

o review and quality assurance of the model, exposing some areas where it was considered that the model did not represent the new policy appropriately – leading, in most cases, to increases in costs;

o review of information leading to refinement of the behavioural assumptions – a slight revision downwards;

o alignment of Ministry of Education forecast figures with forecast figures supplied to BEFU 2005 by Inland Revenue;

o confirmation from the Student Loan Scheme actuarial consultants of the fair valuation of the scheme under the new policy, with that valuation replacing the earlier estimates.

37. The most significant of those changes is the fair valuation change. Earlier costings used as a provisional estimate that the fair valuation of the scheme under the no-interest policy would be about 63% of the scheme face value. The change to a fair valuation of 66.8% has had a substantial effect on the revaluation of the scheme. The difference in the amount of revaluation write-down is over $250 million.

38. The operating provision was particularly affected by the first three of those factors. The alignment of the agencies’ forecast figures, in particular, meant that the revenue baseline, against which the forecasts were compared, was revised upwards.

Appropriation Required

39. From the costings set out in Tables 5 to 7 above, agencies have derived the need for the appropriations set out in the recommendations below.

Recommendations

40. It is recommended that you:

6. note the contents of this report;

7. note that on 14 November 2005 Cabinet authorised the Minister for Tertiary Education, the Associate Minister of Finance (Hon Trevor Mallard), the Minister for Social Development and Employment, and the Minister of Revenue to make changes to appropriations for the move to an interest free student loan policy (CAB Min (05) 38/3 refers);

8. note that the operating provision impacts of the switch to a no-interest student loan policy are:
$ million 2005/06 2006/07 2007/08 2008/09 2009/10
Total impact on operating provision 32 218 256 286 269

9. note that this cost differs from that advised in the Cabinet paper of 14 November 2005, where the operating provision impact in 2008/09 was stated to be $202.22 million;

10. sign and forward the enclosed letters to the Associate Minister of Finance (Hon Trevor Mallard), the Minister for Social Development and Employment, and the Minister of Revenue, asking them to:
6. approve the establishment of “Student Loan Fair Value Recognition” as a new Other expense to be incurred by the Crown appropriation in Vote Revenue from 2005/06;

7. agree that the Vote Social Development “Bad Debt Provisions” appropriation scope statement be:
write-downs of Crown debt administered by MSD due to debt write offs or debt provisions resulting from the need to value debt in accordance with generally accepted accounting practice

8. agree to increase funding for the move to an interest free student loan policy as set out below:

All figures are $m, GST-inclusive where applicable
2005/06 2006/07 2007/08 2008/09 2009/10 & Outyears
Operating Balance Impact 1,648.620 369.451 415.617 442.117 415.156
Debt Impact 156.785 293.466 359.718 366.993 307.712

9. approve the following changes to appropriations to put into effect the decisions in 3 above:

$ million increase / (decrease)
Vote Revenue 2005/06 2006/07 2007/08 2008/09 2009/10 & Outyears
Other Expenses to be Incurred by the Crown:
Student Loan Fair Value Recognition
1,479.435 - - - -

Vote Social Development
Capital Expenditure:
Student Loans 80.534 199.238 266.689 292.912 358.001

Other Expenses to be incurred by the Crown:
Bad Debt Provisions 284.071 418.731 457.393 483.363 504.972

Total Operating 1,763.506 418.731 457.393 483.363 504.972
Total Capital 80.534 199.238 266.689 292.912 358.001

10. agree that the amounts approved under recommendation 4 be included in the 2005/06 Supplementary Estimates and that, in the interim, these expenses be met from Imprest Supply.

Appendix One: The Behavioural Analysis Used in the Costing

41. The modelling of the no-interest policy for student loans is built on an understanding of how students will behave when faced with no-interest loans. The starting point for the behavioural analysis is to look at how an economically rational student might behave, and then consider other factors which may impact on this behaviour – both borrowing behaviour and repayment behaviour.

42. Faced with the need to pay fees and living costs, an economically rational student would choose the most attractive financing option - subject to transaction costs. While not all of options may be available to any one student, the options include:

o using their own savings, the cost of which is the loss of interest they could have earned had they invested these savings or the use of items that those savings might otherwise purchase;

o using someone else’s savings (eg having their parents pay for them) – again, the cost is equivalent to the interest the parents forego;

o borrowing from a lending institution at market rates (although this option may not be available to everyone, given the need to provide the lender with some form of security);

o borrowing from family or friends at a concessionary or zero interest rate; or

o borrowing from the government through the Student Loan scheme.

43. In addition, some students qualify for student allowances, which provides a ‘free’ source of funds for living costs but which reduce the student’s borrowing entitlement.

44. One of the key influences on students’ choices is expected to be advice from parents – with many parents expected to identify the financial advantages of borrowing under the new policy and advising their children to borrow.

45. There is a range of reasons why a student may choose an approach which may not appear to be the most financially attractive option. These are tabulated below:

Table 1: Factors influencing a decision to borrow
Factor Effects on Student Loan Scheme uptake Comment
Transaction costs May dampen response Even if the Student Loan Scheme were the most financially attractive option, the costs (financial and non-financial) associated with applying for, and monitoring and repaying of the loan may be such that alternative options (such as using savings or borrowing from family) may be preferred.
Certainty May dampen response Potential borrowers will look at the extent to which the conditions applying to different options are likely to persist. There will be some who fear that the government’s loan policy might change following a change of government.
Non-financial impediments May dampen response Having a student loan may restrict the opportunities for other borrowing – for instance if repayment arrangements meant the person would be unable to secure enough borrowing from other sources - such as a loan to buy a house.
Repayment arrangements Mixed Some options may have income-contingent repayments, others may offer funding with a “when you can afford it” repayment arrangement.
Inertia Will dampen response While it may be worthwhile financially to adopt a particular strategy, there may be insufficient financial gain for the student to change current arrangements. What may influence this trend is the fact that students (who are the only people who can actually draw down from the loan scheme) will not receive any immediate benefit from the change – as the benefit occurs only post-study.
Lack of awareness or understanding May dampen response, if slightly Some potential borrowers will not have an understanding of the financial advantages of the new policy. The high profile of and the level of debate on the new policy may, however, work to counter this issue, especially as parents with a more sophisticated understanding of finance may influence their children.
Responses to debt May dampen response Some individuals may have an aversion to holding debt (even if there are financial gains and if their net position is positive. There is, however, some growth in social acceptance of debt and an increasing propensity for people to borrow.
Moral or religious grounds Will dampen response, if slightly Some people might not wish to avail themselves of some options – perhaps for religious reasons - even though they may be entitled to them. For example, some may not wish to borrow from family or friends, others may not wish to accept “charity” (by way of concessionary borrowing rates) whether from family, friends, or the government.

46. Taking account of all factors – including the time value of money – the no-interest Student Loan Scheme is likely to be the most attractive option for many or most students seeking to finance their studies. The no-interest policy significantly increases the financial benefit of government student loans compared with alternative sources of finance and compared with the current loan scheme rules. This is because borrowers will not be charged any interest for the life of the loan rather than just while studying (providing they continue to meet the NZ residence criterion). Because the arbitrage opportunity is available for that much longer (ie the money that would otherwise have been spent on fees and other costs can be invested in an interest-earning account), and because it is (probably) the cheapest form of finance they will have access to, there are incentives to take out the loan and to maintain only minimum repayments for as long as it takes to pay off the loan.

47. While it is useful to start from the presumption that many students would seek to take advantage of the arbitrage opportunity offered by the no-interest policy, the factors listed above suggest that there are factors that will dissuade some students from taking up this option. In addition, many of those who stand to gain an advantage from the new policy are already accessing the Student Loan Scheme. And a proportion of those who declined to take advantage of the current no-interest-while-studying policy can be assumed to be debt averse and unlikely to borrow under the new policy.

The Experience of the No-Interest-While-Studying Implementation

48. There is little empirical evidence, either in New Zealand or overseas, which provides guidance on how students or those with student loans might respond to a change to a no-interest student loan policy. The most reliable factual guide to the expected response of students to the no-interest policy comes from looking at what happened in the years following the implementation in 2000 of the ‘no-interest-while-studying’ policy.

49. The Student Loan Scheme take-up rate shifted from 50% in 1999 to 56% in 2001 and 60% in 2003 in response to the no-interest-while-studying policy. These changes, however, were influenced by the trends in enrolments at the time. In particular, it was over that time that enrolments grew very quickly at Te Wānanga o Aotearoa. Given the low propensity of wānanga students to borrow under the Student Loan Scheme, the growth at that institution at that time tended to mute the actual shift in take-up rates. The actual shift in the take-up rate was also moderated by a shift from full-time to part-time students – given that part-timers have a low propensity to borrow.

50. For the purposes of the present modelling, the student population was disaggregated into sub-groups and the behaviour of each sub-group following the 2000 policy was analysed, to provide a base for thinking about the likely responses to the new policy.

51. As an example, we consider changes in behaviour among full-time students in colleges, polytechnics and universities. During the two years following the implementation of the policy, the number of full-time university, college and polytechnic borrowers rose by 21%, despite the fact that the number of loan-eligible full-time students at those institution types fell by 5% over that time. This represented a much larger response that the overall change in take-up rates across the scheme as a whole.

52. While the scale of the response to the 2000 policy was masked by other factors, it needs to be recalled that at the time the no-interest-while-studying policy was approved, officials forecast the overall draw-down rate for fees would rise from 50% to 95%. In other words, officials advised that, of the 50% of students who were eligible to borrow for fees but who were not at that time taking out a student loan, the absence of interest while studying would encourage most of them – around nine in ten - to borrow. The data suggests that only around one-quarter of this group responded in this way, a significantly lower response than expected by officials.

Groups Affected and Key Behavioural Assumptions

53. There are three groups affected by the proposed policy change and whose behaviour could be influenced by the change in policy:

o students currently studying and eligible for a student loan, but who have either not drawn a loan or have not drawn the maximum available to them under the loan;

o new students entering tertiary education and who will become eligible for a student loan for the first time, and

o those with existing student loans but who are no longer studying.

54. The first two groups will drive the loan take-up and draw-down assumptions. The third group will drive assumptions about voluntary repayments (i.e. repayments above the minimum required given their assessable income). The third group will also be affected by the aspects of the whole policy that affect those who travel overseas.

55. The modelling also separated the behaviour of full-time and part-time students. The incentives and disadvantages of the new policy are different for part-time and full-time students.

56. We have no information on the changes in the financial characteristics of students over this period (ie whether the new/additional students are any more or less able to pay their own way, or are any more or less debt averse, etc). We have therefore assumed the profile is unchanged in this respect.

57. While the modelling for the no-interest policy takes account of the 2000 experience, it also recognises that the incentives of the new policy are different. The new no-interest policy offers the no-interest benefit for a longer period and also reduces the risks faced by those who take advantage of the concession for arbitrage. Because the no-interest-while-studying policy offered an interest write-off for a short period (typically 1.5 years ), there are likely to have been many people for whom the financial gain from rearranging their affairs to take advantage of the policy did not outweigh the other factors (such as compliance costs, debt aversion, and moral objections).

58. In particular, the incentives on part-time students to enter the Student Loan Scheme are much stronger under the new no-interest policy than they were under the 2000 no-interest-while-studying arrangements. This is because many part-time students didn’t qualify for the benefits of the 2000 policy and because part-time students tend to have a shorter duration of study than full-timers – and consequently, those who did qualify for the benefit of no interest while studying stood to gain a benefit for a short time only. Further, the majority of full-time students already borrow – there is little ‘head-room’ for further increases in uptake among full-timers, whereas relatively few part-timers borrow . Thus, part-timers might be expected to be more inclined to respond to the new policy .

Assumed Take-up Rates

59. The effective draw-down rates assumed in the modelling for some of the main groups and the main loan components are set out below:

Table 2: Forecast student loan take-up rates
Take-up Rates 2004 2005 2006 2007 2008 2009 2010
Actual Actual* forecast forecast forecast forecast forecast
Course fees - Full-time Students
Forecast under existing policy 67% 65% 65% 66% 66% 66% 67%
Forecast under no-interest policy 67% 65% 71% 75% 77% 77% 77%
Additional Borrowings [$M] 0 0 44 84 101 105 107
Course-related cost - Full-time Students
Forecast under existing policy 46% 43% 43% 44% 44% 45% 45%
Forecast under no-interest policy 46% 43% 53% 60% 62% 62% 62%
Additional Borrowings [$M] 0 0 18 30 36 37 38
Living cost - Full-time Students
Forecast under existing policy 41% 38% 39% 39% 39% 40% 40%
Forecast under no-interest policy 41% 38% 47% 51% 51% 51% 51%
Additional Borrowings [$M] 0 0 59 88 104 108 114
Course fees - Part-time Students
Forecast under existing policy 16% 16% 16% 16% 17% 17% 17%
Forecast under no-interest policy 16% 16% 25% 30% 33% 35% 37%
Additional Borrowings [$M] 0 0 17 34 41 46 52
Course-related cost - Part-time Students
Forecast under existing policy 12% 12% 12% 12% 13% 13% 13%
Forecast under no-interest policy 12% 12% 20% 26% 28% 30% 31%
Additional Borrowings [$M] 0 0 7 12 15 16 18
Full-time Borrowers – Overall
Forecast under existing policy 73% 70% 70% 71% 71% 71% 71%
Forecast under no-interest policy 73% 70% 75% 79% 81% 81% 81%
Additional Borrowings [$M] 0 0 121 202 241 250 259
Part-time Borrowers – Overall
Forecast under existing policy 17% 17% 18% 18% 18% 18% 19%
Forecast under no-interest policy 17% 17% 26% 32% 35% 37% 39%
Additional Borrowings [$M] 0 0 23 46 55 63 70
All Loan Borrowers
Forecast under existing policy 53% 51% 51% 51% 52% 52% 52%
Forecast under no-interest policy 53% 51% 57% 62% 64% 65% 66%
Additional Borrowings [$M] 0 0 144 248 297 313 329

Repayments

60. It is currently estimated that of total repayments of around $655 million per year, approximately $200 million represents payments over and above the minimum required. These are called ‘voluntary repayments’.

61. At the end of the year, Inland Revenue cannot distinguish between over-payments that are intentional and those that are unintentional (for instance, resulting from changes in assessable income). All over-payments are identified and each borrower is asked whether he/she would like to be reimbursed, or whether the payment should be used to reduce the loan. The estimate of $200 million is of those that indicate they would like to reduce their loan.

62. Therefore, individuals will be explicitly asked at the end of the current tax year whether they wish to pay off their loans faster than the minimum rate. Again, applying the behavioural framework described earlier in this paper, there are good reasons for suggesting that some of these individuals would seek to have any payment reimbursed as they could more fruitfully put this money towards some other use (eg saving for a house, etc).

63. This suggestion is reinforced by the most recent voluntary repayments data from the Inland Revenue. These payments have been increasing over recent years, with monthly payments significantly above the same month a year earlier. Over the past four months, there has been a noticeable change in this pattern, with the period since August (following the Labour Party’s announcement of the no-interest proposal) showing a small decline over payments in the previous year and indicating a decrease on the monthly payments in the year to July. It is possible (although not certain) that this reduction in payments is a response to the announcement of the Labour Party’s policy.

64. Voluntary repayments can have a significant impact on loan balances; therefore, voluntary repayments will continue to be attractive to those who would prefer not to hold debt.

65. Because individuals had to make an explicit decision to leave the additional payments with the IRD, officials felt there were grounds for believing many would take advantage of the opportunity to use the funds in some other way.

66. There remain some incentives to make voluntary repayments under the new policy. Those traveling overseas may want to avoid interest through reducing their loans before leaving. And some people will reduce their loans in order to increase their disposable income to qualify for a higher housing loan.

67. On balance, officials have included an assumption that repayments will fall to 20% of the current level over a three year period. It is expected that voluntary repayments will reduce to 90% of the current level in 2005, 50% in 2006, 35% in 2007 and 20% in 2008 and thereafter.

Payments by Those Overseas

68. Currently, around 6% of all those with loans are declared as resident overseas. That is an undercount of those overseas as significant numbers of borrowers live overseas without ever telling Inland Revenue. It is estimated that the number overseas is twice the number declared overseas. Around 20% of the total debt is held overseas.

69. The amnesty on penalty payments for certain borrowers that will apply for the twelve months from 1 April 2006 is expected to have the effect of encouraging significant numbers - the modelling has assumed 7,000 - currently overseas to declare themselves as non-resident. That declaration will provide additional interest revenue as such people now receive substantial interest write-offs.

70. In the longer term, the better policing of the requirement to declare residency through data-matching will provide an assurance that a high proportion of those who are not resident in New Zealand declare themselves as overseas. It is assumed that 28,000 borrowers over five years will be declared as non-resident.

The Effects of the New Policy on Borrowing

71. Under the new policy, it is estimated that the number of borrowers would rise in 2007 to 191,000, an increase of 33,000 or 21% on the forecast under the existing policy. In 2010, the number of borrowers would reach 209,000, up by nearly 44,000 or 26% on the current policy.

72. The amounts borrowed would increase by similar percentages. In 2007, the amount borrowed would rise by $248 million or 23% to reach $1,304 million, while in 2010, the extra borrowing would be $329 million, up 27% on existing forecasts.

73. The face value of aggregate student loan debt would also rise, but less sharply. The forecast for the value of aggregate debt on 30 June 2010 under the existing policy was $10,813 million. Under the new policy that would increase to $11,299 million, an increase of 4.5%.

74. Fuller details of the effects of the policy on borrowing levels are set out in Appendix 3.

Conclusion

75. The cost of the no-interest policy is largely driven by assumptions in three key areas, namely:

o the changes in the take-up and draw-down rates that result from the new policy and especially, the changes in the rates for full-time students

o the rate at which voluntary repayments (above the minimum) are maintained; and

o the rate at which New Zealanders resident overseas declare their residency status and maintain their payment schedules.

76. There is little hard information to inform the judgements required to develop the assumptions about how behaviour might change as a result of the policy change. What information there is largely draws on the changes that were observed following the decision in 2000 to remove interest on loans while studying. At that time, for the key group of full-time students, the percentage of students who were eligible for a loan and took out a loan increased by around 15%, leading to an increase in take-up rates of around 18 percentage points.

77. This response was much lower than officials had projected when that policy was constructed. This suggests that the 1.5 year interest rate concession was sufficient to induce some to take up the option, but that other factors (such as transaction and compliance costs, inertia, debt aversion, etc) were sufficient to outweigh any benefits for a number of students.

78. The no-interest policy presents a more attractive option for students, as it provides for no-interest for the duration of the loan (providing the student maintains the New Zealand residency). Nevertheless there will remain some students for whom the financial incentive may still not be sufficient to warrant taking on this choice of financing, recognising factors such as debt aversion and the moral objection to accepting a subsidy from the government.

79. On balance, officials agree that the take-up rates will increase across all categories of students loans. For the most critical assumptions, the increases are of a lesser scale than those observed at the time interest was removed for the period students were studying.

1 Appendix Two: Costing Assumptions for No-Interest Student Loans

The baseline for this costing is BEFU 2005. The economic assumptions used in BEFU 2005 are unchanged.

Assumptions:
 The no-interest provision applies to all borrowers in New Zealand, not just graduates.
 The interest rate charged to overseas based borrowers is 7 percent in 2006/07 and formula driven after that.
 The number and debt held by those borrowers who are actually overseas is greater than the self-declared non-resident. Results from the integrated dataset on Student Loan Scheme borrowers suggest that 20 percent of aggregate debt is overseas at 31/3/2006, compared with 8 percent of debt non-resident. The amnesty policy and recent operational changes at IRD, together with future data matching will move the status of a number of borrowers to non-resident. It is assumed that 7,000 borrowers holding $120 million of debt will become non-resident as a result of the amnesty, and a further 28,000 borrowers over a five year period making the transition as a result of the other changes. This will significantly increase the interest accrued on the debt.
 It has been assumed that patterns of departure from and return to New Zealand will not change significantly.
 There is no adjustment made for the lower than average collection rates of overseas debt.
 No change in participation in tertiary education is assumed in the baseline.
 No change in eligibility rules for loans and allowances have been factored in.
It is assumed that voluntary repayments will fall. Currently, it is estimated that voluntary repayments account for around 30% of all repayments or $200 million per annum. The new policy has been assumed to reduce voluntary repayments as shown in the table below:

Table 2.1: Projected voluntary repayments as a proportion of the current level
2005 2006 2007 2008 2009 2010
90% 50% 35% 20% 20% 20%

- Based on 2004 data, every dollar spent on allowances will reduce living cost borrowing by 27 cents. No changes to allowances eligibility rules have been factored in to this costing at this stage.

- It is assumed that uptake rates will increase over 2006 – 2009, in the same way as uptake rates grew slowly following the introduction of no interest while studying in 2000. The main driver of increased borrowing is in the increased numbers of borrowers as current borrowers already borrow close to their maximum entitlement.

ENDS

© Scoop Media